NUNC COGNOSCO EX PARTE

THOMAS J. BATA LIBRARY

TRENT UNIVERSITY
 Digitized by the Internet Archive
in 2019 with funding from
Kahle/Austin Foundation

https://archive.org/details/galileoartofreasOOOOfino
GALILEO AND THE ART OF REASONING
BOSTON STUDIES IN THE PHILOSOPHY OF SCIENCE

EDITED BY ROBERT S. COHEN AND MARX W. WARTOFSKY

VOLUME 61
 MAURICE A. FINOCCHIARO
Dept, of Philosophy, University of Nevada, Las Vegas, U.S.A.

GALILEO
AND THE ART OF
REASONING
Rhetorical Foundations of Logic
and Scientific Method

D. REIDEL PUBLISHING COMPANY

DORDRECHT : HOLLAND / BOSTON : U.S.A.

LONDON : ENGLAND
 Library of Congress Cataloging in Publication Data

Finocchiaro, Maurice A. 1942—

Galileo and the art of reasoning.

(Boston studies in the philosophy of science ; v. 61)

Bibliography: p.
Includes index.
1. Galilei, Galileo, 1564-1642—Logic. 2. Logic, Modern—
History. 3. Reasoning. 4. Science—Methodology—History.
5. Galilei, Galileo, 1564-1642. Dialogo . . . dove ... si discorre sopra i
due massimi sistemi del mondo. 6. Solar system—Early works to
1800. 7. Astronomy—Early works to 1800. I. Title. II. Series.
B785.G24F56 160 80-15232
ISBN 90-277-1094-5
ISBN 90-277-1095-3 (pbk.)

Published by D. Reidel Publishing Company,

P.O. Box 17, 3300 AA Dordrecht, Holland.

Sold and distributed in the U.S.A. and Canada

by Kluwer Boston Inc., Lincoln Building,
160 Old Derby Street, Hingham, MA 02043, U.S.A.

In all other countries, sold and distributed

by Kluwer Academic Publishers Group,
P.O. Box 322, 3300 AH Dordrecht, Holland.

D. Reidel Publishing Company is a member of the Kluwer Group.

All Rights Reserved

Copyright © 1980 by D. Reidel Publishing Company, Dordrecht, Holland
No part of the material protected by this copyright notice may be reproduced or
utilized in any form or by any means, electronic or mechanical,
including photocopying, recording or by any informational storage and
retrieval system, without written permission from the copyright owner

Printed in The Netherlands

 TABLE OF CONTENTS

EDITORIAL PREFACE vii

PREFACE AND ACKNOWLEDGMENTS ix

INTRODUCTION / Galileo’s Dialogue and Western Culture xv

PART I: GALILEO’S DIALOGUE

1. Faith Versus Reason: The Rhetorical Form and Content of

Galileo’s Dialogue 3
2. Fact and Reasoning: The Logical Structure of Galileo’s Argument 27
3. Emotion, Aesthetics, and Persuasion: The Rhetorical Force of
Galileo’s Argument 46
4. Truth and Method: The Scientific Content of Galileo’s Dialogue 67
5. Theory and Practice: The Methodological Content of Galileo’s
Science 103

PART II: LOGICAL AND METHODOLOGICAL

CRITIQUES

6. Concreteness and Judgment: The Dialectical Nature of Galileo’s

Methodology 145
7. The Primacy of Reasoning: The Logical Character of Galileo’s
Methodology 167
8. The Rationality of Science and the Science of Rationality:
Critique of Subjectivism 180
9. The History of Science and the Science of History: Critique of
Apriorism 202
10. The Erudition of Logic and the Logic of Erudition: Critique of
Galileo Scholarship 224
11. The Psychology of Logic and the Logic of Psychology: Critique
of the Psychology of Reasoning 256

°>icA o<o
VI TABLE OF CONTENTS

12. The Rhetoric of Logic and the Logic of Rhetoric: Critique of

the New Rhetoric 273
13. The Logic of Science and the Science of Logic: Toward a
Science of Reasoning 293

PART III: THEORY OF REASONING

14. Propositional Structure: The Understanding of Reasoning 311

15. Active Involvement: The Evaluation of Reasoning 332
16. Galileo as a Logician: A Model and a Data Basis 343
17. Criticism, Complexity, and Invalidities: Theoretical Considera¬
tions 413

CONCLUDING REMARKS / Toward a Galilean Theory of Ration¬

ality 432

APPENDIX / Page Concordance of Dialogue Editions 440

SELECTED BIBLIOGRAPHY 454

INDEX 464
 EDITORIAL PREFACE

The work of Galileo has long been important not only as a foundation of
modern physics but also as a model — and perhaps the paradigmatic model
— of scientific method, and therefore as a leading example of scientific
rationality. However, as we know, the matter is not so simple. The range of
Galileo readings is so varied that one may be led to the conclusion that it is a
case of chacun a son Galileo; that here, as with the Bible, or Plato or Kant or
Freud or Finnegan’s Wake, the texts themselves underdetermine just what
moral is to be pointed. But if there is no canonical reading, how can the texts
be taken as evidence or example of a canonical view of scientific rationality,
as in Galileo? Or is it the case, instead, that we decide a priori what the norms
of rationality are and then pick through texts to find those which satisfy
these norms? Specifically, how and on what grounds are we to accept or
reject scientific theories, or scientific reasoning? If we are to do this on the
basis of historical analysis of how, in fact, theories came to be accepted or
rejected, how shall we distinguish ‘is’ from ‘ought’? What follows (if anything
does) from such analysis or reconstruction about how theories ought to be
accepted or rejected?
Maurice Finocchiaro’s study of Galileo brings an important and original
approach to the question of scientific rationality by way of a systematic read¬
ing and reconstruction of Galileo’s project and of the modes of his reasoning.
In effect, Finocchiaro suggests that the standard picture of the scientific revo¬
lution may be fundamentally mistaken. Where, on older views, the scientific
revolution was seen as a radical innovation in method as well as in content,
and even as marginal to the dominant cultural themes of the Renaissance —
largely literary, political, aesthetic, rhetorical — Finocchiaro proposes that the
scientific revolution, in one of its major figures, was characterized fundamen¬
tally by the same emphases on dialectic and rhetoric which marked the
dominant themes of Renaissance humanism. Thus, the scientific revolution
was as much a matter of persuasion as of discovery, its success as science
crucially a matter of the triumphs of practical reasoning and persuasion in
other modes beyond the usually acknowledged one of mathematical rational¬
ity in terms of proofs and demonstrations. If this is correct, not mathematical
rationality alone but scientific judgment, the instruments of persuasion and

vii
viii EDITORIAL PREFACE

of practical argument, the alogical components of discourse, played a major

role in the novel mode of scientific reasoning.
Finocchiaro argues that “rhetoric is sometimes crucial in science; and
hence rhetoric has an important role to play in scientific rationality and the
rhetorical aspects of science should not be neglected”. His argument is based
on a detailed and careful examination of the text of the Dialogues .... What
Finocchiaro argues will, we are sure, become a basic proposal for enlargement
and revision in the canonical program of logical reconstruction which has so
strongly marked and nearly dominated modern philosophy of science. He
offers an alternative rational reconstruction of Galileo’s scientific thought.
Logic is also included, of course, but in a wider sense: “I am advocating”, he
writes, “a practice oriented logic or theory of reasoning, and thus it is logical
principles that are to be tested by and derived from appropriately selected
actual reasoning and not the other way around” (291, fn. 8). And so “a
general logic seems possible, if we concentrate on the critical understanding
of actual reasoning, which concentration carries along with it an empirical,
historical, concrete or contextual, practical and critical appreciation” (305).
In this book, Finocchiaro not only offers us a sustained analysis of the
Dialogues in terms of the structure of the work, Galileo’s reasoning, his uses
of rhetoric. Beyond this, which is the heart of his work, he develops a critical
and appreciative discussion in depth of the principal Galileo studies: those of
Koyr6, Feyerabend, Drake, Clavelin, Kuhn, Shapere. Further, Finocchiaro
gives us a series of sustained studies of topics in the ‘science of rhetoric’ as
an initial sketch of a theory or reasoning which enlarges scientific rationality
beyond the limits of formal or deductive reason. We may say that Finocchiaro
rounds out the traditional approaches to the ‘logical syntax’ and semantics
of scientific discourse with a serious construction of the (long neglected)
pragmatics of this discourse as an instrument of argument and persuasion. In
a sense, his work is an interesting and promising supplement to the syntactic
and semantic reconstruction of the classic Vienna Circle achievement. Nor has
Finocchiaro merely appended ‘externalist’ factors to the ‘internalist’ logical
empiricism; rather, for him the normative is embodied within the actual and
practical reasoning of science, internal to what we could call ‘scientific culture’.
Understanding scientific reasoning must include, then, practical scientific
judgment as part of the ‘science of rhetoric,’ not just the rhetoric of science.

Center for the Philosophy and History of Science ROBERT S. COHEN

Boston University MARX W. WARTOFSKY
May 1980
 PREFACE AND ACKNOWLEDGMENTS

The investigations in this book are interdisciplinary in the sense that different
parts are likely to appeal primarily to individuals in different fields. A layman
will find in the Introduction reasons why he ought to be interested in Galileo
Galilei’s Dialogue Concerning the Two Chief World Systems, and in Chapters
1—5 discussions of the aspects of the book most relevant to him. These
chapters are also addressed to scientists, who of course will find greater mean¬
ing therein. Rhetorical scholars will find Chapters 1—3, 8—9, and 12—17 of
some interest; cognitive psychologists may appreciate Chapters 11 and 16,
linguists Chapters 13 and 16. Religious historians will find here the most
exhaustive study ever published of the book that caused “the greatest scandal
in Christendom”, Chapters 1—5,6 and 16 being of greatest relevance to them.
These Chapters may also interest cultural historians. Logicians will appreciate
Chapters 11 — 17 the most, while philosophers of science can read Chapters
8—9 first and then Chapters 4—7. For historians of science the critique of
Koyr6 in Chapter 9, and of Strauss, Favaro, Drake, and Clavelin in Chapter 10
can serve as the introduction to Chapters 1—5 and 16.
Accordingly this fact multiplies the number of potential misunderstandings
and objections on the part of readers, but this is a risk well worth taking, not
only because of the interdisciplinary nature of Galileo’s own work, but also
because I agree with Karl Popper that a perfect division of labor in scholar¬
ship and scientific research would soon stop progress. I was also encouraged
to pursue these interdisciplinary investigations by the success achieved by my
earlier History of Science as Explanation (1973), which also deals with prob¬
lems at the interface of science, philosophy, and history. If the modesty of
that success is due to the specialism of that subject matter (the historiography
of science), then one may hope that the universal appeal of the present one
(Galileo’s Dialogue and the art of reasoning) will lead to a success commen¬
surate with it. Be that as it may, it will be a useful guide to the reader to
elaborate several themes with respect to which these inquiries may be read.
From a general cultural point of view, this work constitutes a study of a
book that may be listed among the classics and of a number of human activ¬
ities of general interest and relevance, namely the arts of logical reasoning,
rhetorical persuasion, philosophical reflection, and scientific investigation.

IX
X PREFACE AND ACKNOWLEDGMENTS

These four arts are discussed both in their relation to one another, and in
their relation to Galileo’s Dialogue. From this point of view, the Introduction
is a discussion of the general cultural relevance of Galileo’s book, by reference
to which all succeeding chapters concretely study those four activities.
Chapter 8 is a general introduction to the four activities, by means of a dis¬
cussion of the role of reasoning, of rhetoric, and of methodological reflection
in science. Chapter 9 distinguishes reasoning from one of the things with
which it is often confused, namely a priori rationalism. Chapter 1 is a discus¬
sion of the difference between reasoning and relying on religious authority,
and of the problems engendered by this difference in the practice of rhetorical
persuasion. Chapter 2 discusses the role of the logical structure of reasoning
and of the presentation of scientific evidence in the art of rhetorical persua¬
sion. Chapter 3 studies the role of artistic expression and of emotions in
reasoning and in persuasion. Chapter 4 relates several different conceptions of
science to the Dialogue. Chapter 5 discusses the interplay between scientific
practice and methodological reflection. Chapter 6 examines some similarities
and some differences between methodology and philosophy. Chapter 7
studies the connections between methodology and logic by interpreting re¬
flections on knowledge and method in general as aspects of reflection upon
reasoning. Chapter 10 illustrates some problems that the science of reasoning
encounters when it studies the historical material in which human reasoning
leaves a record. Chapters 11—13 discuss various ways of studying reasoning,
namely the science of formal logic, psychology, linguistics, inductive logic,
and rhetoric, as well as the informal involvement in reasoning. Chapter 14
elaborates a number of principles useful for the understanding of reasoning,
the central idea being that reasoning involves the inter-relating of ideas in
accordance with a definite structure. Chapter 15 discusses the problem of
determining how good or bad reasoning is, and it suggests that practical in¬
volvement in and development of the reasoning under consideration is needed.
Chapter 16 is a detailed analysis of many instructive arguments found in the
Dialogue. Chapter 17 discusses the importance in reasoning of criticism,
evaluation, interlocking steps, and several varieties of invalidities.
From the point of view of rhetorical scholarship, these investigations con¬
stitute an attempt to develop concretely a rhetoric of science and a science
of rhetoric. Our example of science consists of Galileo’s Dialogue, and the
concreteness derives from the fact that all discussions refer to it. The rhetoric
of science refers to the rhetorical dimension of science, and the science of
rhetoric refers to the systematic study of rhetorical activities. Chapter 1 then
deals both with the rhetorical dimension of the Dialogue, in the sense of
 PREFACE AND ACKNOWLEDGMENTS XI

rhetoric that refers to external appearance and pretended effects of verbal

discourse, and also with another important distinction for rhetorical science,
namely the difference between relying on authority and engaging in rational
argument. Chapter 2 deals with the book’s rhetoric, in the sense of the term
that refers to argumentation, and also with the importance of the concept of
argument structure in a science of rhetoric. Chapter 3 studies directly the
book’s rhetoric, in the sense of the term that refers to emotional appeals and
beautiful-sounding expressions, and indirectly the role of these concepts in
rhetorical science. Chapters 4—7 deal directly with various aspects of the
scientific methodology of Galileo’s book and indirectly lay the foundations
for determining what it would be for rhetorical phenomena to be studied
scientifically. Chapter 8 discusses in a general way the role of rhetoric in
scientific rationality. Chapter 9 is both a concrete analysis of one rhetorical
aspect of Galileo’s book, in the general sense of rhetoric that refers to argu¬
mentative techniques, and also a concrete analysis of an important feature of
one of the most fundamental concepts in the science of rhetoric, namely
reasoning and how it differs from rationalism. Chapters 10—17 develop a
number of fundamental topics in the science of rhetoric. Chapter 10 is a
critique of a number of scholars, suggesting that the historical record of
reasoning cannot be left entirely in the hands of mere historians. Chapters
11—13 examine the nature of and inter-relationships among the rhetorical
and the psychological, logical, linguistic, and informal study of reasoning.
Chapter 14 is a formulation and illustration of a number of basic principles
for the understanding of reasoning. Chapter 15 is a critique of the fallacy
approach to the evaluation of reasoning and suggests an alternative practical-
oriented approach. Chapter 16 is a detailed analysis of the logical and the
rhetorical features of about half of all the arguments in Galileo’s book. Chap¬
ter 17 draws a number of theoretical conclusions from the data of the pre¬
vious chapter.
From the point of view of philosophy, these investigations constitute
concrete studies in the logic of science and the science of logic. Here the logic
of science may be taken to refer to a distinctive approach to the philosophy
of science, which emphasizes reasoning as opposed to other concepts such as
knowledge, explanation, theory, and progress; while the science of logic refers
to the scientific study of reasoning. The mutual inter-dependence of the two
is obvious: in order to study the nature of science from the point of view of
reasoning, the logic of science must sooner or later give an adequate articula¬
tion of the concept and activity of reasoning; and in order to study reasoning
scientifically, the science of logic must use or develop an adequate view of
 Xll PREFACE AND ACKNOWLEDGMENTS

what science is and of what it is to be scientific. Chapter 8 argues that a

change of emphasis from epistemology (e.g., theory and observation) to logic
(reasoning) is the most fruitful move for contemporary philosophy of science;
the argument is grounded on an elaboration and correction of Paul Feyera-
bend’s main doctrines. Chapter 9 shows that the study of actual scientific
reasoning has not been neglected by the best historians of science (e.g.,
Koyrd), but that important corrections need to be made if their work is to
become useful and acceptable. Chapters 1-7 develop a concrete logic of
science, around the case study of Galileo’s Dialogue. Chapter 1 argues that
this book is a complex multi-dimensional argument in support of Copernican-
ism. Chapter 2 elaborates one aspect of this argument, namely the one
relating to the presentation of the available evidence and the organization of
one’s claims by means of an inferential structure. Chapter 3 discusses the
emotive and aesthetic aspects of Galileo’s argument. Chapters 4-7 discuss its
methodological aspect: first its content for present-day scientists (Chapter 4),
then the methodology inherent in this scientific content (Chapter 5), then the
philosophical significance of this methodology (Chapter 6), and finally the
logical significance of this methodology, that is, its significance for the clarifi¬
cation of the concept of reasoning (Chapter 7). Chapters 10-17 are a series
of studies in the science of logic. Chapter 10 argues that the historical ap¬
proach needed in the science of reasoning requires better scholarship than
that found even among the best historians. Chapter 11 explores the value and
the limitations of formal logic and of experimentation in the study of reason¬
ing; this is done by a detailed discussion of recent work in the psychology of
reasoning. Chapter 12 discusses the importance of logical analysis in relation
to rhetorical analysis, and it clarifies the concept of question-begging; this is
done by a critical elaboration of Perelman’s ‘new rhetoric’ and by using
Galileo’s tower argument as an example. Chapter 13 brings into focus the
problem ot a theory of scientific reasoning and of a scientific theory of
reasoning by discussing certain problems at the foundations of inductive
logic, by discussing some methodological questions in the linguistics of rea¬
soning and elaborating certain suggestions made by Y. Bar-Hillel; by exploring
the possibility of a ‘practical’ approach to the study of reasoning and elab¬
orating certain suggestions by Michael Scriven; and by a critical analysis of
Stephen Toulmin’s views on logic. Chapter 14 formulates and illustrates a
number of basic principles for understanding the structure of reasoning, in
accordance with our approach to logic. Chapter 15 makes a specific sugges¬
tion for the study of the other main dimension of reasoning, namely its worth
or goodness, in accordance with the same approach. Chapter 16 constructs a
 PREFACE AND ACKNOWLEDGMENTS Xlll

data basis by giving detailed reconstructions, analyses, and evaluations of

sixteen complex arguments (consisting of more than 200 simple arguments)
in Galileo’s book. Chapter 17 draws a number of theoretical conclusions
from this data basis, suggesting that the two main types of arguments are
critical and constructive; that several different types of invalidities can be
defined; and that reasoning normally consists of complex rather than of
simple arguments.
From the point of view of historical scholarship and erudition, these in¬
vestigations constitute a commentary to a crucially important work. Each
chapter may then be viewed as studying a particular aspect of Galileo’s book:
the Introduction may be viewed as a discussion of the general cultural signifi¬
cance and impact of the Dialogue', Chapter 1 as an analysis of the book s
religious and theological rhetoric; Chapter 2 as an elaboration of the book s
logical structure; Chapter 3 as an analysis of its emotive, aesthetic, and liter¬
ary content; Chapter 4 as an examination of its content relevant to scientists,
Chapter 5 as a study of its methodological, epistemological, and philosophical
content; Chapter 6 as a discussion of its significance for philosophers; Chapter
7 as an examination of its logical content and significance, in the sense of the
theory and practice of reasoning in general; Chapter 8 as a discussion of how
the book is a fruitful example of scientific rationality; Chapters 9—10 as a
critique of certain aspects of the book’s historiography; Chapters 11-13 as
an argument for its relevance to formal logicians and other students of rea¬
soning, such as linguists, psychologists, rhetorical scholars, inductive logicians,
and scholars who stress informal logic; Chapter 14 as a discussion of the
book’s value for the illustration of a number of basic principles for the under¬
standing of reasoning; Chapter 15 as a suggestion of its value for the illustra¬
tion and formulation of principles for the evaluation of reasoning; Chapter 16
as a detailed analysis of about half of all the arguments in the book; Chapter
17 as a number of theoretical conclusions that can be drawn from the previous
chapter’s study of the book’s logical content. The Appendix is an analysis
of the pagination in several important editions of the book.
Finally, from the editorial-literary point of view, these investigations group
themselves into three somewhat autonomous parts, each of which may per¬
haps be appreciated independently of the others. The first part consists of
Chapters 1—5 and deals with various aspects of Galileo’s Dialogue, the second
consists of Chapters 6-13 and deals with logical and methodological critiques
of various scholars, the third consists of Chapters 14—17 and is a sketch of
various aspects of the theory of reasoning.
My investigations received the support, encouragement, comments, or
XIV PREFACE AND ACKNOWLEDGMENTS

constructive criticism of a number of people and institutions, to whom I

should like to express my gratitude: the National Science Foundation for a
research grant (SOC76—10220), and the University of Nevada at Las Vegas
for a sabbatical leave, in the academic year 1976—77, during which most of
this book was written; and Joseph Agassi (Boston, Bielefeld, and Tel-Aviv),
Harold I. Brown (Northern Illinois), Stillman Drake (Toronto), Paul Feyera-
bend (Berkeley), Mary Hesse (Cambridge), P. N. Johnson-Laird (Sussex),
Henry W. Johnstone, Jr. (Pennsylvania State), Larry Laudan (Pittsburgh),
Marvin Loflin (Alaska), Arthur Millman (Colorado), Michael Scriven (San
Francisco), William A. Wallace (Catholic University), Craig Walton (Nevada,
Las Vegas), and Robert S. Westman (UCLA).
As for claiming responsibility for my own errors, I shall refrain from being
redundant and insulting the reader’s intelligence by paying lip service to the
ritualism that authors usually engage in on such occasions; instead I shall refer
him to what Michael Dummett says on the matter in Frege: Philosophy of
Language.
The book was conceived and written as a single sustained argument, but
publication delays made it inevitable for various parts to find their way
separately into journals and conference proceedings. Thus acknowledgments
are hereby made to the editors and publishers of the journals and books
where those parts first appeared, as follows: the nucleus of Chapter 2, in
Logique et Analyse 22 (1979), 159-80; parts of Chapter 4, in Journal of
College Science Teaching, Sept. 1980; the bulk of Chapter 5, in Scientia
112 (1977), 95-118, and 371—85; the third section of Chapter 6, in PSA
1976, Vol. 1, edited by F. Suppe and P. D. Asquith (East Lansing: Philosophy
of Science Association, 1976), pp. 130-39; substantial parts of Chapter 8, in
PSA 1978, Vol. 1, edited by P. D. Asquith and I. Hacking (East Lansing:
Philosophy of Science Association, 1978), pp. 235-46; the core of Chapter
9, in Physis 19 (1977), 5-27; the first three sections of Chapter 10, in
Synthese 40 (1979); the nucleus of Chapter 11, in Philosophy of the Social
Sciences 9 (1979), 277-91; the last three parts of Chapter 12, in Physis 16
(1974), 129-48; the first section of Chapter 15, in American Philosophical
Quarterly, Library of Philosophy volume on Studies in Epistemology.
Acknowledgments are also due to the publishers for permission to quote
from the following books: G. Galilei, Dialogue Concerning the Two Chief
World Systems, Berkeley: University of California Press, 1967; R. P. Feynman
et al., The Feynman Lectures on Physics, Vol. 1, Reading, Mass.: Addison-
Wesley Publishing Company, © 1963, California Institute of Technology;
B. Croce, History: Its Theory and Practice, 1920; New York: Russell and
Russell, 1960; and A. Koyrb, Etudes Galileennes, Paris: Hermann, 1966.
 INTRODUCTION

GALILEO’S DIALOGUE AND WESTERN CULTURE

Certain individuals and certain books perform a unique role in human civiliza¬
tion. In world history in general, Socrates, Buddha, Confucius, and Jesus may
be regarded as four paradigmatic individuals of incomparable significance, as
Karl Jaspers has argued.1 Restricting ourselves to Western history and to liter¬
ary culture, the works of such men as Plato, Augustine, Galileo, and Marx
emerge as having unique significance and value. This is especially true for
certain particular books of theirs, which have become classics. Thus one does
not have to be a professional philosopher to be able to enjoy, and profit from
Plato’s Socratic dialogues; one does not have to be a Christian in order to ap¬
preciate Augustine’s Confessions; one does not have to be a scientist in order
to understand Galileo’s Dialogue Concerning the Two Chief World Systems;
nor does one need to be a communist in order to value Marx’s Capital. We
may say that the universal relevance of such books derives partly from their
self-contained readibility, and partly from historical circumstances; paradoxi¬
cally, it also derives partly from their individuality, that is, from the fact that
each one of these books plays a unique role in a specific, but fundamental
and primary human endeavor: morality and education, religion, science, and
social theory and practice, respectively. Moreover, it is interesting to reflect
on the pattern of their influence. From the chronological point of view
Socrates’ influence is of course the most firmly established since it has been
operative the longest — twenty four centuries; Augustine’s influence makes
up in intensity what it lacks in age, insofar as his effect on Christianity has
been deeper than Socrates’ effect on any comparable group, culture, or tradi¬
tion; Galileo’s makes up in universality what it lacks in age vis-a-vis Socrates
and what it lacks in intensity vis-a-vis Augustine, insofar as Galileo may be
regarded as the father of modern science and insofar as modern science has
become the most widespread cultural force that the world has ever known
and the only one to have penetrated all other subcultures of the earth; finally,
though Marx’s influence is barely a century old, though Marxism has not (yet)
prevailed the world over, and though in countries where Marxism has prevailed
its influence has not generally reached the depths of the human soul, neverthe¬
less the success and diffusion of Marxism has been the most spectacular, so
that Marx’s influence has made up in rapidity what it lacks in other respects.

xv
XVI INTRODUCTION

Next, restricting our attention to Galileo, I want to argue that, though his
specialty is science, from a more general point of view his work can be taken
as exemplary for the fundamental human desire of the search for truth.
Generalizing in the same manner, Socrates would be a model in the search for
the good and the right; Augustine in the search of the holy and the super¬
natural; Marx in the search for the practical.2
The paradigmatic value of Galileo’s work among scientists is, of course,
well known and will be elaborated later. What needs discussion now is the
attitude of other groups, for example, laymen and professional thinkers.
Regarding Galileo’s influence on laymen, here the effect probably works by
means of the image of Galileo being persecuted and tried by the Catholic
Church. The image was portrayed with classic incisiveness by John Milton,
when in his Areopagitica he recalls his visit to Florence by saying, “There it
was that I found and visited the famous Galileo, grown old, a prisoner to the
Inquisition for thinking in Astronomy otherwise than the Franciscan and
Dominican licensers of thought”.3
To be sure, in 1968 Pope Paul VI announced a revision of Galileo’s 1633
trial by the Inquisition. This is a very interesting cultural-historical fact not
only because it reflects a liberalizing trend within the contemporary Catholic
Church, but also because of the fact that such an announcement strikes us
generally as telling, or promising to tell, more about the Church than about
Galileo. This reaction - typical of layman, scholar, and scientist alike - in
turn is an indication that Western culture has come a long way in the last
three and one-half centuries.
We may agree, as Ortega y Gasset aptly expressed it, that the trial origi¬
nated “more in the small intrigues of private groups than in any dogmatic
reservations of the Church”.4 Nevertheless, the historical significance of an
event depends on the magnitude of its effects, rather than of its causes, and
this is why the whole Galileo affair came to be perceived as “the greatest
scandal in Christendom”.
Since the occasion for the trial was Galileo’s publication of his Dialogue
in 1632, this book would seem to deserve both inspirational reading for the
truth-seekers, and careful scholarly study. The former it certainly has re¬
ceived, but it is one of the greatest scandals of modern scholarship that no
exhaustive, systematic study of the book in all of its aspects and implications
has ever been made. This is not meant to deny the existence of countless
partial studies, which are very informative, insightful, and helpful: the Galileo
literature is so immense, and the Dialogue and the trial so crucial in Galileo’s
life, that it could hardly be otherwise. Nevertheless, it is equally undeniable
 GALILEO’S DIALOGUE AND WESTERN CULTURE xvii

that there is no study of Galileo’s book in all of its aspects, scientific, meth¬
odological, philosophical, logical, rhetorical, theological, religious, literary.
Such a study is undertaken below.
Before doing so, let us look at some general cultural influences on people
other than scientists and laymen.
David Hume’s Dialogues Concerning Natural Religion is a classic discussion
of arguments about the existence and nature of God. The first arguments it
discusses are arguments based on experience: for example, that the universe
must have an intelligent Maker because the universe is like a machine and we
know that machines are made by human beings who have intelligence; or the
universe is an ordered arrangement, and all ordered arrangements can be
observed to derive from human intelligence, and therefore the universe must
have an intelligent cause. Here then the problem arises of whether it is at all
possible to prove the existence or attributes of God with arguments based on
experience, since in order to check such reasoning we would have to have
experience of the origin of universes. To help solve this problem Hume asks
whether to prove by experience the origin of the universe is any different
from proving the earth’s motion by experience. He answers that there is a
difference because we do have observational knowledge of the motion of
other ‘earths’, namely the moon and planets, and these bodies are similar to
the earth. In making this last point Hume adds a discussion about the struc¬
ture of Galileo’s argument in the Dialogue, namely that he first tries to estab¬
lish an analogy between the earth and the heavenly bodies in his attempt to
prove the earth’s motion.5
I am not saying either that Hume’s reference to Galileo is absolutely ac¬
curate, nor that the mere reference to Galileo completely proves his own
critique of the possibility of proving the existence of God by experience. It
suffices that Hume’s reference to Galileo is there, that it is correct to some
extent, and that the reference strengthens Hume’s case. This in turn shows
how Hume used Galileo as a model on one important occasion.
A second example is Immanuel Kant. In the Preface to the Second Edition
of the Critique of Pure Reason he gives a sketch of the main achievement in
his book. He claims to have brought about a revolution in the field of meta¬
physics “in accordance with the example set by geometers and physicists”.6
No geometers are named, but concerning the latter he expresses himself with
words that are so explicit and eloquent as to deserve extended quotation:
When Galileo caused balls, the weights of which he had himself previously determined,
to roll down an inclined plane; when Torricelli . . . ; when Stahl ... a light broke upon
all students of nature. They learned that reason has insight only into that which it
xviii INTRODUCTION

produces after a plan of its own, and that it must not allow itself to be kept, as it were,
in nature’s leading-strings, but must itself show the way with principles of judgment
based upon fixed laws, constraining nature to give answers to questions of reason’s own
determining. . . . Their success should incline us, at least by way of experiment, to imi¬
tate their procedure, so far as the analogy.to metaphysics may permit. Hitherto
it has been assumed that all knowledge must conform to objects. But all attempts to
extend our knowledge of objects by establishing something in regard to them a priori,
by means of concepts, have on this assumption ended in failure. We must therefore make
trial whether we may not have more success in the tasks of metaphysics, if we suppose
that objects must conform to our knowledge.7

As in the case of Hume, Kant’s reference to Galileo need be neither com¬

pletely accurate nor completely decisive in order to have its intended effect.8
Third, in Man and Crisis Jose Ortega y Gasset pays to Galileo a double
homage, so to speak. First, since Ortega’s subject matter is cultural history,
he takes Galileo as the paradigm example of a thinker of the modern age, thus
giving his own book a Galilean dimension in a substantive sense. Second,
Ortega feels that his approach in his historical investigation must be patterned
after Galileo’s, thus making his own book Galilean in a methodological sense.
Ortega’s statement of his first point can hardly be excelled:

The greatest crisis through which the European destiny has ever passed ends with Galileo
and Descartes - a crisis which began at the end of the fourteenth century and did not
taper off until the early years of the seventeenth century. The figure of Galileo appears
at the end of this crisis like a peak between two ages, like a divide that parts the water.
With him modern man enters into the modern age.
It is therefore a matter of supreme interest for us to take this crisis and this entrance
into a new period under very close consideration. Every act of entering into any place,
every coming out from any corner has about it a bit of the dramatic; at times it has a
great deal - hence the rites of the doorway and the lintel. The Romans believed in spe¬
cial gods who presided at that condensation of enigmatic destiny which is the act of
going out or of coming in. The god of going out they called Abeona; the god of coming
in, Adeona. If, in place of a pagan god, we speak in Christian terms of a patron saint,
nothing would seem more justifiable than to make of Galileo both the Abeona, patron
saint of our departure from modernity, and the Adeona, patron saint of our emergence
into a future still palpitant with mystery.9

After an analysis of Galileo’s principle of inertia,10 Ortega applies his con¬

ception of Galileo’s method to his own inquiry, which he calls history, the
science of human lives:

Every science which is concerned with reality, whether of the body or of the spirit, must
be not merely a mirror of the facts, but a genuine construction. Because the science of
physics in the time of Galileo was thus developed, it has endured as a model science and
a norm of knowledge for the whole of the modern age.
 GALILEO’S DIALOGUE AND WESTERN CULTURE XIX

History must adopt a similar decision and prepare itself to construct. It goes without
saying that this comparison between physics as it is and history as it is and as it ought to
be is only valid, for the moment, at this single point — the element of construction as
such. The other characteristics of physics are not such as to be desirable for history.
Take exactitude, for example ... 11

Then by using Galileo’s method in the subject matter of human lives Ortega
arrives at two principles which he takes to be the historical analogues of the
concept of inertia: “first that every man’s life starts with certain basic convic¬
tions about what the world is and what man’s place in it is - it starts from
there and moves within them; second, every life finds itself in surroundings
which include more or less technical skill or control over the material environ¬
ment”.12 For Ortega, this means that life both has a structure and is a drama:
“human life always has a structure - that is to say, it consists in man’s having
to cope with a predetermined world of which we can sketch the profile. The
world presents certain problems which, relatively speaking, are solved, and
raises others, thus giving a definite aspect to man’s struggle for his own
fate”.13 Moreover, “life is not solely man, that is to say, the subject which
lives. It is also the drama which arises when that subject finds himself obliged
to fling his arms about, to swim shipwrecked in that sea which is the world”.14
Then Ortega goes on to apply these ideas to the details of Western cultural
history. As before, one need not agree with Ortega’s reading of Galileo,
though the interpretation is not implausible; still less need one accept his view
of “man and crisis”. Nevertheless,Man and Crisis is perhaps the best example
of the kind of influence and relevance that I am here trying to articulate.

NOTES

1 The Great Philosophers, Vol. I.

2 Benedetto Croce’s elaboration of the universal category of the ‘useful’ may be taken
as such a generalization of Marx’s work. Cf. Croce’s Historical Materialism and the
Economics of Karl Marx and his Philosophy of the Practical, Antoni’s Commento a
Croce, and my review of Gentile’s Filosofia di Marx.
3 Quoted in S. Drake, ‘Galileo in English Literature of the Seventeenth Century’, p. 423.
4 J. Ortega y Gasset, Man and Crisis, p. 9.
5 D. Hume, Dialogues Concerning Natural Religion, pp. 141-152, especially pp. ISO-
152.
6 I. Kant, Critique of Pure Reason, p. 25 (Bxxii).
7 Ibid., pp. 20-22 (Bxii-Bxvi).
8 Kant complicates his discussion by referring to Copernicus and the Copernican revolu¬
tion of changing the point of view from attributing motion to the heavenly bodies to
attributing it to the spectator. This presents no problem for my argument not only
XX INTRODUCTION

because I need not claim exclusive reference to Galileo in this context, but also for the
following reason. Kant uses the reference to Copernicus in two ways, first in order to
elucidate the content of his own metaphysical revolution, insofar as the change from
knowledge conforming to objects, to objects conforming to our knowledge, is analogous
to the change from motion subsisting in heavenly bodies to motion subsisting in the
observer; hence this Copernican reference does not play the role of a methodological
justification in Kant’s discussion. Kant recognizes that Copernicus’ achievement was to
have “dared, in a manner contradictory of the senses, but yet true, to seek the observed
movements not in the heavenly bodies, but in the spectator”. (Ibid., p. 25, footnote a;
Bxxii); hence Kant goes on to explain that the proper methodological similarity between
his own situation and Copernicus’ is insofar as, in the Preface, Kant is conjecturing a
revolutionary metaphysical thesis, which he will then establish in the course of the book
{Ibid.). In short, I would say that Kant refers to Copernicus for an elucidation of his
thesis, and to Galileo for a justification of his approach and consequently for a partial
justification of his thesis.
9 Ortega, op. cit., p. 11.
10 Ibid., pp. 11-19.
11 Ibid.,p. 20.
12 Ibid., p. 27.
13 Ibid., pp. 27—28.
14 Ibid., p. 28.
 PART I

GALILEO’S DIALOGUE
 CHAPTER 1

FAITH VERSUS REASON:

THE RHETORICAL FORM AND CONTENT OF
GALILEO’S DIALOGUE

The Dialogue is written in the form of a dialogue among three characters and
contains numerous indirect references to a fourth person, a Lyncean Acad¬
emician, common friend of the interlocutors, who may be identified with
Galileo himself. Moreover, the book is full of passages containing disclaimers
by the supposedly Copernican spokesman, Salviati, to the effect that his
Copernican beliefs are not necessarily as strong as they may appear when he is
involved in the discussion, so that he really ought to be regarded as being a
good actor who is wearing a Copernican mask. Other passages continuously
remind the reader that the aim of the discussion is not to decide the
Copernican issue one way or the other, but rather to present all the argu¬
ments for and against, so that people in a position of authority, who are
entitled to make such decisions, can do so intelligently. The Dialogue is also
full of emotionally charged passages, where various persons, books, and ideas
are criticized not dispassionately but rather with ridicule, scorn, and con¬
tempt. In other passages the interlocutors play tricks on each other, trying to
trap one another into saying something so that they can win the argument;
such trickery is complicated by the fact that these maneuvers are usually
exposed in the course of the discussion and by the fact that they are (at least
apparently) perpetrated almost always at the expense of the Peripatetic
spokesman, Simplicio. These and other similar features, all of which are
rather obvious even to a casual reader of the Dialogue, may be labeled
‘rhetorical’. The specialist in Geisteswissenschaften will be interested in such
rhetorical matters merely because they exist, and he will try to understand
their nature, how they function, how they relate to other verbal and mental
processes, and to other human activities, etc.
This is not, however, the only reason for studying the rhetoric inherent in
the Dialogue. For even a superficial acquaintance with the book’s origin and
fate makes it obvious that the book had, and was meant to have, more than a
purely intellectual content: in writing it Galileo was also being a man of
action, interested and concerned in bringing about certain practical effects,
namely to persuade the appropriate Church officials of the truth of Coperni-
canism and thereby to cause the repeal of the condemnation of 1616. More¬
over, as everybody knows, what the book caused instead was Galileo’s trial

3
4 CHAPTER 1

and conviction by the Inquisition for heresy and for disobedience. In other
words, the book originated in part as a practical act, and was so judged by the
Church. Hence the study of those parts of the book that have or seem to have
practical import can shed light on what some came to call “the greatest
scandal in Christendom”. This practical dimension of the Dialogue will also
be labeled ‘rhetorical’ since it too, like the dramatic, emotive, and moral
aspects, refers to nonintellectual matters.
A third reason for studying the rhetorical content and structure of the
Dialogue is that nearly all scholars recognize a polemical and propagandists
dimension in the Dialogue, however greatly they may differ about the
importance of this aspect of the book and about its relationships with other
aspects; hardly anyone, however, has seriously studied this topic. It is not
clear why this should be so. The first thought that comes to mind is that
rhetorical analysis (as we may call the scholarly study of verbal propaganda)
requires skills and sensibilities that science-oriented scholars seldom possess,
and, conversely, rhetorically competent scholars lack proficiency and interest
in scientific topics. This, however, cannot be the whole story because there is
an analogous disjointness between science and poetry or literature, and yet
the literary-practical aspects of the Dialogue have received more serious study
than its rhetoric. Part of the rest of the story is that rhetorical studies have
been generally neglected in modern times, whereas aesthetics has gained
increasing popularity and depth. But this raises a new problem, for after all
the last several centuries are also the era when science has gained unparalleled
attention and dominance; the problem is, why is it that in the last three or
four centuries science has flourished but rhetorical studies have decayed? Is
it because there is an intrinsic conflict between the scientific and the rhetori¬
cal points of view? What is clear is that most scholars believe in this conflict;
in fact, when they mention the so-called propagandistic aspect of the
Dialogue, they typically do so by contrast with its scientific value.1
In the last ten or fifteen years the situation has changed somewhat, and
some, e.g. Feyerabend, have not been afraid of concluding that propaganda
and rhetoric play an essential role in science.2 To be sure, Feyerabend’s
language (his ‘rhetoric’, as it were!) is still that of the opposition between
science (or at least reason) and rhetoric; however, the logic of his position is
different, and I would describe it as follows. The context of the discussion
has been the problem of scientific rationality and the question of whether the
transition from one scientific theory to another can be made in a rational
manner.3 The problem arose from the realization that, in the case of very
fundamental scientific developments such as the Copernican revolution and
 FAITH VERSUS REASON 5

the transition from classical to modem physics, purely logical considerations,

rational argumentation, and appeals to the rules of scientific method are not
enough to make a scientist change his mind. This realization was then general¬
ized to conclude that the same limitation applies to all, or at least most, or at
least the important and interesting,scientific developments. The generalization
is I believe illegitimate,4 but it has generated a lot of discussion and con¬
fusion: the more conservative philosophers of science, feeling that the
rationality of science was being threatened, have tended to counterargue that
transitions from one theory to another can be and have historically been
made in a rational manner, using as evidence the admittedly less problematic
minor transitions or at least the less problematic aspects of the major transi¬
tions; on the other hand, the more revolutionary philosophers have tended to
reiterate their arguments. Moreover, these latter philosophers have drawn
further conclusions from their generalization about the limits of reason in
scientific changes. Some have concluded that the only sense in which science
is rational is that the occurrence of scientific developments is exphcable, that
is that we can see how and why it comes about that one theory is abandoned
and another one emerges in the scientific community. Other philosophers
have concluded that the only sense in which science is rational is that scienti¬
fic changes can be justified after the fact, in terms of an abstract ahistorical
scheme which shows how and why the later theory is an improvement of the
older one. Some other philosophers have simply concluded that science is an
irrational enterprise, or at least no better than myth, magic or witchcraft.
It is possible, however, in fact preferable, to draw the following conclusion:6
in addition to argument, rhetoric is sometimes crucial in science; and hence,
rhetoric has an important role to play in scientific rationality and the rhetori¬
cal aspects of science should not be neglected. To what extent this conclusion
can be grounded on an analysis of the rhetoric in the Dialogue is one of the
things we will try to determine here.
To conclude, a rhetorical analysis of the Dialogue is needed because the
rhetoric is there, because its writing was partly a practical act, because this
aspect of the book has never been seriously examined though it has been
universally recognized, and because it can provide a case study in the role of
rhetoric in science; by rhetorical analysis is meant examination of the non¬
intellectual (and non-literary-aesthetic, to be exact) content, structure, and
aspects of the book.
6 CHAPTER 1

RELIGIOUS RHETORIC

Immediately after the title page, the Dialogue has a two-page letter of dedica¬
tion to the Grand Duke of Tuscany, and immediately after this we find a
three-page preface entitled ‘To the Discerning Reader’. Reading the latter by
itself could make one think that the book is an apology of Catholic piety.
One of the most obvious features of this preface is that it was printed in a
type different from that of the book’s text.7 This fact was cited as one of
several accusations in the indictment compiled against Galileo soon after the
book was published,8 the alleged crime being that the different printing had
“rendered it useless inasmuch as it was alienated from the body of the book”.9
Though the fact was not mentioned in the final sentence pronounced against
Galileo,10 its being mentioned in the original indictment is indicative of the
rhetorical effect (or lack of effect) that the preface may have. I am inclined
to agree that the preface is indeed alienated from the text by the printing,
however one cannot conclude that it is thereby rendered rhetorically ‘useless’,
but only that its rhetorical effect is thereby diminished. The reasons for this
are that the print of the preface is of the same type as the letter of dedication,
the postils printed in the margin of almost every page of the text, the two-
page errata section immediately following the last page of the text, and the
30-page appendix which is analogous to the index in a modern book and
which (on the very last page) is referred to as “Table of the most notable
topics which are contained in this book”. Another feature that must be
counted as lessening the rhetorical alienation of the preface is the fact that
the preface was left unsigned, unlike the dedication. It is true, of course, that
Galileo may have left it unsigned for the (otherwise) very good reason that
the preface had been imposed upon him by the church censors;11 on the
other hand, the preface probably incorporated suggestions made by Galileo
himself in the course of his discussions with them. At any rate the lack of a
signature to this section of the book, which after all is not titled ‘preface’ but
merely ‘To the Discerning Reader’, does make it more of an integral part of
the body of the work.12
The preface consists of seven paragraphs13 and contains three elements: a
statement of the aim of the book, a description of its content, and an explan¬
ation of its dialogue form. The last need not concern us here. The stated aim
is a refutation of the rumor that the anti-Copemican decree of 1616 was
based on ignorance, and a proof that Catholic acceptance of the earth’s
motionlessness derives from piety, religiousness, knowledge of divine omnipo¬
tence, and awareness of the weakness of the human mind. The rumor will be
 FAITH VERSUS REASON 7

refuted by showing that Catholics are at least as knowledgeable about the

forbidden topic as Protestants, for Galileo claims that in 1616 the Church had
consulted him before reading its decision. The motive of piety will be
established by showing that, from the point of view of evidence and argu¬
ments, the Copemican position can be made to appear superior; here the text
adds the qualification that this does not mean that the thesis of the earth’s
motion is superior in an absolute sense, but only that the Copernican
supporting arguments are better than the Aristotelian geostatic ones. Then
three of the book’s main results are stated: first, that no terrestrial experi¬
ments can prove its motion; second, that celestial phenomena favor the
Copernican hypothesis; and third, that new light can be shed on the problem
of the cause of tides, if we assume the earth’s motion.
The following points should be noted. The preface gives Galileo partial
responsibility for the decree of 1616; independently of the historical cor¬
rectness of this claim, its rhetorical effect is simultaneously to increase the
intellectual respectability of the decree and the religious credentials of the
present book. The claim has the nature of a compromise, with each side both
gaining and losing a little.
Second, the justification of the motive of piety presupposes a commitment
to a limited rationalism, which it is very easy to confuse with irrationalism.
The presupposition seems to be that the rational acceptability of an idea is
only one consideration for deciding upon beliefs, and that one may decide
to overule reason and believe what religious authorities dictate. However, to
be more exact, one would have to mention another operative presupposition,
namely that religious authorities are usually right; that is, in speaking of
rational acceptability one would have to distinguish between individual
rational acceptability, and rational acceptability to the relevant authorities.
Hence, the words in the text are not really an expression of irrationalism but
merely of piety, in the sense of general willingness to accept the conclusions
that religious authorities deem rationally acceptable, even when this goes
against one’s individual judgment. In this sense the preface is preaching
religious piety.
Third, the summary of the book’s results is expressed with the proper
emphasis. Regarding the terrestrial arguments, the stress is on the fact that
they cannot prove the earth’s motion, though this is only one side of the
coin, the other being that they cannot prove that the earth stands still either.
Regarding celestial evidence, the superiority of Copernicanism is qualified by
the explicit use of the word ‘hypothesis’ in speaking of the ‘Copernican hypo¬
thesis’, and by the qualification that this hypothesis seems to prevail “for
8 CHAPTER 1

facility in astronomy, not for necessity of nature” (F30). Regarding the argu¬
ment from the tides, the qualifications are even more obvious: the text calls it
an ‘ingenious fantasy’, which only ‘sheds light’ on the problem of tides, and
which exhibits “the probability which would render it [the phenomenon of
tides] comprehensible, given that the earth were in motion” (F30).
It is well known, however, that this ‘rhetoric’ fooled no one, or at least no
one besides the two or three censors who gave permission for the book to be
printed. What is not clear is why this rhetoric did not work. Concerning the
book’s content, the preface makes four points: (1) that the Copernican argu¬
ments are shown to be better than the Aristotelian ones; (2) that terrestrial
arguments are shown to be incapable of proving the earth’s motion; (3) that
astronomical arguments favor the Copernican hypothesis; and (4) that a
novel, original argument in favor of Copernicanism can be formulated based
on the cause of tides. Each of these four points is indeed accurate as a
description of the book’s content. Therefore, what all readers must have
refused to do is to draw the conclusions, from these four points, that (A)
therefore geostatic beliefs must be taken on (religious) faith and cannot be
accepted on the basis of reason, and that (B) therefore to the extent that the
decree of 1616 was made in the light of such knowledge it was not arbitrary.
Instead readers must have concluded that (A') therefore geostatic beliefs are
irrational and should be abandoned, and that (B') therefore to the extent that
the decree of 1616 was made in the light of such knowledge it was irrational.
In other words, the religious rhetoric of the preface consists of a logically
invalid justification, and no one accepted this rhetoric because no one was
willing to infer the required apologetic conclusions on the basis of the (more
or less) agreed-upon premises descriptive of the book’s content.
In short, what the preface claims was universally rejected because of its
internal incoherence, and it provides us with evidence that people will not
always believe what they are told to believe, or what they hear, or what they
read; in particular, they will not accept it if it does not make logical sense to
them.
At the other end of the rhetorical spectrum, and it so happens at the other
end of Galileo’s book, we find an example of implicit rhetoric, where it is not
so much the words that are important, but rather the actions, facts, and
deeds. I am referring to the fact that the skeptical argument favored by the
pope is put in the mouth of Simplicio, who is supposedly a simpleton. The
argument is the following:

I know that both of you, asked if God with his infinite power and wisdom could give to
 FAITH VERSUS REASON 9

the element of water the reciprocal [tidal] motion which we observe in it by some other
means than by moving the containing vessel, I say that I know that you would answer
that He would have the power and wisdom to do this in many ways, even if unthinkable
by our intellect. Therefore I immediately conclude that, because of this, it would be
excessively bold if someone wanted to restrict and force divine power and wisdom to a
particular fancy of his [F488, my translation].

Galileo had been required by the censors, as a pre-condition for publication,

that at the end of the book he should include such an argument which the
pope found very impressive.14 Galileo did include the argument, which
appears on the very last page of the text in the 1632 edition.15 Though the
argument does constitute Simplicio’s last word, it is followed by a few short
remarks by Salviati, after which Sagredo in turn utters a few sentences. After
the book was published, some people found fault with the rhetorical force of
such an ending. In fact, part 2 of the sixth accusation of the original indict¬
ment charges Galileo with ‘‘having placed the medicine of the end in the
mouth of a fool and in a section which can be located only with difficulty, in
such a way that it is later approved by the other interlocutor in a cold manner
and only by casually mentioning and not stressing its worth, which is to say
unwillingly”.16 The pope himself seems to have at first accepted this inter¬
pretation and thus felt personally ridiculed, a feeling which helps to explain
his intransigence about the trial, conviction, and sentence.
In this passage the rhetoric is allegedly contained not in what is explicitly
said, but in how it is said; this manner of saying includes: (1) the failure to
state the argument in a physically separate section of the book, under some
such label as ‘Ending’ or ‘Conclusion’; (2) the fact that the argument is given
by the least intelligent of the three interlocutors; and (3) the fact that the
most intelligent of them, Salviati, accepts the argument somewhat incidentally
and casually. These three points are accurate when stated in this more
objective manner, which changes the characterization of Simplicio from ‘fool’
to ‘least intelligent’, and of Salviati’s attitude from ‘unwilling’ to merely
‘casual’ acceptance. There is no question that it is often rhetorically effective
to insinuate claims, without expressing them explicitly. Also there is no
question that Galileo did not like the kind of ending he was forced to include.
However, the rhetorical problem is to determine whether the text is indeed
such as to convey the impression that the skeptical argument is foolish.
In deciding this problem, it seems necessary to say a few words about the
logical merits of the argument. Is the argument per se a foolish argument?
Obviously, if Simplicio is giving a good argument, then to that extent his
intelligence appreciates in the eyes of the reader. In other words, if the argu-
 10 CHAPTER 1

gument is correct, then the ending provides evidence that Simplicio is not the
fool that some would like to believe. I think that the argument is in fact
unanswerable. Though expressed in what might be called the mythological
language of God and divine omnipotence and omniscience, the demytho-
logized version of the argument has become the stock-in-trade of modern
epistemologists and methodologists. In modern terminology we might say: no
scientific theory is ever conclusively proved since a scientific theory can be
proved only relative to a finite amount of data, and for any finite set of data
there is always more than one scientific theory about them, whether we view
the theory as being inductively inferred from the data or as being an hypo-
thetico-deductive explanation of the data. This is true basically because of the
so-called problem of induction, or because of the formal invalidity of affirm¬
ing the consequent.17 Moreover, the force of the argument was felt in
Galileo’s time as well. In fact, a very clear statement of the argument can be
found in a book published in 1629 by Augustinus Oregius, a theologian who
was later (in 1633) one of the three consultants asked by the Inquisition to
report on the content of the Dialogue as it related to the decree of 1616.18
In his book Oregius reports on a discussion between Galileo and the pope and
states that the pope

having agreed with all the arguments presented by that most learned man [Galileo],
asked if God would have had the power and wisdom to arrange differently the orbs and
stars in such a way as to save the phenomena that appear in heaven or that refer to the
motion, order, location, distance, and arrangement of the stars. If you deny this, said
Sanctissimus, then you must prove that for things to happen otherwise than you have
presented implies a contradiction. In fact, God in his infinite power can do anything
which does not imply a contradiction; and since God’s knowledge is not inferior to his
power, if we admit that he could have done so, then we have to affirm that he would
have known how. And if God had the power and knowledge to arrange these things
otherwise than has been presented, while saving all that has been said, then we must not
bind divine power and wisdom in this manner. Having heard these arguments, that most
learned man was quieted, thus deserving praise for his virtue no less than for his
intellect.

Finally, there is evidence that Galileo himself felt the force of the argument.
In fact, in the Third Day discussion of the motion of sunspots (F372-83,
D345-56) after Salviati has given a Copernican explanation, Galileo not only
has Simplicio state the point in logically clear terminology (F379-80) but
takes it seriously enough to have Salviati present an alternative (geostatic)
explanation, though the latter is rejected because of its ad hoc character.20
Galileo shows a similar awareness of the logic of theoretical explanation in
his discussion of the stellar parallax objection (F409-14, D383-7), where it
 FAITH VERSUS REASON 11

is Salviati who uses to his rhetorical advantage these logical facts,21 and in the
Second Day discussion of the arguments in favor of the earth’s diurnal
motion, where Salviati admits that the Copernican side is weakened by the
merely probable force of arguments based on hypothetical explanations
(F139-50, D114—24).22 A fourth relevant discussion occurs in the Fourth
Day, after Salviati has finished his explanation of the tidal diurnal period,
when Simplicio once again gives a relatively clear statement of the logic of
theoretical explanation and applies it to the basic tidal argument: Salviati
feels obliged to give a detailed answer. (F462-70, D436-44).23
Thus according to Galileo, his contemporaries, and modem philosophy of
science, ‘the argument of Urban VIII’ is a very good one. Galileo would not
care for the theological way of expressing the point, any more than we
moderns would. But if someone wants to speak the logical language, then
neither Galileo nor we can object to the substance of the point being made.
It follows that the rhetorical effect of the ending is not that claimed in
Galileo’s original indictment. Actually, the rhetorical import is favorable to
religion insofar as it shows that to think in theological terms leads one to the
same conclusion reached by nontheological, methodological arguments. In
fact, no one believed this charge, except the pope who was personally
involved; instead the fact that no one accepted the conclusiveness of Galileo’s
tidal argument is further evidence the ‘medicine of the end worked, though it
must have worked because of its logical content and not because of its
authoritative source.
If there is any rhetorical problem with this final passage in the book, it
is its pointlessness. In other words, the point has already been made at least
four times in the book (F139-50, F372-82, F409-14, F462-70), including
once in the Fourth Day, so it ought to be logically superfluous. The logically
sensitive reader, like Galileo, feels bored by the passage. Nevertheless.it may
be given the rhetorical function of reemphasizing one last time an important
point, and of doing so in language more easily understood to the man in the
street with whom Galileo wanted to communicate. This may help explain
why Galileo made the final argument part of the text of the Fourth Day,
instead of having a short special section devoted to it. This separate section
which some of the censors favored would then have been rhetorically less
effective from their own point of view, or at least less effective if their aim
had been to impart some genuine understanding, however rudimentary, to
the masses. Unfortunately, their aim was more likely to be one of belief-
control by mere obedience and submission of will. From this point of view,
of course, a separate section, containing a more explicit theological qualific-
 12 CHAPTER 1

ation, would have been better since then people could have just looked at the
‘Ending’ and would have been informed that the motion of the earth had not
been proved in the preceding book; they would have thereby saved them¬
selves the trouble of reading it, and if their understanding of the problem
would have been none the better for it, their actual beliefs would have been
the ‘right’ ones.
Thus, from the rhetorical point of view, the controversy over the book’s
preface and ending is very enlightening. It illustrates two different con¬
ceptions of rhetoric, a Galilean kind according to which one tries to induce
assent by arguments which may have to be simplified and expressed in appro¬
priate language so that it can have an effect on the mind and understanding
of the audience, and an authoritarian kind according to which one’s utmost
concern is to specify clearly what is or is not to be believed so that one’s will
can easily act upon it. Nothing illustrates this better than the contrast
between the logically incoherent and rhetorically ineffective preface drafted
by the pope’s secretary, and the logically superfluous but rhetorically appro¬
priate ending that Galileo came up with. It seems that, in trying to end the
book in accordance with what the Church wanted, Galileo could not help
doing so in a competent manner; whereas in explicitly compiling the preface,
the pope’s secretary could not avoid doing so in an incompetent manner.
The fact that neither the crude rhetoric of the preface, nor the more elegant
one of the ending, helped Galileo in the short run is perhaps an indication
that he was doomed all along, a measure of the magnitude of the tragedy.

THE TITLE, OR THE RHETORIC OF INDECISION

So far I have considered what may be called the religious rhetoric contained
in the Dialogue. Another equally obvious level of rhetoric relates to the
problem of deciding on the relative merits of the two sides, the Copernican
and the Peripatetic. The emphasis here is on decision, rather than on dis¬
cussion; that is, there is throughout the book a recurrent theme that its
author is merely discussing the relative merits of the two positions, rather
than deciding on them, which decision is something to be reached by the
proper authorities, who however may use the book’s discussion as a basis.
This is the rhetoric inherent in the book’s full title, which it is useful to trans¬
late literally. Dialogue by Galileo Galilei, Lyncean Academician, Extra¬
ordinary Mathematician at the University of Pisa, and Philosopher and Chief
Mathematician to the Most Serene Grand Duke of Tuscany; where in
meetings over the course of four days one discusses the two Chief World
 FAITH VERSUS REASON 13

Systems, Proposing indeterminately the Philosophical and Natural reasons for

the one as well as for the other side. The same theme is explicitly stressed on
at least nine occasions, which it is useful to discuss one by one.
At the beginning of the Second Day, Sagredo’s first speech summarizes the
First Day by saying that the day before they had examined which of two
opinions was “more probable and reasonable (FI32): the Aristotelian
division of the universe into a celestial and a terrestrial part, or a second
opinion according to which the earth is a globe very much like the other
planets. Sagredo’s summary ends by saying that the day before they had
“concluded that this second opinion has more verisimilitude than the other’’
(F132). Upon hearing this, Salviati is quick to add the disclaimer that

I did not conclude this, just as I am not deciding upon any other controversial proposi¬
tion. My intention was only to adduce those arguments and replies, as much on one side
as on the other — those questions and solutions which others have thought of up to the
present time (together with a few which have occurred to me after long thought) - and
then leave the decision to the judgement of others. [D157]

A second disclaimer is found in the course of the preliminary statement

of the objections to the earth’s diurnal motion. Here, at one point Sagredo
stresses that Salviati seems to understand very well those geostatic arguments,
this fact has impressed the Aristotelian Simplicio himself, who has let those
arguments be expounded by Salviati even though the latter does not accept
them and will afterwards criticize them. After Sagredo s speech, before going
on to give a statement of some other geostatic arguments, Salviati notes:

Before going further I must tell Sagredo that I act the part of Copernicus in our argu¬
ments and wear his mask. As to the internal effect upon me of the arguments which I
produce in his favor, I want you to be guided not by what I say when we are in the heat
of acting out our play, but after I have put off the costume, for perhaps then you shall
find me different from what you saw of me on the stage. [D131; cf., F157-8.]

Since Salviati is indeed the Copernican spokesman, and since in this discussion
he is acting as the expounder of Aristotelian arguments, it seems at first that
his justification is misplaced, and that instead he should call attention to
the fact that he is now wearing an Aristotelian mask. However, upon reflec¬
tion, since in the present discussion his Aristotelian mask is obvious, what he
is probably doing is to take this opportunity to call attention to the less
obvious fact that his whole Copernican position is an act in a play. Indeed
this is so much less obvious, that it may not be true at all; it certainly is not
convincing, and there is no evidence other than his words that it is so. At any
rate, such is Galileo’s rhetoric.
 14 CHAPTER 1

The third relevant passage is one at the end of the critique of the objection
that if the earth rotated that would mean that our senses would be deceived.
Salviati is trying to show that there would be no deception of the senses,
since we are all agreed upon what the sensory appearances are; the problem
is really with the conclusions we draw therefrom. It is wrong for some Peripa¬
tetics to make it sound as if on a moving earth we would experience terrestrial
phenomena differently, or as if the believers in the earth’s motion are seeing
things differently; in giving a reason for this, Salviati inserts the disclaimer
that “just as I (who am impartial between these two opinions, and masquerade
as Copernicus only as an actor in these plays of ours) have never seen nor ever
expect to see the rock fall any way but perpendicularly, just so do I believe
that it appears to the eyes of everyone else” (D256; cf., F281).
At the end of the Second Day, its main conclusion is summarized.
Simplicio asserts that all that has been shown is that the geostatic arguments
are not compelling, and that so far no compelling argument has been given to
convince one of, or to prove, the earth’s motion. Salviati replies:

I have never taken it upon myself, Simplicio, to alter your opinion; much less should I
desire to pass a definite judgment on such important litigation. My only intention has
been, and will still be in our next debate, to make it evident to you that those who have
believed ... [in the earth’s motion] were not blindly persuaded of the possibility and
necessity of this. Rather they had very well observed, heard, and examined the reasons
for the contrary opinion, and did not airily wave them aside. [D274; cf. F298]

Next, in the Third Day, after Salviati has finished giving both Copernican
and geostatic explanations of the apparent motion of sunspots, each of the
three interlocutors expresses Iris feelings on their relative merits. Simplicio
confesses that the decision between the two is too important for him to make
(F382—3). Sagredo thinks that he has never heard anything more convincing,
other than the demonstrations of pure mathematics. Salviati states:

I do not give these arguments the status of either conclusiveness or of inconclusiveness,

since (as I have said before) my intention has not been to solve anything about this
momentous question, but merely to set forth those physical and astronomical reasons
which the two sides can give me to set forth. I leave to others the decision. [D356]

On the next page, they begin examining the arguments contained in a

book by one of Galileo’s contemporaries, where they find, among others, the
religious and Biblical objections that Copernicans are allegedly forced to say
that Christ ascended into hell and descended into heaven, and that when
Joshua ordered the sun to stand still, it was instead the earth that stood still.
Salviati takes exception to this writer’s “having mixed passages from the ever
 FAITH VERSUS REASON 15

venerable and mighty Holy Scriptures among these apish puerilities, and his
having tried to utilize sacred things for wounding anybody who might, with¬
out either affirming or denying anything, philosophize jokingly and in sport,
having made certain assumptions and desiring to argue about them among
friends” (D357).
Another disclaimer is found at the end of the discussion of the stellar
dimensions required in a Copernican universe by the lack of apparent annual
parallax. Sagredo has just finished a disquisition against the arrogance, indeed
the ‘insanity’ of those who seem to argue that, because we don’t know how
Jupiter or Saturn is of service to us, therefore they are superfluous, or even
they do not exist (F438). Salviati tries to calm him down, saying that “there
is no need, Sagredo, to probe any further into their fruitless exaggerations.
Let us continue our plan, which is to examine the validity of the arguments
brought forward by each side without deciding anything, leaving the decision
to those who know more about it than we” (D369).
At the end of the Third Day, in concluding the discussion of magnetism,
Sagredo refutes the objection that a simple body like the earth cannot have
several natural motions by arguing that the earth is probably not a simple
body, and that anyway it probably is a loadstone, and loadstones have several
natural motions. In his criticism at the very end of that discussion, Sagredo
comes down pretty hard on the Aristotelians. Salviati then makes the follow¬
ing plea for moderation:

Sagredo, please, let us weary ourselves no longer with these particulars, especially since
you know that our goal is not to judge rashly or accept as true either one opinion or the
other but merely to set forth for our own pleasure those arguments and counterargu¬
ments which can be addressed for one side and for the other.. . therefore we shall sus¬
pend judgment, and leave this in the hands of whoever knows more about it than we do.
[D413]

Finally, at the end of the Fourth Day, just before the skeptical “medicine
of the end” is given by Simplicio, Salviati somewhat apologetically asserts
that “I do not claim and have not claimed from others that assent which I
myself do not give to this invention, which may very easily turn out to be a
most foolish hallucination and a majestic paradox” (D463).
This rhetoric of indecision has never impressed anyone. The problem with
it is that it is merely cosmetic. One might say that if Galileo’s arguments were
really equally balanced, there probably would be no need for these constant
explicit disclaimers. I do not think that the task described in the title is
intrinsically impossible; rather it is valid to distinguish between an impartial
 16 CHAPTER 1

presentation of arguments for two opposite sides, and a decision as to which

side is better. The problem with the (full) title of the book is that in the
Dialogue the presentation of the arguments is such that there can be no
question as to which side is more plausible or probable, as Galileo’s enemies
and the inquisitors were quick to point out. It is not so much that the
presentation is biased, but rather that it determines a decision.
It is well known that such rhetoric was imposed on Galileo by the Church
censors. In fact, Galileo would have liked at least to mention the ebb and
flow of tides in the book’s title, as is shown by his correspondence between
the years 1624-1631.24 However, such a title was vetoed by the interpreters
of the pope’s desires.25 What seems to have happened is that Galileo’s better
rhetorical instincts were channeled in the wrong direction by Church officials.
To begin unraveling this problem let us examine the view, widely prevalent
among scholars, that Galileo’s motivation for the original title was his belief
that the tidal argument of the Fourth Day provided a conclusive proof of the
earth’s motion, and that the Church’s motivation for the change was its
unwillingness to put a stamp of approval on the argument.26 Such a view is
(a) based on no direct evidence, (b) rhetorically naive and implausible, and (c)
a misinterpretation of the crucial idea of hypothetical reasoning.
(a) The only scholar who gives any evidence is Shea. He refers to Galileo’s
letter to Elia Diodati of August 16, 1631 and quotes the statement that “I
have not been able to obtain permission to mention the tides in the title of
the book although it is the principal argument that I consider in it”.27 How¬
ever, this quotation is taken out of context, for after a few sentences the
letter ends with the assertion, “I think that, if the book had been entitled on
the ebb and flow, that would have been to the greater profit of the printer.
But after a while the news will spread by means of the first ones who will
have read it.”28 Thus if this letter proves any tiring (and it may not), it is that
he is concerned with the book’s circulation and popularity. He seems to feel
that more people would be likely to buy or read a book whose title claimed
to be dealing primarily with tides. Is Galileo’s judgment right here? A point
in its favor is that whereas the Copernican theory was the subject of the pro¬
hibition of 1616, the tides were not the subject of any such decree.
(b) Second, in the light of the fact that the anti-Copernican decree was
still standing, it seems implausible that Galileo would want to call attention,
in the very title of his book, to an alleged proof of its philosophical incor¬
rectness. Whatever private or verbal assurances may have been given by
Church officials or even the pope himself, Galileo would not have been so
rash as to advertise his violation of the decree, for in 1616 he had wanted to
 FAITH VERSUS REASON 17

be safe enough as to obtain a letter from Cardinal Bellarmine certifying that

Galileo had not been the object of any disciplinary proceedings in connection
with the decree.29 Moreover, this same letter which Galileo had kept stated
explicitly that the Copernican theory could not be ‘held or defended . In
other words, it would have been out of character for Galileo, and intrinsically
imprudent, to want the tides mentioned in the title out of a desire to stress
the argument from tides.
But if we agree that Galileo was far too prudent to have such a wish, isn t
it a fact that it was dangerous for him to mention the tides, in the light of
the tidal argument? It all depends on the nature of this argument. There is no
question that Galileo thought that this was a very good argument, indeed his
strongest one in favor of the earth’s motion. However, there is also no
question of Galileo’s awareness of the hypothetical-explanatory character of
this argument; as mentioned above, he refers to this explicitly twice in the
Fourth Day, once when the logic of reasoning ex suppositione is mentioned
(F462), and once with theological language at the very end; and, as em¬
phasized above, this only reflects an attitude that Galileo expresses on at least
three other occasions in the book. Moreover, the actual structure of the
Fourth Day is hypothetical in the sense that a problem is presented, various
solutions criticized, and an original one proposed. I conclude that his desire
to have the tides mentioned in the title stems from his desire to emphasize
the hypothetical character of his book, and thus conform to the decree.
This conclusion is reinforced by the various descriptions of the title found
in Galileo’s correspondence between 1624 and 1630.
(c) A third point requires discussion now, relating to the distinction among
three different ways of treating a topic, which we may call absolute, hypo¬
thetical and indeterminate. This distinction is mentioned in the original
indictment against Galileo,31 where it is charged that the book’s treatment of
the earth’s motion is either absolute or at least indeterminate, but not hypo¬
thetical. To treat the earth’s motion absolutely means to discuss it in such a
way that its reality is asserted unconditionally; this requires arguments
which are alleged to be deductively valid and based on premises which are
alleged to be indisputably factual. An indeterminate treatment of the earth’s
motion is one where all available arguments for and against it are presented
without a final decision being reached concerning the truth or falsity of the
conclusion. An hypothetical treatment is one where the supporting arguments
are such that the conclusion (earth’s motion) is presented not as a logical
consequence of known truths but as a conjectural explanation of them, m
the sense that if the conclusion is assumed to be true then the known truths
 18 CHAPTER 1

can be seen to follow. The book’s actual title and the above-mentioned
corresponding rhetoric throughout its pages claim that an indeterminate treat¬
ment of the earth’s motion is being given. This is pretty obviously false
because it is immediately clear even to a casual reader that Galileo does show
that the favorable arguments are better than the unfavorable ones; hence he
is in fact deciding the question; hence, the explicit disclaimers to the contrary
are mere rhetoric , that is, they have no real rhetorical force. Unfortunately,
even if the book had really been an indeterminate treatment, it still could
have been damned, and hence the title is still rhetorically infelicitous; for it
was pointed out in the original indictment33 and in the final sentence34 that
it is wrong to regard as undecided the question of whether or not the earth
moves, which is decided in the Bible and anyway had been decided by the
decree of 1616.
The same is not true of an hypothetical treatment, which may not have
violated the spirit of the decree, and which therefore might very well have
been tolerated, though of course it certainly is a way of teaching the subject,
and it also is a weak kind of defending an idea. Therefore, a title mentioning
the tides, which would have stressed the hypothetical character of the discus¬
sion of the earth’s motion, would have been rhetorically superior from
Galileo’s point of view. This is the sense in which his better rhetorical
instincts were misdirected by the censors.

THE RHETORIC OF STRICT DEMONSTRATION

In view of the presence in the Dialogue of such obvious rhetoric of religious

piety and of indeterminancy, it would seem incredible that anyone would
seriously claim to find in it a rhetoric of rigorous proof, from which point of
view the book would be a conscious attempt to give a strict demonstration of
Copernicanism.35 Once again, the emphasis here is on the ‘rhetoric’, for the
same scholars who stress this aspect of the book are the first ones to point out
that it did not succeed in the goal, partly because it is an unrealizable goal,
partly because the book’s arguments are not in fact strictly demonstrative, and
partly because several of the arguments are erroneous in one way or another.
The basic problem here is not so much that the book completely lacks explicit
expressions of the desirability of strict demonstrations and implicit claims of
the conclusiveness of its arguments, but rather that the former are very rare
and the latter never occur in the same context as the former, so much so
that this kind of rhetoric becomes visible only with the tunnel vision which it
is possible to acquire through certain kinds of scholarly discipline.
 FAITH VERSUS REASON 19

By explicit expressions of the desirability of strict demonstrations, I mean

passages like the following:

[1] I take it to be definite and certain that for the proof of a true and necessary con¬
clusion there are in nature not merely one but very many powerful demonstrations ....
I believe on the other hand that to make a false proposition appear true and convincing,
nothing can be adduced but fallacies, sophisms, paralogisms, quibbles, and silly incon¬
sistent arguments full of pitf alls and contradictions. [D130, cf. F 156.]

[2] It is not possible within the bounds of human learning that the reasons adopted by
the right side should be anything but clearly conclusive, and those opposed to them, vain
and ineffective. [D356; cf. F383.]

It is not easy to find any other such passages.

By implicit claim of the conclusiveness of arguments I mean claims,
occurring in The course of the presentation of an argument, emphasizing the
strength of the inferential links, and the certainty of the conclusion. For
example:

[1] Having established, then, that it is impossible to explain the movements perceived in
the waters and at the same time maintain the immobility of the vessel which contains
them, let us pass on to considering whether the mobility of the container could produce
the required effect in the way in which it is observed to take place. [D424]

[2] You see, gentlemen, with what ease and simplicity the annual motion - if made by
the earth - lends itself to supplying reasons for the apparent anomalies which are
observed in the movements of the five planets, Saturn, Jupiter, Mars, Venus, and
Mercury. It removes them all and reduces these movements to equable and regular
motions; and it was Nicholas Copernicus who first clarified for us the reasons for this
marvelous effect.
But another effect, no less wonderful than this, and containing a knot perhaps even
more difficult to untie, forces the human intellect to admit this annual rotation and to
grant it to our terrestrial globe. This is a new and unprecedented theory touching the sun
itself For the sun has shown itself unwilling to stand alone in evading the confirmation
of so important a conclusion, and instead wants to be the greatest witness of all to this,
beyond exception. So hear this new and mighty marvel. [D344-5]

I am stressing here the fact that such remarks occur during the statement of
arguments; it is this that gives these remarks their implicit character. The
contrast to keep in mind is to those remarks made after the argument has
already been stated where, as we saw in part in our discussion of the rhetoric
of indeterminateness, contrasting reflecting judgments are passed upon the
argument: Simplicio remains usually unimpressed and states various reser¬
vations, Sagredo is usually overimpressed and finds the arguments very strong
 20 CHAPTER 1

and compelling, whereas Salviati pays his Up service to the indeterminacy-

claim inscribed in the book’s title.
The important thing to notice about these two kinds of passages is that
they never occur together. Hence, the implicit claims must be regarded as an
integral part of the normal rhetoric of argumentative prose, rather than as
elements of a ‘strict demonstration’ theory of science instantiated by the ac¬
tual arguments in the statement of which these claims are made. On the other
hand, the few explicit claims about strict demonstration occur in other con¬
texts. The first one of these explicit claims is actually made by Simplicio and
is part of a quasi-Socratic, somewhat ‘ad hominem’36 metalogical discussion
trying to show to Simplicio that the consideration of arguments both for and
against a view, far from leading to confusion of mind, is the only way of
arriving at the truth. The ‘strict demonstration’ aspect of Simplicio’s remarks
is completely incidental; it is either an exaggerated way of expressing the con¬
trast between good and bad arguments, or a reflection of the Aristotelian
theory of science of Posterior Analytics. The second explicit passage, on the
other hand, is part of Salviati’s disclaimer at the end of the discussion of sun¬
spots in the Third Day. The remark is actually a disclaimer within a disclaimer,
for Salviati first says that, in accordance with his usual policy, he is not going
to label the argument from sunspots as either conclusive or inconclusive, but
then adds that ultimately the authoritative determination cannot be itself
inconclusive since one of the two sides is true and the other false, and the
arguments for the truth must be as conclusive as those for the falsehood are
inconclusive.
And this brings us to an examination of the argument which leads certain
scholars to find a rhetoric of strict demonstration in the Dialogue; it turns out
(as if Saviati s disclaimer within a disclaimer were true) that this argument is
as unsound as its conclusion is false, which I have just shown. Let us consider
the account given by Ernan McMullin, who is perhaps its leading proponent.37
He begins with the following quotation from the Third Day:

However well the astronomer might be satisfied merely as a calculator, there would be
no satisfaction and peace for the astronomer as a scientist .... (Copernicus) very well
understood that although the celestial appearances might be saved by means of
assumptions essentially false in nature, it would be very much better if he could derive
them from true suppositions .... One of the arrangements must be true and the other
false. Hence it is not possible within the limits of human learning that the reasons
adopted by the right side should be anything less than clearly conclusive, and those
opposed to them vain and ineffective.38

This beginning is an inauspicious one, for the first two sentences of this
 FAITH VERSUS REASON 21

quotation are separated from the last two by tifteen pages of text, as
McMullin’s own footnote indicates.39 This questionable practice has the
effect of connecting by fiat two tilings which do not have the requisite con¬
nection, namely Galileo’s epistemological realism and his alleged rhetoric of
strict demonstration. The first is indeed a basic feature of Galileo s work, but
the second is an insignificant one (at least as far as the Dialogue is concerned).
We may agree that if someone is a ‘strict demonstrationist’ then he would
have to be an epistemological realist; however the converse is not true since
one can be a realist and a probabilist. Hence the quotation as given supports
a rhetoric of demonstration only if its last two sentences do. Unfortunately
these two sentences are injudiciously taken out of context, which is that of
Salviati’s disclaimer of a disclaimer. Moreover, though McMullin’s last two
sentences are an accurate quotation from the translation from which it is
being quoted, it so happens that the translation of this particular passage is
clearly incorrect precisely at the place needed as evidence for strict demon¬
stration’. In Santillana’s revision of Salusbury the last sentence reads, “It is
impossible that (always confining ourselves within the limits of human
doctrine) the reasons alleged for the true hypothesis should not manifest
themselves as conclusive as those for the contrary vain and ineffectual .
The original Italian (F383) makes it even clearer that what is meant is a com¬
parison of the degree of conclusiveness of one side with the degree of in¬
conclusiveness of the other.
A second difficulty with McMullin’s account in general is the unclarity ot
the notion of a demonstration of truth; that is to say, it is difficult to see
what he is talking about in putting forth his claim. It will not do to use the
clear terminology of formal logic and say that it is a formally valid argument
with true premises. In fact, if the argument’s conclusion is an empirical
proposition (as in the present case), then at least one of the premises has to
be empirical, and hence its truth is always questional even if it is unques¬
tioned. Since the success of a demonstration so conceived depends both on
the truth of the premises and on the validity of the inference steps, it follows
that it could always be questioned, which is to say that such demonstrations
are impossible for empirical claims. Now McMullin cannot be excused by
saying that all he is doing is to report Galileo’s beliefs, intentions, and amis;
for if he wants to do this in the sense of ‘merely reporting’, then he is not
conforming to his own aim since it is obvious he is also interpreting and
trying to make sense out of that subject matter; but in doing the latter it is
methodologically undesirable to use an intrinsically incoherent concept.
A third difficulty is that the view is largely based on events and documents
 22 CHAPTER 1

relating to the years 1615—1616, which climaxed in the anti-Copernican

decree, the main document being Galileo’s Letter to the Grand Duchess
Christina. However, this letter is three steps removed from the Dialogue,
first chronologically; second, insofar as it belongs to the context of theory
rather than practice, that is, it contains explicit remarks on science, but no
actual scientific arguments; third, insofar as even within the letter there are
two conflicting theories of science being put forth, and only one of these
includes the ‘strict demonstration’ ideal.41
Fourth, McMullin argues that, since there is evidence that “Galileo was a
thoroughgoing Aristotelian”,42 therefore, as an Aristotelian trying to con¬
vince other Aristotelians, Galileo was required by the theory of science of
Posterior Analytics to present a strict demonstration of his thesis in order to
get a hearing. As a historical interpretation of the Dialogue, McMullin’s argu¬
ment is at best circular (which is not to say that his conclusion is false or his
argument logically incorrect), since what is presently in question is whether
this book is Aristotelian in the required manner. At any rate, the direct
evidence from this work is that Galileo develops, by both words and deeds, a
theory of science different from the Aristotelian one and such as to obviate
the need for a strict demonstration (at least for the moment). Galileo does
this by examining the ‘strict demonstration’ given by the Aristotelians in
support of their geostatic view, finding that their ‘strict demonstration’ is
faulty in several ways, concluding that the Aristotelians have not proved the
geostatic view, presenting and formulating the best arguments he can in
support of the contrary view, and throughout this argumentation never
avoiding but always discussing relevant epistemological and methodological
issues with an eye toward justifying his procedure.43

THE JUSTIFICATION OF COPERNICANISM

If the effective rhetorical content of the Dialogue is not that of an apology

for the decree of 1616, nor that of an indeterminate pre-decision study of the
evidence, nor that of a strict demonstration of Copernicanism, in spite of the
‘rhetoric’ to the contrary, what else is it? It is pretty obvious that it is a
justification of Copernicanism. Agreement concerning this can be found in
the action and words of the theologians and Church officials involved in the
prosecution of Galileo, in the mostly favorable reaction by seventeenth
century scientists and philosophers, and in the judgment of modern scholars.
By a justification of Copernicanism I mean an attempt by verbal means and
techniques to induce or increase adherence to Copernicanism.
 FAITH VERSUS REASON 23

The justification is carried out at three conceptually distinct (though

not completely separable) levels, logical, philosophical, and rhetorical. The
logical aspect of the justification is the one that involves presentation,
analysis, and evaluation of the evidence and arguments for and against Coper-
nicanism; the philosophical aspect is the one that involves the discussion and
clarification of the epistemological and methodological concepts inherent
in this attempt to show the evidentiary superiority of Copernicanism over
geostaticism', the rhetorical aspect, in a more restricted sense of rhetorical
than the general one used so far in this chapter, involves the attempt to view
in a favorable light Copernican theses, concepts, and procedures, and in an
unfavorable light the Aristotelian counterparts, and to do so by means other
than logical, that is primarily by arousing the emotions of the reader in the
appropriate way.
What I am calling here the logical dimension of the justification of
Copernicanism corresponds to what is often called the scientific aspect of
Galileo’s problem. In subsuming the analysis of evidence, under the concept
of ‘logical’, I am using the label in a somewhat wider meaning than is custom¬
ary, but I do so in order not to pre-judge questions like whether there are two
kinds of logic and of reasoning (namely, deductive and inductive), and if so
which is more important and prevalent. I think that ‘logical’ is preferable to
‘scientific’ because of the more time-dependent connotations of the latter,
that is, because ‘scientific’ is often plausibly taken to mean pertaining to
present-day scientific knowledge”. Though it is important and useful to
examine Galileo’s Dialogue from the scientific point of view in this sense
(as I do in Chapter 4), it is equally important to examine it from the point of
view of an objective analysis and presentation of evidence and arguments
(which I do in the next chapter); this is what I am labeling the ‘logical’
dimension.
There is universal recognition of the need for a philosophical discussion,
besides a logical one, in a situation like that faced by Galileo, which involved
very fundamental changes in our cognitive and intellectual structure. For
everybody recognizes that such changes involve not only changes in specific
beliefs and conclusions, but also changes in general, philosophical attitudes
and perspectives. Hence the philosophical dimension of the Dialogue has been
widely studied, though no one has done so systematically (as I do in Chapter
5) with the result that even the best scholars have misinterpreted its philoso¬
phical content (Chapter 10)44 or failed to appreciate its true philosophical
significance (as I argue in Chapter 6). .......
As regards the restrictedly rhetorical level of Galileo’s justification (in an
 24 CHAPTER 1

alogical, but not illogical sense of ‘rhetoric’), this fact is generally admitted
but seldom appreciated. As suggested above, Paul Feyerabend is the only
recent scholar who has perceived the necessity of such (alogical) rhetoric,
though his point is obscured by his own ‘rhetoric’, which is one of irration¬
alism.45 His point is simply that for certain very fundamental intellectual
changes (like the scientific revolution of the 17th century), logic and even
philosophy are not enough, and one needs rhetoric in the sense under dis¬
cussion. However, as I argue later,46 one cannot even begin to understand this
rhetoric unless one has a clear understanding of the book’s logical structure.
Hence we shall proceed to examine the latter before going on to the analysis
of the means, above and beyond the logical-scientific and the philosophical-
methodological, whereby Galileo justifies Copernicanism.

NOTES

1 A. Koyr6, Etudes Galiliennes, pp. 212-215; L. Geymonat, Galileo Galilei, pp. 132-
135; and S. Drake, Galileo Studies, pp. 253—255.
2 P.K. Feyerabend, Against Method, esp. Chapters 1, 12, and 18. Cf. also my discussion
of Feyerabend’s views in Chapter 8 below.
Besides Feyerabend, the problem has been discussed in T. S. Kuhn, The Structure of
Scientific Revolutions, especially pp. 144-160; and in I. Lakatos and A. Musgrave (eds.),
Criticism and the Growth of Knowledge, especially pp. 1-24, 91-196, 197-278.
As I have argued in my History of Science as Explanation, pp. 180-198.
These three approaches are discussed in more detail in my Essay-review of Lakatos’s
Criticism and the Growth of Knowledge. The third conclusion, about the irrationality of
science, corresponds, of course, to Feyerabend’s rhetoric.
This is the conclusion that Feyerabend ought to be drawing, and the one that we may
take him to be really drawing. Another very interesting conclusion to draw would be
that important segments of actual, historical science exhibit features of what is labeled
‘philosophy’ in H. W. Johnstone, Jr., Philosophy and Argument. The coincidences are,
in fact, remarkable: for example, Johnstone speaks of conflicting philosophical state¬
ments lacking the property of being ‘logically commensurate’, whereas Kuhn and
Feyerabend speak of scientific theories being incommensurable; Johnstone holds that
the arguments for and against a philosophical statement are part of its meaning (p. 32),
whereas Kuhn and Feyerabend speak of meaning variance and hold that the meaning of
scientific terms is theory-dependent. This opens the way for vindicating the rationality
of scientific revolutions by making use of Johnstone’s theory of philosophical argument¬
ation.
7
A fascimile of the 1632 edition is now published by Culture et Civilisation, 115
Avenue Gabriel Lebon, Brussels.
8 G. Galilei, Opere, ed. Favaro, 19, 326; cf. G. de Santillana, The Crime of Galileo
£.211.
Opere, 19, 326; my translation.
10
Ibid., pp. 402-426.
 FAITH VERSUS REASON 25

11 Ibid., pp. 327-330.

12 Here one cannot help but commenting on the apparent inconsistency of the charge
numbered (6.2) in the original indictment (Opere 19, 326) for this charge, besides
accusing Galileo of having alienated the preface from the body of the text, also charges
him with having made the prescribed ending (the so-called 'medicine of the end , or the
argument of Urban VIII’) a completely integral part of the text (in the last few pages).
It must be admitted, however, that the inconsistency is more apparent than real, for the
charge is accusing Galileo of excesses in both directions, that is, in the case of the pre¬
face, setting it too much apart from the text, and in the case of the ending, integrating
it so muclTin the text as to be difficult to find. One might add, however, ‘Difficult to
find for whom?’ The answer would seem to be, “For people and readers interested only
in beginnings and endings”. But this means for people primarily interested in form or
appearance or in ‘mere’ rhetoric.
opere 7, 29-31. References to this book will be made in the text by preceding the
page number(s) by ‘F’, short for 'Favaro .
14 Opere 19, 327, 330; cf. G. Galilei, Dialogo sopra i due massimi sistemi, edited by L.
Sosio, pp. 548-549, n. 1.
15 G. Galilei, Dialogo (Florence, 1632), p. 458; cf. F488-489, and G. Galilei, Dialogue
Concerning the Two Chief World Systems, tr. Drake, pp. 464-465. Subsequent
references to the last book will be given in parenthesis within the text, by prefixing the
page number(s) with a ‘D’.
16 Opere 19, 326; my translation.
17 See, for example, C. G. Hempel, Philosophy of Natural Science, pp. 3-18; and A.
Pap, An Introduction to the Philosophy of Science, pp. 139-173.
18 Opere 19, 348; cf. Santillana, The Crime of Galileo, p. 245.
19 Quoted in Galilei, Dialogo, edited by L. Sosio, pp. 548-5 95, n. 1 from Augustinus
Oregius, De Deo uno (Rome, 1629), pp. 194-195.
20 Cf. Chapter 5, passage F372-383.

21 F413-414; cf. my reconstruction in Chapter 5.

22 Cf. my reconstruction in Chapter 5.
23 Cf. my reconstruction in Chapter 5.
24
Opere 13 and 14.
25
26 E^McMullin, 'Introduction: Galileo, Man of Science’, p. 34; A. Koestler The Sleep¬
walkers, p. 480; W. R. Shea, Galileo’s Intellectual Revolution, pp. 173-174; and G. de
Santillana, The Crime of Galileo, p. 183.
27
" Quoted by Shea, Galileo's Intellectual Revolution, p. 187, n. 12. Cf. Opere 14, 289.
28 Opere 14, 289, my translation.
29 Santillana, The Crime of Galileo, pp. 125-144; cf. Opere 19, 348.
30 Opere 13, 236, and 282; 14, 49, 54, 66, and 130.
31 Opere 19, 325, lines 38—48; and 19, 326, lines 95 — 100.
32 This conception can be found in the report of one of the consultants to the In-

quisition: Opere 19, 356, lines 24 28.

83 Opere 19, 326, lines 99-100.
34 Ibid., p. 404, lines 62-68.
31-35; and idem,
35 See, e.g., E. McMullin, 'Introduction: Galileo, Man of Science , pp.
‘The Conception of Science in Galileo’s Work’, pp. 247-51.
26 CHAPTER 1

36
In the old, seventeenth century meaning of this term.
McMullin, ‘Introduction: Galileo, Man of Science’.
39 Quoted McMullin, pp. 31-32 from Galilei, Dialogue, tr. Drake, pp. 341, 356.
4q McMullin, op. cit., p. 40, n. 70.
Galilei, Dialogue on the Great World Systems, tr. Salusbury-Santillana p 366
Cf. McMullin, pp. 33-34.
42 Ibid., p. 32.
43
For more details, see Chapters 2 and 5.
44
In the part dealing with Clavelin.
Cf. Against Method, and Chapter 8.
6 See Chapter 8.
 CHAPTER 2

FACT AND REASONING:

THE LOGICAL STRUCTURE OF GALILEO’S ARGUMENT

Here I shall provide a solution to a problem which very much deserves discus¬
sion, which I shall call the problem of the structure of the Dialogue. The first
aspect of this problem derives from the fact that there is a scholarly need for
a descriptive outline of the book. Such an outline would facilitate communi¬
cation among scholars insofar as it would provide a standard way of referring
to a particular passage or discussion. Moreover, it would provide an easy way
of locating a particular passage that one may be looking for; at present this is
rather difficult because, though the book is very long, the only signposts are
the division into four ‘Days’, the few diagrams which are interspersed in the
text, and the numerous marginal notes by the author himself. The usefulness
of the latter is seriously impaired by the fact that they are not formulated
from a single point of view. In fact, what one reads in these marginal notes is
not always an idea propounded by Salviati and opposed by Simplicio, but
sometimes one propounded by the latter and opposed by the former, some¬
times an idea accepted by both, sometimes one rejected by both, and some¬
times a speculative idea being considered only hypothetically. Now, in order
to provide an adequate descriptive outline of the whole book, one needs to
study and ascertain its structure, since one needs to consider the relationships
among the various sections.
The second aspect of the problem is the book’s apparent lack of structure,
or lack of explicit structure. The Dialogue is such that it is very natural and
easy for the reader to feel that, in de Santillana s eloquent words, it meanders
at ease across the whole cultural landscape of the time, carrying in its broad
sweep strange material of various origin. As a composition, it looks unfinish¬
ed, unpolished, at times inconsistent. This is partly nature, partly art. It has
no’ unity except that of life itself”.1 This leads one to ask whether the
apparent lack of structure, which is undeniable, is real; whether the book
really lacks unity; or, if its unity is ‘that of life itself, what this amounts to.
It will be shown below both that the proposition ‘the earth moves’ provides
the book with unity and exactly how this is so.
The third aspect of the problem is the book’s wealth of content; the con¬
tent is so rich that it creates confusion. By wealth of content I mean the fact
that the book deals explicitly with and has obvious implications for philos-

27
 28 CHAPTER 2

ophy, physics, astronomy, literature, practical polemics, and pedagogy.

Koyr6 is making this point when he says at the beginning of his discussion of
Galileo’s Dialogue, in the third part of Etudes Galileennes:

The Dialogue on the Two Chief World Systems pretends to be an exposition of two rival
astronomical systems. But, in fact, it is not an astronomy book, nor even one of physics.
It is above all a book of criticism, a work of polemic and struggle; it is at the same time a
pedagogical work, and a philosophical work; it is finally a history book, ‘The History of
Galileo’s Mind’.2

This remark needs qualification as regards the astronomical and physical

content of the book; nevertheless it conveys the right idea. This wealth of
content is, of course, one of the things that makes the book uniquely
important and importantly unique in world history, but it can be confusing.
The wealth of content is pretty well displayed by Shea in his book Galileo’s
Intellectual Revolution,3 Shea’s account is adequate, competent, and helpful
and it probably represents the best that can be accomplished by following the
procedure of retaining in one’s account the mixture of topics and aspects
found in the original work. Yet a mere mixture has no structure. To find
structure along these lines one must do the following; one must begin by
distinguishing several points of view; physical, philosophical, astronomical,
literary, rhetorical, dramatic, pedagogic; then one must choose one point of
view, and then reconstruct (reconstruct, not merely summarize) the whole
book from that point of view. A very good example of a reconstruction from
a single point of view is Clavelin’s account in Chapters 4 and 5 of his book
The Natural Philosophy of Galileo;4 and in spite of the title of Clavelin’s
book, the point of view in these chapters is that of physics. There are two
problems with this approach to the study of structure. The first is that one is
likely to carry out the analysis from only one or a few of the possible points
of view, the most common being physics, philosophy, and astronomy. Of
course, this is not an insurmountable problem; there is no reason why the
literary, polemical, and pedagogic aspects should be neglected; yet it is a
historical fact that no one account of the Dialogue takes all the points of view
in turn. The second problem with this multifaceted approach is that the
book’s structure from the various points of view is different. The question
then arises whether these different structures are really in the book or
whether they are being imposed on it. I do not think that this problem is
insurmountable either, but I see no way of solving it other than dismissing it
as a problem.
This may motivate one to follow another approach. The only other way of
 FACT AND REASONING 29

finding structure, and thus avoiding confusion, and hence gaining under¬
standing, is to examine the interrelationships of the various parts and various
statements found in the book, and to keep as close as possible to the original
wording in one’s reformulation. In a sense this, too, is a point of view, namely
the internal point of view; that is, the point of view of the book itself. This is
not the point of view of the title of the book, or of the author’s preface or
introduction, or of the author’s stated intention as found in other documents;
it is not even the point of view of the author’s table of contents, if there is
one, though it can be very helpful. It is the point of view of the whole book
in its entirety, and of the interrelationships of all its parts. The following
analysis of Galileo’s Dialogue is offered in part as an attempt to delineate its
internal structure.
The logical structure of the Dialogue is that of an argument designed to
show that the earth moves. The critiques of the various geostatic arguments
can be easily integrated into such an argument as follows. Those critiques are
designed to show that there are no sound reasons for thinking it false that the
earth moves, or ontologically expressed, that there is no real evidence against
the earth’s motion. The reconstructed argument would then start as follows:
We may conclude that the earth moves since there is no real evidence or sound
reasons against its motion while there is considerable evidence and various
reasons in its favor; the former is true because all arguments against the earth’s
motion are incorrect; the latter is true because there are sound arguments for
the earth’s motion. Then would come all the details of the arguments for and
against the earth’s motion, in such a way that virtually every main topic men¬
tioned in the book would be integrated into this main argument.
However it will be almost impossible to appreciate this integration, or to
understand the details of the argument, or to give a statement of it, or to
check its accuracy without some fairly precise way of referring to the various
portions of the book. What is needed is an outline, something that historians
and scholars also need, for their own reasons, as mentioned above. So, 1 have
constructed the outline given below with page references to the standard
scholarly edition by Antonio Favaro,5 to Stillman Drake’s English transla¬
tion,6 and to Pietro Pagnini’s edition,7 which is the most valuable and easily
available one in Italian. The various outline subdivisions are partly grounded
on the book’s literary structure and so will remain unaffected by possible
changes in the outline headings; good examples of this are IIC and IIIA. The
subdivisions are also grounded on the subject matter or topical unity of a
given section, and the various headings are usually descriptions of this subject
matter; a good example of this is IC1, 2, 3, and 4. Moreover, both the headings
30 CHAPTER 2

and the subdivisions are partly grounded on the logical unity, clarity, or
beauty of the various paragraphs making up what I call below the Analytical
Summary. However, it is primarily the analytical summaries that originate
from the outline subdivisions and descriptions. It should be noted that the
outline headings are in every instance the grammatical subjects of the first
(or only) sentence in the corresponding paragraph of the Summary. The
Analytical Summary constitutes my reconstruction of the main steps of the
argument in the Dialogue. It thus defines what may be called the logical
structure of this work. Of course, many paragraphs of the summary could
themselves be expanded in such a way as to integrate (and thus number)
almost literally every proposition in the book.
My Analytical Summary is thus open-ended in a downward direction, so
to speak. That is, there are many subarguments (long, complex, important,
and interesting arguments) of which I have stated only the conclusion in
order to use it as a premise of the subarguments that I did state. It is clear,
however, that this open-endedness does not make the summary incomplete
but is a necessity because a summary should be a summary and not a com¬
plete reconstruction. An Analytical Reconstruction of the Dialogue would
include the whole iceberg of which my Analytical Summary is the tip. Since
it is obvious that the various paragraphs of my Analytical Summary vary in
logical complexity (some even being mere propositions, not arguments)
I should mention that I have followed rather rigorously a criterion of compre¬
hensibility in deciding how much to include. That is, I have left out the
reasoning and stated merely the conclusion in those sections where the
reasoning is either too complicated or too controversial or relatively easy.
Examples of excessive complexity are IA3,IB3,IIB6; examples of controversy
are IA3, IIB2c; examples of relative ease are IB2, IC1, IVB3. In some cases
(e.g., IIB2a, IIB3e), I have given reasoning which is relatively complex
because it is particularly beautiful and it is in part this logical beauty that
justifies making that portion of the text into a distinct outline subdivision.
Besides being thus open-ended in a downward direction my Analytical
Summary is open-ended in an upwards direction, namely insofar as it leaves
undetermined other steps in the over-all argument which are not too explicit
in the text and which would serve to integrate into it those sections which
are not as obviously relevant as the examination of the objections and the
statement of the favorable arguments. The sections I am referring to primarily
are IIIA and I. In summarizing IIIA1, IIIA2, and IIIA3 as I have in IIIA, I
have already taken a step toward integrating into the main argument this
section of the Dialogue which at first looks like a digression. However, it
 FACT AND REASONING 31

would have been inaccurate as a summary to go too far in that direction; but
as part of what might be called the latent structure we could add the follow¬
ing intended consequence: Hence it is not improper to consider astronomical
evidence nonquantitatively (as it is done in IIIB1, IIIB2, and IIIB3). The
relevance of the First Day (I) is more or less obvious. It is an examination of
the main conceptual objections to the earth’s motion, and this can be made
clear by the following argument: The earth’s motion is not a conceptual
impossibility because neither the objection from natural motion nor the
objection from the earth-heaven distinction is sound. The objection from
natural motion is that the natural motion of the earth parts is straight toward
the center of the universe (and hence the whole earth stands still therein); this
objection is groundless because the empirical argument that the earth parts
move toward the center of the universe is circular (IA4) and because any
conceptual justification of the same premise would involve all the problems
besetting the concept of natural straight motion (IA2, IA3). The objection
from the earth-heaven distinction is that the earth’s annual motion would
involve the conceptual absurdity of placing the earth in (the third) heaven;
this objection is groundless because the earth-heaven distinction is untenable
and false (IB, IC).
Finally, in my Analytical Summary I have some semi-technical terminology
which it will be good to explain briefly. The objection from A (where A is a
noun phrase) refers to the argument fromyl against B (where it is contextually
obvious what B is). The argument from A refers to the argument whose main
premise is some proposition p(A) constructed from A (in a contextually
obvious manner) and whose conclusion is some proposition p(B) whose
identity is contextually obvious and unproblematic. For example, in the
argument from the motion of sunspots: A is the motion of sunspots; B is
the earth’s annual motion; p(A) is the proposition that sunspots move as
described in (a), (b), and (c) of IIIB3 in the Summary; and p(B) is the pro¬
position that the earth has the annual motion. The argument from A is false
means that the proposition p(A), constructed from A in a contextually
obvious and unproblematic way, is false. The argument from A is groundless
means that the proposition p(A) is not in the context supported by any
sound argument. The argument from A is invalid means that, whether or not
p(A) is true, it does not support the conclusion of the argument. The argu¬
ment from A is unsound means that the argument fromyl is false, groundless,
invalid, and/or subject to some other problem.
32 CHAPTER 2

OUTLINE

(Favaro) (Drake) (Pagnini)

First Day : The Unity of the World 33 9 93

A. Natural motions 33 9 93
1. The perfection of the world 33 9 93
2. Aristotle’s classification of motions 38 14 101
3. Straight and circular motions 43 19 108
4. The evidence from the senses 57 32 129
B. The terrestrial-celestial dichotomy 62 38 138
1. The argument from contrariety 62 38 138
2. The a posteriori justification 71 47 151
3. The teleological argument 83 58 169
C. The moon and the earth 87 62 175
1. Similarities 87 62 175
2. The roughness of the lunar surface 95 71 189
3. The reflecting power of the earth 112 87 213
4. Differences 124 98 229

Second Day: The Earth’s Diurnal Motion 132 106 255

A. The problem 132 106 255
1. Aristotle’s authority 132 106 255
2. The arguments in favor 139 114 266
3. The objections 150 124 283
B. Examination of the classical objections 159 133 296
1. Aristotle’s first two arguments 159 133 296
2. Falling bodies 164 138 303
a. The objection from vertical fall 164 138 303
b. The ship analogy argument 169 143 310
c. Conservation and composition of motion 175 149 318
3. Projectiles 193 167 345
a. East-west gunshots 193 167 345
b. The relativity of motion 197 171 350
c. Vertical gunshots 200 174 354
d. North-south gunshots 203 178 359
e. Point-blank gunshots 205 180 362
4. The flight of birds 209 183 367
5. A crucial experiment 212 186 372
6. The extruding power of whirling 214 188 374
C. Examination of contemporary authors 244 218 417
1. The time of fall from the moon 244 218 417
2. The objection from the inexplicability of
the earth’s motion 260 233 439
3. The objection from the deception of the senses 272 247 455
4. The objection from the impossibility of
multiple natural motions 281 256 468
 FACT AND REASONING 33

(Favaro) (Drake) (Pagnini)

5. The similarity of motions of similar

substances 289 264 480
6. The objection that motion causes tiring 293 269 487

III. Third Day: The Earth’s Annual Motion 299 276 5

A. The 1572 nova 299 276 5
1. Preliminary discussions 299 276 5
2. The evidence from parallax difference 303 280 11
3. The poiar and stellar distances 337 309 54
B. The favorable arguments 346 318 66
1. The heliocentrism of planetary motions 346 318 66
2. Retrograde planetary motion 368 340 100
3. The motion of sunspots 372 345 106
C. The objections 383 356 122
1. Biblical passages 383 356 122
2. The stellar dimensions 385 358 124
3. Tycho’s objection 399 372 146
4. The celestial polar elevation 400 373 148
5. The changes in stellar elevations 403 376 151
6. The annual constancy in stellar appearances 404 377 154
7. The sun’s apparent motion 416 389 172
8. The properties of loadstones 423 397 182

IV. Fourth Day: The Cause of the Tides 442 416 223
A. Previous theories 442 416 223
B. The geokinetic theory 449 423 235
1. The primary cause 449 423 235
2. The fluid properties of water 454 428 241
3. The basic tidal effects 457 431 247
C. The behavior of winds 462 436 254
D. The monthly and annual periods 470 444 266

analytical summary

I. First Day: The Unity of the World

A. Natural motions are the same for all bodies.
1. The perfection of the world is better grounded on its being the
work of God rather than on its three-dimensionality.
2. Aristotle’s classification of motions into straight and circular,
simple and mixed, and natural and violent is untenable because
(a) his equation of natural circular motion with motion around
the center, and of natural straight motion with motion toward
or away from the center, is conceptually unjustified (and hence
34 CHAPTER 2

prejudicial); (b) his idea of mixed motion is incoherent; and (c)

his distinction between simple and mixed motion is viciously
circular.
3. Straight and circular motions are not two distinct instances of
(simple) natural motion but rather two different stages of natural
motion: straight motion can be acquired naturally but cannot
naturally continue forever, whereas circular motion can naturally
continue forever but cannot be acquired naturally without
straight motion.
4. The evidence from the senses, namely the up and down motion of
terrestrial bodies, constitutes a methodologically significant but
ultimately unsound objection to the idea of natural circular
motion; it is methodologically significant because the idea seems
to conflict with that evidence (and hence to violate one of the
most fundamental principles of philosophizing); and it is
ultimately unsound because that up and down motion is likely to
be either not straight or not natural or not peculiar to the earth.
It may not be straight because the fact that it appears to be
straight does not imply that it really is straight (as it will be
shown later). It may not be natural because it has not been shown
to be toward and away from the center of the universe, and it is
more likely to be primarily toward and away from the center of
the earth. And it may not be peculiar to the earth because that
motion is likely to exist on each of the celestial bodies.
B. The terrestrial-celestial dichotomy is untenable and false.
1. The argument from contrariety is that celestial and terrestrial
bodies are very different because change derives only from con¬
trariety, contrariety exists only among terrestrial bodies, and
hence change exists only among terrestrial bodies. It uses an
assumption (namely, the connection between change and con¬
trariety) which is more questionable than its main intended con¬
sequence (namely, the motionlessness of the earth). It may be
self-contradictory because it implies that celestial bodies are
changeable as well. It is groundless to the extent that the con¬
trariety of rarity and density exists among celestial bodies and
insofar as the contrariety of straight-up and straight-down does
not exist among terrestrial bodies only. And it is ambiguous
because the various mentioned ‘bodies’ sometimes refer to whole
bodies, sometimes to parts of bodies.
 FACT AND REASONING 35

2. The a posteriori justification of the unchangeability of the sky is

that no celestial changes have ever been observed, and it is both
invalid and factually false.
3. The teleological argument for the unchangeability of the sky is
that celestial changes would be superfluous and useless, and it is
unsound.
C. The moon and the earth do not differ in their nature.
1. Similarities between the moon and the earth include: shape, dark¬
ness and opacity, solidity, unevenness of apparent brightness,
phases, reciprocal illumination, and reciprocal eclipsing.
2. The roughness of the lunar surface may be justified by the fact
that it is visible at all and by its mountainous appearance through
the telescope.
3. The reflecting power of the earth may be justified by the fact
that during daylight both terrestrial objects and the moon appear
equally bright and that during the night the moon has a secondary
light (whose cause can only be the light reflected by certain parts
of the earth).
4. Differences between the moon and the earth do exist because the
moon has (a) no water, (b) a night-and-day period of one month,
(c) no significant seasons, and (d) no rain, and hence no life
similar to ours. Other unimaginable differences are bound to exist
because the human mind cannot pretend to be a measure of what
can occur in nature.

I. Second Day: The Earth’s Diurnal Motion

A. The problem with the idea of the earth’s motion is that it goes
counter to Aristotle’s authority, that the arguments favoring it
though plausible are indirect and only probable, and that there are
apparently insuperable objections to it.
1. Aristotle’s authority deserves respect but is abused and harmed
by his followers because they accept blindly and construe pre¬
judicially his words; what is needed in the philosophical search
for truth is not an authority but rather independent-mindedness.
2. The arguments in favor of the earth’s diurnal motion are plausible
but merely probable because they derive from the problems
besetting the contrary view, namely that a celestial diurnal
motion would (a) violate the principle of simplicity, (b) imply
that each planet has two contrary circular motions, (c) violate
36 CHAPTER 2

the law of periods of revolution, (d) imply that the fixed stars
have incongrously unequal orbits and velocities, (e) imply that
each fixed star keeps on changing its orbit and velocity, (f) make
it inexplicable why the circular motion transferred from one
celestial sphere to the one below it is not transferred to the earth,
and (g) force the Aristotelians to postulate the existence of a
fictitious primum mobile.
3. The objections to the earth’s diurnal motion are numerous and
apparently conclusive, and hence great open-mindedness and
great rational-mindedness are required for their refutation.
B. Examination of the classical objections to the earth’s diurnal motion
shows that all the phenomena alleged as counterevidence would
happen the same way whether the earth is rotating or standing still.
1. Aristotle’s first two arguments (from violent motion and from the
doubleness of circular motions) are equivocations, the first
because the clause “the parts of the earth would also move
circularly” can mean either that those parts would move around
their own centers or else that they would move around the earth’s
center, the second because its conclusion could be a denial either
of the diurnal motion or of the annual motion.
2. Falling bodies provide no evidence against the motion of the
earth but rather provide the basis for a new concept of motion,
according to which motion has the properties of conservation and
composition.
a. The objection from vertical fall is either circular or invalid
because if it is stated in terms of actual vertical fall then it
depends on the assumption that apparent vertical fall implies
actual vertical fall (which is true if and only if the earth stands
still), and if it is stated in terms of apparent vertical fall then it
depends on the invalid argument that the nonoccurrence of
mixed motion on a moving ship implies its physical impos¬
sibility.
b. The ship analogy argument is false because the body falls at
the foot of the mast even when the ship is moving.
c. Conservation and composition of motion are two properties
suggested by the criticism of the objection from falling bodies;
they mean that motion is conserved if undisturbed and that it
may be mixed without its components interfering with each
other; and such a concept of motion can be further justified as
 FACT AND REASONING 37

follows: (1) it is the one required for adequately solving the

problem of the cause of projectile motion, (2) it has interesting
and novel consequences concerning the motion of projectiles,
(3) it can explain certain puzzling facts about projectiles, and
(4) it fits well with the idea of natural circular motion.
3. Projectiles provide no evidence against the motion of the earth
but rather illustrations of the new concept of motion with the
properties of conservation, composition, and relativity.
a. East-west gunshots provide no evidence against the earth’s
motion because on a rotating earth the range of gunshots in
one such direction would still be equal to the range in the
opposite direction.
b. The relativity of motion is the concept which shows that there
would be no denial of the senses if bodies were regarded to
move transversally in reality while they were seen to fall
vertically.
c. Vertical gunshots provide no evidence against the earth’s
motion because their horizontal motion on a rotating earth
would be conserved; rather they provide a clear illustration of
the composition of motion.
d. North-south gunshots provide no evidence against the earth’s
motion, but not because shooting on a rotating earth is
analogous to hunters’ shooting at birds, rather because on a
rotating earth the cannon ball would have the same eastward
speed as the target.
e. Point-blank gunshots provide no evidence against the earth’s
motion primarily because computation shows that the alleged
deviation would be imperceptibly small (and hence there is no
way of knowing that such gunshots are not in fact high toward
the east and low toward the west), but also because on a
rotating earth the gun as well as the target is rising (or falling)
at the same rate (and hence the cannon ball would have the
same motion, up or down with respect to the fixed tangent, as
the target), and because one could argue equally plausibly that
if the earth stood still such gunshots would then be high
toward the west and low toward the east.
4. The flight of birds is the basis of a distinct argument against the
motion of the earth because birds unlike projectiles have the
power of self-movement (and hence they could not follow the
38 CHAPTER 2

earth’s rotation naturally but would have to do so through their

own efforts). This argument is groundless because the air within
which birds fly would be following the earth’s motion (and hence
they would not have to do so by their own efforts).
5. A crucial experiment (to nullify all the evidence alleged against
the earth’s motion from falling bodies, projectiles, and the flight
of birds) could be made below decks on a ship by observing the
flight of flies and butterflies, the swimming of fish in an aquarium,
the dripping of water, the motion of smoke from incense, and the
effort required to jump or throw objects in different directions.
You would notice that everything would happen the same way
when the ship is moving uniformly as when it is standing still.
6. The extruding power of whirling provides no evidence against the
earth’s motion because (a) the argument as ordinarily stated
would prove at best only that the earth did not at one time start
rotating after having been at rest; (b) the extruding motion would
be tangential, and the downward tendency due to the weight is
always large enough to overcome the tangential tendency; and (c)
the extruding tendency depends not on the linear speed at tne
circumference, which is large, but rather on the angular speed,
which is small.
C. Examination of contemporary authors opposing the earth’s motion
shows that none of their objections has any force.
1. The time of fall from the moon to a rotating earth constitutes no
objection to the earth’s motion because the objection conflicts
with (a) the mathematical fact that the radius of a circle is only a
fraction of its circumference, (b) the law of squares which yields
a time of less than a day rather than six days, and (c) the double¬
distance rule which yields a much greater terminal speed.
2. The objection from the inexplicability of the earth’s motion is
invalid because our relative ignorance of a cause does not imply
the non-existence of the effect, and groundless because it cannot
be shown that the cause of the earth’s rotation is neither external
nor internal.
3. The objection from the deception of the senses is false (a)
because shared motion is imperceptible (and hence there is
nothing for our eyes to see about falling bodies besides their
downward motion); (b) because wind is air moving relative to us
(and hence there is no perpetual wind due to the earth’s rotation
 FACT AND REASONING 39

for us to feel); and (c) because experience with navigation shows

that we can only feel changes in motion (and hence the earth’s
rotation is not something susceptible of being felt). Moreover, the
objection is invalid because the fact that the senses are to some ex¬
tent deceived in this and other cases implies that one has to be care¬
ful about what the senses tell us rather than that they are useless.
4. The objection from the impossibility of multiple natural motions
in simple bodies is (a) groundless because the function of joints
in animals is not to allow multiple motions but rather to allow
some parts to be moved while others are not (and hence it is
unjustified to say that bodies without joints cannot have multiple
natural motions); (b) irrelevant because there is no way for the
earth to have joints which would enable it to have its three
multiple motions (and hence even if it did have joints, it could
not have the types of motion it has, and hence there is no point
in saying that bodies without joints cannot have multiple natural
motions); and (c) false because Jupiter’s satellites and the sun
have multiple motions but no joints.
5. The similarity of motions of similar substances does not constitute
a valid objection to the earth’s rotation because (a) dissimilar sub¬
stances like water and air need not have dissimilar motions com¬
pletely but only to the extent that their dissimilar natures can be
inferred from differences in motion or other behavior (and hence
if the earth rotates the common diurnal motion of water and air
would not conflict with their dissimilar natures); and (b) similar
are the earth and the planets on the one hand and the sun and the
fixed stars on the other, due to their darkness and luminosity
(and hence the rotation of the earth would not imply that similar
bodies - the fixed stars, planets, and the sun - were having
dissimilar motions).
6. The objection that motion causes tiring is false and invalid: false
because the cause of animal tiring is the use of parts to move the
whole, and because much animal motion is violent rather than
natural; invalid because even if motion caused tiring the earth
would not tire any more than the primum mobile or stellar sphere
do in the Peripatetic system.

III. Third Day: The Earth’s Annual Motion

A. The 1572 nova provides a good example of how unreliable quantita-
40 CHAPTER 2

tive astronomical data can be and of how careful and critical one
must be in drawing conclusions from them.
1. Preliminary discussions point out that though certain arguments
are fallacious they deserve discussion for reasons other than their
logical merit.
2. The evidence from parallax difference examined by Chiaramonti
does not imply that the 1572 nova was sublunary because his 12
computations yield distances differing by as much as a factor of
1500 and because those computations are a biased fraction of all
those possible from his own evidence. This evidence, if a con¬
clusion must be drawn, supports rather the superlunary location
of the nova because when all his data are taken into account some
corrections are necessary, and fewer are needed to harmonize the
data yielding a superlunary location than those yielding a
sublunary location.
3. The polar and stellar distances of the nova can be used to argue
that it is superlunary because the needed observations are very
simple and are not invalidated by, respectively, the effects of
refraction and the instrumental difficulties of using the sextant.
B. The favorable arguments for the earth’s annual motion are very
cogent.
1. The heliocentrism of planetary motions supports the earth’s
annual motion {because the earth is located between bodies that
go around the sun) and is supported by the planets’ pattern of
changes in their apparent size and shape because-, (a) each planet’s
distance from the earth varies greatly (and hence the earth is not
the center of their motion); (b) the outer planets are close to the
earth when they are in opposition, and they are distant and look
round when they are in conjunction; (c) Venus always stays close
to the sun and appears homed in shape when large and round
when small; and (d) Mercury stays even closer to the sun and is
brighter than Venus.
2. Retrograde planetary motion supports the earth’s annual motion
because it is best explained as resulting from the earth’s annual
motion.
3. The motion of sunspots is best explained in terms of the earth’s
annual motion because they appear to move across the solar disc
along paths which exhibit the following features: (a) they curve
upwards in the solar disc for half a year and downwards for the
 FACT AND REASONING 41

other half; (b) they also slant upwards in the solar disc for half
a year and downwards for the other half; and finally (c) both
the curvature and the slant are continuously changing in such a
way that the paths are straight twice a year when the slant is
greatest, and the slant is absent twice a year when the curvature
is greatest.
C. The objections to the earth’s annual motion are inconclusive though
they are instructive and cannot be dismissed.
1. Biblical passages cannot properly be used in hypothetical reason¬
ing about natural phenomena.
2. The stellar dimensions implied by the earth’s motion and by the
lack of annual parallax would not be absurdly great because
apparent stellar diameters have been universally overestimated by
a factor of about 30 (and hence stellar sizes have been universally
overestimated), and because the required stellar sizes and dis¬
tances, though very great, are not absurd; this is so because (a)
some Ptolemaic estimates of distances are of the same order of
magnitude, (b) size and distance are concepts such that all large
ones after a certain amount are unimaginable, (c) there is no
justification for saying that the space between Saturn and the
fixed stars is useless, and (d) size and distance are relative
concepts.
3. Tycho’s objection from stellar dimensions is groundless because
he assumes without investigation that stellar positions show no
annual change and because he is unclear about the exact changes
implied by the earth’s annual motion.
4. The celestial polar elevation would not show any annual change if
the earth had the annual motion because in that case the celestial
pole would be defined by the terrestrial one and the elevation of
the latter can change only by moving on the earth s surface, not
by moving the whole earth as in the annual motion; hence the
objection that the celestial pole shows no annual change in
elevation is invalid.
5. The changes in stellar elevations implied by the earth’s annual
motion would not be at all comparable to the changes resulting
from moving along the earth’s surface because the former motion
occurs on a plane surface while the latter occurs on a relatively
highly curved one; hence the objection that stellar elevations
show no great changes is invalid.
42 CHAPTER 2

6. The annual constancy in stellar appearances constitutes an invalid

objection because the changes implied by the earth’s motion are
complicated and no one has systematically tried to observe them,
and they are very small and no instruments are available for
detecting them (and hence the apparent constancy does not
imply that the changes do not exist).
7. The sun’s apparent motion provides no evidence against the
earth’s annual motion because (a) the sun’s apparent motion in
the order of the signs of the zodiac would be a simple conse¬
quence of the earth’s own motion in the same direction; (b) the
significant seasonal changes in solar elevation and length of nights
and days would be consequences of the inclination of the earth’s
axis and of its unchanging direction; and (c) there is no conflict
between the large differences in solar elevation and the small ones
in stellar elevations implied by the earth’s annual motion.
8. The properties of loadstones invalidate the objection that a
simple body like the earth cannot have three or four natural
motions because (a) loadstones have several natural motions; and
(b) they have the property of always pointing toward the same
place, and there is evidence that the earth is a loadstone (and
hence its axis is always parallel to itself, and the third motion
attributed to the earth by Copernicus does not exist).

IV. Fourth Day: The Cause of the Tides

A. Previous theories about the cause of the tides must all be rejected
because (1) differences in sea depth cannot produce and sustain the
motion of the water, (2) lunar attraction could not produce tides
in only certain parts of a given sea and not in others, (3) the water in
a tide has the same temperature and density as ordinary water (and
hence lunar heat is inoperative), (4) the miracle explanation is not to
be invoked unless one can find no other cause, and (5) the periodic
attraction and expulsion of water by the earth through undersea
caves could not produce tides in only certain parts of a given sea and
not in others.
B. The geokinetic theory explains the main features of the tides as
resulting from the interaction of two causes: the primary cause is the
combination of the earth’s diurnal and annual motions, the second¬
ary cause consists of the fluid properties of water.
1. The primary cause of the tides is the daily accelerations and
 FACT AND REASONING 43

retardations produced in every part of the earth as the diurnal

component is added or subtracted from the annual component of
the earth’s motion; this is so because water in a container can be
made to move like the tides by accelerating or retarding the
container.
2. The fluid properties of water which act as a secondary cause of
the tides are that (a) it tends to oscillate before reaching equilib¬
rium; (b) its oscillations take less, the smaller the length of the
basin; (c) the deeper the water, the shorter the period of oscilla¬
tion; (d) it moves vertically at the extremities and horizontally at
the middle of the basin; and (e) different parts of the same body
of water can move at different speeds simultaneously.
3. The basic tidal effects that can be explained as resulting from the
interaction of the primary and secondary causes are (a) the
absence of tides in lakes and small seas, (b) the six-hour tidal
interval in the Mediterranean and presumably different periods in
other seas, (c) the absence of tides in seas that are narrow in an
east-west direction, (d) that tides are greatest at the extremities
and least at the middle of a gulf, (e) the great currents through
certain straits, (f) the violent agitations and vortices in certain
straits, and (g) the unidirectional flow of currents through certain
straits.
C. The behavior of winds provides no evidence against the geokinetic
explanation of the tides because air unlike water does not have the
property of retaining acquired motion (and hence the earth’s motion
could cause the tides without causing a perpetual wind); and because
the turning of the lunar orb could not produce the prevailing west¬
ward winds that do exist, and these could not produce the back and
forth motion of the tides (and hence these winds are not the link
between the diurnal motion of the lunar orb and the tides).
D. The monthly and annual periods of the tides must be caused by
variations in the velocity changes that cause the diurnal period
(namely, variations in the speed of the earth’s annual or diurnal
motion or both); hence, the monthly period is caused by the varia¬
tions in the speed of the earth’s annual motion that occur monthly
as the earth-moon system goes around the sun in a circle whose
effective radius undergoes monthly variations as a result of the
changing relative positions of the earth, sun, and moon; hence also,
the annual period is caused by the variations in the effective speed of
44 CHAPTER 2

the earth’s unchanging diurnal motion resulting from the inclination

of the earth’s axis to the plane of its orbit.

The fact that it is possible to reconstruct the Dialogue in the way just
described suggests the following historical and philosophical theses. First, it
is a mistake to regard the book as primarily a defense of the whole Coper-
nican system and then to blame Galileo for neglecting the details of technical
astronomy; instead, the detailed examination of the whole book shows that
it is primarily a defense of Copernicanism only to the extent that the pro¬
position ‘the earth moves’ (with the diurnal and annual motions) is part of
Copernicanism. The book is rich enough in its content, as it is; there is no
good reason why Galileo should have discussed the details of technical
astronomy.
Second, it is unfair to fault Galileo for not explicitly discussing the
Tychonic system since from the point of view of the proposition ‘the earth
moves’ there is no difference between it and the Ptolemaic system. Tycho’s
arguments are discussed in IIIC insofar as they are relevant to Galileo’s
purpose.
Third, Galileo’s alleged commitment to ‘natural circular motion’, if not
taken out of context, is part of a critique of Aristotle’s concept of natural
motion (in IA) and of an elucidation of a new concept of motion (IIB2c).
When seen in this light, there is nothing very obviously wrong with it.
Fourth, it is a mistake to regard Galileo as inimical to or unappreciative of
logic.8 The possibility and accuracy of the above reconstruction makes
Galileo a logician-in-action or applied logician, that is a skillful practitioner
of logical analysis and explicitly formulated argumentation.
Fifth, if it should turn out, from another type of analysis of the Dialogue,
that the book has no unity or prevalent characteristic in terms of some other
epistemological-methodological idea or practice, then we would have to
conclude that Galileo is first and foremost a logician. At present I state this
thesis as a conditional, though my conjecture is that the book lacks any other
prevailing methodological characteristic; and this conjecture is quite consistent
with the claim that the book is full of philosophy and methodology. In fact,
I believe that the full extent of the wealth of its philosophical content has not
so far been appreciated. What we seem to have in the Dialogue is a philoso¬
phical goldmine, but one lacking a unifying theme, other than that of applied
logic exhibited in this chapter.
Finally, there is no conflict between this reconstruction of the Dialogue
and the rhetorical interpretation, popularized by Feyerabend,9 that the book
 FACT AND REASONING 45

is a piece of propaganda aimed at winning the argument and at persuading the

opponents at all costs, and exploiting their weaknesses. On the contrary, a
reconstruction of the internal structure of the Dialogue is logically prior to
the rhetorical interpretation since the formulation of the latter amounts to
claiming that certain arguments actually given in the book are fallacious or
deceptive in various ways. The rhetorical interpretation loses its relevance if
and to the extent that its reconstructions are inaccurate, or taken out of
context.

NOTES

1 G. de Santillana, The Crime of Galileo, p. 174.

2 A. Koyrb, Etudes Galiiennes, p. 212.
3 W .Shea, Galileo’s Intellectual Revolution, pp. 109-186.
4 M. Clavelin, The Natural Philosophy of Galileo, pp. 183-267.
5 G. Galilei, Opere, edited by Favaro, Volume 7.
6 G. Galilei, Dialogue Concerning the Two Chief World Systems, tr. S. Drake.
7 G. Galilei, Opere, edited by P. Pagnini, Vols. 2 and 3.
8 See, for example, Shea, Galileo's Intellectual Revolution, p. 88.
9 P. K. Feyerabend, 'Problem of Empiricism IF, pp. 275-353; idem, Against Method.
 CHAPTER 3

EMOTION, AESTHETICS, AND PERSUASION:

THE RHETORICAL FORCE OF GALILEO’S ARGUMENT

It was argued in a previous chapter that a book like Galileo’s Dialogue must
have a type of content that can best be appreciated after its scientific-logical
content has been analyzed. This is what may be called its rhetorical content,
in a restricted sense of ‘rhetoric’, according to which the term refers to actual
substance present in the book, and not to merely cosmetic verbal expressions
of desires and intentions. The substance actually present in the book from
this rhetorical point of view is a type of intellectual content, but one that
plays upon feelings and emotions, either directly and explicitly by verbal
expressions that have the desired emotive effect, or else indirectly and im¬
plicitly by emphatic identification with what is explicitly said or done.
Some of this type of rhetorical content derives from the book’s dramatic
structure. However, it is an oversimplification to follow the typical scholarly
opinion1 that Salviati is a Copernican spokesman; that Simplicio is an Arist¬
otelian spokesman; and that Sagredo is an intelligent layman, attentive
listener to both sides, curious and open-minded, and unwilling to use any¬
thing but his own judgment in making up his mind. It is supposed to be more
or less obvious that Simplicio is something of a simpleton, as his name
suggests; that Salviati is Galileo’s mouthpiece; and that Sagredo is a neutral,
objective, uncommitted judge.
Such an interpretation would not do justice to the complexity of the
book’s dramatic structure. The thesis that Salviati is Galileo’s only spokesman
is at least complicated and perhaps invalidated by the following: (1) there is
in the Dialogue a fourth character, referred to as ‘the (Lyncean) Academician’
or ‘our common friend’,2 who is obviously Galileo, so that if he is Galileo
then Salviati cannot be; (2) frequently it is Salviati who expounds Aristotelian
views (e.g., F38-9, F356-63), though the text sometimes indicates that
these are not his own beliefs, and that he is doing a job that belongs to
Simplicio (e.g., FI51-3).
The thesis about Sagredo’s objectivity is faced by the difficulties that he
always sides against the Aristotelian position; that he tends to exaggerate the
strength of the Copernican arguments (e.g., FI 32-3, F383, F485-6); that he
sometimes expresses his judgments somewhat prematurely (e.g., F98); and
that occasionally it is he who obviously expresses Galilean views, rather

46
 EMOTION, AESTHETICS, AND PERSUASION 47

than Salviati, whose statements are less correct than his (e.g., F202-3).
The thesis about Simplicio’s simple-mindedness is made problematic by
the fact that he often expresses flashes of great logical acuity (e.g., F379-80).
In view of these facts, it is best to regard all three interlocutors as being
partial spokesmen for Galileo. The important question is not which speaker
speaks for Galileo, but rather how the specific details of the dramatic
structure of the dialogue contribute to his main purpose, which is that of
justifying Copernicanism. Let us then examine all the passages that seem
relevant to the concept of rhetoric under investigation.
In the course of the critique of the earth/heaven distinction, Simplicio
makes an appeal to fear when he says that a manner of philosophizing that
does away with this distinction, tends to the subversion of natural philosophy
and of nature herself (F62, D37-8). It is interesting to point out that
Simplicio expresses this fear when he is unable to make an intellectual
response to Salviati’s argument that since it does not make sense to distinguish
between two kinds of natural motions (straight and circular), and since all
natural motions must be circular, it follows either that there is no difference
between heaven and earth from the point of view of perfection, or else that
circular natural motion has nothing to do with perfection. It is certainly quite
legitimate to express one’s emotional feelings about a certain state of affairs
in nature; if the resulting emotions are negative, then the responsible philoso¬
pher-scientist will not dismiss the matter. In general a fear can be allayed
partly by a reiteration that there is no (‘rational) reason for the fear, in an
attempt to let reason conquer one’s emotions, and partly by pointing out an
aspect of the situation which would elicit a positive emotional reaction, in an
attempt to fight one emotion with another. Galileo does both of these things,
with respect to natural philosophy, Salviati says that there is no reason for
the fear because natural philosophy can only benefit by such an attempt to
subvert, either by making new discoveries,or by strengthening and reinforcing
the old doctrines; with respect to nature herself, Salviati points out that in
abolishing the earth/heaven distinction the Copernicans are, so to speak,
placing the earth in the heavens and hence are giving it more nobility and
perfection. This last point obviously has the intent of eliciting some kind of
rejoicing, which might make one forget about the previously expressed fear.
In the discussion of the existence and significance of sunspots in the First
Day (F77, D56), Galileo does not miss the opportunity of describing the
situation with emotive language. Salviati calls sunspots ‘importunate and says
that they have come to render the heavens and especially Peripatetic philoso¬
phy ‘turbid’. Does such name-calling have any function? I think what is
 48 CHAPTER 3

happening here is that Galileo is appealing to our aesthetic-Unguistic sensibility

to score some points against the Aristotelians. Galileo is calling the Peripatetic
heavens dirty and the Peripatetic philosophy muddled. Given that from the
Peripatetic point of view sunspots are impurities, the temptation to say that,
from that point of view, the heavens are dirty, is irresistible. Since, as Galileo
argues, there is no way of removing the dirt as long as one remains Arist¬
otelian, the need for a change becomes more intense. Moreover, the linguistic
possibility of using the epithet (‘turbid’) with the figurative meaning of
‘muddled’ makes irresistible the inclination of saying that the Peripatetic
philosophy is muddled. Thus, though few would be likely to be swayed
merely by such name-calling, few could resist the temptation of so using
language at the expense of Aristotelianism.
In the First Day, at the end of a discussion on the role of authority in
natural philosophy, Salviati argues very cogently that it would be not only
more correct, but even more in the spirit of Aristotle to argue that heaven is
alterable because the senses so indicate, rather than to argue that heaven is
inalterable because so reasoned Aristotle (F80—81). Galileo is here advocating
a significant change in the practice of natural philosophy. The discussion ends
with a very long passage, full of images, in which Sagredo describes what it
would be like to do without the security of an authority. The fact that the
description is put in the mouth of Sagredo (who asserts that he is thereby
portraying Simplicio’s attitude), together with the length of the passage and
the wealth of its images, can only be interpreted as an attempt by Galileo to
show that he can sympathize with the predicament of a committed Arist¬
otelian. This is a very effective and quite legitimate rhetorical technique,
for the pain of unpleasant experiences tends to be softened by the proof
that other people understand us; this is especially true when the expression
of sympathy comes from the people who have something to do with caus¬
ing the pain, which they regard as a necessary evil. Though there are traces
of irony in Sagredo’s speech (F81, lines 21-22), the over-all tone is one of
sympathy.
A few pages afterwards, in the discussion of the teleological argument,
in opposition to the Peripatetic idea that change makes things imperfect,
Sagredo says that he thinks that those who value unchangeability so much,
do so out of a fear of death and a desire to live a long time, and that they
deserve to encounter a Medusa’s head which would turn them into statues,
so as to become more perfect than they are (F84, D59). This is a good
example of what some would nowadays call an ad hominem argument,
namely an argument which tries to refute an idea by discussing the (usually
 EMOTION, AESTHETICS, AND PERSUASION 49

unfavorable) motives of the person holding the idea, rather than by discussing
the issue on its own merits. Though many logic textbooks regard such argu¬
ments as fallacies,3 such a view is at best superficial.4 It would indeed be
wrong to regard such motivational analysis as a conclusive argument against
the truth of the idea, but this is not to say that the argument lacks any force
whatever. In fact, such arguments are especially effective when values are
involved, as in the present case. The issue here seems to be whether or not
change is a good thing. How does one discuss this on its own merits ? If
people approve unchangeability because they want to remain alive, that is
logically relevant to the issue because one can then question their approval
by questioning their motivating desire. Would it really be good to live
forever? Galileo seems to be casting doubts on this by suggesting that this
would cause overpopulation problems, the solution of which might require
an end to births, with the consequence that it would be doubtful whether
most of us who are now alive would have been born in the first place. More¬
over, in his reference to Medusa, Galileo seems to suggest that an immortal
life might be a kind of death, and thus a conceptual contradiction. When seen
in this light, Sagredo’s remarks, though not decisive, acquire some logical
value, in the broad sense of logical’ that includes persuasive force. If there is
a problem with Sagredo’s remarks, it is their irrelevance, at least contextually
speaking; for it emerges from the discussion that the issue is not one of
valuing or not valuing unchangeability, but rather whether changes in the
heavens would be superfluous and therefore are to be ruled out. However,
Sagredo’s argument against unchangeability is irrelevant only when viewed
from the point of view of the fully articulated teleological argument (i.e. that
there cannot be any changes in the heavens because change would make
celestial bodies imperfect; and this is so because changes in celestial bodies
would be superfluous; and, in turn, this is true because changes in celestial
bodies would be of no use or benefit to man);s for, from the point of view
of coming to discover that unchangeability may not be an intrinsic Peripatetic
value, Sagredo’s argument has the methodological function of facilitating this
discovery.
According to Peripatetic philosophy, some heavenly bodies, especially the
sun caused changes on the earth, while they remained themselves unchanged
and’ unchangeable. This offered Galileo the opportunity for a memorable
comparison: he says that this is like placing a marble statue next to a bride
and expecting children from such a union (F84-5, D60). This critique could
be interpreted as an argument from analogy, possessing a certain degree of
plausibility, depending on the strength of the analogy between the two
50 CHAPTER 3

unions. However, such an interpretation would hardly do justice to it, for I

believe the argument has a persuasive force much greater than that deriving
from the inherent analogical reasoning. The reason is that one of the things
being compared is not merely abstractly and objectively false, but emotionally
absurd as well; we might say that to expect children from the union is doubly
unnatural, that is, against the nature of (physical) things, and against human
nature in the sense of human psychology, human wishes, desires, and dreams.
In short, not only no children will in fact come from such a union, but no
children should come from it, in the sense that it would be much better if
they came in the usual way. To be able to argue on the basis of comparisons
that not only have some factual basis but which also strike a resonant chord
in human feelings involves more than being merely a logician, or merely a
poet; what is involved is the combination of the two which is what is done by
the good rhetorician (in the sense presently under discussion).
Another Peripatetic doctrine stated that the heavens consisted of a fifth
element, different from the four terrestrial ones, namely earth, water, air, and
fire. This fifth element was thought to pervade the whole celestial region,
where the heavenly bodies were merely its denser parts; it was also thought to
be impenetrably hard but to lack all other ordinary qualities such as visibility
and weight. This doctrine had become highly discredited in Galileo’s time,
primarily because the motion of comets seemed to make impossible the
existence of any hard, impenetrable, crystalline substance. Galileo ridicules
the doctrine by means of the following exchange:

SAGR. What excellent stuff, the sky, for anyone who could get hold of it for building a
palace! So hard, and yet so transparent!
SALV. Rather, what terrible stuff, being completely invisible because of its extreme
transparency. One could not move about the rooms, without grave danger of running
into the doorposts and breaking one’s head.
SAGR. There would be no such danger if, as some of the Peripatetics say, it is intangible;
it cannot even be touched, let along bumped into.
SALV. That would be no comfort, inasmuch as celestial material, though indeed it
cannot be touched (on account of lacking the tangible quality), may very well touch
elemental bodies; and by striking upon us it would injure us as much, and more, as it
would if we had run against it.
But let us forsake these palaces, or more appropriately these castles in the air, and
not hinder Simplicio. [D69; cf. F94]

Though this wit may be recognized as brilliant, one may question its logical
force as an argument against the impenetrable hardness of the fifth element.
I believe that the argument lacks any intrinsic logical force, but that it
acquires some from the inherent wit; what happens is that one acquires a
 EMOTION, AESTHETICS, AND PERSUASION 51

negative emotional attitude toward this substance because it lends itself to

such ridicule, and such an attitude then translates cognitively into the belief
that one ought not to have the concept of such a substance in one’s system,
and that it probably is an unhappy invention, and that it probably does not
exist.
At the beginning of the Second Day, Sagredo asserts that he finds the daily
rotation of the universe around a motionless earth as absurd as if someone
went on top of the cupola of a cathedral to have a view of a city and its
surroundings, and then he expected that the city be made to turn around him
so that he would not have to make the effort of moving his head (F414,
D115). Here we have another passage that could be interpreted as an argu¬
ment from analogy, with some degree of plausibility, but whose persuasive
force is increased beyond the purely logical strength of the analogical
reasoning. The increase derives from the emotive repulsiveness of the image of
the lazy, arrogant observer.
In the preliminary exposition of the geo static arguments Salviati is the
one who gives a statement of several ones based on gunshots. After Salviati
is finished Simplicio expresses his joy at the fact that the truth (that the earth
stands still) should be supportable by such ‘unconquerable’ arguments and by
such new evidence (as compared to that available to Aristotle). At this point
Sagredo interjects the exclamation, “What a shame that there were no
cannons in Aristotle’s time! With them he would indeed have battered down
ignorance ...” (D127; cf. F153). This is certainly a witty remark, and it
gives the reader great aesthetic pleasure. In this case, however, the wit does
not perform any logical-rhetorical function; the gunshot arguments are
neither strengthened nor weakened. A rhetorical service is performed by the
fact that the anti-Aristotelian Salviati shows such understanding of the
Aristotelian position as to be able eloquently to present new evidence in its
favor, for this inclines the reader to think more highly of Salviati and his
judgment, and indirectly and ultimately of Copernicanism. However, this is
a service performed by the dramatic features of the present dialogue, rather
than by Sagredo’s witticism, which must therefore be attributed a purely
literary-aesthetic value.
In the critique of the ship analogy argument (FI69-75), Salviati argues
that the rock falls at the foot of the mast regardless of whether or not the
ship is in motion. He feels that it is unnecessary to appeal to direct sense-
experience because he is sure the fact can be proved from other more easily
accessible observations; he thinks he can even convince Simplicio, who is sure
that the rock falls away from the foot of the mast when the ship is in motion.
52 CHAPTER 3

Salviati’s display of confidence reaches a climax when he boasts, “I am so

handy at picking people’s brains that I shall make you confess this in spite of
yourself’ (D145; cf. FI71). This arouses even Sagredo, who was about to
make a comment, but who now remarks, “the interest aroused in me by
hearing you threaten Simplicio with this sort of violence . . . has deprived me
of any other desire; I beg you to make good your boast” (D145; cf. F171). It
is obvious that the reader’s interest and curiosity are aroused as much as
Sagredo’s, hence the passage has literary-dramatic value. But does its worth
extend beyond that? It is true, of course, that displays of confidence are
standard rhetorical devices designed to insinuate logical force and thus
achieve persuasion, or at least increase assent. It is also true that such displays
sometimes have the desired effect, the operative assumption on the part of
the audience being that, because of the inherent risk of failure to make good
one’s boast, people do not normally claim more than they can prove. All the
same, it is also common knowledge that for some people, their boasting is
mostly an indication of their confident or over-confident nature. Since the
general question of the logical worth of displays of confidence is thus some¬
what debatable, let us see about the present case. It should be noted that this
question of the rhetorical worth of the present boast is independent of the
question to what extent Galileo’s ensuing argument justifies the boast, for
what we are looking for here is some alogical worth, above and beyond the
aesthetic, and pertaining to persuasion. I believe that by looking at the
dialogue just preceding Salviati’s present speech, we can see why there is a
rhetorical need for some such boast. In fact, to a naive, uncritical empiricist
like Simplicio, the question of what happens on the moving ship is a paradigm
example of one to be decided by the senses, and so he does not see any need
for further discussion of the matter and is about to give up the dialogue at
this point (F170, D145). The only way to prevent this, the only way to
prevent his Aristotelian readers from closing the book at this point, is for
Galileo to make a boast that would otherwise seem excessive. Whether or not
he proves his point to their satisfaction, he has at least held their attention
and convinced them that the matter is not so simple, that it does require
discussion.
In the discussion of the cause of projectile motion (F175-80, D149-54),
Galileo criticizes the Peripatetic idea that when air is in its own region it has
the disposition to receive and conserve an impulse (F178, D152). Among the
evidence Galileo presents against this idea, there is the following very quaint
experiment: “Let us go into that room and agitate the air as much as possible
with a towel, then stopping the cloth, have a little candle flame brought
 EMOTION, AESTHETICS, AND PERSUASION 53

immediately into the room, and you will see from [its] quite wandering . . .
that the air has been instantly restored to tranquillity” (D152; cf. FI78).
Here it is the familiarity and accessibility of the experience that gives the
experiment a special rhetorical force. Though we may agree that, logically,
the force of this counter-evidence is no greater than that of the ‘thousand
other’ experiments that Salviati claims could be made, from a psychological
point of view there is a great difference. The ability to choose appropriately
familiar examples is a valuable skill in the art of persuasion, of which purely
logical reasoning is a part, an important part to be sure, but only a part.
The discussion of conservation and composition of motion (F175-93,
D149-67), after the critique of the ship analogy argument, is an elaboration
of the concept of motion presupposed by Galileo’s critique of this and of the
other arguments from fading bodies. It is a relatively long passage, and lacks
the relatively more explicit structure provided by the pattern of argument
statement, argument analysis, and argument evaluation that characterizes so
much of the book. Consequently, this passage might strike the layman or the
Aristotelian not interested in the topic for its own sake as being relatively
dull, and he might very well omit it. However, in the middle of the discussion
we find a very interesting exchange between Simplicio and Sagredo, which
has the rhetorical function of preventing an onset of boredom. Sagredo has
just finished saying that the doctrine of conservation of motion allows him to
understand the puzzling fact that the hoops used by players are sometimes
seen to move faster over a later portion of their path than over a previous
one. He ends his speech with the question, “Now what would Simplicio say
to that?” (D157; cf. FI83). The latter replies, “I should say that in the first
place I have not observed such things; second, that I do not believe them; and
then, in the third place, if you should assure me of them and show me proofs
of them, that you would be a veritable demon” (Ibid.). To this Sagredo adds,
“One like Socrates’s, though; not one from hell” (Ibid). This is indeed a very
effective way of arousing our curiosity and anticipation, besides being, of
course, very clever from a literary point of view.
In the discussion of semicircular fall (FI88—93) Galileo refrains from
considering quantitatively the acceleration of falling bodies, giving as a reason
that it would not be relevant for present purposes, i.e., would be a “digression
within a digression” (F190, D164). He wants to make sure that the reader
does not get the impression that his reason is his acceptance of the Peripatetic
view of the relationship between science and philosophy, and so he gives a
very brief critique of this idea. This critique consists of a very eloquent but
arrogant, and consequently self-defeating statement by Simplicio, togethei
54 CHAPTER 3

with a remark by Sagredo that this is like putting a philosopher on a royal

throne, which is a rather repulsive image, and which therefore makes the
view implicitly undesirable. The passage deserves literal translation since the
two available ones miss some rhetorically significant phrases:

SIMP. Philosophers occupy themselves principally with universal; they discover defini¬
tions and the most common symptoms, then leaving to mathematicians certain subtleties
and trivia, which are rather curiosities. In fact, Aristotle was content with defining what
is motion in general, and with showing the principal attributes of local motion, namely
the differences among natural, violent, simple, compound, uniform, and accelerated; and
for accelerated motion he was content with explaining the cause of acceleration, leaving
then the investigation of the proportion of such acceleration and of other more specific
accidents to the mechanic or to some other inferior artist.
SAGR. All right, my dear Simplicio. But you, Salviati, descending occasionally from the
throne of Peripatetic majesty, have you ever played with the investigation of this propor¬
tion of the acceleration of the motion of falling bodies? [FI 89-90; cf. D189-90]6

I believe that, here, the persuasive force of the criticism derives from the kind
of visual image we get of the Peripatetic philosopher, namely of someone
removed from ordinary reality. I do not think the description is biased; the
emphasis is even on qualities which would otherwise be favorable. The prob¬
lem is with the inappropriateness of these qualities.
In the critique of the objections from gunshots there is a passage which
absorbs and captivates the reader like few others. It is one where Sagredo
imagines an artist painting a scene on a ship going from Venice to Alexandretta
and notes that the tip of his pen would describe a line which over its total
length would vary almost insignificantly from a smooth or straight fine
(FI97-200, D171-4). Yet from such an (almost) smooth line we get a
painting; the only part of the actual line that is effective is what from another
point of view are insignificant variations. Simplicio describes himself as
‘bewitched’ by the image (F199, D173), which indeed is bewitching. The
example is given as illustration and justification of the idea of the relativity of
motion, and the aesthetic magic works in the sense that it enhances the pos¬
sibility of the doctrine and of its application to the situation on a rotating
earth. To emphasize the rhetorical effect of the example, and its logical
import, Galileo has the speakers also mention a number of what he calls
“very insipid subtleties” (F199, cf. D173), allegedly elicited by Sagredo’s
description: these are the facts that when a ship is in motion the top of its
mast moves more than its bottom, and the fact that when a person walks his
head moves more than his feet, in both cases because of the earth’s spherical
shape.
 EMOTION, AESTHETICS, AND PERSUASION 55

The next rhetorically significant passage offers a contrast between the

Peripatetic approach to natural philosophy and the one Galileo is advocating.
In order to appreciate it properly, a literal translation is needed:

SALV. ... I am beginning to realize that you [Simplicio] have so far been of the herd of
those who, to understand how such things occur and to learn the details of natural
phenomena, do not go onto boats or around crossbows and cannons, but rather retire
into their study to shuffle indices and outlines to find out whether Aristotle has said
anything about them, and who, having ascertained the true meaning of the text, neither
desire anything else nor believe that anything else could be known.
SAGR. Great happiness, and much to be envied; for if knowledge is naturally desired by
all and if it is the same to be as to claim to be, they enjoy a very great benefit and can
persuade themselves to know and understand everything, by contrast with those who,
knowing not to know what they don’t know, and consequently saying that they know
only a very small part of what there is to know, kill themselves with lack of sleep and
with studies, and waste away around experiments and observations. [F211, my trans¬
lation; cf. D185]

To be sure, the contrast is exaggerated; however, since it is not a fabrication

and does have a grain of truth, it does have the desired effect of arousing our
sympathy for the Galilean type of natural philosopher. This emotion is
indeed sympathy, for, as Sagredo says, as far as envy goes, it is the Peripatetic
philosopher who is to be envied. However, such envy can be felt only at a
superficial level, for at a deeper level one feels repugnance for the ‘herd’ of
people who shuffle papers in order to leam about the real world. Here this
criticism consists in the mere description, and the strength of the criticism in
the intensity of the contrast. How valid is the criticism? Obviously, it applies
to some Peripatetics but not to others; Galileo thought that it certainly did
not apply to Aristotle himself (F75, FI36—7). The caricature does not lose
its rhetorical force, however, for once created it acquires a life of its own.
In the critique of the centrifugal force objection (F214-44, D188-218),
there is a Socratic discussion between Salviati and Simplicio attempting to
show that, though terrestrial rotation would give objects a centrifugal
tangential impetus, their weight however small would be sufficient to keep
them on the surface. Just before this Socratic discussion is concluded,
Simplicio gives an interpretation of the situation according to which a feather
would have to fly away from the surface, and “since this does not happen, I
say that the earth does not move” (F223, D196). Salviati replies, Oh,
Simplicio, you yourself rise up so fast that I begin to fear more for you than
for the feather. Relax a little, and listen (D197, cf. F223). This is a very
pleasing and clever aesthetic image, but I do not think that it has any
rhetorical force. It does have, to be sure, what might be called a rhetorical
56 CHAPTER 3

import, insofar as it wants to convey the idea that the Aristotelians often
draw hasty generalizations. However, the incident is too parenthetical, and the
image of Simplicio flying away faster than a feather on a rotating earth
according to the Aristotelian conception, is short-lived. What this means is
that the remark has only aesthetic-literary value, but no logical-rhetorical
worth; but to say the latter is not to say that it has logical-rhetorical demerit.
What I am saying is that in this case the value exists only in the aesthetic
domain; this would entail rhetorical demerit only if it could be proved that
everything in the text indicates that at precisely this place, Galileo is
attempting to score a rhetorical point. Since there is no such evidence, the
error would be committed by the critic, if he were to see rhetorical failure.
In short, nonrhetoric is not bad rhetoric.
One difficulty that some people found with attributing the daily rotation
to the earth was the problem of how the diurnal motion could inhere in both
living and inanimate beings. Besides giving a more or less technical answer to
the problem through Salviati (F264-5, D238—9), Galileo has Sagredo say
that the people who feel this difficulty “must believe that if a dead cat falls
out of a window, a live one cannot possibly fall too, since it is not a proper
thing for a corpse to share in qualities that are suitable for the living” (D239;
cf. F265). This may be interpreted as a reductio ad absurdum argument; the
rhetorical component here is in the absurdity being referred to, which is not,
or not merely, an abstract impossibility, but primarily an emotively ridiculous
situation. For those who cannot follow the technical refutation, or for those
who want to add ridicule to refutation, the present image is at their service.
Next, we should not neglect a passage at the beginning of the discussion of
Chiaramonti’s book on the new stars; the wit and sarcasm are memorable,
while the fact that the passage seems to lack a rhetorical function may serve
as a reminder that we should not confuse lack of rhetoric with rhetorical
failure. Galileo’s remark is a comment to Chiaramonti’s attempt to show that
the so-called new stars were sublunary. Chiaramonti’s argument was based on
an analysis of parallax data, the very same data that had convinced most
astronomers that the new stars were superlunary. Galileo says that “since this
author brings about such an undertaking, namely to remove the new stars
from the heavens into inside the elemental sphere, he deserves to be greatly
exalted and himself transferred among the stars, or at least that his name be
made eternally famous among them” (F272, my translation; cf. D248).
A few pages afterwards we do have what must be regarded as a rhetorical
failure, a small one to be sure, but an error nonetheless. The passage begins
with Sagredo remarking that the previous discussion had somewhat settled his
 EMOTION, AESTHETICS, AND PERSUASION 57

stomach which had been upset by the fish and snails he had eaten (F275,
D250). This is a humorous remark from an aesthetic point of view; and from
a literary point of view it is a clever and successful device to introduce a new
topic into the previous discussion which was a critique of the geostatic objec¬
tion from the deception of the senses, as found in Chiaramonti’s book; in
fact, Sagredo says that besides settling his stomach the previous discussion has
made him think of correcting the erroneous view that the telescope cannot be
used as effectively at the top of the mast of a ship as on the deck. The correc¬
tion proceeds by Sagredo arguing that the only thing that can affect the view
is angular displacements of the telescope, which are the same on the mast as
on the deck, not parallel displacements which are too small even on the mast.
Throughout the discussion Salviati plays the role of someone who finds the
erroneous view initially very plausible, and who is receiving enlightenment by
Sagredo in a semi-Socratic discussion; Salviati ends the discussion by saying
that Sagredo’s correction illustrates how very careful one must be before
affirming or denying a proposition, and that just as this erroneous belief is
excusable because of its prima facie plausibility, so is the belief of those who
cannot understand how the apparently vertical path of a freely falling body
might not really be straight (F278, D253). Thus, because of Salviati’s role and
his conclusion, the rhetorical intention and pretension of this passage is to
soften Galileo’s criticism of the objection from the deception of the senses
and to make us sympathize with the author presenting the objection. How¬
ever, most of the evidence from the rest of the passage criticizing this objection
(F272-81, D247-56) points in the opposite direction, namely that Chiara-
monti is not only wrong, but foolish in putting forth this objection. The most
that can be said in favor of Galileo’s rhetorical attempt here is that it
indicates ambivalence, since there are traces of a respectful attitude toward
Chiaramonti; for example, at the beginning of the passage Simplicio is made
to introduce him as “a consummate philosopher and also a great mathe¬
matician” (F272, my translation; cf. D247), and at the end Sagredo is made
to describe him as a “philosopher, who seems to me a cut above most of the
followers of these doctrines” (D256; cf. F281). However, the ambivalence is
not sustained either, and in the rest of the discussion of Chiaramonti
(F281-346, D256-318) the criticism becomes more intense, and the con¬
tempt more apparent. Therefore, in any case, Sagredo s correction must be
regarded as a rhetorically ineffective digression.
The next rhetorically striking passage occurs toward the end of the Second
Day, at the end of the critique of the objection that similar substances must
have similar motions. The geocentric universe, together with the idea that the
58 CHAPTER 3

heavens are noble, perfect, and incorruptible, whereas the earth is base and
full of imperfections and impurities, lends itself to the criticism that this is
like having a lazaretto in the center of town rather than away from it in the
country (F292—3, D268). The effect of this comparison is very strong; it
creates in us a very powerful feeling of the inappropriateness of a geocentric
arrangement, probably powerful enough to overcome the positive feeling of
its being cozy. Realizing that the geocentric universe was not only support¬
able by evidence and arguments but also emotionally satisfying, Galileo is
here calling attention to a feature that ought to make it emotionally dissatis¬
fying.
At the end of the Second Day, the objection is stated and criticized that a
moving earth could not move forever without tiring, since animals become
tired even though they move by an internal principle (F293—9, D269—76).
Besides answering this objection with serious criticism, Galileo feels that it
also deserves rhetorical ridicule, for it seems to Sagredo “that there are also
animals which refresh themselves from weariness by rolling on the ground,
and that hence there is no need to fear that the earth will tire; it may even be
reasonably said that it enjoys a perpetual and tranquil repose by keeping itself
in an eternal rolling about” (D271; cf. F295). This is ad hominem reasoning
(in the 17th century sense of the term),7 where the Aristotelian principles
used for the formulation of the objection are used to answer it and show that
it is not impossible for the earth to move. However, the total rhetorical force
of the reply is much greater than its logical strength; it is the possibility of
creating such wit that adds the extra strength.
At the beginning of the Third Day we find an explicit statement that,
besides trying to enlighten the reader, Galileo wants to arouse his feelings.
The occasion is a discussion of the location of the 1572 nova and a critique of
Chiaramonti’s attempt to show that it was sublunary. Salviati tells Simplicio
that, once he has understood Chiaramonti’s error, he can feel indignation if
he notices that Chiaramonti, “covering his cunning with the veil of your
naivete and that of other mere philosophers, tries to insinuate himself into
your good graces by gratifying your ear and puffing up your ambition, pre¬
tending to have convicted and silenced these trifling astronomers who wanted
to assail the ineradicable inalterability of the Peripatetic heavens, and what is
more, to have struck them dumb and overpowered them with their own
weapons” (D285; cf. F309). Galileo is charging Chiaramonti not only with a
logical-scientific error but also with a morally reprehensible deception; he is
not saying that this deception is intentional, but only that it is real. I believe
that the rhetorical effect of this charge is to arouse the curiosity of Simplicio
 EMOTION, AESTHETICS, AND PERSUASION 59

and of the Peripatetic readers of the book, and to motivate them to under¬
take the effort to understand the time-consuming and elaborate calculations
and analyses of quantitative data required to see the error in Chiaramonti’s
argument. What we have here, then, is an instance of the technique of making
a strong claim in order to arouse interest and curiosity.
A few pages later (F313, D289), Salviati calls Simplicio an Archimedes
after Simplicio has shown his understanding of a geometrical point in the
theory of parallax; Salviati says that he had almost despaired of being able to
explain it to a pure Peripatetic philosopher. These remarks obviously have the
function of encouraging ordinary Aristotelians that the computations in
Galileo’s critique are not beyond their intellectual powers, and that they will
be able to follow them if they only put their minds to it.
After Galileo has criticized Chiaramonti’s examination of the evidence
from parallax difference, he goes on to criticize how Chiaramonti explains
away the evidence from polar and stellar distances, and the occasion comes
for the creation of a poetic image:

When a person finds no defense to be of any avail against his mistake and produces a
frivolous excuse, people say that he is reaching for ropes from the sky. This author
grasps not at ropes, but at spiderwebs from the sky, as you will plainly see upon
examining these two points just mentioned. [D311;cf. F339]

Though the value of this passage is primarily aesthetic, from the rhetorical
point of view it functions to hold the reader’s attention.
The critique ends with a series of three images:

SAGR. This is as if 1 were watching some unfortunate farmer who, after having all his
expected harvest beaten down and destroyed by a tempest, goes about with pallid and
downcast face, gathering up such poor gleanings as would not serve to feed a chicken for
one day.
SALV. Truly, it was with too scant a store of ammunition that this author rose up
against the assailers of the sky’s inalterability, and it is with chains too fragile that he has
attempted to pull the new star down from Cassiopeia in the highest heavens to these base
and elemental regions. [D318; cf. F346]

From a rhetorical point of view Sagredo’s description might arouse sympathy

and therefore might be taken as an indication of Galileo’s ambivalent attitude
toward Chiaramonti that we have previously mentioned. However, Salviati’s
description attributes a somewhat bellicose attitude to Chiaramonti, and thus
destroys that feeling of sympathy. So, rhetorically speaking, the two descrip¬
tions undo each other. This time, the aesthetic value is also dubious; for,
though each description has value by itself, and though Salviati s two images
60 CHAPTER 3

go well together, it is difficult to picture Chiaramonti, or anyone, both as the

‘unfortunate farmer’ and as the soldier of heavenly inalterability. The
aesthetic worth might be rescued by emphasizing the dramatic situation, that
is, by taking Sagredo’s image as a description of his own attitude toward
Chiaramonti, and the other as being Salviati’s own state of mind. This would
be confirmed by the fact that Salviati has earlier had a similar image (D311,
F339), namely of Chiaramonti grasping for spiderwebs from the sky, and the
fact that Sagredo had earlier expressed respect for Chiaramonti, in describing
him as a cut above most other Peripatetic philosophers (D256, F281).
In the middle of the discussion of the heliocentrism of planetary motions,
after the basic Copernican arrangement has been sketched, we find Salviati
dismissing certain arguments as being for people whose definition (‘rational
animals’) contains only the genus (‘animals’) but lacks the species (‘rational’)
[F355] .8 This is great wit, though incredibly subtle, and an unprosaic way of
saying what Galileo himself says in a marginal note, namely that these are
“utterly childish reasons sufficing] to keep imbeciles believing in the fixity
of the earth” (D327; cf. F355). The reasons being referred to are: (1) that it
does not happen that one is in Constantinople at lunch time and in Japan at
supper time; and (2) that the earth is so heavy that it cannot climb up over
the sun in order later to roll down again (Ibid.). To mention and then
summarily dismiss an argument is of course a rhetorical technique to express
a negative judgment; however, here Galileo seems to be overdoing it. This is
especially unfortunate because Salviati’s main point in this context (F354—6,
D327—8) is to emphasize the great initial cogency of the objections to the
earth’s annual motion. It seems that Galileo here was carried away by the
literary brilliance of his witticism and became temporarily forgetful of
rhetorical effects. In fact, he later, after the book had been published,
decided to insert at this point a passage which discusses some of these pre¬
viously dismissed arguments (F356-62, D328-33), I believe in order to
soften the literary excess and rhetorical (counter) effect of his witticism.
Galileo ends his discussion of the heliocentrism of planetary motions with
the remark, “Now you see how admirably these three chords, which at first
seemed so dissonant, accord to the Copernican system” (F368, my transla¬
tion; cf. D340). The three ‘chords’ are the apparent sizes of Mars and Venus,
the apparent shapes of Venus, and the geocentrism of the moon’s orbit. At
first these seemed dissonant with Copernicanism in the sense that with the
naked eye no significant variation is observable of planetary sizes or of the
shape of Venus, as it is required by the Copernican system; moreover, in
being around the earth, the moon’s proper motion seemed to go against the
 EMOTION, AESTHETICS, AND PERSUASION 61

general trend alleged for the other planets. The accord is the result of the
telescopic observations that Venus and Mars show the required periodic
change in apparent size, that Venus shows phases, and that Jupiter’s satellites
have a proper motion analogous to the moon. Though Galileo’s eloquent
musical metaphor has great aesthetic value, it is doubtful that it has any
rhetorical force; it may make us feel somewhat reluctant to disturb the
harmony it creates, but since such reluctance has a purely aesthetic origin, it
should be possible to overcome it.
At the beginning of the critique of the objection from the annual con¬
stancy in the stellar appearances, Sagredo says to Salviati, “This is a knot
which has never passed through my comb, and if you untie it for me I shall
consider you greater than Alexander” (D378; cf. F405). This reference to the
Gordian knot is a rhetorically effective way of emphasizing the importance
of the objection and of its resolution. It is an unprosaic way of making the
same point that Galileo makes in a marginal note when he describes it as the
“chief objection against Copernicus” (D378; cf. F405). We also have here a
suggestion that the Galilean answer to the objection may be more analogous
to a cutting than to an untying of the knot, that is, that it constitutes not an
actual solution but rather an effective response. In fact, Galileo’s answer
consists of an admission that no change can be observed and a rather detailed
outline of a program of theoretical and experimental research that would be
likely to lead to the observation of some stellar variations.
Near the beginning of the discussion of the properties of loadstones,
Galileo expresses his criticism of the Peripatetic dogmatic attitude toward
new ideas, such as those contained in Gilbert’s De Magnete. He also ridicules
that attitude by asserting that “perhaps Gilbert’s book would never have
come into my hands if a famous Peripatetic philosopher had not made me a
present of it, I think in order to protect his library from its contagion” (D400;
cf. F426). The absurdity of the imputed motive, together with the cleverness
of the caricature, contributes to instilling an anti-Peripatetic attitude.
On the next page, in order to arouse the curosity of his Peripatetic readers,
and to let this curiosity prevail over their fears, Galileo resorts to one of his
characteristic exaggerations. Salviati proposes to Simplicio:

If you like, I can make it evident to you that you are creating the darkness for yourself,
and feeling a horror of things which are not in themselves dreadful - like a little boy
who is afraid of bugaboos without knowing anything about them except their name,
since nothing else exists beyond the name. [D401; cf. F427]

Later on, in the discussion of magnetism, Galileo gives an argument to

62 CHAPTER 3

show that the earth is de facto a loadstone. The argument involves a compari¬
son which is rhetorically so effective that its persuasive force becomes greater
than its purely logical strength:

SALV. Now, Simplicio, suppose that a thousand pieces of different materials were set
before you, each one covered and enclosed in cloth under which it was hidden, and that
you were asked to find out from external indications the material of each one without
uncovering it. If, in attempting to do this, you should hit upon one which plainly
showed itself to have all the properties which you had already recognized as residing
only in lodestone and not in any other material, what would you judge to be the essence
of that material? Would you say that it might be a piece of ebony, or alabaster, or tin?
SIMP. There is no question at all that I should say it was a piece of lodestone.
SALV. In that case, declare boldly that under this covering or wrapper of earth, stone,
metal, water, etc. there is concealed a huge lodestone. [D404; cf. F430—1]

At the end of Salviati’s explanation of why an armature greatly increases

the amount of iron a loadstone can hold, Sagredo expresses his usual highly
favorable evaluation of the argument. He also expresses the feeling that our
intellect could desire no tastier food than to have equally clear explanations
of the many other magnetic phenomena (F434; cf. D408). Though such
language does not add anything to the cogency of the explanation, it conveys
very well Galileo’s excitement and enthusiasm. The description leaves as strong
an impression as that left, two pages later, by Sagredo’s application of the
label ‘magnetic’ (F436; cf. D410) to the preceding discussion of magnetism.
In this same context there is a passage of great rhetorical value in which
Galileo criticizes the practice of explaining by naming. Simplicio asserts that
Peripatetics would account for the effect just explained by Salviati, in terms
of sympathy, which is a mutual appetite between similar substances (F436,
D410). Sagredo replies that there is sympathy between this manner of philo¬
sophizing and the technique of painting by verbal description which was
practiced by a friend of his, who would write in various spots on the canvas
descriptions of the figures and scenes he wanted there and would then leave
to a painter the actual disposition of colors and shapes. The ingeniousness
of the comparison and the play on the term ‘sympathy’ is the source of the
rhetorical effect here.
At the beginning of the Fourth Day Galileo mentions several available
explanations of the tides. One of these consisted of an account in terms of
differences in the depth of the various seas. In the course of giving a brief
statement of this cause, Galileo describes it as having recently been ‘fished
out’ (F445; cf. D419) from a certain Aristotelian text, previously not
properly interpreted. This is one of the many instances where Galileo is able
 EMOTION, AESTHETICS, AND PERSUASION 63

to use language metaphorically by applying in a critical context terms whose

literal meaning refers to the physical phenomena under investigation. The
description here seems to have a negative connotation, in spite of the fact
that it is given by Simplicio; this is due perhaps to the inappropriateness of
the implied image of the philosopher as fisherman, or more probably to the
implied image of the Aristotelian texts as an ocean with the same fluidity,
formlessness, and susceptibility to taking an almost infinite number of shapes.
On the next page Galileo begins to criticize the various available explana¬
tions of the tides. One of these attributed the tides to lunar heat which would
v/arm the water and cause it to expand and thus rise. Besides claiming that this
explanation is false, insofar as tidal waters can be observed to have the same
density and temperature as ordinary water (F448—9, D422—3), Galileo also
gives the following rhetorical refutation: have the people who hold this
theory “put a fire under a kettle of water, hold their right hands in this until
the heat raises the water a single inch, and then take them out to write about
the swelling of the seas” (D420; cf. F446). The persuasive force of this
rhetorical criticism is much greater than the logical force of the objective
counterevidence also presented; whereas the objective criticism affects only
our intellect, the rhetorical one affects our whole being.
Later in the Fourth Day, just before giving his account of the monthly
and annual periods, he summarizes some of his criticism of the alternative
theories. Here he also makes a new attempt to score a rhetorical point when
he says that the various alleged causes of the tides, far from really causing the
tides, are instead caused by them, in the sense that the existence of the tides
makes the corresponding thoughts arise in frivolous brains (F470, D445).
This is a clever linguistic maneuver and may reflect Galileo’s real opinion on
the matter, but by itself it hardly possesses any rhetorical force. What we
have here is a statement of criticism, a rhetorically interesting such statement,
but the statement contains little if any self-validating force; instead it needs a
supporting argument, which of course Galileo gives elsewhere in the Fourth
Day. So it is not the present statement by itself which possesses persuasive
force.
After Salviati has outlined the general approach he is going to follow in
explaining the monthly and annual periods, there is a rather long pause
(F472, D446-7) in the dialogue, occasioned by Sagredo’s statement that,
though he understands what has been said so far, he does not know where
this is going to lead. Salviati says that he is not surprised that even Sagredo’s
usually quick mind fails to see the aim of the discussion, for he himself
required a very long time before he was able to master these ideas, so much so
64 CHAPTER 3

that he often despaired of ever being able to do so. Then Salviati makes a
reference to Roland from Ariosto’s poem Orlando Furioso, a reference con¬
tinued in the response of Sagredo, who says he is glad that Salviati’s despair
did not have the same effect as Roland’s. This passage is striking both because
of the length of the pause, when the exposition has just started, and also
because of the reference to Roland. I suppose that the rhetorical effect of the
pause and of Salviati’s confession is to emphasize the difficulty of the topic.
In order to understand the reference to Roland, it should be mentioned that
in the poem this character, while being in love with a woman, fmds clear
evidence that she has deceived him with another man; at first Roland refuses
to accept the evidence, and then he becomes insane. Obviously Sagredo is
glad that Salviati’s despair did not lead to madness. What Salviati says is not
so obvious; his words are worth translating literally; “despairing of being able
to unravel it, as a consolation of myself, I forced myself to be persuaded, like
the unhappy Roland, that it might not be true, what however the testimony
of so many trustworthy men presented before my eyes” (F472, my transla¬
tion; cf. D447). The thing that Roland refuses to believe is obviously the
infidelity of his paramour, in spite of the evidence. If the analogy is to hold,
Salviati’s situation must be one where he refuses to believe the falsity of the
geokinetic explanation of the monthly and annual periods in spite of some
counter-evidence, which should be relatively clear evidence. Galileo discusses
and seems to have a plausible answer to the objection that the monthly
change in the earth’s annual motion presupposed by his theory has not been
observed (F479—82, D454—7); however, he does not explicitly discuss or
mention the objection that the tides are greater at the equinoxes than at the
solstices, instead of the other way around as required by his theory.9 There¬
fore, the present analogy to Roland’s situation must be a subtle way of
recognizing such counter-evidence, for there is no other way of making sense
out of the reference. If so, however, the recognition is perhaps too subtle to
be rhetorically effective. Of course, it is rhetorically appropriate not to
emphasize the weaknesses of one’s theory, though at the same time, if the
weakness is very obvious, then it is rhetorically inappropriate not to mention
it at all. So perhaps the subtle recognition does constitute a sufficient qualific¬
ation after all.
The last rhetorically significant passage is contained in a speech by
Sagredo, given after Salviati has expounded the details of the geokinetic
explanation of the monthly period. The passage is a simile, and I think that
its aesthetic value inclines us favorably toward Galileo’s argument, in the
sense that the feeling expressed by Sagredo becomes contagious. This is
 EMOTION, AESTHETICS, AND PERSUASION 65

especially true since, besides giving his usual praise, this time Sagredo prosa¬
ically expresses a difficulty at the end of the same speech, immediately aftei
the poetic image. The simile is one of which Homer himself could be proud:

SAGR. If a very high tower were shown to someone who had no knowledge of any kind
of staircase, and he were asked whether he dared to scale such a supreme height, I believe
he would surely say no, failing to understand that it could be done in any way except by
flying. But being shown a stone no more than half a yard high and asked whether he
thought he could climb up on it, he would answer yes, I am sure; nor would he deny that
he could easily climb up not once but ten, twenty, or a hundred times. Hence if he were
shown the stairs by which one might just as easily arrive at the place he had adjudged
impossible to reach, I believe he would laugh at himself and confess his lack of
imagination.
You, Salviati, have guided me step by step so gently that I am astonished to find I
have arrived with so little effort at a height which I believed impossible to attain. It is
certainly true that the staircase was so dark that I was not aware of my approach to or
arrival at the summit, until I have come out into the bright open air and discovered a
great sea and a broad plain. And just as climbing step by step is no trouble, so one by
one your propositions appeared so clear to me, little or nothing new being added, that
I thought little or nothing was being gained. So much the more is my wonder at the
unexpected outcome of this argument, which has led me to a comprehension of things
I believed inexplicable. [D454; cf. F479]

To conclude, we have attempted to define a dimension of Galileo’s argu¬

ment distinct from its purely logical one and from its purely rhetorical one.
We have found it in its combined emotional appeal and literary-aesthetic
value, and we may label it its rhetorical force. This is distinct from the argu¬
ment’s logical structure, studied in the last chapter, insofar as the latter refers
to argument and appeal to evidence, whereas rhetorical force involves aesthetic
images and appeal to emotions. This rhetorical force is also distinct from
rhetorical appearance, studied in an earlier chapter, insofar as the latter refers
merely to communicative expressions, whereas the former pertains to persua¬
sive effectiveness.
We have also found no necessary conflict between reasoning and emotions,
or between reasoning and aesthetic expression; for, in taking the form of
appeals to emotions, some arguments derive force from the emotions
involved, and the logical element is present in connecting the idea being
justified to those emotions. Similarly, some arguments derive force from the
beauty and pleasantness of the linguistic expressions used and from the
aesthetic images involved. As long as one does not confuse an emotional
appeal or a brilliant literary expression with an appeal to evidence, there is no
problem. Moreover, it is useful to remember that these nonlogical rhetorical
66 CHAPTER 3

devices have their own standards of value; hence, though the pure logician
may act as if they did not exist, the concrete logician, or theorist of reasoning,
cannot do so.
Finally, our study suggests, as envisaged by Feyerabend, that both literary
art and rhetorical persuasion have some role to play in science, for we have
seen that in his justification of Copernicanism Galileo displays the expressive
ingenuity of the artist and the rhetorical skill of the orator. Of course, this is
to say neither that rhetoric and literary art are a guarantee of scientific
success, nor even that they are always relevant. As elements of scientific
method, they are not universally valid, any more than any other elements, as
will be argued exhaustively in later chapters. Nevertheless, they are elements,
and both scientists and scientific methodologists need to note the fact.

NOTES

1 See, for example, W. R. Shea, Galileo’s Intellectual Revolution, p. 117; and L.

Geymonat, Galileo Galilei, p. 126.
2 G. Galilei, Opere, edited by Favaro, 7: 44, 51, 53, 79, 96, 190, 248, 275, 302, 303,
363, 372, 373, 380, 382, 388, 431,477. As usual, subsequent references to this edition
of the Dialogue will be made in the text by prefixing the page number by ‘F\ Similarly,
references to Drake’s translation will be denoted by ‘D’.
3 See, for example, I. M. Copi, Introduction to Logic, pp. 75-6.
4 Cf. M. Scriven, Reasoning, p. xvi; and C. L. Hamblin, Fallacies, pp. 41—42.
5 Cf. the discussion of this argument in Chapters 14 and 16 (Section 6).
6 This translation is my own. To see that the Salusbury-Santillana translation of this
passage, like Drake’s, misses some nuances, cf. G. Galilei, Dialogue on the Great World
Systems, tr. Salusbury-Santillana, pp. 177-178.
7 See Chapter 16, Section 4.
8 For more details see Chapter 10, Section on Strauss, Paragraph 6.
9 Cf. Shea, Galileo’s Intellectual Revolution, pp. 182-3; and Galilei, Dialogue, tr.
Drake, pp. 490-491, note to p. 457.
 CHAPTER 4

TRUTH AND METHOD:

THE SCIENTIFIC CONTENT OF GALILEO’S DIALOGUE

Scientists from Newton to Einstein have found Galileo’s Dialogue on the

Two Chief World Systems to be rich in scientific content. It is well known,
for example, that in the Principia Newton attributed to Galileo a knowledge
of the law of inertia, the law of force, the principle of superposition, the law
of squares, and the parabolic path of projectiles;1 what is not well known is
that, as I. B. Cohen has argued, “Newton almost certainly did not read the
Discorsi - if, indeed, he ever did at all - until some considerable time after
he had published the Principia",2 whereas “the evidence is certain and un¬
mistakable that Newton, early in his scientific career, had read the great
Dialogo . . . sopra i due massimi sistemi".3 It is also well known that, in his
Foreword to Drake’s translation of the Dialogue, Einstein summarizes its
scientific content as being a result about the nonexistence of an abstract
center of the universe, with analogies to his own work. According to Einstein,
“Galileo opposes the introduction of this ‘nothing’ (center of the universe)
that is yet supposed to act on material bodies; he considers this quite un¬
satisfactory . . . [because] although it accounts for the spherical shape of the
earth it does not explain the spherical shape of the other heavenly bodies”.4
Einstein then suggests

that a close analogy exists between Galileo’s rejection of the hypothesis of a center of
the universe for the explanation of the fall of heavy bodies, and the rejection of the
hypothesis of an inertial system for the explanation of the inertial behavior of matter.
(The latter is the basis of the theory of general relativity.) Common to both hypotheses
is the introduction of a conceptual object with the following properties:
(1) . It is not assumed to be real, like ponderable matter (or a “field”).
(2) . It determines the behavior of real objects, but it is in no way affected by them.
The introduction of such conceptual elements, though not exactly inadmissable from a
purely logical point of view, is repugnant to the scientific instinct.5

Typical of scientists’ attitude is perhaps the judgment expressed by Arthur

Schuster in the 1916 Presidential Address to the British Association for the
Advancement of Science: “Modern science began not at the date of this or
that discovery, but on the day that Galileo decided to publish his Dialogues
(1632)”.6 The exact reason for this scientific popularity is not clear. In part,
it may be that, to use Boris Kouznetsov’s eloquent phrase, “an eternal dawn

67
68 CHAPTER 4

reigns in science”,7 and that therefore scientists feel very much at home in a
work characterized by fluid and plastic thought, with which they can identify
in some way. Such scientific intuitions, besides having value in themselves, fit
very well with the latest biographical evidence, which suggests that the book
constitutes Galileo’s mature synthesis of physics and astronomy.8

SCIENTIFIC RELEVANCE VS HISTORICAL ACCURACY

If scientists’ attitudes are so universally positive, the same cannot be said for
historians of science. For one thing, many of them tend to pay more attention
to the Two New Sciences, though in so doing they often engage in a self-
defeating exercise insofar as they begin by choosing it for its seemingly more
scientific character and then end up concluding that its scientific content is
rather meager after all. For example, W. Wisan ends her otherwise excellent
study of the Two New Sciences by saying:

Most of his propositions were of little subsequent interest and only a very few contri¬
buted to an advance in scientific knowledge. His mathematics was unoriginal and seems
clumsy and obsolete for a work published the year after Descartes’ new analytic
geometry, and it is difficult to see precisely how the De motu locali influenced later
treatments of motion. Galileo’s laws of motion are, of course, an obvious contribution,
and the resolution of projectile motion into its components was an important innovation.
But the treatise as a whole seems hopelessly medieval in its separation of ‘natural’ and
‘violent’ motions and the failure to give an analysis of circular motion. There is relatively
9
little here that suggests later studies of the motion of rigid bodies.

Other historians are misled by the fact that the Dialogue has several other
aspects (rhetorical, philosophical, cultural) into thinking that its scientific
content is not noteworthy, as if it were impossible for a work to have all
these dimensions simultaneously; for example, according to Koyrb,

the Dialogue on the Two Chief World Systems pretends to be an exposition of two rival
astonomical systems. But, in fact, it is not a book of astronomy, nor even of physics.
It is above all a book of criticism; a work of polemic and struggle; it is at the same time a
pedagogical work, and a philosophical work; lastly, it is a book of history, “the history
of Mr. Galileo’s mind”.10

Most historians of science regard the book as a chapter in the history of

the struggle for Copernicanism; two attitudes are prevalent here. Some look
for a demonstration of the truth of the Copernican system; when they fail
to find a scientifically valid one, the scientific worth of the Dialogue depre¬
ciates in their eyes. For example, Ernan McMullin feels that “in evaluating
 TRUTH AND METHOD 69

Galileo’s contributions as a ‘man of science’ ”,n one must include the ques¬
tion of whether or to what extent Galileo proved that “the earth is really in
motion”.12 McMullin is unimpressed13 by the proof provided in the Dialogue,
though he admits that “most scientists among Galileo’s readers were per¬
suaded that the Copernican view had been adequately validated by the argu¬
ments of the Dialogo",14 and though he is not insensitive to the fact
“Galileo’s magnificent effort to move the earth has always had and will
always have an even greater power to move men”.15 From the point of view
of the scientific content of the Dialogue, and the point of view of the history
of science in general, one feels like asking rhetorically, “But why should one
want to equate science with demonstrative proof?”, or like quoting Richard
P. Feynman’s view:

In learning any subject of a technical nature where mathematics plays a role, one is
confronted with the task of understanding and storing away in the memory a huge body
of facts and ideas, held together by certain relationships which can be ‘proved’ or
‘shown’ to exist between them. It is easy to confuse the proof itself with the relationship
it establishes. Clearly, the important thing to learn and to remember is the relationship,
not the proof. In any particular circumstance we can either say “it can be shown that”
such and such is true, or we can show it. In almost all cases, the particular proof that is
used is concocted, first of all, in such form that it can be written quickly and easily on
the chalkboard or on paper, and so that it will be as smooth-looking as possible. Conse¬
quently, the proof may look deceptively simple, when in fact, the author might have
worked for hours trying different ways of calculating the same thing until he has found
the neatest way, so as to be able to show that it can be shown in the shortest amount
of time! The thing to be remembered, when seeing a proof, is not the proof itself, but
rather that it can be shown that such and such is true. Of course, if the proof involves
some mathematical procedures or ‘tricks’ that one has not seen before, attention should
be given not to the trick exactly, but to the mathematical idea involved.
It is certain that in all the demonstrations that are made in a course such as this, not
one has been remembered from the time when the author studied freshman physics.
Quite the contrary: he merely remembers that such and such is true, and to explain how
it can be shown he invents a demonstration at the moment it is needed. Anyone who has
really learned a subject should be able to follow a similar procedure, but it is no use
remembering the proofs.16

In answer to my rhetorical question, McMullin might answer that it is proper

in this context to equate science with demonstrative proof because Galileo
himself equated the two.17 It would not be difficult to show that such a
philosophy of science is not at all representative of Galileo’s remarks on
scientific inquiry (see Chapters 5 and 6); however, for the present purpose,
that fact, if and to the extent that it is a fact, is not relevant. For we are not
examining here the philosophical content of Galileo’s Dialogue, but rather its
70 CHAPTER 4

scientific content; our own context is here one of history of science not his¬
tory of the philosophy of science. If we were doing the latter, then it would
be proper first to determine what Galileo’s philosophy was, as McMullin
does,18 and then to look in Galileo’s works for those elements which can be
regarded to constitute science according to that conception of science.19 It
would be clearly necessary to do the latter, as well as the former, for it would
be unfair to extract someone’s philosophy of science by merely examining his
reflections upon science without also looking at his scientific practice. So it
may be said as a point in favor of McMullin’s account that he does look at
both, thus filling with concrete content his account of Galileo’s philosophy
of science. Nevertheless, such an account is criticizable in its own terms, and
moreover, we do not get into science and scientific context by his route.
The second ‘Copernican’ approach to Galileo’s Dialogue defines its content
to be essentially propaganda for Copernicanism. In recent times, this view has
originated from Arthur Koestler,20 propagandized by Paul Feyerabend,21 and
traces of it can be found in Dudley Shapere,22 William Shea,23 and Giorgio de
Santillana.24 The truth behind this interpetation is that the book had con¬
siderable practical impact, that, in fact, it is full of rhetorically significant
passages, and that practical considerations lurk everywhere in its conception
and composition. However, the only responsible way of defining its scientific
content in terms of the rhetoric of the earth’s motion is in the context of the
science of rhetoric; for after all the study of the art of persuasion and of
achieving practical effects by verbal means is the subject of a discipline
which is at least as old as Aristotle’s rhetoric.25 So the only business that a
historian of science has meddling into the rhetorical analysis of the Dialogue
is if he is writing the history of the science of rhetoric, not if he is concerned
with the history of natural science. Let me make it clear that I think that
studies of the rhetorical content of Galileo’s Dialogue are quite proper; I
myself believe that the book is a classic in the art of rhetoric, besides being
one in the literature of science. However, these scholars do not provide a
serious rhetorical examination of the book, the foundation of such an under¬
taking being that there can be good as well as bad rhetoric, a point that the
above mentioned writers tend to ignore when they speak of propaganda. In
fact, given that the context of their own investigations is an ambiguous one
between science and rhetoric, when they speak of propaganda, they are
contrasting this to valid science, so that the term has for them an implicit
negative connotation. Partly to remedy this situation, in previous chapters I
undertook the rhetorical analysis of Galileo’s book, and I attempted to
extract its rhetorical force, structure, and content. However, here I am
 TRUTH AND METHOD 71

concerned with its scientific content, scientific in the sense of natural science.
The divergence between scientists’ and historians’ perception of the scien¬
tific content of the Dialogue acquires its clearest expression in connection
with the interpretations emanating from the medievalist historians. Naturally
enough, when these scholars read the Dialogue the main thing they perceive
is similarities and differences between what’s contained therein and the views
of various medieval thinkers.26 Though this kind of exercise informs us about
the medieval content, if any, of Galileo’s Dialogue, it is not clear that it tells
us anything about the nature of its scientific content. The less tenable form
of the medievalist interpretation would speak of a medievalist origin of, or
influence upon, Galileo’s book; I believe that this version of the medievalist
thesis is presently being rendered of historical interest only (if I may be
allowed the pun) through the efforts of Stillman Drake and his followers.27
Let us then formulate the thesis in the more tenable form according to which
the Dialogue is claimed to contain views which de facto happen to be similar
to those found in various medieval texts. If true, this might be very interesting
from a number of points of view, which it is not my job to elaborate. For my
job here is to ask of what relevance would the alleged medieval aspects of the
Dialogue be to understanding its scientific content? Here no one, not even the
medievalists themselves, would accept the relevant equation (scientific=
medieval); so there is no question of confusion, as there is for the identifica¬
tion of science with demonstrative proof, presupposed in one of the other
interpretations. But if we avoid the relevant equation, then the relevance can
only be defended by the influence thesis; but if we accept the latter, then at
least for the present case of Galileo, we draw fire from the guns of Drake and
his followers, a fire which is very difficult to resist and which may therefore
very well consume us. In the light of the continued resistance by scientific
intuition28 to the infusion of medieval elements into modern science, we may
set aside the work of these historians in our attempt to examine the scientific
content of Galileo’s Dialogue.
Another common approach to the analysis of the scientific content of this
book and of Galileo’s work in general, is to examine it from the point of view
of Newtonian science. This is of course more satisfactory, but only slightly so.
The practice was more justified before the advent of post-classical physics,
but nowadays it is no longer proper to equate science with Newton’s work.
There is no question that the detection of what might be called the
Newtonian content of Galileo’s book is a highly relevant exercise, of the
approaches so far considered it is perhaps the most relevant. Nevertheless the
Newtonian approach breaks down completely when the Newtonian content
72 CHAPTER 4

blinds one to the more-modern scientific content,29 and it can be shown still
to be inadequate when there is no conflict.30
A conflict exists, then, between scientists’ and historians’ perceptions of
the scientific content of the Dialogue. It won’t do to point out that historians
can often show that what scientists perceive as being in the Dialogue is not
there in some sense, and that by and large scientists find in this work what¬
ever they want to find. For, even when inaccurate, scientists’interpretations
are suggestive of the way the scientific mind works and hence relevant to the
analysis of the book’s scientific content. Moreover, though historians tend to
be more accurate, what they sometimes perceive in the Dialogue is not there
either, as I show in detail in later chapters for the cases of Koyr6, Strauss,
Favaro, Drake, and Clavelin;31 and, more importantly, one can often show, as
I have done above, that what historians of science perceive in the Dialogue is
often not science, but something else. The real problem is, then, that
historians’ interpretations tend to be accurate but irrelevant, whereas
scientists’ interpretations tend to be relevant but inaccurate. Is it possible to
combine relevance to science and accuracy to the text? Unfortunately, it may
not be possible because on the one hand the historical attitude as such is
bound to introduce irrelevancies, whereas the scientific attitude as such is
bound to introduce inaccuracies. Why is this so? I shall first give an abstract
general argument and then a concrete one grounded on the particular
example of Galileo’s tidal theory in the Fourth Day.
A historian of science is usually someone who is acquainted with the deve¬
lopment of science, and hence with scientific ideas and facts of various time
periods. However great his acquaintance with contemporary science and the
latest textbooks and journals is, this is for him merely the science of the
present period. The minute he holds the latter with some special reverence or
gives it special status, he thereby abandons the historical attitude, which
requires of him a period-free neutrality or objectivity. Of course, his
specialized area of competence is likely to be a particular period, but the
historical sensibility does not allow him to magnify this accident of his bio¬
graphy into a historical or historiographical fact. What happens, then, when a
historian opens a book like Galileo’s Dialogue and we ask him to tell us about
its scientific content? When the question is asked in the more or less loaded
terminology of ‘scientific’ content, perhaps he will refuse to answer it, and
confess that he doesn’t know what is being asked in asking for the specifically
scientific content of the book. We can make the question less loaded and ask
about the intellectual content of the book. The intellectual elements that he
is going to come up with will be ones involving similarities and differences
 TRUTH AND METHOD 73

with other such elements found in previous and in subsequent periods,

thinkers, texts. The relevance of such comparisons and contrasts, the rele¬
vance, that is, to science or at least to the interests and concerns of (living)
scientists, this relevance is exactly what is hard to see.
Let us look at the scientists’ situation now. A natural scientist is one who
is making more or less original contributions to the understanding of nature.
Science is for him what he himself (and his peers) do. When he opens the
Dialogue the intellectual elements that are likely to attract his attention are
those that have similarity to his own scientific involvements. Because of the
growth of scientific knowledge he will necessarily be blind to problems and
ideas that may have been central to the book’s author but which do not relate
to present day scientific research; this relation has to be one identifiable by
his scientific intuition, not one demonstrable by historiographical, philoso¬
phical, or logical means. I have argued elsewhere that what Joseph Agassi calls
inductivism is the typical and inescapable attitude of the working scientist.32
The growth of scientific knowledge makes it also unlikely, in principle, that
the elements detected by the scientist are really in the book; for this would
be like saying that Galileo was concerned with problems whose content and
substance, as opposed to method and general character, were similar to those
of the contemporary scientist; since the latter’s problems are usually
definable only by reference, at least implicit, to past scientific achievements,
it follows that the actual similarity of content and substance is excluded
almost in principle. Of course, this becomes less true as one gets closer to
Galileo’s own period; so that, for example, Newton could accurately claim
more similarity of content. However, the real question is whether the extent
of this similarity is significant. The fact is that, almost as a matter of defini¬
tion, the more creative a scientist is, the more his achievements surpass his
predecessors (in content, we must remember, not necessarily in approach).
It follows that he is less likely to understand them, that is to say understand
them in the historical sense of accuracy. Of course, in another sense, the
scientific sense, it may be said that he is the only one who really understands
his predecessors, since he is the one who superseded them by using their
results and building upon them. However, this superior scientific understand¬
ing is merely another way of saying that the scientific relevance of what he
finds in his predecessors is guaranteed, a point I have already conceded, but
what 1 am discussing here is his historical accuracy, which remains necessarily
problematic.
74 CHAPTER 4

TIDES: THE FOURTH DAY

A beautiful example of this tension between scientific relevance and historical

accuracy is provided by available accounts of the tidal theory of the Fourth
Day. As is well known, in the last ‘Day’ of the Dialogue, Galileo justifies the
earth’s motion by arguing that only its combined daily axial rotation and
annual orbital revolution can explain the existence of tides.
Scientific readers of Galileo’s tidal theory, whether their attitude is critical
or apologetic, have perceived a number of scientific elements in it. For
example, Ernst Mach feels that “the principle of relative motion is a correct
feature of this theory, but it is so infelicitously applied that only an extremely
illusive theory could result”;33 he also states that “it is noteworthy that
Galileo in his theory of the tides treats the first dynamic problem of space
without troubling himself about the new system of coordinates. In the most
naive manner he considers the fixed stars as the new system of reference”.34
In 1954, V. Nobile gave a mathematical argument designed to show the
existence of what he calls a tidal ‘Galilean effect’, independent of gravitation,
and deriving from the combination of the earth’s rotation and revolution.35
In 1962, H. Burstyn36 argued that the following scientific concepts can be
found in the Fourth Day: (1) an imperfect but fundamentally correct theory
of the so-called ‘tide of reaction’;37 (2) the insight that once the ocean has
been set in motion by the primary cause of the tides, it tends to oscillate
from its own inertia with a period which may differ from that of this primary
cause, since this period depends also on such factors as the size of the basin
and the depth of the water;38 and (3) the fact that “the earth’s orbit around
the sun is described not by the center of the earth but by the center of mass
of the earth-moon system”.39
The historical accuracy of Burstyn’s account has already been questioned.
Concerning the point about the earth-moon system, E. J. Aiton has shown
that the context of Galileo’s argument is such that he is not thinking that the
distance between the earth and the sun changes, as the earth-moon system
goes around it, but rather only that their orbital speed changes as the sun-
moon distance changes due to the moon going around the earth.40 Though
significant, Aiton’s argument is merely an expression of what a careful reader
of the Dialogue readily notices. In his response to Aiton, Burstyn all but
concedes the textual incorrectness of his own interpretation when he is
willing to admit that his view depends on an anachronism.41
Burstyn’s second one of the above attributions is indeed correct, and he
does deserve great credit for having emphasized this aspect of Galileo’s
 TRUTH AND METHOD 75

theory, which historians usually neglect. This is one of those things appre¬
ciated much more readily by a scientifically minded reader like Burstyn,
who is very much aware that the scientific problem of the tides is extremely
complex, that Newton’s gravitational theory is merely the beginning, that it
was not until 1960 that a general numerical solution to the equations of tidal
motion was found,42 and that “in only a very few cases we can carry out the
theoretical calculations necessary to approximate closely the tide of a given
region of the ocean; in practice we predict the tide for various coastal points
by harmonic analysis of records of tidal observations”43
As for Burstyn’s first point, Aiton also criticizes his interpretation of
Galileo’s theory as an approximation to the ‘tide of reaction’.44 However, in
the light of Burstyn’s reply 45 I find the scientific debate inconclusive. Never¬
theless, the textual accuracy of Burstyn’s interpretation can be effectively
decided by pointing out that Galileo is not considering centrifugal accelera¬
tion; he is not adding and subtracting the rotational and the orbital accelera¬
tion, but the speeds. This is clear from the text:

in coupling the diurnal motion with the annual, there results an absolute motion of the
parts of the surface which is at one time very much accelerated and at another retarded
by the same amount. This is evident from considering first the parts around D, whose
absolute motion will be very swift, resulting from two motions made in the same direct-
tion; that is, toward the left. The first of these is part of the annual motion, common to
all parts of the globe; the other is that of this same point D, carried also to the left by
the diurnal whirling, so that in this case the diurnal motion increases and accelerates the
annual motion.
It is quite the opposite with the part across from D, at F. This, while the common
annual motion is carrying it toward the left together with the whole globe, is carried to
the right by the diurnal rotation, so that the diurnal motion detracts from the annual. In
this way the absolute motion - the resultant of the composition of these two - is much
retarded 46
D
76 CHAPTER 4

It would be wrong, however, to argue as some do47 that Galileo was not
thinking of centrifugal acceleration because he could not have been so think¬
ing; and he could not, because it would have been inconsistent with his own
principle of mechanical relativity, according to which all terrestrial phenomena
would happen the same way whether or not the earth is moving, so that there
is no way of testing its motion with a mechanical experiment within the
earth. The reason is that there is evidence that Galileo did not hold the prin¬
ciple of relativity as an absolute truth but rather as a first approximation.48
In summary, what one can say about Burstyn’s account is that though it is
full of brilliant scientific suggestions, it develops a tidal theory which is more
correctly regarded as Galilean rather than Galileo’s. It is indeed in accordance
with knowledge that Galileo shows elsewhere to possess, and it may be
regarded as a plausible development and elaboration of what is in his mind.
However, it is not actually in his mind (in the text); in this sense it is not
Galileo’s theory.
By contrast, the most recent and elaborate account by a historian does not
attribute to Galileo beliefs not present in the text, but neither, as we shall
see, any beliefs present in the body of scientific knowledge, namely, any
scientific content. The account is part of W. R. Shea’s Galileo’s Intellectual
Revolution, and it is typical of the rest of the book, which may be regarded
as the most extensive historical treatment recently undertaken of Galileo’s
Dialogue,49 A lack of scientific content of the Fourth Day is being claimed
when, after giving a critical summary of Galileo’s theory, Shea concludes:

There can be no doubt that Galileo’s theory of the tides opened no new scientific vista
to his successors. He neglected to take cognizance of the four well known periods of the
tides, he rode roughshod over the discrepancies between his theory and experience, he
did not investigate striking observational consequences entailed by his explanation, and
he brushed aside contemptuously any appeal to the influence of the moon.50

It must be said to Shea’s credit that he is aware that it is questionable how his
analysis is relevant to contemporary science. In fact, in advocating the
‘contextual method’ as opposed to the “linear view of the history of science”,
he is explicit that questions about scientific content (in the sense presently
under discussion) are not the only ones for historians of science to deal with.
They should also examine the blind alleys of the past if this turns out to be
necessary to “assess old theories in terms of the conceptual framework of the
scientists who held them, and judge them against the background of the
world picture of their age”.52 The condition just mentioned is very important
and I shall discuss it shortly. For now, it should be noted that Shea’s espousal
 TRUTH AND METHOD 77

of the ‘contextual method’ is tantamount to admitting that he is not

analyzing what I have been calling here the scientific content of the Dialogue.
The fact that Shea is aware of what he is doing does not make it any more
relevant to contemporary science.
To be sure, Shea would argue for the indirect relevance of his analysis to
contemporary science. For the contextual method puts us in touch with the
actual thought-processes of past scientists and hence with the actual process
of scientific discovery.53 For example, when we realize that Galileo’s dis¬
cussion of the tides fails to make sense from the scientific point of view,54
we begin to understand that “he was a natural philosopher who saw beyond
the problem of determining the periods of the tides, about which he did not
feel strongly, to the great vision of a science in which the real is described by
the ideal, the physical by the mathematical, matter by mind”55; and this leads
to the realization that “Galilean science was not so much an experimental
game as a Platonic gamble”.56 In other words, the indirect scientific content
will be what it tells us about the nature of science, which we may call the
book’s methodological content.
I believe that this is the right direction to move toward. However, in trying
to extract the methodological content one must use the contextual method
quite rigorously. Unfortunately, Shea smuggles in contemporary science, or
at least his conception of it, by the back door in his examination when he
uses science-dependent evaluations of some of what he attributes to Galileo.
For example, Shea is eager to refer to a criticism of Galileo’s theory given
by Mach.57 Shea reports Mach’s criticism as amounting to the following:

Galileo makes the error of mixing two different frames of reference. Whereas the motion
of the earth is considered relative to the sun, the motion of the water is considered
relative to the earth. But relative to the earth, the water can receive no acceleration due
to the earth’s annual motion, and the water must therefore be at rest relative to the
earth.58

Unfortunately, though this criticism can be found in Aiton59 and in Arthur

Koestler,60 I do not find it in Mach.61 Mach does say that “the principle of
relative motion is a correct feature of this theory, but it is so infelicitously
applied that only an extremely illusive theory could result”62; to show the
misapplication Mach describes a thought-experiment in which a rotating
sphere is set in uniform motion; for Mach no tides can occur during the
uniform progressive motion any more than while it was merely rotating since
“the case in question does not differ, according to our view, in any essential
respect from the preceding, inasmuch as the progressive motion of the sphere
78 CHAPTER 4

may be conceived to be replaced by a motion in the opposite direction of all

surrounding bodies”.63 Second, Mach claims that Galileo’s error lies in what
he calls a “negative conception of the law of inertia”64; though obscure, this
has nothing to do with mixing reference frames. Mach then goes on to explain
that Galileo did not apply the concept of acceleration properly.
Shea’s eagerness to misinterpret Mach, or to invoke his authority, repre¬
sents a violation of the contextual method. Moreover, even when correctly
interpreted, such an appeal would be anachronistic. To defend himself against
this latter criticism Shea claims that Mach’s criticism was expressed by some
of Galileo’s contemporaries by quoting from a 1633 letter in which the
criticism of a group of French physicists is reported.65 Though it is difficult
to read into this letter Mach’s criticism, one can find in it traces of Shea’s
own criticism (from mixing). In the letter it is stated that, in Galileo’s theory
the various parts of the globe are accelerated relative to the sun but not
relative to each other, hence it is difficult to see how the oceans could acquire
motion relative to the land. No charge of confusion is argued for or even
voiced. Nor does Shea give a justification. Since there is a difference between
saying that it is difficult to see how A can follow from B, and saying that A
and B are being confused, the contextual nature of Shea’s criticism is un¬
supported.
Next, Shea refers to “Galileo’s failure to distinguish centripetal accelera¬
tion from linear acceleration”.66 The reality of this failure is undeniable,
but so is the uncontextual character of criticism based on it.
Shea’s criticism so far has been that Galileo’s theory is supported by a
fallacious argument, which would be a fair criticism only if it really follows
the contextual method, which it does not. His next criticism is that Galileo’s
theory is inconsistent with the facts insofar as it implies a 12-hour lapse
between high and low tide, rather than a 6-hour lapse, which was allegedly
well known.67 What Shea ignores here is the context of Galileo’s own discus¬
sion68 which emphasizes that the tides are not due to the primary cause alone
(the earth’s double motion) but to the interaction between it and various
secondary factors (such as water depth and size of basin) which affect the
period of oscillation of the water.
In conclusion, if there is no way of analyzing the (substantive) scientific
content of the Dialogue which is both scientifically relevant and historically
accurate, then we will have to be satisfied with analyzing its methodological
content as its indirect scientific content. In the latter investigation we will
have to use what Shea calls the contextual method, but in a more rigorous
manner than he does. For we have seen that he has the tendency to mix
 TRUTH AND METHOD 79

contextual evaluations with ones based on contemporary (or at least subse¬

quent) scientific knowledge; and such an impure contextual method would
seem even less adequate than the procedure of the so-called linear historian of
science, who uses contemporary science primarily as a criterion of choice of
topics.

SCIENCE: THE FIRST DAY

Is there really no way of combining textual accuracy and scientific relevance?

Should it not be possible to give an account of the Dialogue that keeps an eye
both on the text and on concepts that are likely to be of interest to the work¬
ing scientist? To explore further this possibility I have attempted a reconstruc¬
tion of the scientific content of the First Day. One reason for choosing this
section of the Dialogue is that it is often alleged to be more philosophical
than the other three Days, philosophical in a sense in which this term is
contrasted to scientific; hence, our attempt will also be a test of this belief.
The analysis given below has resulted in the First Day being divided into
eleven sections, each of which discusses a substantive scientific idea, problem,
instrument, fact, observation, or experiment. My interpretation of each
section attempts to integrate all its substantive scientific topics around a suit¬
able theme. Whenever a particular claim in my interpretations can be traced
directly to a given place of the text, I have indicated the page number in
parenthesis. It has turned out that the topic of acceleration predominates in
the first third of the ‘Day’, and that the topic of the moon predominates in the
last third; the former fact is somewhat surprising, though the latter obviously
is not. The various sections and their scientific topics are as follows:69

The importance of acceleration: F33-45.10; D9-21.

The continuity of acceleration: F45.ll—47.10 & F51.18—5 3.13; D21—23 & D27—29.
Acceleration vs. speed: F47.10—51.18; D23—27.
An acceleration problem: F53.13—54; D29-30.
The isotropy of space: F54-61.28; D30—37.
The primacy of mechanical properties: F61.28-75.8; D37-50.
Sunspots and the telescope: F75.9-87.14; D50-62.
Amount of visible lunar surface: F87.15-95.30; D62-70.
The moon’s mountainousness: F95.31 —112.28; D71-87.
The moon’s secondary light: F112.29—122.29;D87—97.
The problem of moon spots: F122.30—131; D97-105.

The details of the scientific content of these various sections are, respec¬
tively, as follows.
 80 CHAPTER 4

The importance of acceleration (F33—45.10; D9—21). Geometrical

(spatial) properties can act as the foundation for cosmological facts, but not
for the physics of motion; for the latter one needs to combine spatial
properties with temporal ones, such as temporal uniformity; this combination
leads to the concept of acceleration.
For example, the three-dimensionality of the world is based on the possi¬
bility of constructing three and only three mutually perpendicular straight
lines through a given point (F33—8); and the orderliness of the world is based
on the existence of circular motions (F43-5). But the geometrical simplicity
of straight and circular motions is useless in physics since circularity around
one center is as good as circularity around another center (F40), and a
straight line off the center is as good as one going through it (F40). Accelera¬
tion is the important property in the physics of motion because accelerated
straight motion is the natural means of acquiring uniform circular motion
(F43—5), which alone retains cosmological orderliness and is in this sense
natural (F43—5). And acceleration is a temporal phenomenon in the sense
that it takes place in time, i.e. is not instantaneous (F44, 45), and that it
depends on the concept of uniformity.
The continuity of acceleration (F45.11-47.10 & F51.18-53.13; D21-23
& D27-29). The continuity of motion implies that a body falling from rest
has at some point in its path such a small degree of speed that at this speed
it would not have traversed a distance of two hundred braccia in a thousand
years. This is difficult to understand in the light of the fact that the body can
be seen to acquire very quickly a high degree of speed. The problem can be
solved as follows.
First, it seems plausible that the degree of speed at any point of the fall is
such as to allow the body to go back to the height from which it started to
fall. Evidence for this is the plausibility of imagining that a body falling
toward the center in a tunneled earth would continue to move past the center
for a distance equal to that through which it had gone in falling to the center.
Evidence is also provided by (1) the fact that a pendulum removed from the
vertical swings back to the other side with equal amplitude, except for the
small disturbance of the air and string, and (2) the fact that water seeks its
own level. It follows that bodies acquire the same degree of speed when they
fall the same amount, regardless of their paths. For example, in fading from
c to A by CA the speed at A would be the same as that at B in falling from C
to B by CB.
Second, it seems plausible that the less inclined a plane is, the more slowly
a body moves in falling by that path. For example, a body will move from D
 TRUTH AND METHOD 81

to A much more slowly than from C to A. Since over the horizontal BA a

body will not move naturally at all, it is easy to imagine an inclined plane,
between DA and BA, that is, whose top is so close to B, that the body in
falling to A over that path will take an arbitrarily long time, for example a
thousand years.
Third, it is a fact that a body falling through a given distance, has at the
end of that distance a speed which if kept constant would allow it to traverse
double that distance in the same time.
It follows that one could drop a body from a point above B so close to it
that at B the body would have a speed as slow as you like. Similarly, dropping
the body from C, there are points so close to C, that at those points the body
has a speed as slow as you like.
Acceleration vs speed (F47.10-51.18; D23-27). It is important to distin-
82 CHAPTER 4

guish acceleration from speed, otherwise one gets into unnecessary problems.
For example, let AB be horizontal, CB perpendicular to it, and compare the
motion of a ball falling freely from C to B, with its motion along on the
inclined plane from C to A. Clearly, in one sense, the perpendicular motion
will be faster than the inclined one, and yet, in another sense, the speed of
the body at A will be the same as at B, since they have descended the same
amount and hence should have acquired a degree of speed which would
enable them to go back to the same height. The apparent contradiction can
be removed since the quantity which is the same from C to B and from C to
A is the speed, whereas the quantity which is greater along CB than along CA
is the acceleration. This can be shown as follows.
To say that two speeds are equal means that the distances traversed are
proportional to the times elapsed, and to say that one body moves faster than
another means that in the same time the first traverses a greater distance than
the second, or that in traversing the same distance it takes less time. Now it is
easy to visualize that in the time that the body falls from C to B, it would
move along CA up to a point T such that CT < CB, or in the time that the
body falls from C to A, it would have fallen from C past B to some point O
such that CA < CO. In either case, in the perpendicular motion, greater
distances are being covered in the same time. Hence at corresponding times
the speeds along the perpendicular are always greater, which is to say that the
speed is increasing faster along the perpendicular than along the incline. That
is, the acceleration is greater.
On the other hand, in moving from C to A the body takes longer than in
falling from C to B. Hence a greater distance is covered in a longer time. All
that is required for the speeds to be equal is that the ratio of the time along
CA to that of the time along CB should be the same as the ratio of CA to
CB. Clearly this can happen in spite of the fact that the perpendicular motion
(i.e. the increase of speeds along the perpendicular) is faster than the inclined
one. For, by the time the body falls from C to B, it would have moved to
some point T on the incline, and it is plausible to think that in continuing to
move toward A, the excess of the time will equal the excess of CA over CB.
An acceleration problem (F53.13-54; D29-30). Imagine the solar system
to have originated in such a way that the planets first moved with straight
accelerated motion toward the sun and then, after each had acquired its
proper speed, it started to move at this speed in a circular orbit around the
sun. The question is whether all planets could have originated from the same
place and what was the distance from the sun of this place of origin. To solve
the problem one can choose two planets (e.g., Saturn and Jupiter) and from
 TRUTH AND METHOD 83

the ratio between their velocities, the distance between their orbits, and the
proportion of acceleration for natural motion, one can calculate the distance
of their place of origin. Then one checks whether each of the other planets,
by falling from this place toward the sun would have acquired the speed and
orbit that it actually has. It turns out that the correspondence is amazingly
close. The initial calculation involving a pair of planets such as Saturn and
Jupiter, would be the following.70

S = sun
O = place of origin of fall
Rj - radius of Jupiter’s orbit
Rs = radius of Saturn’s orbit
Pj = period of Jupiter’s revolution
Ps - period of Saturn’s revolution
Vj = orbital speed of Jupiter
Vs = orbital speed of Saturn
Dj = distance fallen by Jupiter
Ds = distance fallen by Saturn
Tj = time of fall for Jupiter
Ts = time of fall for Saturn

(] \ Ji = 2riRi/Pl = RJPs . woui^ be known from astronomical data.

V ; Vs 2ttRs/Ps RsPj
(2) Rs - Rj would be known from astronomical data.

(3) ZL = iz_ ; this would be what Galileo calls “the proportion of accel-
Ds Ps eration of natural motion”.

(4) Since — = It therefore > by (3).

Vs Ts Ds Vs2
84 CHAPTER 4

(5) Since Rs - Rj = Dj - Ds, therefore = 1 +— -- .

Rs-Rj_RlPl
(6) From (5), (4), and (1): 1 +
Dc

TVi f
(7) Therefore, */2V
—-—- = --- ,, „ 0 — i1.
D,

— Rj
(8) Therefore, Ds =
(Rf2Ps2/Rs2Pj2) - 1 '

Here, all the quantities on the right side of the equation would be known
from astronomical data. The other calculations would involve checking
whether for each planet n,

_ (2ttRn/Pn)2 _ Ds + (Rs-Rn)
Vs2 (2irRs/Ps)2 Ds
The isotropy of space (F54—61.28; D30—37). There is no preferred direc¬
tion in space. In moving downwards, heavy bodies are moving towards the
center of the whole of which they are parts, rather than toward the center of
the universe. In the vicinity of other celestial bodies, such as the moon or
the sun, their respective parts would fall toward the centers of their respective
wholes.
The primacy of mechanical properties (F61.28—75.8; D37—50). The
mechanical properties of bodies are the most basic in nature. For example,
in the Aristotelian system, the all-pervasive distinction between the terrestrial
and the celestial worlds is based on the difference between straight and
circular natural motions (F61—3). Or compare the proposition that the earth
rotates with the proposition that change derives from contraries; the former is
much easier to ascertain and understand (F64). The latter is full of empirical
difficulties (F64-5) and of conceptual ones such as: (1) if there exist both
changeable and unchangeable bodies in nature, then given that change derives
from contraries, the unchangeable bodies should be changeable after all
because of their contrariety with unchangeable bodies (F65-9); (2) some¬
thing (e.g. a whole) can be unchangeable while its parts change (F69-71);
(3) if a body has never been observed to change, is it incapable of change or
merely unchanged-so-far (F71—5)?
Sunspots and the telescope (F75.9-87.14; D50-62). The telescope makes
possible a revolution in astronomy by bringing heavenly objects thirty or
forty times closer; for example it led to the discovery of the most significant
 TRUTH AND METHOD 85

piece of evidence against the earth/heaven dichotomy, namely sunspots

(F80—1). The force of their evidence contrasts with that from comets, which
may very well be atmospheric phenomena (F76—7), and with that from the
new stars, which might be claimed to constitute changes in the heavens but
not changes in some heavenly body (F82—3). The sunspots are significant
because they involve what may be regarded as the noblest of the heavenly
bodies (F82) and because the changes in their shape and in their speed across
the solar disc allow no other interpretation but that they are contiguous to
and part of the spherical body of the sun (F77-80). Finally it is incorrect to
object (F83-7) that sunspots would be of no use or benefit to man, and
hence superfluous, thus making the sun an imperfect body, which cannot be;
such anthropocentric teleology is without any justification.
Amount of visible lunar surface (F87.15—95.30; D62—70). A number of
facts about the moon are ascertainable relatively easily, for example its
spherical shape, opacity, solidity, monthly axial rotation (F90), mutual
eclipsing, and mutual showing of phases with the earth. Others require careful
reasoning and observation: for example, a little more than half of the lunar
surface is visible and the moon’s motion is tied to the earth’s center (F90).
This is shown in the following two figures.71
In Figure 1, as the relative position of the moon and a point on the earth’s
surface S changes through Positions 1,2, and 3, from the earth’s center O one

Figure 1.
86 CHAPTER 4

would always see the same lunar hemisphere ABC; in each case the line con¬
necting the centers of the earth and of the moon intersects the lunar surface
at the same point B. However from S one would see ABC only when the
moon is directly overhead (Position 2); when it is in Position 1, one would see
A iBiCi; when it is in Position 3, one would see A3B3C3. That is, from S one
canseeCCj andAA3 besides the hemisphere ABC (¥90—\).
In Figure 2, line SOS is in the plane of the ecliptic, LOL in the plane of
the moon’s orbit. When the moon is north of the ecliptic (left), it becomes
possible to see the region A Y near the moon’s south pole (if the moon and
sun are in opposition); similarly, when the moon is south of the ecliptic
(right) it becomes possible to see region CZ near the north pole (F90-1).

Figure 2.

This can be confirmed by observation. It is possible with the telescope to

see on the lunar disc two diametrically opposite spots, in a northwest, south¬
west direction. These spots, which are very near the edge of the disc, can be
observed to vary by a factor of two or three in their apparent distance from
that edge (F90-1).
The moon’s mountainousness (F95.31—112.28; D71-78). The mountain¬
ousness of the moon can be observed rather directly by the telescope (FI 12).
The observable details cannot be explained in terms of variations of opacity
such as are found in mother of pearl (Fill); nor can mountainousness be
ruled out of existence by the doctrine that heavenly bodies are perfect and
hence perfectly spherical, since shape has nothing to do with perfection
either in fact (FI09-10) or even according to the Aristotelian supporting
argument (F109). Moreover, if the moon were perfectly spherical, it would
reflect light in all directions and would transmit only a ray to an observer on
 TRUTH AND METHOD 87

earth, for whom it would be invisible because of the distance; whereas its
roughness would allow its being seen the way it is (F95—109).
The moon’s secondary light (FI 12.29—122.29; D87—97). The reflecting
power of the earth is something that can be established by a judicious com¬
parison of observations, namely of the moon and of clouds or mountains
during the day and of the moon during the day and at night (FI 12—5). So,
the moon’s secondary light is to be explained as coming from the earth
(FI 15), and not as being the moon’s own light (FI 15—16), or as deriving
from others stars (FI 16-17), or as coming (directly) from the sun
(F117—21).
The problem of moon spots (F 122.30— 131;D97—105). What is the nature
of the moon’s large dark spots visible with the naked eye (FI24)? Though
we can observe with the telescope that the moon surface in those spots is
rather flat, we have no way of knowing whether this is sufficient to make
them appear darker than the surrounding regions (F125). The effect would
result if these dark regions were covered with water (F123—4, 125) or with
forests, or with rocks of a naturally darker color (FI25). However, there
seems to be no water and no life similar to ours on the moon since the lunar
day is one month long, the seasons are very insignificant, and no clouds are
visible with the telescope (F125—6). Moreover, one should not expect that
the nature of these spots is the same as the darker appearances which the
oceans would show to someone observing the earth from the moon; for so
requires the richness of nature and the omnipotence of God.

INERTIA AND CIRCULARITY

One could proceed with this type of analysis for the rest of the book. How¬
ever, before doing so we should ask whether it is giving us what we want, in
particular whether we are overcoming the tension between relevance and
accuracy. Generally speaking I am pretty well satisfied with the accuracy,
but I wonder whether it is relevant enough, or at least whether it has the
right kind of relevance.
Let me begin with accuracy. Part of the justification lies in the page
references that I inserted in my account. Such references could not, however,
be exhaustive; there simply is not a one-to-one correspondence between every
one of my interpretative claims and the sentences in the original text. But
there is no reason why there should be, for what is being sought here is not a
copy, or a repetition, or even a summary of the original. In an abstract-outline
one could provide that sort of one-to-one summary. Here, instead, we are
88 CHAPTER 4

seeking an analysis from a definite point of view, the substantive-scientific. So

part of the justification of the above account lies in the nature of the present
undertaking.
The accuracy of the above account can also be justified by comparison
with other available interpretations. It is not worthwhile to do this for every
one of the eleven passages above. Let us select some of the most controversial
ones, such as the very first one.
The scientific content of this first passage is usually discussed in terms of
the pre-history of the concept of inertia, rather than, as I have done, in terms
of the concept of acceleration. This applies both to the negatively critical
accounts as well as to the apologetic ones. I will discuss one of each.
By far the most adequate account of the inertial content of Galileo’s work
has been given by Stillman Drake;72 his account constitutes an admirable
synthesis of textual accuracy and substantive scientific relevance. Drake
admits that in this passage Galileo argues explicitly that rectilinear motions
cannot be perpetual, only circular ones can; however, he seems to agree that
this fact provides some support for the view that in the Dialogue the inertial
motions of bodies are circular,73 even though he argues cogently and decisive¬
ly that the other passages commonly given as evidence for the same view do
not in fact support it.74 The way in which Drake criticizes the alleged
scientific import of the present passage is to emphasize its rhetorical content:
Drake says that the passage is merely propaganda and that in it Galileo is
trying to appear to be more Aristotelian than the Peripatetics, or at least that
he begins by accepting some Aristotelian principles so that he can derive un-
Aristotelian conclusions therefrom.75 In short, for Drake “Galileo certainly
did not state or believe that the celestial motions would perpetuate them¬
selves merely by being circular in fonn, but this does not mean that he was
averse from letting the philosophers believe that if they wished to”.76
In criticism of Drake, it may be said that the structure of his argument
seems to be the following: the evidence for Galileo’s belief in ‘circular inertia’
based on other passages can be directly refuted;77 the evidence from the
present passage can be criticized only indirectly by arguing that here Galileo
is acting not as a physicist but as a propagandist; and Drake’s indirect argu¬
ment (1) implicitly admits that the scientific content of this passage is the
erroneous view that inertial motions are circular, and (2) requires that one
attribute to Galileo the intention of wanting his opponents to reach right
conclusions on the basis of wrong reasons. However, to attribute such an
intention to Galileo is somewhat far-fetched; its only supporting argument is
really that only by such an attribution can one vindicate the scientific content
 TRUTH AND METHOD 89

of the present passage; or to be more exact, such an attribution is no better

grounded than the one I gave above centering around the concept of accelera¬
tion, which is the only thing I wish to show in this context. Moreover,
Drake’s implicit admission about the specifically scientific content of the
passage must be rejected because it lends authority to the following fallacious
manner of reasoning: only circular motions can be perpetual, therefore
inertial motions are circular. But there is no reason to accept such an argu¬
ment since it involves an illegitimate equation of perpetual and inertial, or to
be more exact, the equation of natural in the sense of ontologically ever¬
lasting with natural in the sense of logically prior in the conceptual frame¬
work of classical mechanics.78 In other words, everyone must, and does,
admit that this passage contains the idea that rectilinear motions cannot but
circular motions can be perpetual. In my interpretation, I conclude (partly)
from this that therefore the passage contains the idea that accelerated
motions are important. Koyr£ and his followers79 criticized by Drake have
concluded that the passage contains the idea that inertial motions are circular;
Drake seems to conclude this also, but he destroys the point of this con¬
clusion by concluding further that therefore the passage is just propaganda.
In the light of the additional fact that accelerated straight motion is explicitly
discussed in the passage (F43-45), and that acceleration is the topic of
several other immediately following passages, the comparative accuracy of
my interpretation is thereby vindicated.
Let me justify this further by examining the latest of the negatively
critical accounts, namely Dudley Shapere’s.80 Though his book has a number
of very serious faults,81 it does have the merit of discussing explicitly the
scientific content of Galileo’s work and of explicitly recognizing the danger
of reaching simplistic conclusions. So it includes an analysis of the inertial
content of Galileo’s work82 and a criticism both of Drake’s affirmation, and
of Koyrd’s denial, that Galileo understood the ‘essential core’ of the inertial
concept.83 Shapere’s main conclusion here is that Galileo’s relevant views
have some similarities and some differences with the concept of inertia of
classical physics, but that it is arbitrary to regard any one, or any few ones,
of the several features of the concept of inertia as constituting its ‘essence’.84
Part of this claim seems judicious enough and it constitutes a claim about the
substantive scientific relevance of Galileo’s views as interpreted by Shapere.
However, there are other parts to his claim which must be judged in terms of
textual accuracy; these are his interpretative claims, to which we now turn.
Shapere considers Galileo’s inertial views from several of his works. In his
discussion of the Dialogue, he reconstructs the ‘inertial’ content of the latter
90 CHAPTER 4

in terms of five propositions. Of these, only the first four relate to the passage
being considered here. They are: (1) “anything but rest or circular motion is
contrary to the orderliness of the universe” ;85 (2) “natural circular motion is
uniform and perpetual”f6 (3) “the only purpose that might possibly be
attributed to straight-line motion is that it could be used to create or restore
orderliness” ;87 and (4) “if non-circular order-restoring motion does indeed
exist, it would be accelerated and rectilinear”,88
Shapere’s attribution (2) may be accepted, at least as long as it is inter¬
preted to mean “when motion is both natural and circular, then it is uniform
and perpetual”, for the evidence is from passages containing arguments about
cases of natural circular motion.89 This is to be contrasted to interpreting
Shapere’s (2) to mean “natural, i.e. circular, motion is uniform and
perpetual”, which would be textually groundless. But since Shapere means
the former, there is no problem.
The situation is different with his thesis (1). From Shapere’s discussion it
is clear that this thesis has two parts: the ‘anything but’ refers to straight
motion, which is alleged to be contrary to the orderliness of the universe,
while rest and circular motion are alleged not to be contrary to it. From the
same discussion, it is also clear that, for Shapere, Galileo equates “contrary to
the orderliness of the universe” with ‘unnatural’, and ‘not contrary’ with
‘natural’. So Shapere is saying that Galileo believes that motion is unnatural
when straight, and natural when circular. The problem is with the second part
of this thesis, that is with attributing to Galileo a belief in “the universality
of circular motion as natural”.90 In fact, though the text does warrant the
attribution that straight motion is unnatural,91 it warrants only the attribu¬
tion that circular motion can be natural.92 In short, Galileo believes that the
only type of natural motion is circular, that if motion is natural then it is
circular, not that if it is circular then it is natural; and Shapere has simply
misinterpreted “not contrary to the orderliness of the universe” as “in
accordance with the actual order of the universe”.
Shapere’s proposition (3) may be accepted, but his (4) is not accurate. His
evidence is the passage between the first and the second reference to the
allegedly ‘Platonic’ creation theory.93 Here he distinguishes two arguments,
a general one94 and a specific one dealing with the continuity of acceleration
of falling bodies.95 Shapere’s proposition (4) is a conditional, and he tries to
interpret his evidence as two arguments where the consequent clause
(acceleration and rectilinearity) is derived from the antecedent clause (non¬
circular order-restoring motion). That is, for Shapere, the passage he uses is
one where the main topic is non-circular order-restoring motion, and the
 TRUTH AND METHOD 91

claim being made is that this motion would be accelerated and straight.
However, it is really the latter that is the main topic, and the claim being
made is that such motion would not be everlasting. In other words, Galileo’s
point is not that if there is non-circular order-restoring motion then it would
be accelerated and rectilinear but rather that if there is accelerated rectilinear
motion then it is not perpetual. This is so because the straight accelerated
motion must have a source-origin of the force field (“luogo desiderato, cioe
dove 1’ inclinazione naturale lo tira”96) so that when this place is reached the
accelerated motion stops or there is a change to circular uniform motion, as
in a body rolling down an inclined plane and then moving uniformly along
the horizontal.
The accuracy of Shapere’s interpretation is prima facie suspect when the
reader notices that he supports his attribution of proposition (4) to Galileo
by attributing to him arguments that are described by Shapere as being
fallacious in several important ways. This is initially suspicious in the sense
that if one cannot interpret some passages as providing actual argumentative
support for a certain conclusion, then how is one going to support one’s
interpretative claim that the author held that conclusion? Even if one could
find an explicit assertion by the author stating that a certain proposition is
his conclusion, why should one attach more weight to this evidence than to
the evidence of the actual content of the passage? It would seem that the
actual content of a passage is what is really contained therein, not what the
author may have said that it contains. At any rate, since in this passage
Galileo nowhere makes the needed explicit assertion, there would seem to be
no reason at all for attributing to him Shapere’s proposition (4).
In fact, when one examines Shapere’s evidence more closely one finds that
it is based on several textual misunderstandings. For example, in his discus¬
sion of Galileo’s general argument97 Shapere claims that

the argument does not prove that a body moving in natural straight-line motion will
continually accelerate; all it shows (assuming the Principle of Continuity, of course) is
that, if the body is to arrive at any particular finite non-zero velocity, it will have to pass
successively through all lesser velocities; it does not show that when any given speed is
reached the body will continue to increase (rather than maintain or decrease) its speed
beyond that until it arrives at its goal.

The passage Shapere refers to is the following:

Every body constituted in a state of rest but naturally capable of motion will move when
set at liberty only if it has a natural tendency toward some particular place; for if it were
indifferent to all places it would remain at rest, having no more cause to move one way
92 CHAPTER 4

than another. Having such a tendency, it naturally follows that in its motion it will be
continually accelerating. Beginning with the slowest motion, it will never acquire any
degree of speed (velocita) without first having passed through all the gradations of lesser
speed - or should I say of greater slowness? For, leaving a state of rest, which is the
infinite degree of slowness, there is no way whatever for it to enter into a definite degree
of speed before having entered into a lesser, and another still less before that. It seems
much more reasonable for it to pass first through those degrees nearest to that from
99
which it set out, and from this to those farther on.

Shapere apparently has taken the phrase ‘continually accelerating’ to mean

‘everlastingly accelerating’ rather than ‘nondiscontinuously accelerating’.
Though in the abstract these are possible meanings of the adverb ‘continually’,
in the context in which the above quoted argument occurs100 Galileo’s main
claim is that straight accelerated motion is both nondiscontinuous and non-
perpetual, but rather transitory and instrumental for achieving a given ‘deter¬
minate velocity’101 in uniform, circular motion which alone can last forever.
Moreover, if one checks the original Italian text, one finds that the phrase
used is icontinuamente accelerando’,102 which is much less ambiguous than
‘continually accelerating’ and closer to ‘continuously accelerating’.

METHODOLOGY

On the basis of the critical comparison just given, and of others that could be
made for the other passages, I feel that the accuracy of my account of the
scientific content of the First Day is vindicated. As I suggested earlier, what I
am dissatisfied with is its relevance. Let us examine in turn the various
passages as I have reconstructed them. The first (F23—45)103 stresses the
importance of acceleration; however, though it is indeed a fact of contem¬
porary science that acceleration is a phenomenon and concept of crucial
importance, it is clear that the Galilean justification given in the passage is
of merely historical interest; that is not the way a contemporary scientist
would justify its importance. Similar remarks apply to the second (F45—7,
51—3) and third passages (F47—51): though it is true that acceleration is con¬
tinuous, and that it must be distinguished from speed, the contemporary
scientist would not appreciate the justifications given by Galileo. It is different
with the next passage (F53—4) dealing with the acceleration problem of
whether it is possible to originate the solar system by letting the planets fall
toward the sun and then changing their accelerated, straight motion into a
uniform, circular one: the problem is one that would still interest a contem¬
porary scientist, at least in the sense that he would regard it as a useful
 TRUTH AND METHOD 93

exercise for beginning students; moreover, Galileo’s method of solution

would be regarded as acceptable, but several important details of the sub¬
stance of his solution would be found incorrect. The next two passages, on
the isotropy of space (F54—61) and on the primacy of mechanical properties
(F61—75), would be found as being correct in their main claim, but un¬
interesting as regards the justification of those claims. For the next passage on
the telescope and sunspots (F75—87), a scientist would agree with the
historical significance of this instrument and of this phenomenon, and he
could even appreciate Galileo’s justification of this significance; however, for
him nowadays, though telescopes are still used, and though sunspots are still
studied, their present significance is not as great as what it was in the seven¬
teenth century. The next passage (F87—95) emphasizes the fact that slightly
more than 50% of the lunar surface is visible from the earth; I believe that
both the argument and the conclusion would be of some interest to a contem¬
porary scientist. Similar remarks apply to the next two passages on the
moon’s mountainousness (F95—112) and on its secondary light (FI 12—22).
The problem of moon spots (F122—31) would be regarded by a scientist as
being exactly the right type of problem studied in science; Galileo’s remarks
on the solution would be appreciated more for their tentativeness and
judiciousness, than for the correctness of their details.
In summary, the scientific content of the First Day consists of (1) the
at least implicit assertion of a number of facts that would be still accepted
as such and used by a contemporary scientist, (2) the explicit formulation of
a number of problems recognizably scientific in their character due to their
subject matter or the concepts involved, (3) the basic correctness and scien¬
tific character of a number of arguments given in support of scientifically
accepted facts, and (4) the basic correctness and scientific character of a
number of methods or approaches used or suggested for the solution of
recognizable scientific problems. It should be noted that not all of these four
kinds of scientific content are present in each passage, but rather that each
passage contains one or more of them. Of course, it would be no problem
for the textual analyst to point out portions within most of these eleven
passages which are devoid of scientific content, but this fact may be judged as
irrelevant. The important fact is that every plausibly self-contained discussion
in that First Day has scientific content. This is important because it helps to
explain the persistent appeal that the Dialogue has for scientists. For what is
true of the First Day is even more so of the other three: we have already seen
above that the allegedly ill-fated Fourth Day addresses itself to scientific
problems that may not be completely resolved even today; the wealth of
94 CHAPTER 4

scientific content of the Second Day is usually admitted; the Third Day is
equally rich when examined from the point of view of the theory of parallax
and the practice of the evaluation of parallax measurements.104
Generalizing, we may say that the scientific content of the Dialogue con¬
sists of the specific facts, problems, arguments, and methods mentioned
therein which have some substantive correspondence with the facts, prob¬
lems, arguments, and methods found in the latest science textbooks or con¬
temporary scientific practice. This has interesting implications for the theory
of science and the question, What is science? For what it means is that science
is a collection of facts, problems, arguments, and methods which have sub¬
stantive inter-historical similarities. To say this involves denying that science
is a purely formal enterprise in the sense that it would be possible for a
theory to be scientific if it was held for valid (though factually incorrect, or
somewhat incomplete) reasons or had been arrived at in a proper manner.
This is a very plausible and widespread view; it corresponds to what Joseph
Agassi calls ‘conventionalism’,105 and one version is Thomas Kuhn’s philo¬
sophy of science;106 nevertheless, it does not receive support from our
examination of Galileo’s Dialogue coupled with scientists’ reaction to this
book.
Another consequence of our analysis is a denial of the view that the
essence of science is its factual truth-claims, so that those people and theories
are scientific whose claims happen to correspond (perhaps approximately) to
the substantive content of present-day scientific knowledge. This is the
philosophy of science that Agassi calls ‘inductivism’.107 According to this
view, what matters most is that one’s results be right, independently of the
rationale or methods underlying them. Against this view our analysis suggests
that truth-content must be combined with proper method and/or rationale.
Third, this study of the scientific content of Galileo’s Dialogue suggests
certain refinements to the distinction between the context of discovery and
the context of justification. I have discussed elsewhere108 the contextual
though nonarbitrary character of this distinction. What needs to be noted
here is that it would be a mistake to regard a book as belonging solely to the
context of justification. For it has been shown above that, besides the
justifications found in a book, what can make an impression on scientific
judgment is its problems or its mere claims. Moreover, even when the context
is obviously one of justification, the most striking aspect of such a justifica¬
tion may be the fact that certain ideas were formulated in the course of the
justification. In other words, in the course of proving a proposition p one
may find himself formulating some other proposition q such that though p is
 TRUTH AND METHOD 95

false, and the proof incorrect, nevertheless the conception of q is scientifically

significant. Or to give a concrete application, suppose that the whole Dialogue
is categorized as an attempt to prove Copernicanism, and hence placed in the
context of justification; in the course of such an attempted proof one may
find himself formulating the principle of mechanical relativity, or of con¬
servation of motion. Then the same book constitutes context of discovery,
from the point of view of these principles. Or, suppose that we regard the
Fourth Day as a justification of the proposition that the earth’s (double)
motion is the (primary) cause of the tides; even if one regards such a pro¬
position false, and the justification fallacious, one can be struck by the dis¬
covery inherent therein of the necessary role of secondary factors relating to
the natural period of the local sea basin and to the differential effects of
water depth. In other words, from the point of view of the distinction
between the context of discovery and the context of justification, a book like
the Dialogue belongs at least as much in the former as in the latter. This is
so because the book though not unconcerned with justification, is full of that
fluidity and plasticity of thought that characterizes the context of discovery.
Having said this much, my dissatisfaction with the scientific relevance of
the scientific content so far considered lingers on. It is not the case that such
substantive scientific content is not relevant, but rather that there is another
point of view from which the book’s scientific relevance is so much greater
that one sees little point in continuing to elaborate this first more direct kind
of scientific content; that point of view is that of methodology. That is, the
book is full of methodological discussions about scientific inquiry. Such
methodological content might be regarded as a second, indirect, higher-level
kind of scientific content. So far we have examined only the book’s scientific
content at the level of substantive scientific concepts, problems, and ideas.
However, scientists are even more interested in extracting methodological
lessons about the proper scientific procedure and the nature of scientific
knowledge. In fact, as I shall soon show, even when a scientist refers to the
substantive scientific content, he usually does so for the purpose of making a
methodological point. Since the book’s methodological content would be of
interest to philosophers as well as scientists, it seems advisable to concentrate
on it. In fact, since discussions about the nature of knowledge and the means
for acquiring it are also in the domain of what is ordinarily called ‘philo¬
sophy’, we may also say that the book has great philosophical content and
that this content is its methodology. Thus, the book’s true (i.e. more
important) scientific content is its methodological content, which coincides
with its (only type of) philosophical content. Let us not forget, however, that
96 CHAPTER 4

it is a very important fact that the book possesses great scientific content of
the first kind. For it is this kind of content that makes it possible to regard
its methodological content as dealing with scientific methodology. Before
examining the scientific-methodological-philosophical content (next chapter),
let us however show that even for a scientist the substantive scientific content
often has primarily a methodological significance. I shall take the examples
of Newton and Einstein with which this chapter began.
In the Principia, Newton’s reference to Galileo occurs in the scholium that
follows the statement of the three laws of motion and six corollaries;109 the
scholium begins by saying, “Hitherto I have laid down such principles as have
been received by mathematicians, and are confirmed by abundance of experi¬
ments. By the first two Laws and the first two Corollaries, Galileo discovered
. .. ,”110 The section referring to Galileo ends as follows: “ ... On the same
Laws and Corollaries depend those things which have been demonstrated con¬
cerning the times of the vibration of pendulums, and are confirmed by the
daily experiments of pendulum clocks”.111 Then Newton goes on to say, “By
the same, together with Law III, Sir Christofer Wren, Dr. Wallis, and Mr.
Huygens, the greatest geometers of our times, did severally determine . . . .”112
The first two laws are the law of inertia and the law of force; the first corollary
is the principle of the composition of velocities, the second is the principle of
the composition and resolution of forces. I suggest that the function of
Newton’s reference to Galileo is to serve as a methodological argument in
support of the first two laws and corollaries. The structure of Newton’s argu¬
ment is the following: In his work on the law of squares, the parabolic
trajectory of projectiles, and the period of vibration of pendulums, Galileo
used and accepted the first two laws and corollaries. Therefore, in dealing
with the problems discussed in this book, we should accept the first two laws
and corollaries. The argument, of course, presupposes two things: that Galileo
is an appropriate scientific model, and that the situation he was in is appro¬
priately similar to Newton’s. That is, the argument is simultaneously an argu¬
ment from authority and from analogy. The truth of Newton’s premise here
is the sort of thing that historians have disputed. At any rate, there is no
question that historical investigations can be of great service to a scientist for
the establishment of such propositions on which such arguments can be
grounded. The plausibility of the inference of course depends on the two
presuppositions. In Newton’s time the appropriateness of using Galileo as a
scientific model may have been a matter of some controversy, nowadays it
no longer is; whereas the appropriateness of the analogy appealed to by
Newton neither was nor is disputable, since the two situations involve
 TRUTH AND METHOD 97

obviously similar problems about the motion of bodies. However, for other
such arguments that a scientist may want to give, the analogy between the
two situations may be subject to question.
Turning now to Einstein’s Foreword to Drake’s translation of the
Dialogue, Einstein’s argument quoted above113 may be reconstructed as
follows:
(El) In the Dialogue, Galileo rejected the hypothesis of a center of the
universe for the explanation of the fall of heavy bodies.
(E2) The hypothesis of an inertial system for the explanation of the
inertial behavior of matter is analogous to the one being rejected in Galileo’s
Dialogue (because both hypotheses introduce a conceptual object which (1)
does not have the same kind of reality that matter or fields do, and (2)
affects, without being affected by, the behavior of real objects).
(E3) Therefore, this hypothesis about an inertial system should be re¬
jected (insofar as it is as unscientific as the one rejected by Galileo).
This argument is very explicit in what Einstein says, in fact he even gives a
subargument to justify the appropriateness of the analogy (E2). He takes for
granted the propriety of Galileo as a model, and his historical claim (El)
seems to contain no obvious textual inaccuracies.
Such methodological arguments are common in scientific research, and
they have an obvious ‘relevance’. In examining the methodological content of
a book like the Dialogue one cannot, however, provide specific instances of
such methodological arguments; but one can provide the groundwork for
them. That is, one can describe or reconstruct the various scientific discus¬
sions and simultaneously formulate in general (though judiciously qualified)
terms a methodological claim which each scientific situation illustrates. The
scientist can then use such a series of methodological reconstructions by
determining which one (or which ones) of them applies (or apply) to the
situation or problem he himself faces. In such a determination he can be
assisted by deciding whether his own case is an instance of the general claim
formulated in the methodologist’s reconstruction. A book like the Dialogue
lends itself very well to this kind of methodological analysis because it is full
of explicit methodological and epistemological remarks; hence we have two
ways of checking the reliability of our reconstructions, namely whether they
do justice to Galileo’s explicit methodological remarks, and whether they do
justice to the concrete scientific discussions in the course of which such
remarks occur. Such a methodological reconstruction for everything discussed
in the Dialogue is provided in the next chapter.
98 CHAPTER 4

NOTES

1 E Newton, Mathematical Principles of Natural Philosophy, pp. 21-22.

2 E B. Cohen, ‘Newton’s Attribution of the First Two Laws of Motion to Galileo ,
p. XXVI.
5 Ibid., p. XXVIII.
4 A. Einstein, ‘Foreword’, in G. Galilei, Dialogue Concerning the Two Chief World
Systems, tr. Drake, pp. xi-xiii.
5 Ibid., p. xiii.
6 Quoted in S. M. Uzdilek, ‘Galileo Galilei, The Founder of Experimental Philosophy
and . .. ’, p. 230.
7 B. Kouznetsov, ‘Galilee et Einstein. Prologue et Epilogue de la Science Classique’,
p. 63.
8 S. Drake, editor and tr., Galileo Against the Philosophers, p. xi: “since it can now be
shown beyond reasonable doubt that Galileo’s important work on terrestrial physics was
essentially completed before he ever applied the telescope to astronomy, it would follow
that . . . not an unreasonable zeal for a system incapable of scientific demonstration at
the time, but patient study of observational data, first in physics and then in astronomy,
led Galileo to effect his synthesis of a new physics and the new astronomy of his famous
Dialogue of 1632.” For the exhaustive evidence, see Drake’s Galileo at Work.
9 W. L. Wisan, The New Science of Motion: A Study of Galileo’s De motu locali', p. 298.
10 A. Koyrd, Etudes Galil&ennes, p. 212.
11
E. McMullin, ‘Introduction: Galileo, Man of Science’, p. 3 .
12
Ibid.
13
Ibid., pp. 35- -42.
14
Ibid., p. 35.
15
Ibid., p. 43.
16
R. P. Feynman, R. B. Leighton, and M. Sands, The Feynman Lectures on Physics,
Vol. I, p. 14-1.
17
McMullin, ‘Galileo, Man of Science’, pp. 31-35.
18
Ibid.
19
Ibid., pp. 35-42.
20
A. Koestler, The Sleepwalkers, pp. 473-479.
21
P. Feyerabend, Against Method.
22
D. Shapere, Galileo: A Philosophical Study, p. 105.
23
W. R. Shea, Galileo’s Intellectual Revolution, p. 117.
24
G. de Santillana, ‘Introduction’, in G. Galilei, Dialogue on the Great World Systems,
tr. Salusbury-Santillana, p. XXX; cf. p. 349, n. 34. See also his Crime of Galileo, pp.
174-182.
25 For recent statements see C. Perelman and L. Olbrechts-Tyteca, The New Rhetoric;
also Benedetto Croce, Philosophy of the Practical; idem, La Poesia; also M. Natanson
and H. W. Johnstone, Jr. (Eds.), Philosophy, Rhetoric, and Argumentation.
26 The classic locus of such an approach is the works of Pierre Duhem. The best recent
elaboration, partly grounded on new textual evidence, is the work of W. A. Wallace, such
as ‘Mechanics from Bradwardine to Galileo’; ‘Galileo and Reasoning Ex Suppositione’;
and ‘Galileo and the Doctores Parisienses'.
 TRUTH AND METHOD 99

27
See, for example, S. Drake’s ‘Galileo’s Discovery of the Law of Free Fall’; ‘Galileo’s
Experimental Confirmation of Horizontal Inertia: Unpublished Manuscripts’; ‘The
Uniform Motion Equivalent to a Uniformly Accelerated Motion from Rest’; ‘Galileo’s
New Science of Motion’; Galileo Studies, and Galileo at Work.
28
See, for example, R. J. Seeger, Men of Physics: Galileo Galilei, His Life and Works,
pp. 36-37; and idem, ‘On the Role of Galileo in Physics’, p. 36.
29 Examples of attempts to examine the post-classical physics scientific content of
Galileo’s work are B. G. Kouznetsov’s ‘L’id^e d’homog^ndit^ de l’espace dans le Dialogo
de Galilee’; ‘Galilde et Einstein. Prologue et Epilogue de la Science Classique’; ‘Style et
Contenu de la Science’; and ‘Le soleil comme centre du monde, et l’homog^ndit^ de
l’espace chez Galilde’. Also, B. M. Kedrov and B. H. Kouznetsov, ‘La logique de Galilde
et la logique de la physique actuelle’. Also E. Agazzi, ‘Fisica galileiana e fisica contem-
poranea’. The passage on semicircular fall in the Dialogue (pp. 162-67 in Drake’s trans¬
lation) is a good example of how concern with classical physics can blind one to its
content from the point of view of modern physics. One can very easily read into it the
kind of force-free geometrical physics that is characteristic of kinematical general
relativity, whereas it is commonly taken to indicate that Galileo’s view of ‘inertia’ did
not coincide with Newton’s.
30 For example, at the beginning of the First Day Galileo suggests that the solar system
may have originated by letting the planets fall with uniform acceleration from their place
of creation toward the sun until they reached their respective orbits, at which time and
place the motion they had acquired must have been changed from straight and
accelerated to circular and uniform. In this case there is no conflict between classical and
modem physics, and the suggestion is in some sense wrong. However, in trying to under¬
stand what Galileo has in mind in this passage, Newton-minded scholars uncritically use
Kepler’s third law in reconstructing Galileo’s reasoning. For example, in ‘Galileo,
Newton, and the divine order of the solar system’, I. B. Cohen considers no fewer than
five different reconstructions each of which, however, presupposes Kepler’s third law
(pp. 212-218). I believe Cohen’s justification speaks for itself: “In this presentation I
have, of course, introduced Kepler’s third or harmonic law, which seems not to have
been known to Galileo. This anachronistic procedure is wholly justifiable, however, since
the third law is a quite accurate representation of the relation among planetary data as
known to Galileo and other Copernicans” (p. 213)! The same approach is taken by S.
Nakayama, in ‘Galileo and Newton’s Problem of World-Formation’. By contrast, in
‘Galileo’s “Platonic Cosmogony” and Kepler’s Prodromus', S. Drake, with more
adequacy, interprets Galileo’s cosmogony as actually involving a groping toward Kepler’s
law (pp. 185-187); Drake’s article is primarily an analysis of previously unavailable
documents. However, there is no documentary excuse for the Newtonian readings of the
passage in the Dialogue since my reconstruction below shows that by merely considering
this passage it is possible to give an interpretation which may be said to correspond to
the tip of the iceberg uncovered by Drake.
31 See Chapter 9 for a critique of Koyrd, and Chapter 10 for the others.
32 In my History of Science as Explanation, especially pp. 131-141, and 229-238.
Cf. J. Agassi, Towards an Historiography of Science, esp. pp. 1-28. I might add that
what Agassi calls ‘conventionalism’ (pp. 28-48) is analogous to the historical attitude as
I have just described it; cf. my History of Science as Explanation, pp. 131-141.
33 E. Mach, The Science of Mechanics, p. 263.
100 CHAPTER 4

34
Ibid., p. 264.
35 V. Nobile, ‘Sull’ argomento galileiano della quarta giornata dei Dialoghi e sue
attinenze col problema fondamentale della geodesia’, especially p. 432.
36
H. L. Burstyn, ‘Galileo’s Attempt to Prove that the Earth Moves’.
37
Ibid., p. 182.
38
Ibid., p. 164.
39
Ibid., p. 167.
40
E. J. Aiton, ‘On Galileo and the Earth-Moon System’, pp. 265-266.
41 H. L. Burstyn, ‘Galileo and the Earth-Moon System: Reply to Dr. Aiton’, p. 401.
42
H. L. Burstyn, ‘Galileo’s Attempt to Prove that the Earth Moves’, pp. 168, 181.
43
Ibid., p. 174.
44
E. J. Aiton, ‘Galileo and the Theory of the Tides: Comments by E. J. Aiton’.
45
H. L. Burstyn, ‘Galileo and the Theory of the Tides: Reply by Harold L. Burstyn’.
46
G. Galilei, Dialogue Concerning the Two Chief World Systems, tr. S. Drake, p. 427.
Cf. Galilei, edited by A. Favaro, Opere 7, 453.
47 E. J. Aiton, ‘Galileo’s Theory of the Tides’, pp. 5 7—58.
48 Cf. H. L. Burstyn, ‘Galileo’s Attempt to Prove that the Earth Moves’, p. 169.
49 W. R. Shea, Galileo’s Intellectual Revolution. About one-half of the book deals with
the Dialogue. More recently, a more adequate and original account of the tidal argument
has been given by H. I. Brown’s ‘Galileo, the Elements, and the Tides’.
so
Shea, Galileo’s Intellectual Revolution, p. 184.
51
Ibid., p. vii.
52
Ibid.
53
Ibid.
54
Ibid., p. 185.
55
Ibid.
56
Ibid., p. 186.
57
Ibid., p. 175.
58
Ibid.
59 E. J. Aiton, ‘Galileo’s Theory of the Tides’, p. 46.
60
A. Koestler, The Sleepwalkers, pp. 464-466.
61
Shea refers to E. Mach, The Science of Mechanics, pp. 262-264, which corresponds
to Chapter II, Part IV, Section II of the ninth German edition.
62
Mach, Science of Mechanics, p. 263.
63
Ibid.
64
Ibid., p. 264.
65
Shea, Galileo’s Intellectual Revolution, pp. 175-176. The letter is the one of
September 5, 1633 found in G. Galilei, edited by Favaro, Opere 15, 251-252.
66
Shea, p. 176.
67
Ibid., pp. 177-178.
68
G. Galilei, Dialogue, tr. Drake, pp. 428-436; idem, edited by Favaro, Opere 7,
457-462.
69 Page numbers within the text of my interpretations refer only to Vol. 7 of Favaro’s
edition of Galileo’s Opere, whether or not the numerals are prefixed by ‘F’, which is
short for ‘Favaro’. References for the main subdivisions are given immediately after my
chosen titles for these sections; note that for these section titles two sets of page
numbers are given: the ones denoted by ‘F’ refer to Galileo’s Opere, Vol. 7, and contain
 TRUTH AND METHOD 101

line references as well; the ones denoted by ‘D’ refer to Drake’s translation. Thus, for
example, the references to the third section (on acceleration vs speed) is ‘F47.10—51.18;
D23—27’ and means that the section goes from p. 47, line 10 to p. 51, line 18 in
Favaro’s edition, and from p. 23 to p. 27 in Drake’s translation.
70
The figure and the computation are not, of course, explicitly in the text.
71
The figures and the arguments are, of course, not explicitly in the text.
72
S. Drake, ‘Galileo and the Law of Inertia’; idem, ‘Semi-circular Fall in the Dialogue'-,
idem, Galileo Studies, pp. 240—278.
73
Drake, Galileo Studies, pp. 252-53.
74
Ibid., pp. 253, 257-278.
75
Ibid., pp. 253-254.
76
Ibid., p. 254.
77
Ibid., pp. 256-278.
78
My view here is similar to the one held by Libero Sosio when he stresses the
importance of distinguishing between ‘cosmological-architectonic’ and dynamical con¬
siderations; see ‘Galileo e la cosmologia’ in G. Galilei, Dialogo sopra i due massimi
sistemi, edited by L. Sosio, pp. IX-LXXXVII, especially pp. L-L1;J. A. Coffa is making
essentially the same point when he states that “the heavy task facing the circularist
trying to rest his case on the present passage is to show that it has inertial implications at
all”; see his ‘Galileo’s Concept of Inertia’, p. 281.
79 A. Koyr£, Etudes GaliUennes, pp. 161-341, esp. 205—211. W. R. Shea, Galileo’s
Intellectual Revolution, pp. 116-121. Though Shea does not really discuss the scientific
content of the passage, but rather approaches it from the point of view of the history of
Platonist anti-Aristotelianism, he does not give up the opportunity of stating in the
course of his discussion that “the triumph of circularity is complete, but it is purchased
at the price of excluding the possibility of rectilinear inertia” (p. 118).
80
D. Shapere, Galileo, pp. 87-125, esp. pp. 87-98.
81
See M. S. Mahoney, ‘Galileo’s Thought’; C. S. Schmitt, ‘Review of Shapere’s Galileo'-,
P. Str^mholm, ‘Galileo and the Scientific Revolution’; and my ‘Philosophizing About
Galileo’. Strtfmholm goes so far as saying that Shapere’s book deserves ‘condemnation’.
My own analysis, while severely critical is simultaneously an explanation and implicitly
a defense from certain kinds of criticism.
82 Shapere, Galileo, pp. 87-121.
83
Ibid., pp. 121-125.
84
Ibid.
85
Ibid., p. 87; italics in original,
86
Ibid., p. 89; italics in original,
87
Ibid., p. 91; italics in the original,
88
Ibid., p. 93; italics in the original.
89
Galilei, Dialogue, tr. Drake, pp. 31-32.
90
Shapere, Galileo, p. 89; italics in the original.
91
Galilei, Dialogue, tr. Drake, pp. 19-21, and 31-32.
92
Ibid., pp. 31-32.
93
Galilei, Dialogue, tr. Drake, pp. 20-28.
94 Shapere, Galileo, pp. 93-95; cf. Galilei,Dialogue, tr. Drake, pp. 20-21.
95
Shapere, pp. 95-97; Dialogue, tr. Drake, pp. 22-28.
96
Galilei, edited by Favaro, Opere, 7, 44.
102 CHAPTER 4

97 Dialogue, tr. Drake, pp. 20—21.

98 Shapere, p. 94.
99 Shapere, pp. 93, 94; Dialogue, tr. Drake, p. 20.
100 Dialogue, tr. Drake, pp. 19-21.
101 Galilei, edited by Favaro, Opere, 7,45; the Italian phrase is determinata velocita.
102 Ibid., p. 44.
103 In acdordance with the conventions of footnote 69 above, references consisting of
page number preceded by ‘F’ denote pages from Galilei, edited by A. Favaro, Opere, 7.
104 Concerning the central section of the Third Day (F369ff.) W. Hartner says that
“this long and very clear discussion of parallax is a masterpiece”. See his ‘Galileo’s
Contribution to Astronomy’, p. 190.
105 J. Agassi, Towards an Historiography of Science, pp. 1-28; cf. my History of
Science as Explanation, p. 131—141.
106 T. S. Kuhn, The Structure of Scientific Revolutions.
107 Agassi, Historiography; cf. my History of Science as Explanation, pp. 131-141.
108 See my History of Science as Explanation, pp. 229-238.
109 I. Newton, Mathematical Principles of Natural Philosophy, pp. 21-22. I have
suggested above that since Newton almost certainly had read the Dialogue but not the
Two New Sciences, he may be taken to have ‘read’ these things into the former.
110 Newton, p. 21.
111 Ibid., p. 22.
112 Ibid.
113 Galilei, Dialogue, tr. Drake, p. xiii.
 CHAPTER 5

THEORY AND PRACTICE:

THE METHODOLOGICAL CONTENT OF
GALILEO’S SCIENCE

Galileo’s Dialogue is full of explicit remarks about topics such as: (1) the role
played in physical inquiry by experience, authority, mathematics, logical
analysis, conceptualization, human interests and capacities, causal investiga¬
tion, explanations, comprehensibility, criticism, simplicity, quantitative and
qualitative considerations, open-mindedness, and ignorance; (2) the nature of,
and interrelationships among, these things and others, such as the Socratic
method, unconscious knowledge, and astronomical investigation; and (3) the
distinction between method and results, between discovery and justification,
and between theory and practice. Remarks about such matters are universally
regarded to be methodological remarks, and when a book contains them with
the frequency and intensity that the Dialogue does, one has to conclude that
such a book is a work on method, whatever its title may be and whatever
other dimensions it may possess.
How is one to understand the methodological content of the Dialogue? A
systematic attempt to understand must be distinguished from the attempt to
systematize this methodological content. The former is prior to the latter
since systematization is a refinement in understanding, and before a refine¬
ment is possible one must have a basic understanding of the fundamentals.
Moreover, in order to systematize a set of methodological claims one must
first be in possession of them; hence their detection and extraction is prior.
Finally, just as it is obvious that the Dialogue contains a wealth of methodo¬
logical remarks, it is equally obvious that no systematization of them is
present in the text; hence such a systematization would not be the explana¬
tion of the methodological content present therein, but rather the theoretical
elaboration of it.
A systematic attempt to understand the book’s methodological content
should be first an attempt to understand every methodological remark made
therein; second one should try to understand the remark in the context in
which it is made. It seems redundant to state such a requirement except that
the context is almost always a nonphilosophical but a scientific one, namely
the discussion of a concrete physical problem, and because of this the context
has seldom been taken into account by interpreters. Such a contextual
analysis has the double advantage that sometimes one can use a very explicit

103
104 CHAPTER 5

methodological remark to throw light on a concrete scientific discussion or

series of discussions, and sometimes one can use the clarity of a particular
scientific discussion to make sense of a number of relatively unexplicit, but
present, methodological remarks.
Next, a systematic analysis of methodological content must not be con¬
fused with the analysis of the philosophical significance. Such significance can
be present without methodological awareness, e.g., in the work and accom¬
plishment of such people as Jesus or Euclid, or in the occurrence of such
historical events as the French or industrial revolution. In other words, by
methodological content I do not mean the philosophy implicit in the book,
but rather the philosophy explicitly discussed in it, that is, the philosophy
which is both more or less explicitly stated and illustrated with examples of
scientific discussions. In speaking of the greater or lesser degree of explicit¬
ness of statement, I am admitting that Galileo’s methodological remarks are
not always equally explicit; I would also admit the existence of borderline
cases difficult to decide. Nevertheless, an analysis of methodological content
should always have some evidence from Galileo’s words as a reason for
attributing to him a methodological claim.
To explain further this distinction between methodological content and
philosophical significance, i.e., between explicit and implicit philosophy, I
will call attention to one feature of the investigation undertaken below. This
investigation consists of an analysis of the entire Dialogue into 55 contiguous
passages each of which can be reconstructed as the illustration by scientific
example(s) of some methodological statement(s). Each passage is evidence
that Galileo achieved a very judicious combination of scientific practice and
philosophical reflection. Such a science-philosophy synthesis is, to repeat,
substantiated 55 times in the book, every section of which can be integrated
into it. Thus we have overwhelming evidence of the implicit presence of the
philosophy that one of the highest values is to synthesize theory and practice,
that is, to enlighten one’s experience and to act on one’s thoughts. Such a
philosophy is a central doctrine of both Socrates’s moral philosophy and of
Benedetto Croce’s purified pragmatism. This means that it helps us to under¬
stand Galileo’s Dialogue by regarding it partly Socratic and partly Crocean.
However, in Galileo such a philosophy is totally unexplicit; it would be com¬
pletely wrong to attribute to him such a doctrine. Unlike this doctrine, all
the methodological theses attributed to him in the analysis below are held by
him with some awareness. This is the sense in which this analysis provides the
methodological content of the Dialogue.
 THEORY AND PRACTICE 105

Outline

Pages1 Methodological Topic

FIRST DAY
33-8/9-14 Mathematical reasoning and number mysticism
38-57/14-32 Conceptual frameworks
57-9/32-5 Sense experience
59-71/35-47 Aristotle’s authority as a logician
71-82/47-57 The role of authority in natural philosophy
82-95/57-70 Teleology and anthropocentrism
95-112/71-87 The necessity of experiments
112-7/87-91 Cognitive awareness
117-24/91-98 Different categories of error
124-31/98-105 The powers of human understanding
SECOND DAY
132-9/106-14 Independent-mindedness and the misuse of authority
139-50/114-24 Simplicity and probability
150-9/124-33 Open-mindedness and rational-mindedness
159-69/133-43 Logical analysis
169-80/143-54 Reasoning vs experience, and the Socratic method
180-8/154-62 The Socratic method: understanding vs information-
knowledge
188-93/162-7 Fanciful vs relevant digressions
193-6/167-71 Unconscious knowledge
196-206/171-180 Hypothetical and circular reasoning
206-9/180-3 Critical experimentalism
209-14/183-8 The proper time for experiments
214-7/188-90 Criticizing and understanding
217-23/190-7 Knowledge and recollection
223-9/197-203 Geometry and philosophy
229-37/203-10 Mathematics and physical reality
237-44/210-8 Causal investigations
244-60/218-33 Physico-mathematical synthesis
260-72/233-47 Comprehensibility and truth
272-81/247-56 The deception of the senses
281-8/256-64 Inconceivability-claims and facts
289- 90/264-6 The primacy of nature, over man
290- 3/266-9 Logical distinctions
293-8/269-75 The superficiality of abstract answers
106 CHAPTER 5

THIRD DAY
299-337/276-309 Quantitative data and qualitative conclusions in astronomy
337-46/309-18 The reliability of astronomical measurements
346-9/318-21 Procedure vs results, and Aristotle’s authority
349-68/321-40 Simplicity, insight, and sense experience
368-72/340-5 Computational vs philosophical astronomy
372-83/345-56 Logic vs methodology in theoretical explanation
383-5/386-8 The Bible
385-99/358-71 Objectivity and the concept of size
399-406/371-9 Ad hominem arguments
406-9/379-83 The value of criticism
409-14/383-7 Theories and observations
414-16/387-9 The inexactness of astronomical instruments
416-25/389-99 The complexity of simplicity
425-31/399-405 Open-mindedness, intellectual cowardice, and curiosity
431-6/405-10 Causes: role, discovery, and justification
436-41/410-15 Parts vs wholes
FOURTH DAY
442-5/416-9 Choosing the effect in a causal investigation
445-62/419-36 The artificial reproduction of cause and effect
462-70/436-44 Reasoning ex suppositione
470- 1/444-5 Confessed ignorance and occult qualities
471- 84/445-60 The method of concomitant variations
484-9/460-5 Searching vs results: the supremacy of method

METHODOLOGICAL ANALYSIS AND RECONSTRUCTION2

33—8: Mathematical reasoning and number mysticism

Mathematical reasoning should not be confused with number mysticism any
more than the latter should be confused with Pythagorean or Platonic mathe-
maticism; hence mathematical reasoning should not be excluded in principle
from the study of nature. An example of number mysticism is the Aristotelian
justification of the three-dimensionality of the world based on the perfection
of the number three. An example of genuine mathematical reasoning is the
proof that the world has three dimensions because there can be at most three
mutually perpendicular lines through a given point. The intuitive clarity
(F36) of the latter example shows that it is not impossible to use mathe¬
matical arguments in natural philosophy.
 THEORY AND PRACTICE 107

38—57: Conceptual frameworks

Conceptual frameworks fprecetti d’architettura' [F40, F42—3]) should be

formulated independently of empirical considerations. Aristotle’s theory of
simple (natural) motion (F38—43) is a conceptual framework because it is the
conceptual foundation of the earth-heaven distinction. It depends on
empirical considerations insofar as it equates (1) simple lines with straight and
circular thus excluding the cylindrical helix; (2) straight with up or down; (3)
circular with around the center of the universe; (4) different simple motions
with natural motions of different simple bodies; and (5) bodies having simple
natural straight motion with earth and fire. A more correct conceptual frame¬
work is the theory of universal circular motion (F43—57) according to which,
in a well-ordered universe, all movable bodies must move with circular
motion. The basis of this theory is the following conceptual distinction:
straight and circular motions are two different stages of natural motion;
straight motion can be acquired naturally but cannot naturally continue
forever, whereas circular motion can naturally continue forever but cannot
be acquired naturally without straight motion.

57—9: Sense experience

Sense experience cannot be dismissed, but it should be treated critically, in
the sense that it is subject to different interpretations and one should not
assume that his own interpretation is the only possible one. Hence it would
be naive to reject the theory of universal circular motion on the grounds that
it conflicts with the natural up and down motion of terrestrial bodies and
thus violates the principle that sense experience should prevail over any
reasoning. In fact, sense experience tells us merely that this up and down
motion appears to be straight; to interpret it as really being straight involves
the uncritical acceptance of the principle that apparent vertical motion
implies actual vertical motion; and this implication does not hold as it will be
shown later. In other words, naive empiricism is to be rejected in favor of
critical empiricism, which involves logical analysis and logical synthesis
(reasoning, logical practice) in an essential way.

59—71: Aristotle’s authority as a logician

Aristotle’s authority as a logician is not sacrosanct and at any rate pertains
primarily to the domain of logical theory rather than logical practice. His
authority as a logical practitioner cannot be grounded on his undisputed
authority as a logical theorist, but only on the soundness of his concrete
108 CHAPTER 5

reasoning. The latter can only be tested by the logical analysis of his actual
arguments, which analysis one should therefore be free to carry out. If you
do this you will discover that his logical practice leaves much to be desired.
For example, consider Aristotle’s argument that the point toward which
heavy bodies fall is the center of the universe. Logical analysis reveals that
this argument begs the question because it assumes that, in moving away from
the earth’s circumference, light bodies are moving either (1) toward a greater
circle concentric with the center of the universe, or (2) along lines which
extended in the opposite direction pass through the center of the universe.
Or, consider Aristotle’s argument for the immutability of the heavens based
on the natural circular motion of heavenly bodies; the analysis of the con¬
ceptual status of natural and of circular motions (which conceptual analysis
is a kind of logical analysis) has revealed that all movable bodies must move
with circular motions; hence the only conclusion that follows in this context
is either that the earth like heavenly bodies is unchangeable or that change
has nothing to do with difference in motion. Or consider the argument from
contrariety; logical analysis reveals several faults with this argument: (1) it
uses an assumption (the premise asserting a connection between change and
contrariety) which is more problematic and questionable than its main
intended consequence (namely, that the earth stands still); (2) it may be self¬
contradictory because it suggests a parallel argument concluding that
heavenly bodies are changeable as well; (3) it is groundless to the extent that
the contrariety of rarity and density exists in the heavens; and (4) it is
ambiguous because the various mentioned “bodies” sometimes refer to whole
bodies and sometimes to parts of bodies.

71 —82: The role of authority in natural philosophy

Aristotle’s authority in natural philosophy has little if any function because
acceptance of it implies its rejection in favor of direct sense experience. Both
his example and his own words show this. Consider the question of celestial
unchangeability. Aristotle’s a priori arguments in its support have already
been examined and found to be inconclusive (F76); moreover, they probably
reflect his method of justification rather than his method of discovery, which
is probably reflected in his a posteriori argument. So consider this argument.
The argument is logically invalid (F71-5); but Aristotle’s procedure (‘modo
di filosofare’, F75) is such that, if followed today, it leads one to believe that
the heavens are changeable. For, his procedure is to conclude that the heavens
are unchangeable on the basis of seeing no changes in the heavens; hence, the
same procedure would force one to conclude that the heavens are changeable
 THEORY AND PRACTICE 109

on the basis of seeing some changes; now it so happens that today one can
observe changes in the heavens (F76—80); hence acceptance of Aristotle’s
example in matters of procedure leads one to reject his authority concerning
the substantive claim of celestial inalterability, and to rely on sense experience
rather than on an authority. As for his words, one of his most important
principles is that sense experience should prevail over any reasoning, and
the principle is more fundamental than the claim that the heavens are un¬
changeable. Now, applied to the sense experience possible today and to
Aristotle’s own reasoning, this principle implies that it is more in conformity
with Aristotle to argue “the sky is changeable because so shows us sense
experience” than to argue “the sky is unchangeable because reasoning so
persuaded Aristotle”. Hence acceptance of Aristotle’s words in methodo¬
logical matters leads one to let sense experience prevail over Aristotle’s own
substantive conclusions, and hence to the rejection of his authority in sub¬
stantive matters.
It may also be concluded that no authority has any methodological role in
natural philosophy, however understandable the security of relying on it is
(F81), since Aristotle’s procedure in his a posteriori argument for celestial
immutability is not an isolated example, but rather a posteriori arguments are
methodologically prior to a priori ones, since the former reflect the method
of discovery and the latter the method of justification (F75-6).
In other words, the concrete example of Aristotle and the inalterability of
the sky shows that if we take the methodological point of view, if we dis¬
tinguish between an author’s methodological claims and his substantive
claims, if we distinguish between methodological theory and methodological
practice and do not neglect the example of the latter besides the use of the
former, if we distinguish between methods of discovery and methods of
justification within methodological practice, and if we believe that a posteriori
reasoning is the typical method of discovery and a priori reasoning the typical
method of justification, then it is impossible for an authority to have a
methodological function.

82-95: Teleology and anthropocentrism

Teleological anthropocentrism is the thesis that the whole universe exists for
the sake and benefit of man (F84, 85). A good example of reasoning which is
both teleological and anthropocentric is the following Aristotelian argument
(F83-85; cf. F109): (1) changes would make celestial bodies imperfect, since
such changes would not be of any use or benefit to man, and hence they
would be superfluous; therefore, (2) for celestial bodies, being unchangeable
 110 CHAPTER 5

would be their way of being perfect; (3) but celestial bodies are unchangeable,
because celestial changes would be vain and useless, and hence there cannot
be any; it follows that (4) celestial bodies are perfect, that is have all perfec¬
tions. Here, teleological anthropocentrism is presupposed in steps (1) and (3).
The doctrine is objectionable for at least two reasons. First, one valid reason
for teleological anthropocentrism would be if the place where man lives — the
earth — were appropriately special and different from the other bodies in the
universe. But, using the nearest celestial body as an example — the moon — it
can be shown that it and the earth are rather similar (F87—95): in fact they
are both spherical, opaque, lacking their own light, solid, and irregular in their
surface; and each illuminates and eclipses the other and shows phases and
uneven brightness to it. Second, teleological anthropocentrism leads one to
conclusions, such as the perfection of celestial bodies mentioned above,
which can be shown to be false. For the moon has a rough rather than per¬
fectly smooth surface, because of the fact that it is visible at all and because
of its mountainous appearance through the telescope (F87-8; cf. F95-112).

95—112: The necessity of experiments

Experiments are sometimes necessary (F100,101,108) in natural philosophy.
One such situation occurs when the facts involved are counterintuitive (F96),
as in the question of the manner the moon reflects sunlight. For example, (1)
if you hang a mirror on a wall and compare the way the mirror and the wall
reflect the light, you will see that the wall looks brighter than the mirror from
all places except one, where the mirror’s reflection goes (F96-8). Or, (2)
compare the reflection of sunlight from a flat and from a spherical mirror by
comparing their effect on the brightness of a neighboring wall; you will see
that the spherical mirror has no noticeable effect on the illumination of
neighboring objects (FI00-1). Or, (3) fold a sheet of paper and let one part
receive light perpendicularly, and the other part obliquely; you will see the
latter to be less bright than the former (F105). Or, (4) fold a sheet of paper
into two unequal parts, let some light shine perpendicularly on the smaller
part and obliquely on the larger, and look at the paper from an angle such
that the two unequal parts appear of equal size; you will notice that the
obliquely-lit part still looks less bright than the other (F108-9). From the
first two experiments one can argue that if the moon’s surface were not
rough, it would be invisible (F96-105); from the latter two, that if its surface
were not rough, a full moon would appear brighter at the center and darker
at the edges (FI05—9).
Another example of the necessity of experiments in natural philosophy
 THEORY AND PRACTICE 111

occurs when the relevant facts need to be ascertained with some precision, or
when a man-made model of the natural situation is possible, as in trying to
explain the uneven appearance of the lunar surface to the telescope by
differences of opacity and transparency. Such an explanation cannot be
faulted by means of general considerations, but only by the specific details
of the telescopic appearances, all of which are reproducible with a solid
opaque ball subject to varying illumination, though none with a solid ball of
nonuniform transparency (Fill): you can see (FI 12) that the various lighted
bulges cast shadows; that the length of these shadows varies depending on the
distance of the bulge from the boundary of light and darkness; that this
boundary is very irregular; that in the dark region beyond this boundary
many bulges are already lit; that the length of the various shadows of the
lighted region decreases until there are no more shadows when the whole
hemisphere is lit; and that then as the darkness starts growing the same bulges
can be recognized and cast shadows whose length increases.

112-7: Cognitive awareness

The intellectual contents of the human mind are not always open to intro¬
spection or known best to the person involved; this applies both to the real
reasons underlying one’s beliefs and to propositions that one may know with¬
out knowing that he knows them. For example, someone may say that the
reason why he thinks that the earth cannot reflect sunlight as well as the
moon is that the earth is so dark and opaque (FI 12-3); however, cross-
examination reveals that his real reason is that whereas the earth can be seen
lit only during the day, the moon can be seen lit also during the night, for
from this he is led to the error of comparing the moonlight as seen from a
dark place (night) with the earth-light as seen from a lighted place (day), and
he naturally concludes that the moon reflects light much more easily that the
earth. An example of knowledge without awareness is the proposition that
the earth can reflect light as well as the moon (FI 15); someone may think
that he believes the opposite, but cross-examination can show that he knows
all the facts from which the correct conclusion easily follows: that the moon
appears more shiny at night than during the day, that during the day it
appears about as shiny as clouds, and that the earth-light cannot be seen at
night. Another example of unconscious knowledge is the proposition that the
moon’s secondary light is light reflected by the earth onto the moon
(FI 15-6).
These examples also implicitly illustrate the importance of the rationale
underlying beliefs for the resolution of disagreements. For when that rationale
 112 CHAPTER 5

is seen to be wrong, as in the first example above, it is often an easy matter to

see what is the proper conclusion to be drawn, as in the second example
above.

117—24: Different categories of error

Different categories of errors exist; they are not all equally culpable. The
acceptance of an erroneous theory by its originator or at a time when no
better theory exists is a relatively minor methodological transgression. How¬
ever, if a follower defends a grossly erroneous theory at a time when a better
one has been published, that can only be an appeal to the ignorant masses. An
example of such a methodologically objectionable appeal is the recently
resuscitated theory that the moon’s secondary light is sunlight which pene¬
trates the somewhat transparent lunar body. Glaring falsehoods asserted by
this author are that the secondary light is greater at the edges than in the
middle, and that in solar eclipses the portion of the lunar disc outside the
solar disc is darker than the rest.

124—31: The powers of human understanding

The powers of human understanding are likely to disappoint both the
optimists and the pessimists. On the one hand, the human mind is not a
measure of what can occur in nature since if you ever try to understand fully
some one thing you realize how little you know (as the example of Socrates
shows). Moreover, the human mind is infinitely inferior to the divine mind
both in number of propositions known and in manner of knowing them; for
God keeps infinitely many propositions simultaneously in His mind without
having to use a step-by-step knowing process. On the other hand, the human
mind can know a few propositions (such as those of mathematics) with an
‘objective certainty’ (FI29) which equals that which God has; and moreover,
one cannot but marvel at the various artistic and scientific achievements of
man. An example of this limited but real power of the human mind is our
knowledge of the moon: though we can think of a number of conditions that
might cause the big dark spots on the moon, we have no way of knowing
which one it is; and though we can have no idea of what kind of life there is,
we know that it cannot be anything like on earth because the moon has very
long days and nights (one month combined), no significant seasons, and no
rain.

132-9: Independent-mindedness and the misuse of authority

The misuse of Aristotle’s authority by his followers is laughable and ridiculous.
 THEORY AND PRACTICE 113

They often oppose to the evidence of sense experience, Aristotle’s mere

authority or mere words, rather than other experiences or Aristotle’s reasons;
an example would be for someone who has been shown experimentally that
the nerves originate from the brain, to say that he would have to accept it as
true if Aristotle had not said that the nerves originate from the heart
(F133-4). They also often claim that, and act as if, all knowledge is contain¬
ed in his books: an example is the Peripatetic who tried to show that the tele¬
scope was known to Aristotle in its fundamentals by suitably interpreting, ex
post facto, the passage where he describes the effect of looking at stars from
the bottom of a well (FI34—6).
Aristotle’s authority results only partly from the strength of his demon¬
strations, but primarily from lack of independent-mindedness, as well as
academic opportunism and the failure to appreciate what might be called
epistemological modesty. This last attitude means the failure to see that not
only is there nothing wrong with claiming ignorance about certain things, but
it is actually better to do so since if one does this then he will be taken more
seriously (at least by serious philosophers) concerning the things that he does
profess to know (FI37). An example of academic opportunism is the Peri¬
patetic who had written a book on the soul in which he denied that Aristotle
regarded it as immortal on the basis of certain less well-known passages; when
he was told that it would have been difficult for him to get permission to
publish, he decided to use certain other Aristotelian passages favorable to
immortality (FI37-8). It is primarily lack of independent-mindedness which
generates the weight of Aristotle’s authority because, no matter how strong
his arguments, it is we who must evaluate this strength; it follows that in
philosophy at least, authorities play no essential role (FI38-9). It also fol¬
lows that it is imperative to examine carefully his arguments concerning the
question of the earth’s motion.

139-50: Simplicity and probability

Arguments based on the principle of simplicity are neither worthless nor
conclusive, but rather probable. They are not worthless because, in the
absence of conclusive arguments, the simplest idea is the most acceptable,
they are not conclusive because a single piece of counterevidence or con¬
clusive counterargument is sufficient to refute an idea based on simplicity,
and they are probable because nature tends to act by means of the fewest
possible operations. For example, the arguments in favor of the earth’s
diurnal motion are based on the principle of simplicity insofar as they
amount to saying that it is simpler (1) to let the earth rotate on its axis at a
 114 CHAPTER 5

relatively low speed rather than to let innumerable celestial bodies revolve
around the earth at incredibly higher speeds; (2) to let each planer nave only
one motion rather than two contrary ones; (3) to let the law of periods of
revolution be universally valid, rather than to make an exception for the case
of the motion of the uppermost sphere; (4) to let all the fixed stars do one
thing — stand still — rather than to let them revolve over unequal orbits; (5)
to let the earth change its axis of rotation than to let all fixed stars change
their orbits in unison; (6) to let the earth rotate than to coordinate the
motions of all fixed stars in such a way as to keep their relative positions
unchanged; (7) to prevent the earth’s rotation from being transferred to the
rest of the universe, rather than to prevent the diurnal motion of the heavens
which is transferred through all the spheres from being transferred to the
earth; and (8) to let the earth rather than an ad hoc highest sphere be the
primum mobile. These arguments are obviously inconclusive but they do lend
some initial probability to the earth’s diurnal motion (F148). For this pro¬
bability to remain, one must show that the apparent counterevidence does
not really exist and that the apparently conclusive objections are in fact
unsound.

150—9: Open-mindedness and rational-mindedness

It is very important to know and understand the arguments, reasons, and
alleged evidence against one’s view (FI53), since its justification is streng¬
thened if undertaken in the light of the awareness of counterarguments
(FI 54-5), and since truth emerges from the conflict of opposing views if this
conflict takes the form of the assessment and evaluation of the logical
strength of the relevant arguments (FI55—8). Let us call open-mindedness
the willingness and capacity to know and understand the arguments against
one’s view, and rational-mindedness the willingness and capacity to accept the
views which are supported by the best arguments; and let us call logical
analysis the (art of) understanding the structure of arguments and evaluating
and comparing their strength. Then we may say that open-mindedness and
rational-mindedness are essential in the search of truth, and logical analysis
is in turn essential for both of these traits.
The motion of the earth provides a very good example of all this. First
the methodological superiority of the proponents as compared to the
opponents is obvious insofar as the former are usually knowledgeable of the
counterarguments, whereas the opponents are usually not knowledgeable
about the favorable arguments (FI54-5); moreover, the proponents have
evenimproved the clarity and cogency of the counterarguments (F153,F159).
 THEORY AND PRACTICE 115

Second, these counterarguments against the earth’s motion are so numerous

and so undeniably convincing and conclusive in appearance that it will be a
very good test of the power and usefulness of logical analysis to see what
results it yields when applied to them.

159—69: Logical analysis

The logical analysis of the arguments against the earth’s motion yields the fol¬
lowing results: (1) (F159-62) Aristotle’s first objection is an argument from
violent motion and has the following structure: the earth cannot move circu¬
larly because such motion would be violent, and hence not perpetual; it
would be violent because if it were natural, the parts of the earth would also
move circularly by nature, which is impossible since the natural motion of its
parts is straight downwards. This argument is a fallacy of equivocation since
the clause “the parts of the earth would also move circularly” can mean
either that these parts would move around their own centers or that they
would move around the earth’s center. In the first case, the corresponding
proposition, namely that “if the earth’s motion were natural, then its parts
would also move circularly by nature”, would be obviously false, and hence
the argument would be unsound by reasons of falsehood of one premise. In
the second case, the argument would be unsound by reason of invalidity,
since it would contain two incompatible steps; for one step claims that the
earth’s circular motion would not be perpetual because it would not be
natural, thus equating the natural with the perpetual, whereas another step
would claim that the earth’s parts do not naturally move around its center
because their natural motion is straight downwards, and this is so because
their spontaneous motion when free to move is straight downwards, thus
equating the natural with the spontaneous under certain supposedly natural
conditions. The invalidity derives from the fact that this second notion of
natural is both inconsistent with the first, and implausible because the
conditions under which the spontaneous motion is straight downwards are
not really natural but rather contrived artificially or seldom occuring in
nature. (2) (FI62-4) Aristotle’s second argument is another fallacy of equi¬
vocation because its conclusion could mean either a denial of the diurnal
motion or of the annual motion. (3) Aristotle’s third argument is the
argument from natural motion and has the following structure: the natural
motion of the earth’s parts is toward the center of the universe, and there¬
fore the whole earth stands still therein. This argument is groundless since the
empirical argument that the earth’s parts move toward the center of
the universe is circular (as it is shown in F58-61), and since the a prion
 116 CHAPTER 5

justification of the same premise presupposes an untenable, confused, or

inconsistent concept of what it is for motion to be natural (as it is argued in
F33—62 and F159—62). (4) (F164—9) Aristotle’s fourth argument is the
argument from vertical fall, and it is circular when stated in terms of actual
vertical fall, and invalid when stated in terms of apparent vertical fall. The
former version has the following structure: The earth must stand still because
bodies fall vertically, which they could not do if it moved; and bodies fall
vertically, since they are seen to fall vertically and since apparent vertical fall
implies actual vertical fall. Now, it turns out that this implication holds if and
only if the earth stands still; hence one would have to assume that the earth
stands still to justify the corresponding premise in the argument, namely the
proposition that if bodies are seen to fall vertically, they really do fall verti¬
cally.
The apparent vertical fall version of the argument has the following struc¬
ture: The earth stands still because bodies are seen to fall vertically; now this
could not happen if the earth were moving since if the earth were moving and
bodies were seen to fall vertically, then bodies would be moving with mixed
motion, around and tow'ard the center; but such mixed motion is impossible
since bodies dropped from the mast of a moving ship fall behind. The invalid
step in this argument is the last one, which presupposes that the nonoccur¬
rence of mixed motion on a moving ship would imply physical impossibility
(or nonoccurrence on a moving earth). Such an implication is invalid because
the motion of the ship is violent whereas the earth’s diurnal motion would be
natural, and because the relevant portion of air does not move along with the
ship but would do so (near the earth’s surface) if it moved.

169-80: Reasoning vs experience, and the Socratic method

Experiments are sometimes unnecessary to ascertain the results of a test, for

sometimes it can be argued on the basis of known or more easily ascertain¬
able facts, what these results must be. For example, consider dropping a rock
from the mast of a moving ship; the Aristotelians believe that the rock will be
left behind, and from this, by analogy, argue that the vertical fall of a rock
dropped from a tower implies that the earth stands still. It is doubtful
whether they ever made the experiment; but at any rate, it can be shown that
on the moving ship the rock will fall at the foot of the mast. Here, the more
easily ascertainable facts are that (a) the undisturbed downward motion of
bodies on an inclined plane is accelerated, (b) their undisturbed motion up
an inclined plane is decelerated (F171-3), and (c) the cause of the motion
of projectiles is not the motion of the air surrounding them (FI75—80).
 THEORY AND PRACTICE 117

From (a) and (b) one may conclude that (1) the motion of bodies on an
horizontal plane is conserved if undisturbed, and consequently that (2) the
horizontal motion of the rock on the moving ship, even after being dropped,
continues if undisturbed. Now from (c) one can conlude that (3) the cause of
the motion of projectiles is the ‘virtue’ impressed on them by the projector,
and consequently that (4) the cause of the horizontal motion of the rock,
after it has been dropped, would be the horizontal ‘virtue’ impressed on it by
the hand holding it before dropping. Since there is no way in which this
horizontal impressed virtue could be disturbed by the vertically downward
tendency due to weight, it follows that (5) the horizontal motion of the
dropped rock is undisturbed, and hence by (2) that (6) the horizontal motion
of the dropped rock will continue, and therefore that (7) the rock will end up
at the foot of the mast. It should be noted that such a justification can be
effectively carried out by means of the Socratic method of questioning
(F171, 176-7).

180-8: The Socratic method: understanding vs information-knowledge

The Socratic method (F183) is very effective in making explicit, knowledge

which is implicit in the mind. This is not teaching in the ordinary sense, that
is in the sense of imparting or conveying new information, but rather in the
sense of deriving what was previously not accepted from propositions that are
accepted. Alternatively we may say that the knowledge in question is not
information or mere knowledge, but understanding.
For example, someone may be skeptical about whether certain projectiles
(1) can move faster after they hit the ground than in the air and (2) can move
faster over a later than over an earlier segment of their whole path. (1) can be
made simultaneously certain and comprehensible by appropriately question¬
ing the skeptic and making him realize(F 186) that when the projectile is some¬
thing like a hoop, that can spin, then the spinning does not add to its forward
motion as long as it is in the air, but it does so when it hits the ground if the
spinning motion is appropriately in the same direction as the forward motion.
(2) can be demonstrated to the skeptic by making him realize that, on falling
to the ground, a hoop may hit a rock slanting in the direction of the forward
motion and thus may acquire more spinning which is added to the forward
motion. Thus from the doctrine of the conservation of motion one can
understand the apparently paradoxical consequence that the motion of
certain projectiles can sometimes increase, when a basic part of their motion
is conserved and other amounts are added by various secondary factors.
 118 CHAPTER 5

188—93: Fanciful vs relevant digressions

There are two kinds of digressions: some are justified by their relevance to
the main argument and by the logic of the discussion, which is a complicated
one where several people are involved. Others are justified by the whim of the
writer or persons involved or by the intrinsic beauty rather than by the
reasoned tenability of the ideas involved (FI88). The digressions up to this
point have been typically of the first type; an example of the latter type of
digression is the following fanciful (F192) speculation concerning the actual
path of a falling body on a rotating earth: that the path is an arc of a circle
having as a diameter the line going from the earth’s center to the point from
which the body was dropped. This is a path that is almost too good to be true
(FI92) insofar as it means that whether or not a body is dropped or held still
on a rotating earth, it would be moving (a) circularly, (b) along distances of
the same length, and (c) with the same uniform motion.

193—6: Unconscious knowledge

Some knowledge is unconscious (FI94) in the sense that it is possible to

arrive at it by reasoning (F196), though experience may be necessary for
those who are unwilling or incapable (F196). For example, by means of
Socratic questioning and hypothetical reasoning involving an analogy with
arrows shot from a crossbow mounted on a moving carriage, it is possible to
show to someone who initially believes otherwise, that on a rotating earth the
range of eastward and of westward gunshots would not differ (F194-6).

196—206: Hypothetical and circular reasoning

hypothetical reasoning is a very valuable skill, but a difficult one to master;

deficiency in this skill often results in circular reasoning or arguments that
beg the question. For example, when people argue that east and west gun¬
shots would have unequal ranges if the earth were rotating, they are reasoning
as if the state of the cannon ball before firing would be one of rest even on a
rotating earth, this amounts to a failure to take seriously such rotation when
examining the consequences of such rotation. The same is true for the argu¬
ments from vertical fall, from vertical gunshots, from north-south gunshots,
and from point-blank gunshots. In all these cases, the opponents of the
earth s motion pretend to be engaging in hypothetical reasoning examining
the consequences of the hypothesis that the earth moves, and yet there is
one crucial step where they disregard this assumption; it is the step where the
initial state of the projectile or falling body is taken to be rest.
 THEORY AND PRACTICE 119

206—9: Critical experimentalism

The uncritical acceptance of alleged experimental results is to be avoided.

One way of being critical is to watch out for the possibility that the results
have been merely hypothesized in conformity with some theory and for the
possibility that the effect is so small that one cannot check its presence
experimentally. For example, consider the question of whether point-blank
gunshots hit high toward the east and low toward the west. The Aristotelians
claim that experience shows this does not happen (and since they think it
would have to happen if the earth rotated, they conclude that the earth
stands still). Now, by means of a rough calculation based on certain (other¬
wise objectionable) Aristotelian principles, one can show that the magnitude
of the effect would be about 1/25 of a ‘braccio’, which is experimentally
undetectable.

209-14: The proper time for experiments

The proper time to make an experiment is when reasoning is incapable of

deciding whether a phenomenon would happen, or why it does happen, or
how it can happen. For example, consider the phenomenon of the flying of
birds on a rotating earth. Even someone who understands very well the
motion of projectiles on a rotating earth (cf. Sagredo’s superiority over
Salviati in F196-206, especially F202 and F204-5), cannot help but think
that flying birds are essentially different, for just as a (live) bird does not do
the same thing a rock does in the case of the natural motion of falling, so the
bird would not do the same thing as inanimate bodies in the case of the
natural circular motion along with a rotating earth (F212). Experiment can
show that, for the latter type of motion, there is in reality no essential dif¬
ference between an inanimate body and a bird, and that it is an irrelevant
difference for them to have different sources of motion, an external projector
for the projectile and an internal principle for the bird. The experiment is
made under deck on a ship: one observes the behavior of flies, Fish, butter¬
flies, dripping water, rising incense, and projectiles when the ship is standing
still and when it is moving. One observes no difference deriving from the state
of rest or of uniform motion of the ship, and that the flies, fish, and butter¬
flies share the common motion of the moving ship just as easily, naturally,
and effortlessly as the inanimate bodies. The experiment is also very effective
in answering the geostatic arguments from falling bodies and projectiles,
which are criticizable in other ways as well.
 120 CHAPTER 5

214—7: Criticizing and understanding

Destructive criticism is more effective when preceded by constructive elabora¬
tion which strengthens the idea being criticized. For example, consider the
objection to the earth’s motion which claims that if the earth rotated, then all
bodies on its surface would fly off into the heavens, because turning gives a
body an impetus away from the center of the whirling. Before answering this
objection it is better to make a logical clarification and a substantive addition.
The elucidation is that the objection should properly be stated by saying that
if the earth rotated, then the unattached bodies or creatures on its surface
could not have come into existence on its surface; this corrects the usual
version that if the earth rotated then the unattached bodies would move away
from it, which is a virtually universally shared formulation, and which is a
misstatement since it would show at best that the earth cannot begin rotating.
The substantive addition is the evidence of a pail full of water, tied at one end
of a rope, and swirled from the other end; the water will not fall out of the
pail regardless of the angle of the plane of its circular motion; moreover, if a
hole is made in its bottom, the water will come out through the hole in all
directions, but always away from the center of the swirl.

217—23: Knowledge and recollection

Plato’s doctrine that knowledge is a kind of recollection is in a sense true. The
sense in which it is true is that knowledge is sometimes acquired by reasoning
based on facts ascertainable by reflection upon certain chosen aspects of
one’s sense experience. For example, one can easily conclude that bodies on
the surface of a rotating earth would not be extruded into the heavens after
having been made, by questioning, to reflect that: (1) when a body which has
been made to move along a circle is let go, it acquires a motion tangential to
the point of release; (2) as the body starts moving along the tangent, it moves
away from the point of tangency much more than away from the circum¬
ference, so that to prevent the body from moving away along the tangent all
that is needed is for it to move a very small amount toward the circumference;
and (3) bodies, however light, do have a tendency to move toward the surface
(circumference) of the earth due to their weight.

223—9: Geometry and philosophy

Geometry is indispensable in natural philosophy. For example, only by means
of geometrical considerations is it possible to understand how and why the
falling tendency of a terrestrial body, however small one chooses to make it,
is always sufficient to overcome any centrifugal tendency on a rotating earth
 THEORY AND PRACTICE 121

and thus to keep it on the surface. The main geometrical consideration is

that, given an arbitrarily large ratio, one can fmd a point outside a circle
such that the tangent from that point bears that ratio to the portion of the
secant outside the circle; other reinforcing considerations are that (1) the
geometry of the situation is such that the tangential tendency can be over¬
come by an arbitrarily small weight, even though the tangential tendency
were diminishable to infinity in only one way, whereas the downward
tendency were diminishable to infinity in two ways arising from the decrease
in weight and from the approach to the point of tangency; and (2) the
decrease of the downward tendency due to decreasing weight is not only not
faster than linear but much slower.

229-37: Mathematics and physical reality

Mathematical truths are about abstract entities in the sense that they are state¬
ments about the necessary consequences of certain definitions and axioms.
For example, the proposition that a sphere touches a plane in only one point
is about abstract spheres and planes in that it is a necessary consequence of
the definition of a sphere and the principle that a straight line is the shortest
distance between two points. Mathematical truths are also about physical
reality, though only conditionally; that is, a mathematical proposition is
physically true if the abstract entities it is about happen to exist as material
entities in physical reality. For example, the proposition that a sphere touches
a plane in only one point is true of physical reality in the sense that if there
happen to be material spheres and planes, then they touch in only one point.
Third, mathematical truths are applicable to physical reality because and
insofar as material entities approach or approximate abstract ones; for when
material entities do not approximate one type of abstract entity, they approx¬
imate another. For example, if and to the extent that material spheres and
planes touch in more than one point-; they instantiate abstract spheres and
planes that are imperfect, and of these it is equally true in mathematics that
they touch in more than one point. Finally, the real problem is to use the
proper type of abstract entity in terms of which to interpret physical entities
and processes, for though one is sure that the latter must correspond to some
type of abstract entity or situation which is treatable mathematically, one
cannot be sure of which one. It is along these lines that one could properly
question the relevance of the geometrical proof given in the previous section.

237-44: Causal investigations

Causal investigations are important in natural philosophy. Sometimes one
122 CHAPTER 5

may want to investigate how the cause of a certain effect varies. For example,
it is important to realize that in circular motion, the cause of extrusion or
projection varies directly as the circumference speed when the radius is
constant (F238 begin.) and inversely as the radius when the circumference
speed is constant (F239, 243). Sometimes one is certain about a fact but may
want to know the causal explanation of it. For example, there is no doubt
that equal circumference speeds produce a greater tendency to extrusion
when the radius is smaller; but it is important to know ‘the cause why’
(F238) or reason for it (F239); this is to be found by considering that the
force needed to keep a given body from escaping along the tangent will be
greater when greater is the speed it would need to move from the tangent
back onto the circumference; and such speed is greater for the smaller circle
since the circumference speeds, and hence the relevant times, are equal.
Finally, sometimes a cause can be discovered by comparing its effect to a
second effect whose cause is known, and using the way in which the given
effect differs from the second one as a clue to how the unknown cause of the
given effect differs from the known cause of the second effect. For example,
in a steelyard, why does a lighter weight balance a heavier one? The cause is
that the lighter weight moves more than the other. For compare the steel¬
yard and the equal-armed balance; in the latter equal weights balance each
other because they are moving equally; in the steelyard it is unequal weights
that are balancing each other; so this must happen because they move
unequally.

244—60: Physico-mathematical synthesis

It is not enough that the questions treated be important and worthwhile; it is
essential that the method of treatment be worthwhile (F246). To be worth¬
while the method must be both mindful of physical facts and sophisticated
mathematically. For example, the problem of a body falling from the moon
to the earth is a very interesting one. In dealing with it, however, a recent
author has not treated it properly: first, in computing the time of fall he has
arbitrarily assumed that its speed of fall would be equal to the speed of
diurnal rotation at the moon, instead of determining the actual speed with
which bodies actually fall (F246); second, his answer that the time would be
six days seems to involve the mathematical absurdity that the radius of a
circle is six times longer than its circumference (F247, 258-9). A proper
treatment of the question would begin by noting the fact that falling bodies
accelerate (F248), showing the mathematical sensitivity of asking by what
proportion the velocity increases, ascertaining that the distances from rest
 THEORY AND PRACTICE 123

vary as the square of the times, that a body falls 100 braccia in 5 seconds, and
then using these values together with the law of squares to compute the time
of fall from the moon as being about 3Vi hours (F248—52).
A second example (F252—56) to illustrate the proper combination of
physical and mathematical considerations is the double-distance law, which
states that a body would move double the distance through which it has
fallen from rest if it were allowed to move uniformly at the acquired speed
for a time equal to that of fall. The physical considerations involve both a
pendulum and a body oscillating about the earth’s center in a tunnel going
from a point on the surface to the point diametrically opposite to it; here,
observation in the case of the pendulum, and physical intuition in the case of
the tunneled earth, tell us that the impetus acquired by the body while it
undergoes acceleration is such as to enable it to move an equal distance while
being decelerated by the same amount. The mathematical considerations
involve summing up all the speeds during both acceleration and deceleration
and equating this sum to the sum-total of constant speeds during an
imaginary second half of the motion if it had continued uniformly instead of
decelerating; another mathematical consideration involves summing up all
the speeds during a given accelerated motion and comparing this sum
(measured by a right triangle) to the sum-total for a uniform motion at a
speed equal to that acquired during acceleration (measured by a rectangle
twice the area of the triangle).
A third example (F256-8) of physical-mathematical synthesis involves the
question of whether a pendulum would come to rest even if air resistance is
disregarded. The answer is that it would because any physical pendulum is
actually a series of pendulums of different lengths, due to the matter in the
string, and all these pendulums are trying to vibrate at different rates and
hence interfere with each other, with the slower continually slowing down
the faster. The physical observation is that pendulums of different lengths
vibrate at different rates, the rate being faster the shorter the pendulum; the
mathematical consideration is the analysis of a single material pendulum into
a series of pendulums of different lengths.

260—72: Comprehensibility and truth

The aim of natural philosophy is understanding, not mere knowledge; not
merely truth, but comprehensible truth. Hence the discussion of the compre¬
hensibility of an idea is always relevant and important; this is especially true
when it is not known whether the idea corresponds to reality, for at this stage
incomprehensibility constitutes a strong reason for rejection; when the idea is
124 CHAPTER 5

known to be true, incomprehensibility is merely a reason for doing further

work. In other words, comprehensibility and truth are distinct: if an idea is
comprehensible, that does not make it true, nor if it is incomprehensible,
does this make it false;nevertheless natural philosophy aims at both.
A good illustration of all this is the idea of the earth’s rotation and the
difficulties that have been raised concerning its cause. First (F260), it would
be invahd to argue that the earth cannot move because there would be no
way to explain the rotation in terms of either an internal or an external
principle; the reason is that ignorance of the explanation does not imply non¬
existence of the effect; in other words, inexplicability does not imply false¬
hood.
Second (F260—1), a rotating earth would be no more inexplicable than
the fall of heavy bodies, the motion of projectiles, or the alleged diurnal
motion of heavenly bodies; for it is no explanation to call their respective
causes ‘gravity’, ‘impressed virtue’, and ‘informing or assisting intelligence’; in
other words, knowledge of the name does not imply knowledge of the
essence of a cause, or, naming is not the same as understanding.
Third (F262—72), it cannot be argued that an explanation of the earth’s
rotation would be inconceivable, i.e., that this rotation would be inexplicable-
in-principle; an explanation could always be given in terms of an internal
principle, by attributing the rotation to nature and eliminating the need for
an external cause. The reason is that the idea of the earth’s rotation is at least
as comprehensible as the diurnal motion of the heavens (which is regarded as
comprehensible); moreover, there are no conceptual difficulties with the
earth’s rotation; what are usually mentioned as difficulties turn out to be
nothing but unproblematic consequences of it. In other words, causal in¬
explicability, even in principle, does not imply incomprehensibility, for
intrinsic comprehensibility is possible.

272—81: The deception of the senses

The deception of the senses can be instructive rather than a cause for despair.
For many alleged deceptions are not deceptions at all, while others are
deceptions of reason rather than of the senses. The main lesson to be learned
is to be critical toward the senses, i.e., to realize that there is a difference
between appearance and reality and that it is by reason that we go from
sensory appearances to claims about reality (F280-1).
For example (F272—8), it would be no deception of the senses if on a
rotating earth we perceived bodies falling vertically and failed to perceive
any circular component in their motion; for motion exists only relative to
 THEORY AND PRACTICE 125

things that do not share it, hence motion shared by an observer and an
observed object does not exist for him and is imperceptible to him; on a
rotating earth, the circular component of the motion of the falling body
would be shared by the observer and hence would not be there to be per¬
ceived. Second (F278—9), the failure to perceive a wind due to the annual
motion would not be a deception of the senses because wind is perceivable
only when air moves relative to the observer. Third (F279—80), the failure to
feel any motion on a rotating earth would by itself be merely a sensory
appearance; it would generate a deception only if one went on to say that
therefore the earth stands still since if it moved one would be able to feel its
motion; but this would be a deception of reasoning, not of perception, since
the error would be in assuming that one can feel mere motion, whereas
experience with navigation shows that one can only feel changes of motion.

281—8: Inconceivability-claims and facts

It is wrong to argue that a doctrine is false because it has conceptual diffi¬
culties, that is difficulties concerning the way in which it could conceivably
be true (F287); in other words, it is wrong to say that something is not the
case because there is no way to conceive how it could be the case. It is wrong
to draw factual conclusions from inconceivability claims. The problem is
that inconceivability claims frequently end up being falsified by factual
discoveries. For example, it is wrong to argue that the earth does not move on
the grounds that there is no way to conceive how a simple body like the earth
could have the three distinct motions that it is supposed to have according to
Copernicus. For the only factual content of this inconceivability claim is that
there is no way for the earth to have its different motions the way that an
animal can move in various ways; on the other hand, the earth could move
with more than one motion the way that Saturn, the sun, Jupiter’s satellites,
and loadstones have been discovered to do.

289—90: The primacy of nature over man

It is wrong to argue that something is the case because it fits well our under¬
standing, for nature is prior to the human mind. For example, it is wrong to
argue that different substances must behave differently otherwise we could
not come to know them; what’s wrong with this is that it is empty of factual
content; insofar as it does refer to natural reality the principle must be
qualified to specify the difference in behavior. In other words, though we
know a priori that different substances must behave differently in some way
(in order to be different), it does not follow from this that they must behave
126 CHAPTER 5

differently in a particular given manner. For example, there is no necessity

for water and air to have to behave differently with respect to their circular
diurnal motion on a rotating earth; for even on a rotating earth they would
retain enough differences.

290—3: Logical distinctions

One should not confuse (F293) rhetorical appeals with arguments any more
than sound with unsound arguments, or the correctness of an argument-form
with the truth of premises. Examples of mere rhetorical appeals are the
following: to say, as the Aristotelians do, that it is more fitting to have the
earth at the center of the universe (rather than between Mars and Venus)
because that way there is a better separation of the pure (heavenly bodies)
from the impure (earth); or to say, more plausibly, that it is worse to have the
(impure) earth at the center because this would be like having a lazaretto in
the heart of a city; or to say, with Copernicus, that it is more fitting to have
the sun at the center because this would be like having a great lamp in the
middle of a temple so that it can best illuminate all places. An example of an
unsound argument is to say that the sun and stars move because they are
more like the planets than the earth is, and this is so because the sun, stars,
and planets are all incorruptible and unchangeable, while the earth is not; the
Aristotelians beg the question here since they base their last premise (the
corruptible/incorruptible distinction) on the difference with respect to
motion of the earth on the one hand and all the other bodies on the other.
An example of an argument with valid form but false premise is to say that
the sun and stars move because they are more similar to the planets than the
earth is, and this is so because the sun, stars, and planets are all luminous,
whereas the earth is not; here the last premise is not true; the planets are not
luminous. Thus the following has both valid argument-form and true pre¬
mises: the planets and the earth are alike in that they all lack their own light,
and they all differ from the sun and stars, which have their own light; since
the planets obviously move, the earth probably does also, but the sun and
stars do not.

293—8: The superficiality of abstract answers

It can be a sign of superficiality (F295) to ignore the empirical-factual aspect
of a question and deal only with its conceptual-verbal aspect in cases when
the two approaches suggest conflicting answers. For example, consider the
question of which is more probable and credible: the Copernican commitment
to immense stellar distances, or the Ptolemaic commitment to immense stellar
 THEORY AND PRACTICE 127

speeds of diurnal motion;and consider Kepler’s argument that the Copernican

view is more probable because it is harder to conceive a property beyond the
model of the subject than to increase the subject without the property. It is
true but irrelevant to object that Ptolemy’s immense diurnal stellar speeds are
not inappropriate, in the sense that they are just the right ones needed to
complete merely one revolution per day at their distance from the center. It
is much more important that those speeds are inappropriate in the sense that
they go against the norm, observable in the revolutions of the planets and of
Jupiter’s satellites, namely that the speed of revolution decreases and the
period increases as the distance from the center increases.

299—337: Quantitative data and qualitative conclusions in astronomy

Quantitative astronomical data are so problematic that at best they yield only
qualitative conclusions. For example, consider the observational data avail¬
able for the 1572 nova; when the distance is computed from the difference
between the nova and the polar parallaxes due to the change in observer
position, the data yield physically impossible distances in most cases and
widely different distances in the rest. The physically impossible distances are
those implied by negative parallax difference; the other distances vary from
1/48 to at least 716 earth radii. This may lead one to reject all the data as
unreliable, or else to correct them appropriately. The failure to do either was
Chiaramonti’s main mistake. One way to correct the observational data would
be from the point of view of the qualitative question whether the nova is
sublunary or superlunary; the observational data could be divided into those
that yield sublunary distances and into those yielding superlunary distances;
one could then try to make consistent the sublunary set of data on the one
hand, and the superlunary set on the other, by changing the observational
values. When this is done, the corrections needed to harmonize the super¬
lunary data are very much smaller than those needed for the sublunary data,
from which one might conclude that the nova was probably superlunary.
Another way to correct the observational data would be from the point of
view of the range of distance suggested by most observations; then only the
data yielding impossible distances would be changed; since almost all of the
data yield impossibly great distances and constitute a majority of the data,
the great majority of the observational data may be said to suggest a super¬
lunary distance.

337—46; The reliability of astronomical measurements

Quantitative astronomical data are not equally reliable; what’s needed is a
128 CHAPTER 5

critical evaluation of the methods for collecting them. For example, the data
concerning the elevations of the 1572 nova on positions off the meridian are
practically worthless because such positions are much more difficult to
measure than its meridian positions, whose measurements have been shown to
be quantitatively worthless and usable only for qualitative conclusions. The
most reliable data are from the measurements of distance between the nova
and the pole or some neighboring star; such measurements are the simplest
ones to make and neither the effects of refraction nor the use of the sextant
present any real problem.

346—9: Procedure vs results, and Aristotle’s authority

Aristotle’s authority is misused by the Peripatetics when they follow him in
a specific conclusion he reached rather than in the manner whereby he arrived
at it. The advantage of following his procedure is that one can apply it to a
novel situation to reach novel conclusions when conditions have changed. For
example, since Aristotle argues that the earth is at the center of the world
because it is the center of celestial revolutions, it is reasonable to conclude
that, if it could be shown that the sun and not the earth is at the center of
celestial revolutions, then he would conclude that the sun and not the earth is
at the center of the world; the novel situation here would be one where the
available evidence made it more likely that the sun rather than the earth was
at the center of celestial revolutions. In other words, the follower of the spirit
of Aristotle, as opposed to the blind follower of his letter, should be able to
see that the proposition that the center of the world is the center of celestial
revolutions is more basic than the proposition that the earth is the center of
the world, and from this to prefer the former in cases of conflict.

349—68: Simplicity, insight, and sense experience

Insight is admirable, but the only reliable procedure is to take sense experience
seriously and combine it with reason (F355-6, 363), that is, to infer truth
from rational simplicity only where there is no disagreement with sense
experience. For example, one cannot but marvel at the insight of Aristarchus
and Copernicus who became convinced of the truth of the heliocentric
system, because of its rational simplicity, despite the insuperable empirical
difficulty that the planets did not show the necessary changes in size and
shape (F355, 362—3, 367, 368); the Copernican attitude was insight rather
than mere intellectualistic stubbornness because the telescope has nowadays
made possible the observation of the necessary changes, which constitute
direct evidence from sense experience of the truth of the heliocentric system.
 THEORY AND PRACTICE 129

However, though admirable, Copernicus’s procedure is not rationally justi¬

fiable because he was inferring truth from mere simplicity and disregarding
clear empirical counterevidence (F354-368). Nevertheless, Copernicus’s
conclusions are the most probable ones nowadays (F349—354) when we can
observe with the telescope phases in Venus and great changes in the apparent
size of the planets.

368—72: Computational vs philosophical astronomy

There are two distinct attitudes possible in astronomy, a merely computa¬
tional attitude and a philosophical attitude. The merely computational
astronomer tries to give the reasons for celestial appearances in terms of
devices from which those appearances can be calculated, without worrying
about the physical reality of these devices; the philosophical astronomer tries
to give the reasons for celestial appearances only in terms of devices that can
be physically real. Ptolemaic astronomy is a good example of merely compu¬
tational astronomy; its physically questionable devices are: (1) circular
motions which are uniform with respect not to the center but to a point off
the center; (2) circular motions which are simultaneously from east to west
and from west to east; (3) planetary motions which are retrograde; and (4)
epicycles which are so large that they cross the orbits of other bodies.
Copernican astronomy, on the other hand, is a good example of philosophical
astronomy; its primary assumption is the physical reality of the earth’s annual
motion, which gives such a simple explanation of retrograde planetary motion
that this alone would be conclusive to an unbiased mind.

372—83: Logic vs methodology in theoretical explanation

Logically speaking, whenever certain observable facts are shown to follow
from a theory, this does not constitute a conclusive argument for the truth of
the theory, unless one shows that there is no other way of explaining those
facts (F377-80). Nevertheless, methodologically speaking, given two dif¬
ferent theories each of which explains the relevant facts, the theory which
has greater explanatory power and which is less ad hoc is the better and more
probable one. For example, consider the intricate way in which sunspots can
be observed to move across the solar disc, namely along curved and slanting
paths whose curvature and slant are continuously changing in such a way that
they curve downwards and slant upwards for half a year, they curve upwards
and slant downwards for the other half a year, they are straight twice a year
when the slant is greatest, and their slant is absent twice a year when their
curvature is greatest. This motion of sunspots can be shown to follow very
130 CHAPTER 5

easily by assuming that the earth has the annual and the diurnal motion and
that the sun rotates around an axis which is inclined to the plane of the
ecliptic. The same observations can also be explained by assuming that, in
addition to the diurnal and annual motions, and in addition to an axial
rotation around an axis inclined on the ecliptic, the sun has a fourth motion,
namely that its inclined axis of rotation itself rotates once a year around the
axis of the ecliptic. Now aside from the obvious greater simplicity, the
geokinetic theory has greater explanatory power insofar as it can explain the
yearly period of the motion of sunspots, whereas there is no reason why this
should be so in the geostatic theory (F382).
Moreover, and more importantly, the geostatic theory is completely ad
hoc (F382); the sun’s fourth motion is added merely out of the necessity of
accounting for the observations under consideration. By contrast, the
geokinetic theory not only does not have any such ad hoc element, but it
made possible or at least facilitated Galileo’s discovery of the relevant facts;
the geokinetic theory is freed of an ad hoc character by the mere fact that the
discovery of the full details of the motion of sunspots could have been made
by predicting (F379) these on the basis of it (F372—80). The geokinetic
theory is therefore better and more probable.

383-5: The Bible

It is in bad taste and disrespectful to inject Biblical references in a scientific
discussion. That is what the author of the “booklet of conclusions” does
when he says that, according to the Copernicans, Christ ascended into hell
and descended into heaven, and when Joshua ordered the sun to stand still
it was the earth that stood still.

3 85—99: Objectivity and the concept of size

The concept of size is in some ways subjective and relative, but it is not tele¬
ologically anthropocentric.
Our ideas about the size of fixed stars are subjective in the sense of being
dependent on our perception of apparent stellar diameters (F387—91); now,
under ordinary conditions apparent stellar diameters are perceived much
greater than they are by factors ranging from 24 (F389) to more than one
hundred (F388), due to the irradiation of the eye which makes us see them
with a halo; it is possible, however, to reduce or eliminate this observational
error by observing them with a telescope, or during the day, or with the help
of a fine string behind which they can be made to hide. Moreover (F393—4),
our ideas about very large sizes are unclear and confused in the sense that our
 THEORY AND PRACTICE 131

imagination cannot distinguish between one such size and another; an

example would be the size of the universe, as measured by stellar distances.
Finally (F396), the concepts of ‘large’ and ‘small’ are relational in the sense
that to say that something is large (or small) makes sense only relative to a
reference class; hence, to say that the stellar sphere of the Copernican system
is too large means that it is too large as compared to other celestial spheres;
but this is not so because if we compare the Copernican stellar sphere to the
lunar sphere, the ratio is much smaller than that obtained from a comparison
of an elephant to an ant or of a whale to a gudgeon. Though the concept of
size is subjective and relative in these ways, it is not, however, teleologically
anthropocentric (F394—96, 397—98); that is, it is wrong to think that the
size of things fulfills a purpose definable in terms of human interests; to think
so would be as absurd and arrogant as it would be for a grape to think that,
just because the sun acts on it in a way which is perfectly suited to cause its
ripening, the sun exists and acts only for its benefit; hence it is wrong to
argue that the universe cannot be as large as required in the Copernican
system because the space between Saturn and the fixed stars would be super¬
fluous and useless; this argument involves several nonsequiturs: to ground
nonexistence on superfluousness, superfluousness on lack of purpose for man,
and lack of purpose for man on ignorance of the purpose for man.

399—406: Ad Hominem arguments

Ad hominem arguments must be used with care if they are to have a method¬
ological function, in addition to their rhetorical one. That is, they are easily
effective for the purpose of discrediting the person whose beliefs are used
as premises, but such arguments are insufficient to resolve the issue of the
factual truth of these beliefs (F399). For an ad hominem argument is one in
which the arguer derives a conclusion not acceptable to an opponent from
premises which are accepted by him (the opponent), but which may not be
too well grounded; hence, such an argument does not establish its conclusion,
but only that one cannot both accept the premises and deny the conclusion.
For example (F399—400), consider the Tychonic argument that if the
earth has the annual motion, since there is no annual change in the stellar
appearances, then the fixed stars are at such an immense distance from the
earth that, in order to be visible as they are, their size must be of the order of
planetary orbits. This argument is ad hominem against Copernicus in the
sense that the proposition that there is no annual change in stellar appear¬
ances, though accepted by him, has not been supported by evidence or
argument. In fact, no one has tried to observe such changes in stellar appear-
132 CHAPTER 5

ances, or even inquired into exactly how stellar appearances should change,
instead there are a lot of misconceptions about the matter. One misconception
(F400-3) is that the elevation of the celestial pole should change, which is
absurd because if the earth had the annual motion then the celestial pole
would be defined by the terrestrial one, whose elevation can change only by
moving on the earth’s surface; Tycho himself (F400) seems to have made this
error since, in discussing the matter, he claims that no change in polar
elevation is in fact observed and fails to note that the point is misconceived.
Another misconception (F403—■4) is that the annual changes in the elevation
of the fixed stars should be comparable to those resulting by moving on the
earth’s surface, which is wrong because the latter is highly curved, whereas
the annual motion takes place on a plane. A third misconception (F405—6)
is that the annual changes for fixed stars would be comparable to those
observable in the sun, which though more difficult to clear up is also wrong,
as it will be shown later (F406—16, and F416—23).

406—9: The value of criticism

It is advisable to test one’s arguments and theories against the criticism of
men of sound judgment and sharp intellect; for if they survive the criticism,
we can have great confidence in their correctness (F408). For example, con¬
sider the argument that, for fixed stars lying in the plane of the ecliptic, the
earth’s annual motion would not cause an annual change in their apparent
position, since their elevation relative to the ecliptic would stay constant
regardless of how the earth moved in that plane. If one presents this argument
to an intelligent person like Sagredo, his responses lead to the following
qualifications. That is, the conclusion would be right if all fixed stars were
equidistant from the center; but if their distances vary, as they probably do,
then some changes should appear in the relative positions of those lying in
the ecliptic plane.

409—14: Theories and observations

When one succeeds in detecting certain specific observational consequences
of an hypothesis, and there is no other way to explain these observations,
then one must accept the hypothesis; but when one fails to detect these
consequences, and there is another way to explain this failure, then one need
not reject the hypothesis (F413—14).
For example, the observational consequences of the earth’s annual motion,
as regards the appearance of the fixed stars, are the following (F409—12):
annual changes in the apparent magnitude of the stars, the change being
 THEORY AND PRACTICE 133

greatest for those in the plane of the ecliptic and decreasing to nothing for
those near the pole of the ecliptic; annual changes in elevation above the
ecliptic, the change being greatest for stars near its pole and decreasing to
nothing for those in its plane; differences in both these changes (in magnitude
and in elevation) due to the distance of the stars, the changes being greater
for nearer stars. Now, if one were able to detect these variations in stellar
appearances, since there would be no way to explain them other than by the
earth’s annual motion, then we would have to conclude that the earth had
this motion. But the fact that no such variations are generally known or have
generally been observed does not force us to conclude that the earth lacks the
annual motion, because such general ignorance is explicable by the fact that
no one has searched for them seriously or systematically enough, or with the
appropriate instruments, or with the necessary skill (F413 —14). Even if we
had seriously searched and failed to observe these variations, we still would
not have to reject the earth’s annual motion, since our failure could be
explained by stellar distances so great, and hence changes so small, as to be
undetectable with our available instruments (F413).

414—6: The inexactness of astronomical instruments

Available astronomical instruments cannot be trusted much for the obser¬
vation of very small effects, such as stellar parallax, even though some instru¬
ments are better than others. In fact, it is common for astronomers to dis¬
agree by several minutes in the values they get for the elevation of fixed stars,
and even of the pole, let alone for the comets or new stars. The fact that
Tycho’s instruments are the best does not mean that they are good enough
for the very exact observations needed to observe annual parallaxes. What is
needed is instruments whose sides are miles long, so that differences of
seconds in celestial elevation correspond to distances of the order of braccia
along the instrument. For this purpose one can use topographic features on
the earth’s surface such as mountains. An example of such an instrument to
try to observe stellar parallaxes, would be to observe the changes in the way
that a particular star would be hidden by a beam on top of a building, at the
top of a mountain, when one carried out the observations from the valley
below, at different times of the year.

416-25: The complexity of simplicity

The concept of simplicity is not simple; it has at least two aspects, mathe¬
matical and physical. For example, consider the Copernican explanation of
the seasonal variations in solar elevations and in length of nights and days
134 CHAPTER 5

(F416-22). Someone (Salviati, F422) may be impressed by the fact that this
explanation uses two terrestrial motions which are (1) both in the same
direction, namely west to east, and (2) performed in periods corresponding to
their magnitudes. He may then (F422-3) contrast this to the fact that the
geostatic explanation requires celestial motions which (1) violate the
symmetric character of the relationship between speeds and size of the thing
moved, by having the very large highest sphere move at a very high diurnal
speed, and the very much smaller planetary spheres move at a very much
slower characteristic speed; (2) are performed in opposite directions; and
(3) are performed against one’s individual inclination, in the case of the
participation by the planetary spheres in the diurnal motion. He may con¬
clude (F423) from this that the Copernican explanation is superior in
simplicity, and hence more probable. Another person (Sagredo, F423) may
agree with the conclusion but give different reasons for the greater simplicity,
namely its being more in accordance with the widely accepted philosophical
principles that nature (a) does not multiply entities without necessity, (b)
uses the simplest and easiest means in producing its effects, and (c) does
nothing in vain. A third person (Simplicio, F423) might object to the first
person’s concept of simplicity as being excessively mathematical in its stress
on the correspondence between periods of revolution and magnitude of orbit,
or symmetry between speed and size of body moved, which involve quantita¬
tive relationships. Instead he may be impressed (F424) by the fact that the
geostatic explanation attributes a single natural motion to each simple body,
whereas the Copernican explanation violates this principle; and from this he
may conclude that the Copernican explanation is inferior in simplicity. The
difference in their concepts of simplicity would remain even if it were shown
to him that, in a sense, the Copernican explanation does not violate the
principle that a simple body must have only one natural motion (F424—5),
for the three or four terrestrial motions supposedly required can be reduced
to one as follows: the straight-downwards motion of terrestrial bodies is a
motion that belongs to the parts of the earth, not to the whole earth; the
earth’s diurnal and annual motions since they are in the same direction, are
two only in a special sense, a sense related to that in which a ball rolling down
an inclined plane has two motions; and finally, Copernicus’s “third motion”
does not exist, the phenomenon is really a kind of rest.

425—31: Open-mindedness, intellectual cowardice, and curiosity

Curiosity and awareness of the infinitesimal extent of human knowledge tend
to make one willing to pay attention to new ideas, whereas intellectual
 THEORY AND PRACTICE 135

cowardice hampers open-mindedness (F426). For example, consider Gilbert’s

idea that the primary substance of the terrestrial globe is loadstone. Peri¬
patetics have generally dismissed this idea as chimerical and insane, even
though neither Aristotle nor anyone else has refuted it. Such dismissal is as
cowardly as being afraid of one’s shadow (F426—7). For it is easy to show
that the idea is initially as plausible as other theories about the primary sub¬
stance of the earth (F427—9). In particular, to think that this primary sub¬
stance is earth in the sense of topsoil is implausible, since digging into the
ground reveals hard substances like rocks, since weight and pressure would
harden underground substances even if they were like soil, and since this
theory is probably the result of the linguistic accident that the word ‘earth’
denotes both the terrestrial globe and the substance of topsoil. Moreover,
without actually repeating Gilbert’s proof, a description of his procedure
ought to be sufficient to awaken one’s curiosity and take his idea seriously
(F429) (which incidentally illustrates the fact that the method of arriving at
an idea as well as its content may motivate the open-minded individual).
Gilbert’s procedure is to infer the nature of the substance from knowledge of
its properties (F429-30) and may be compared to the following: suppose
(F430) you were given samples of a large number of substances all covered
by a cloth so that you could not see what they were; suppose further that
your problem was to guess what the substances were; finally, suppose that
one of these covered samples exhibited all the properties known to be present
only in loadstones; then, clearly, you could infer that the sample in question
was loadstone. Similarly, Gilbert reaches his conclusion (F430-31) after
finding that the earth shows all the properties of a loadstone: for example, a
magnetic needle tilts from the horizontal as it is moved away from the
equator; and the poles of a magnet behave as if they were in the presence of a
bigger magnet, namely they have equal strength only at the equator, while its
south pole is stronger in the northern hemisphere, and the north in the
southern.

431-6: Causes: role, discovery, and justification

The development of a new field by the conception of new ideas, by the
observation of new effects, and by the initial investigations of causes requires
uncommon originality and so is much more admirable than the subsequent
perfecting of the field; for example, Gilbert is to be envied for his magnetic
philosophy (F432-3). Nevertheless, the explanation of the causes of puzzling
and curious known effects is an obvious second step as well as one of the
most intellectually rewarding activities (F432; F434, Sagredo); one such
136 CHAPTER 5

effect is the fact that an armatured loadstone can sustain much more weight
(in some cases 80 times more) than without the armature. Now, in causal
investigations, one thing to avoid is explaining by naming, for example, to
explain magnetic phenomena in terms of the ‘sympathy’ between qualitatively
similar substances; to do so would be like painting by giving verbal descrip¬
tions (F436). Another thing to avoid is excessive readiness to accept a causal
explanation as conclusively proved; practice with mathematical demonstra¬
tions makes one less willing to do so (F432). Many of Gilbert’s explanations
show this excess (F432), whereas the explanation here to be given for the
greater weights sustained by armatured loadstones is an example where the
conclusiveness is greater than any other to be found in all magnetic philo¬
sophy (F434); it is, in fact, almost as great as the conclusiveness of mathe¬
matical demonstrations (F434, F435). What can give a causal explanation a
certainty comparable to the mathematical one is the direct confirmation of
the presence of the cause, after one has been lucky enough to think of it
from a knowledge of the effect (F434—5, Salviati). A quasi-method to acquire
this luck is to compare the effect to be explained with another significantly
similar but appropriately different effect whose cause is known (or con¬
textually unproblematic); then check whether an appropriate variation in the
known cause is responsible for a corresponding variation in the effect equiva¬
lent to the difference between the two effects being compared; if not, check
what new condition is present in the effect to be explained, which is not
present in the effect whose cause is known; this new condition must be the
cause we are looking for (F433). For example, let us compare the loadstone
with the armature to the loadstone without the armature; let us check
whether an increase in the force of the loadstone is responsible for the
increase in the weight that the armatured loadstone can sustain; we find that
this is not so, since the armatured loadstone does not act at a greater distance,
and since it does not attract more strongly through an interposed paper or
goldleaf; next we notice that the only variation introduced by the armature
is the difference in the contacts with the sustained weights; so we conclude
that this difference is the cause of the increase in the weight sustained with
the armature (F433). This difference in contacts amounts to a greater contact
area between iron and armature than between iron and loadstone, that is a
greater number of lines of force (so to speak, “filamenti.. . che collegano i
due ferri" [F434]) uniting the two surfaces. Fortunately, the actual presence
of this cause can be confirmed by a number of facts and experiments (and
this is what gives the explanation almost mathematical certainty): the texture
of iron is much finer than that of loadstone (F433-4); loadstones contain
 THEORY AND PRACTICE 137

many impurities, and one can observe that a needle is not attracted by them
but slips onto the surrounding loadstone substance (F434); the sharp point of
a needle is not attracted any more strongly by the armature than by the bare
loadstone (F434); if you let a needle be suspended between a bare loadstone
and a nail, with one end touching the loadstone and the other the nail, when
you separate the two the needle will remain attached to the one which it was
touching with its eye (F434); finally, if you take a loadstone and smoothen
and polish it on one side, you will clearly see on the resulting surface many
bright spots on a dark background, and you can observe that iron filings are
not attracted to these spots but to the surrounding surface only (F435—6).

436—41: Parts vs wholes

The principle that “the same is true of the whole as of the parts” is often mis¬
applied and misunderstood (F440). Sometimes it is ignored when it should be
applied. For example (F437—9), consider what Peripatetics are likely to say
when confronted with the fact that loadstones have at least three natural
motions. They would say that this does not refute the doctrine that a simple
body like the earth cannot have the several natural motions attributed to it in
the Copernican system, because loadstones are not simple bodies; however,
if loadstones are not simple, since they are parts of the earth, then the prin¬
ciple in question could be properly applied, and it would follow that the
earth is not simple either, thus allowing it to have its Copernican motions
without violation of the doctrine (of one natural motion per simple body).
Moreover, they would try to account for the multiple motions of loadstones
in terms of the different motions of their simple constituent parts; however,
if these parts are elementary bodies, then their natural motions would pre¬
sumably be straight (up or down), but unfortunately you cannot mix straight
motions and get a circular motion for the result (and here too we would have
a proper application of the principle being discussed); hence Peripatetics
would be led to the absurd consequence of allowing celestial substances as
component parts of loadstones.
Sometimes, the principle in question is applied when it shouldn’t. For
example, consider (F439-40) Gilbert’s own assertion that if a small load¬
stone sphere were balanced properly and free to move, it would turn on its
axis; however, there would be no reason for this to happen even on the
assumption that the earth is a loadstone and has the diurnal motion, for such
a spherical loadstone would be doing the same as the whole earth by turning
around the center of the earth rather than around its own center; there would
be a reason for it to happen only if the small sphere were prevented from
138 CHAPTER 5

turning around the center of the whole. Another example of a misapplication

is (F440—1) Sacrobosco’s argument that since small drops of water can be
seen to arrange themselves in a round shape, that shows that the whole
element water on the globe forms a sphere; by such reasoning one would have
to conclude the absurdity that any body of water forms a sphere; whereas the
relevant sense in which the same is true of the whole as of the parts concerns
the tendency of water to form a sphere around the center of gravity of the
whole (the earth’s center).

442—5: Choosing the effect in a causal investigation

Knowledge of the effects is helpful and necessary for the investigation and
discovery of causes, but such knowledge is always limited both in extent and
reliability; hence the best way to proceed is to start from those features of
the effects that are certain and principal, and to try to arrive at a cause which
is primary and which would serve as a pointer for the direction to follow in
extending and deepening our knowledge in a consistent manner (F443—4).
For example (F444—5), in the investigation of the cause of the tides, the
principal and best known effects are the fact that the time interval between
high and low tide is six hours, and the fact that the motion of the water is
merely upwards and downwards in some places, merely eastwards and west¬
wards in others, and a combination in still others. The primary cause that will
be proposed is the motions of the earth (F445), to which the necessary
additions and articulation will later be made, by taking the fluid properties
of water as secondary causes (F454—62), and by explaining the monthly and
annual periods as due, respectively, to actual variations in the earth’s annual
motion and to merely effective variations in its unchanging diurnal one.

445—62: The artificial reproduction of cause and effect

In causal investigation, a proposed cause should be rejected if it does not
allow an artificial reproduction of the effect, and it should be accepted if
such an artificial reproduction is possible (F447, 456; cf. 471—2). For
example, the only cause of the tides that passes this test is the motion of the
earth. The failure to pass this test is the real reason why one finds implausible
the theories of lunar attraction, lunar heat, differences in water depth, and
seabed caves; none of these conditions would produce up and down, or back
and forth motion in only some places of a container of water (F447-9). On
the other hand, by accelerating or retarding a container of water one can
make the water rise and fall at its extremities, and move back and forth at the
middle, as one can observe on boats that carry water (F450-1); and the
 THEORY AND PRACTICE 139

motions of the earth would produce a daily acceleration and a daily retarda¬
tion at every point, as the diurnal motion component of this point is respec¬
tively added to or subtracted from the annual motion component (F452—4);
moreover, the various fluid properties of water, which can be investigated by
means of artificial models (F456), are bound to act as secondary causes
interacting with the primary one (F454—6), to produce the complexity of
effects observable in the motion of sea water (F457—62), some of which
effects are observable in artificial models (F457, F459).

462—70: Reasoning ex suppositione

Reasoning ex suppositione involves two steps: a known effect is explained

by making a supposition, and then one takes the effect as evidence and con¬
firmation for the supposition; such reasoning is pointless if the supposition is
false, and invalid if there are in the context reasons to reject the supposition
(F462). For example, consider the argument for the earth’s motion based on
the tides (F462, postil). This argument is ex suppositione in the above sense,
and hence must be defended from any contextual objections to the supposi¬
tion that the earth moves, which would render the argument invalid. One
such objection (F363-4) is that if the earth’s motion causes the tides then it
would cause a perpetual westward wind, which however does not exist. The
answer to this is twofold: first, the earth’s motion could cause the tides
without causing the perpetual wind because air unlike water does not have
the propensity of retaining acquired motion, which is the property that
makes water move back and forth when accelerated or retarded; second,
prevailing westward winds do exist, especially in equatorial regions. Another
objection (F467-8) would be that since prevailing westward winds do exist,
and since they could be explained as caused by the diurnal motion of the
lunar orb, and since this explanation would be more economical and less
innovating (F467) than a geokinetic explanation, therefore we ought to reject
the earth’s motion. The problem with this objection is that (F468-70) the
prevailing winds could not be produced by the turning of the lunar orb since
it is supposed to have a very smooth surface, and hence its motion could not
be transmitted to the element fire; moreover, the turning of this element
could not produce the turning of the denser element air. On the other hand
(F465-6), these winds could be easily produced by the earth’s diurnal
motion, as the air lying over its larger smoother sections would be less easily
carried along than that trapped closely among mountains, which would
constitute a further reason favoring the supposition in question.
140 CHAPTER 5

470—1: Confessed ignorance and occult qualities

The confession of ignorance is sometimes preferable to the claim to know, for
such claims are sometimes a mere figment of the imagination (F470, F446)
and the result of vanity or verbosity (F470). For example, it would be better
to say that you don’t know the cause of the tides, rather than explaining
them in terms of attraction by occult qualities, which is intellectually
repugnant (F470), or in terms of lunar or solar heat and light, which is
empirically ridiculous and absurd (F470—1, F445—7).

471 —84: The method of concomitant variations

Whenever there is a regular variation in the effect, there must be a corres¬
ponding variation in the cause (F471). Or to be more exact, whenever an
effect varies in a regular way, it is best to explain this variation in terms of
the same primary cause, by means of corresponding variations in it (F471—2),
rather than in terms of a novel cause (F474), which would run the risk of
being ad hoc. For example, consider the monthly and annual variations in the
magnitude of the daily tides. It would be best to account for such monthly
and annual periods in terms of variations in the accelerations and retardations
that cause the diurnal period, namely variations in the speed of the earth’s
annual or diurnal motions or both (F471—4). Now, there are independent
reasons to believe that the earth’s annual motion would undergo monthly
variations: the evidence comes from what may be regarded as a highly con¬
firmed law of revolution, namely that whenever the cause of circular motion
remains unchanged, such motion is slower along larger circles and faster along
smaller ones (F474—8); for the earth-moon system would go around the sun
in a circle whose effective radius would undergo monthly variations as a result
of the moon being now in opposition and now in conjunction with the sun
(F477—8). The fact that there is such independent evidence for variations in
the earth’s annual motion is more important than the fact there is no direct
evidence for them from astronomical observations; for such a lack may be
due to the failure to search for the variations, or to their very small
magnitude (F479-82); hence we are entitled to conclude that the monthly
period of the tides is caused by these lunar variations in the earth’s annual
motion (F479). Regarding the annual period, the inclination of the earth’s
axis would cause annual variations in the effective speed of the diurnal
motion of points on the surface of the earth, the effective speed being the
speed projected on the plane of the ecliptic (F482^4). Hence, by the same
principle, we may conclude that the annual period is caused by such annual
variations in the earth’s actually unchanging diurnal motion.
 THEORY AND PRACTICE 141

484-9: Searching vs results: the supremacy of method

An idea may need refinement (F484-5), and yet it may be very valuable if
it is essentially correct and points in the right direction (F485—6); conversely,
a theory may be highly supported (F487), and yet it may be false (487—8), if
it is possible for God to have created a world not in accordance with it
(F488); this is so because the mental exercise required in the investigation is
very valuable and probably the most important thing (F489). In other words,
the important thing in natural philosophy is that the approach or method be
right.
The geokinetic theory of the tides is a good example of this. It provides
only a general account of the basic diurnal, monthly, and annual periods, and
only a sketch of the complicated and somewhat irregular details, not all of
which are well known anyway (F484-5); yet it represents a breakthrough,
especially insofar as it suggests a mechanical approach (F485-6). It con¬
stitutes a very strong argument in favor of the earth’s motion (F487), yet it
could be mistaken (F487—8), because it is possible for God to cause the
motion of tides in some other way.

NOTES

1 Two sets of page numbers are provided here: those before the slash refer to G. Galilei,
Opere, edited by Favaro, Vol. 7; those after the slash refer to G. Galilei, Dialogue, tr. S.
Drake.
2 The page references here are all to Favaro’s edition, both those in the various sub¬
headings, and those in parenthesis in the text; the latter numbers are prefixed by an ‘F’,
which is short for ‘Favaro’, as a reminder that his edition is being referred to.
 PART II

LOGICAL AND METHODOLOGICAL CRITIQUES

*

*
 CHAPTER 6

CONCRETENESS AND JUDGMENT:

THE DIALECTICAL NATURE OF
GALILEO’S METHODOLOGY

IMPLICIT PHILOSOPHY

The first thing to note about the methodological content of Galileo’s Dialogue
is that it is more or less evenly spread throughout the book. As Table I
(below) indicates, the First Day contains on the average a distinct method¬
ological discussion about every 10 pages,1 the Second Day about every llA
pages, the Third Day every 9 pages, and the Fourth every 8 pages. In other
words, the methodological intensity of the book does not vary much from
beginning to end.

TABLE I

Frequency of philosophical discussions

Number of

Subdivisions Text pages Commentary pages

DAY Total Average Total Average

I 10 99 10 6.3 0.63
II 23 167 7 1/4 14.4 0.63
III 16 143 9 11.0 0.69
IV 6 48 8 3.3 0.55

Second, one cannot help but be impressed by the quantity and variety of
philosophical topics which are discussed. Though the number of philosophical
passages distinguished in my reconstruction is 55, and though many of them
overlap, the number of individual philosophical topics touched upon is at
least 92, as an index of these topics shows (Table II).
This large number of individual topics cluster around a much smaller
number of themes (15), as Table III shows. Nevertheless, there is no one
philosophical thesis that predominates, though some are discussed more
frequently than others. It turns out that there are three topics which are
much more frequently discussed than others. The topic of sense experience,
and the topic of causes, explanation, and understanding occur each in 11
different passages; logic and reasoning in 10. The topics of unconscious

145
 146 CHAPTER 6

TABLE II

Index of individual philosophical topics

At least 92 individual topics are mentioned as indicated in this table.

Abstract entities, 229-37 Insight, 349-68

Anthropocentrism, 82-95, 289-90, Introspection, 112—7
385-99 Knowledge, and information, 180-8
A priori vs a posteriori, 71-82 Knowledge, and recollection, 217-23
Ad hominem arguments, 399-406 Knowledge, limitations of, 425-31
Aristotle’s authority, 59-71, 71-82, Knowledge, unconscious, 193-6
132-9,346-9 Logic, 57-9, 59-71
Astronomical instruments, 414-6 Logical analysis, 150-9, 159-69,
Astronomy, 299-337, 368-72 290-3
Authority, 59-71, 71-82,132-9, Logical distinctions, 290-3
346-9 Logic vs methodology, 372-83
Bible, 383-5 Mathematical practice, 431-6
Causal investigation, 436-41,471-84 Mathematical reasoning, 33-8
Causes, 237-44, 260-72, 431-6, Mathematics, 33-8, 229-37
445-62 Mathematics and physics, 244-60
Certainty, 431-6 Mathematics, applicability of, 229—37
Conceptual frameworks, 38-57 Measurement, astronomical, 337-46
Cognitive awareness, 112-7 Methodological point of view, 71—82
Comprehensibility, 260-72, 281-88 Method of concomitant variations,
Concepts and facts, 293—8 237-44,471-84
Concrete entities, 229-37 Method, importance of, 244-60, 484-9
Cowardice, intellectual, 425-31 Naming, as a way of explaining, 260-72,
Criticism, 206-9, 214-7,406-9 431-6
Curiosity, 425-31 Nature, primacy over man, 289-90
Data, 337-46, 299-337 Number mysticism, 33-8
Deception of the senses, 272-81 Objectivity, 385-99
Digressions, 188-93 Observations and theories, 409-14
Discovery vs justification, 71-82 Occult qualities, 470-1
Empiricism, 57-9, 206-9 Open-mindedness, 150-9, 425-31
Errors, 117-24 Opportunism, academic, 132-9
Experience, 169-80 Optimism, 124-31
Experimentalism, 206-9 Originality, 431-6
Experiments, 95-112, 169-80, 206-9, Parts vs wholes, 436-41
209-14 Physico-mathematical synthesis, 244—60
Explanation, 260-72 Physical reality, and mathematics,
Explanation, causal, 237-44,431-6 229-37
Explanation, theoretical, 372-83 Platonic mathematicism, 33-8
Geometry, and philosophy, 223-9 Practice vs theory, 59-71, 71-82
Human understanding, 124-31 Probability, 139-50
Ignorance, 132-9,470-1 Probable arguments, 139-50
Inconceivability, 281-8 Procedure vs results, 346-9
Independent-mindedness, 132-9 Pythagorean mathematicism, 33-8
 CONCRETENESS AND JUDGMENT 147

TABLE II (Con’t)

Qualitative and quantitative, 299-337 Searching, vs results, 484-9

Quantitative data, 337—46 Sense experience, 57-9, 349-68
Rational-mindedness, 150-9 Simplicity, 139-50, 349-68
Reason, 349—68 Simplicity, complexity of, 416-25
Reasoning, 169—80 Socratic ignorance, 124-31
Reasoning, circular, 196-206 Socratic method, 169—80,180-8,139—6
Reasoning, ex suppositione, 462-70 Teleology, 82-95
Reasoning, hypothetical, 196-206 Understanding, 180-8, 214-7, 260-72

TABLE III

Index of main philosophical topics

The large number of individual philosophical topics cluster around a much smaller
number of themes.

Anthropocentrism: teleology, 82-95; human understanding, 124-31; primacy of nature

over man, 289-90; objectivity, 385-99
Astronomical investigation: quantitative vs qualitative considerations, 299—337,
measurement and reliability of data, 337—46; computational vs philosophical
attitude, 368-72; instruments, 414-6
Authority: 59-71,71-82, 132-9, 346-9; Bible, 383-5
Causes, explanation, and understanding: understanding, 180-8, 214-7, 260-72;
causal explanation, 237—44, 431—6, 471—84; causes, 237—44, 260—72, 431—6,
442-5, 445-62; method of concomitant variations, 237-44, 471-84; explanation,
260—72; naming as a way of explaining, 260—72, 431—6; truth and comprehens-
sibility, 260-72, 281-8; theoretical explanation, 372-83; occult qualities, 470-1
Conceptualization: conceptual frameworks, 38-57; inconceivability, 281-8; abstract
answers, 293-8; parts and wholes, 436-41
Criticism: 206-9; errors, 117-24; criticizing, 214-7; value of criticism, 406-9
Experience: conceptual frameworks, 38—57; empiricism, 57—9, 206—9; experiments,
95-112, 169-80, 206-9, 209-14; physico-mathematical synthesis, 244-60;
deception of the senses, 272-81; and reason, 349-68; theoretical explanation,
372-83; theories and observations, 409-14
Ignorance: 132-9; Socratic, 124-31; limitations of knowledge, 425-31; confession of,
470-1 .
Logic: 57-9, 59-71; probable arguments, 139-50; rational-mindedness, 150-9, logical
analysis, 150—9, 159-69, 290-3; reasoning, 169-80; circular reasoning, 196-206;
hypothetical reasoning, 196—206; logical distinctions, 290—3; ad hominem argu¬
ments, 399-406; reasoning ex suppositione, 462-70
Mathematics: reasoning, 33-8; number mysticism, 33-8; Platonic mathematicism,
33-8; Pythagorean mathematicism, 33-8; geometry and philosophy, 223-9;
applicability to physical reality, 229—37; physico-mathematical synthesis, 244-60,
mathematical practice, 431-6
Method: methodological point of view, 71-82; importance of, 244-60, 484-9; results
vs procedure, 346-9, 484-9; logic vs methodology, 372-83 _.
148 CHAPTER 6

TABLE III (Con’t)

Open-mindedness: 150-9; curiosity and intellectual cowardice, 425-31

Practice vs theory, 59-71, 71-82
Simplicity: 139-50, 349-68; different concepts of, 416-25
Socratic method and unconscious knowledge: cognitive awareness, 112-7; Socratic
method, 169-80, 180-8, 193-6; unconscious knowledge, 193-6; knowledge and
recollection, 217—23

knowledge and the Socratic method, of authority, of mathematics, and of

method occur each 5 times; the doctrine of anthropocentrism, the nature of
astronomical investigation, the activity of criticism, and the topic of ignorance
are each discussed in 4 different passages; simplicity and the activity of con¬
ceptualization each 3 times; open-mindedness (in a sense distinct from but
related to independent-mindedness) and the distinction between practice and
theory are each discussed in two different passages.
The next interesting question to ask is whether the primary topics tend to
predominate in certain parts of the book. Table IV shows that, except for the
extremely short Fourth Day, the range of topics discussed in each day is
about the same. In other words, except for the preponderance of causal
discussions in the Fourth Day, no one major topic predominates for very long
in the rest of the book. It is true, however, that most major topics have places
of high concentration, with two or more consecutive passages devoted to the
same general topics; for example, astronomy on pages 299-346, authority on
pages 59—82, cause on pages 442—62 and 470—84, experience on pages
38-59 and 206-14, logic on pages 57-71 and 139-80, mathematics on
pages 223-37, the practice/theory distinction on pages 59-82, and the
Socratic method on pages 169—188.
One could go on with such attempts at “data analysis.” They are legitimate
insofar as, in part, they attempt to discern the structure of, or patterns in¬
herent in, the methodological content, and in part they attempt to system¬
atize this content. In other words, after having analyzed the (relatively)
explicit methodological content, a next step would be to study the implicit
philosophy. Now, this implicit philosophy can be of two kinds, namely the
relatively systematized philosophical doctrine implied by the relatively un¬
systematized methodological claims extracted so far, and the one (or the few)
pattern(s) exhibited by and inherent in the methodological content extracted
so far. Though these two types of implicit philosophy are abstractly possible,
it so happens that for the Dialogue, the philosophy clearly inherent in the
structure of the philosophical content precludes an implied philosophical
system. This can be seen as follows.
 CONCRETENESS aND JUDGMENT 149

TABLE IV

Distribution of philosophical topics

Day: I: 33-131 II: 132-298 III: 299-441 IV: 442-89

Anthropocentrism 82-95,124-31 289-90 385-99

Astronomy 299-337,
337-46,
368-72, 414-6

Authority 59-71,71-82 132-9 346-9,383-5

Cause 180-8,214-7 372-83, 442-5,

237-44, 260-72, 431-6 445-62
281-8 470- 1,
471- 84

Conceptualization 38-57 281-8, 293-8

Criticism 117-24 206-9, 214-7 406-9

Experience 38-57,57-9, 169-80, 206-9, 349-68,

95-112 209-14, 244-60 372-83
272-81 409-14

Ignorance 124-31 132-9 425-31 470-1

57-9,59-71 139-50, 150-9, 399-406 462-70

Logic
159-69, 169-80,
196-206, 290-3

Mathematics 33-8 223-9, 229-37, 431-6

244-60

71-82 244-60 346-9, 484-9

Method
372-83

Open-mindedness 150-9 425-31

Practice/theory 59-71,71-82

Simplicity 139-50 349-68,

416-25

Socratic method 112-7 169-80, 180-8,

193-6,217-23
 150 CHAPTER 6

The overwhelmingly obvious structural feature of the Dialogue was already

mentioned at the beginning of last chapter; it is the synthesis of practice and
theory, or experience and self-awareness, or to be more specific, scientific
practice and philosophizing about science. Every single passage in the book
supports such skillful combination by Galileo. Such characteristic may be
given the single label of ‘concreteness’. We may say then that the philosophy
exemplified by the Dialogue is a concrete philosophy of science. The second
general feature of its explicit methodological content is really a consequence
of the first, but it deserves special mention; it is judgment. That is, we find
Galileo emphasizing on various occasions different and in some cases opposite
things; for example, sometimes the need, and sometimes the superfluousness
of experiments; sometimes quantitative considerations, sometimes qualitative
considerations; sometimes antiverbalistic mathematical analysis, sometimes
verbal-oriented logical analysis; sometimes causal explanation, other times
phenomenological description. There is no inconsistency here, except when
the methodological claim is improperly generalized, something I have always
tried to avoid in my reconstructions above.
If concreteness and judgment are the two essential features of the book’s
methodological content, then it would be un-Galilean to try to arrive at a
philosophical system implied by it. The reason is that the systematization
would involve two things, namely generalizing some of the methodological
claims, and analyzing their interrelations in the abstract, that is in abstraction
of the scientific context; but the generalizing would conflict with the judg¬
mental feature, whereas the abstract interrelating would conflict with the
concreteness. This character of the philosophy implicit in the Dialogue is
not only clearly present, but also helps to understand two otherwise rather
puzzling facts, namely (1) the discrepancy between the place of Galileo in the
actual history of philosophy and in history-of-philosophy books, and (2) the
uniqueness of his role in the philosophy of science.

GALILEO AND THE HISTORY OF PHILOSOPHY

It is rather obvious that the kind of history of philosophy one writes depends
very much on what one conceives philosophy to be. Let me give a personal
anecdote to provide a very simple but incisive illustration of my point. A few
years ago I happened to have some discussions with a philosopher, now
deceased, who believed that the history of philosophy was almost a desert
between Aristotle and Frege, whom he regarded as the greatest philosopher
since Aristotle. It is easy to see what his concept of philosophy was: he
 CONCRETENESS AND JUDGMENT 151

equated philosophy with formal logic. From this point of view his historical
interpretation makes sense, just as it perhaps makes sense to regard him as
“the most brilliant American philosopher of this era”, a claim made by the
publisher of his posthumous book entitled Formal Philosophy.2
Let me now inject a little more substance as well as history into such an
illustration and step back to the year 1917, the year of publication of
Benedetto Croce’s Theory and History of Historiography,3 The theoretical
part of this treatise concludes with a chapter entitled ‘Philosophy and
Methodology’. In it Croce formulates the concept of philosophy as method¬
ology, according to which philosophical inquiry is the methodological aspect
of historical understanding, in the broad sense of the term ‘historical’ so that
historical understanding is any kind of concrete analysis, or applied know¬
ledge, or inquiry into specific problems. The implications for the historio¬
graphy of philosophy were expressed by Croce as follows:

Even the history itself of philosophy has hitherto been renewed only to a small extent,
in conformity with the new conception of philosophy. This new conception invites us to
direct our attention to thoughts and thinkers, long neglected or placed in the second
rank and not considered to be truly philosophers because they did not treat directly the
‘fundamental problem’ of philosophy or the great peut-etre, but were occupied with
‘particular problems’. These particular problems, however, were destined to produce
eventually a change of view as regards the ‘general problem’, which emerged itself
reduced to the rank of a ‘particular’ problem. It is simply the result of prejudice to look
upon a Machiavelli, who posited the conception of the modern state, a Baltasar Gracidn,
who examined the question of acuteness in practical matters, a Pascal, who criticized the
spirit of Jesuitry, a Vico, who renewed all the sciences of the spirit, or a Hamann, with
his keen sense of the value of tradition, as minor philosophers, I do not say in com¬
parison with some metaphysician of little originality, but even when compared with a
Descartes or a Spinoza, who dealt with other but not superior problems. A schematic
and bloodless history of philosophy corresponded in fact with the philosophy of the
‘fundamental problem’. A far richer, more varied and pliant [history of] philosophy
should correspond with philosophy as methodology, which holds to be philosophy not
only what appertains to the problems of immanency, of transcendency, of this world
and the next, but everything that has been of avail in increasing the patrimony of guiding
conceptions, the understanding of actual history, and the formation of the reality of
thought in which we live.4

To be sure, Croce did not practice what he here preached, and though his
works on Hegel, Vico, and Marx assure him a significant place in the history
of the history of philosophy, he did not interpret them as being essentially
methodologists; instead he gave critical explanations centered around, respec¬
tively, the concept of knowledge as concrete-universal, the concept of art as
expression, and the philosophy of the practical. Nor has a history of philo-
 152 CHAPTER 6

sophy as methodology been written by any of his followers, who do not even
interpret Croce himself as a methodologist, though in the abstract they pay
lip service to the doctrine; instead, they interpret him, in effect, as a meta¬
physician of mental activity, when the methodological interpretation would
be to regard him as a methodologist of criticism. I should mention that on
one occasion Croce did take seriously his methodological conception of
philosophy; this occured when he interpreted Galileo as a philosopher,5 in¬
sofar as he was a methodologist, that is as a philosopher-pure-and-simple,
which is not to say a pure or a simple philosopher.
However, if one looks at history-of-philosophy books, the most striking
fact about Galileo is the neglect he receives. Ueberweg devotes to him four
lines6, Br^hier one page7, Copleston about three8. W. T. Jones has probably
the longest and most detailed account9 to be found in a general, comprehen¬
sive history of philosophy; yet his 11 pages should be contrasted with the 15
on Bacon, and with the complete chapters on Hobbes (25 pages), Descartes
(27 pages), Spinoza (22 pages), Leibniz (12 pages) and Locke (34 pages).
This neglect would create no problem if the place of Galileo in the actual
history of philosophy were as insignificant as it is in history-of-philosophy
books. But this is far from being the case, and one could document this fact
in a number of ways: by exhibiting the wealth of philosophical content of
Galileo’s books (as I have done above); by discussing the philosophical sensa¬
tion that he caused among 17th century philosophers; by tracing and defining
the nature of his influence on important subsequent philosophers such as
Hume10, Kant11, Husserl,12 and Ortega;13 by authoritative quotations from
more knowledgeable historians, such as the following from Sir David
Brewster:

Had Bacon never lived, the student of nature would have found in the writings and
works of Galileo not only the principles of inductive philosophy, but also its practical
application to the noblest efforts of invention and discovery;14

or the following from E. A. Burtt:

It is difficult indeed to leave Galileo without pausing a moment to reflect on the simply
stupendous achievements of the man ... and then, as if these accomplishments were
not enough, we must turn to him likewise as the philosopher who sufficiently per¬
ceived the larger implications of his postulates and methods to present in outline a
new metaphysics - a mathematical interpretation of the universe - to furnish the final
justification of the onward march of mechanical knowledge. ... In view of these mani¬
fold and radical performances Galileo must be regarded as one of the massive intellects
of all time.15
 CONCRETENESS AND JUDGMENT 153

However such documentation would really be unnecessary because I regard it

as uncontroversial that Galileo has a very significant place in the actual
history of philosophy, for even those philosophers who neglect him in their
own accounts pay lip service to his philosophical greatness. For example,
Ueberweg’s four lines on Galileo belie his own practice:

Galileo Galilei (1564-1641) [sic]) acquired by his investigation of the laws of falling
bodies a lasting title to esteem not only as a physicist, but also as a speculative philo¬
sopher. Worthy of note are his maxims of method: independence of authority in matters
of science, doubt, and the founding of inferences on observations and experiments.16

Thus the neglect is not justified as being a deserved one; nor could it be
justified by saying that Galileo’s place is in the history of science and not in
the history of philosophy. This would amount to equating science with
natural philosophy and then excluding the latter from philosophy; now, to do
this, whatever its theoretical or present day justification, would be un-
historical since neither the exclusion nor the equation characterize the
relevant historical period. I don’t think the neglect is justifiable at all.
Let me therefore try to find an explanation for it. This will give us some
insight into the nature of the study of the history of philosophy, or at least
into the nature, however perverse, of general histories of philosophy. The
raison d’etre of such books is the attempt to find common topics and themes
so that one can discern some kind of development in human thought about
such topics and themes. A thinker whose philosophical ideas cannot be easily
related to the others has no place in such a development. In other words, the
principles underlying such books are the following: (1) history of philosophy
is the study of the development of philosophical ideas; (2) philosophical ideas
are ideas about such topics as God, the soul, immortality, free will, goodness,
justice, truth, knowledge, and beauty; and (3) philosophical ideas originate
from other philosophical ideas in the double sense that (a) ideas about one
philosophical topic originate partly from ideas about other philosophical
topics, and (b) the philosophical ideas of one philosopher originate from the
philosophical ideas of earlier philosophers. (3a) requires that an immense
amount of integration be present in order to qualify as philosophy, and (3b)
requires that one have the appropriate connections to qualify as a philosopher.
Such a concept of philosophy, or philosophy of the history of philosophy,
excludes those thinkers who, though they had significant ideas about some
philosophical topics, acquired these ideas from their experience, and hence
were unconcerned about developing a system of philosophical ideas, and
about justifying them on the basis of other philosophical ideas; that is, their
154 CHAPTER 6

philosophical ideas are unsystematic and justified primarily on the basis of

practical considerations, where ‘unsystematic’ means relatively unintegrated
at an abstract level, and where ‘practical’ refers to any general human activity
upon which one may reflect from a metalevel. As we have seen, Galileo was
such an unsystematic, practice-oriented philosopher.
The way to criticize such a concept of philosophy presupposed by general
histories is partly by showing that it leads to the neglect of philosophers who
ought not to be neglected (as in the case of Galileo), and partly by showing
the inaccuracies of specific historical claims made about the origin of the
ideas of the various philosophers that are discussed in such books. For
example, I believe it could be shown that Hume derived many of his philo¬
sophical ideas from Galileo, in the following indirect but precise sense,
namely by taking Galileo as an example, that is by trying to use in his own
area of concern the approach that Galileo followed in his; here a comparison
of the Dialogues Concerning Natural Religion and of Galileo’s Dialogue Con¬
cerning the Two Chief World Systems would be a very relevant exercise,
though beyond the scope of the present investigation. A third way to criticize
the above mentioned philosophy of the history of philosophy would be to
oppose to it a different one and to show the novel insights to which it leads;
one of these alternatives would be to adopt Croce’s concept of philosophy as
methodology and to begin writing the history of philosophy accordingly.
This I propose to illustrate by arguing that there is a very interesting method¬
ological similarity between Galileo and Socrates, that is a similarity at the
level of the method followed by each in their respective and very different
fields. This discussion will also illustrate the kind of integration and unity
that such a ‘Crocean’ history of philosophy would have; it would be unity at
the level of method followed. The developments in such a history would be
restricted to the following types: application of the same basic method to a
new domain, and/or refinement of the method as applied in an old domain.17
For the purpose of my argument I shall accept Vlastos’s general interpretation
of Socrates’s philosophy.18
According to Vlastos, Socrates’s greatest accomplishment, “among the
greatest achievements of humanity”,19 is his method, i.e., his method of
moral inquiry; only this interpretation makes sense of what Vlastos calls the
Socratic paradox, which is the central fact of his life. In other words, Socrates
was first and foremost a methodologist, a moral methodologist (Vlastos’s own
word being ‘searcher’); virtually all of Socrates’s activities can be understood
as instances of preaching or practicing this method.
The ‘Socratic paradox’ consists of two groups of apparent inconsistencies.
 CONCRETENESS AND JUDGMENT 155

First, on the one hand Socrates claims that the improvement of your soul is
the highest value; on the other he does not seem to care about improving the
soul of his fellow men since he is such a tireless critic, such a ‘despotic logi¬
cian’ (to use Nietzsche’s phrase), and so unwilling to give the answers and
solutions. Second, on the one hand he explicitly asserts that knowledge is
virtue, implicitly exhibits in his behavior supreme confidence in being right,
and in general feels that he has a mission to fulfill; on the other hand, he also
explicitly asserts that he is not wise and implicitly exhibits in concrete dis¬
cussions ignorance of the answers and solutions.
The methodological interpretation solves the paradox as follows. His tire¬
less criticism, despotic logic, and unwillingness to give the answers derive
from the nature of his method of moral inquiry and from his conviction that
the improvement of the soul consists in the mastery of the method, or to be
more exact, in the improvement of your ability to use the method, which to
some extent is a universal human patrimony.20 Here, the relevant essential
characteristics of the method are the following: it is critical in the sense that
the role of the teacher or expert is to offer criticism; it is logical or rational in
the sense that it makes essential use of logic and reasoning and rigorously
follows the argument wherever it leads; and the method is maieutic, or
“Socratic” in the usual more restricted sense, insofar as it aims to make the
person involved arrive at the solution himself. This is so in part because the
Socratic teacher, according to Vlastos, does not really know the answer in
advance, does not possess knowledge of the result, but only or primarily
knowledge of the method of discovering it. And this brings us to the solution
of the second element of the paradox: when Socrates declares that virtue is
knowledge, this virtuous knowledge is knowledge of the method of inquiry
and not knowledge of the answers or results, of which he is genuinely
ignorant until the inquiry is completed, i.e., temporarily completed; for such
virtue-knowledge is open-ended and always ready and willing to reexamine
itself. In other words, Socratic knowledge is really methodological insight,
not knowledge in the usual sense.
What are the essential features of the Socratic method of moral inquiry?
I have already mentioned that it is critical, logical-rational, maieutic, and
open-ended. Vlastos also stresses21 that it is difficult in the sense that it
requires sincerity, humility, and courage; that is, it requires commitment and
involvement on the part of the inquirer. Other terms that come to my mind
are seriousness, intellectual honesty, or ‘concreteness in a Crocean-Hegelian
sense. What this really amounts to is that the method requires a give-and-take
between action and thought, or experience and reflection, or theory and
156 CHAPTER 6

practice (in a Crocean sense). The answers that the inquirer is to test must be
ones that really inhere in his experience. Finally, and perhaps most im¬
portantly, the method is judgmental; it is not a method in the sense in which
some people conceive of method, namely as an infallible rule. Instead, ‘it
offers no guarantee’,22 but it is what the teacher aims at improving.23
In speaking of Socrates’s method of moral inquiry, Vlastos distinguishes
it from other specific Socratic achievements or doctrines, e.g., that it is
always wrong to repay evil with evil, or that piety could amount to nothing
but the improvement of your soul, or that no harm can come to a good man.
These maxims alone would rank Socrates as a great moralist, but it is
important to make the distinction between them and the method.
Finally, a correction should be made to one of Vlastos’s critiques of
Socrates, namely of the principle that virtue is knowledge. He argues plausibly
that, contra Socrates, knowledge is neither a necessary nor a sufficient con¬
dition for virtue. However, this argument uses a concept of knowledge which
is both un-Socratic and different from the one that Vlastos himself has attri¬
buted to Socrates, in order to solve the Socratic paradox. Socratic knowledge
is methodological insight and not information stored in one’s head and
immediately available to recall; and it is concrete, i.e., serious, i.e., sincere-
humble-courageous; it is not divorced or abstracted from one’s experience or
practice; hence it is to be found as much in what one explicitly says, as in
what inheres implicitly in action. Therefore, I do not accept Vlastos’s counter¬
example of the courageous person who cannot give a general explanation of
the concept, or the one of the person who fears death in spite of his
ignorance about it; that is, I do not accept them as proper examples involving
the concept of knowledge that he himself is articulating and suggesting as
being the Socratic one.
In summary, Socrates’s greatest achievement is his method of moral
inquiry, which is distinct from his specific moral maxims, and which can be
characterized as being critical, logical-rational, maieutic, open-ended, con¬
crete, and judgmental.
In the case of Galileo, it is a commonplace that his greatest achievement is
his method; it is also obvious that this is a method of physical investigation.
Moreover, it is easy to see that his method is critical, logical-rational,
maieutic, and open-ended. The novelty is to suggest that Galileo’s method of
physical inquiry is concrete and judgmental; yet I believe I have shown this
already, though the proof was neither short nor easy. So let me repeat what it
is that was shown. Concreteness is my Crocean label to refer to what Vlastos
calls sincerity-humility-courage in the case of Socrates. This involves primarily
 CONCRETENESS AND JUDGMENT 157

using one’s own experience to test one’s abstract ideas, and in turn using the
latter to enlighten one’s own actual experience. It is a synthesis of two
aspects of human activity: the involvement in a particular situation, and the
reflection on this specific involvement; I speak of synthesis in such away as
to oppose it to an eclectic mixture of the two; and moreover I speak of two
aspects of human activity to indicate that they are not normally separate or
separable in reality, but rather merely distinguishable in the mind. In the case
of Galileo the two relevant aspects are scientific practice and philosophizing.
Scientific practice is the element where Galileo is involved in the discussion
and solution of some scientific problem or topic, for example concerning the
motion of the earth; philosophizing is the element where Galileo remarks
about such things as the relationship between theorizing and observation, the
role of mathematics and of authority in physical inquiry, the origin, justific¬
ation, and limitations of physical knowledge, etc.
The judgmental character of Galileo’s method refers to the fact that
Galileo’s work is characterized by a proper balance (1) between physical in¬
quiry and philosophical awareness, and (2) among the following: observation
and speculation, mathematical (quantitative) analysis and qualitative con¬
siderations, causal investigation and positivism, negative criticism and con¬
structive articulation, anti-authoritarianism and respect for original thinkers,
anti-anthropocentrism and humanism, and anti-verbalism and logical analysis.
The balancing of all these requirements is the most fundamental feature of
Galileo’s work in the sense that it alone does justice to the complexity and
wealth of philosophical remarks that we find in his writings. This is not to say
that I have portrayed Galileo as an eclectic, or that I have followed an eclectic
approach in analyzing his method. The reason for this is that the balancing of
the numerous activities mentioned in (2) above is a consequence of the
theory-practice synthesis, mentioned in (1) above, which therefore must be
regarded as an essential feature of Galileo’s method, especially since it is
merely another aspect of the concreteness mentioned earlier. Finally, such
philosophical-scientific synthesis would alone insure Galileo’s originality in
the history of thought, which originality may otherwise be difficult to see in
the light of the existence of precursors among previous students of nature and
among previous epistemologists and theorists of method, none of whom
combined the two types or levels of inquiry in a nontrivial way.

GALILEO AND THE PHILOSOPHY OF SCIENCE

Galileo occupies a unique position in the philosophy of science. He has a

158 CHAPTER 6

significant place both in the history of the philosophy of science and in the
history of science, that is, both in the theory of scientific practice and in the
practice of which philosophy of science is the theory. Moreover, virtually
every philosopher of science has felt or feels the need to come to grips with
Galileo, in the sense that he either derives his theories from his analysis of
Galileo, or he tests the theories he has otherwise formulated by applying
them to the case of Galileo.24 Finally, it has turned out to be possible for
philosophers of science of almost any persuasion to use Galileo for their pur¬
poses, in the sense of finding evidence from Galileo to support their theories.
It is this last fact to which I wish to call attention. I believe it has no counter¬
part with any other scientist. There are indeed plenty of scientists who are
more highly regarded by some one school of philosophers; for example,
Newton by inductivists and positivists, Einstein by hypothetico-deductivists
and explanationists. There is no one scientist, however, in whom almost any
philosopher, no matter how idiosyncratic, can with some semblance of
plausibility find his own predilections or prejudices. This fact, which I shall
document below, constitutes both a problem and a singular characteristic of
Galileo. It is a problem from the point of view of the current state of the
philosophy of science, since the existence of these conflicting interpretations
represents an unsatisfactory state of affairs. At the same time the existence of
such a spectrum can be regarded as a uniquely important and importantly
unique feature of Galileo’s work, indeed its most basic and general aspect.
For there is no other scientist for whom such a spectrum is possible.
Let us begin our list of interpretations with what is perhaps the most un¬
expected one. According to Feyerabend, “Galileo violates important rules of
scientific method which were invented by Aristotle, improved by Grosseteste
(among others) and canonized by logical positivists such as Popper. Galileo
succeeds because he does not invariably follow these rules”.25 This inter¬
pretation is used by him to support what may be called a counterinductivist
methodology of science, which asserts that “in order to progress, we fre¬
quently have to step back from the evidence, reduce the degree of empirical
adequacy of our theories, abandon what we have already achieved, and start
afresh”.26
Another philosopher has recently found evidence that Galileo was a
progressive, somewhat eclectic, scholastic, Thomistic Aristotelian27 using the
methodology of demonstration ex suppositione. This is reasoning of the form
“p; if p, then q\ q” where “p stands for a result that is attained in nature
regularly or for the most part, whereas q states an antecedent cause or con¬
dition necessary to produce that result”.28
 CONCRETENESS AND JUDGMENT 159

The Platonist interpretation of Galileo also continues to receive support,

and the latest word on it is perhaps that of Howard Stein, who has given a
new twist to the interpretation. He claims that “the notion of ‘inquiry’
suggested by Galileo’s exposition is one which it is not far-fetched at all to
compare with Plato’s notion of ‘dialectic’; and the movement, through
inquiry, from fallacies and paradoxes to science, invites comparison (for
instance) with the ‘divided line’ of the Republic”,29
Empiricists, too, continue to find evidence that Galileo was a sophisticated
experimentalist; it must be admitted, in fact, that they have effectively
refuted Koyrd’s original claims to the contrary.30 But that is not to say that
the ‘rationalist’ interpretation has not unearthed its own evidence. In fact,
recent studies have added a new twist to the interpretation by suggesting that,
though Galileo did depend on actual experiments, and though he usually had
actual experimental evidence supporting his theories, he often concealed this
empirical underpinning, out of rationalist commitments.31
So Galileo has been variously portrayed as counterinductivist, Aristote¬
lian, Platonist, empiricist, and rationalist. Some of these differences derive
from, and would be resolvable in terms of, the different kinds of evidence
examined; the main distinction here is between published works and un¬
published manuscripts and documents. Some of the differences are due to
emphasizing different aspects of Galileo’s work; two distinctions would be
relevant here: context of discovery vs context of justification, and scientific
practice vs reflections on this practice. However, I shall follow another
approach.
One may find useful in this context the cliche that each interpretation is
right in what it asserts and wrong in what it denies. In other words, I think it
is true that there is evidence that Galileo was a counterinductivist, and an
Aristotelian, and a Platonist, and an empiricist, and a rationalist. What is
wrong is to attribute to Galileo one of these characteristics and to deny the
presence or significance of the others. This tendency toward exclusiveness
and intolerance can be illustrated by reference to the Feyerabend counter¬
inductivist interpretation. Actually, it is clear from the above quotation that
Feyerabend is making three distinct claims. The first is the historical thesis
that Galileo’s work displays the feature of counterinductivism, i.e., violation
of important methodological rules such as to take the evidence seriously, and
to accept the most empirically adequate theory. Feyerabend’s second claim is
a causal explanation of Galileo’s success, namely that it was because he was
such a counterinductivist that he succeeded. Feyerabend’s third point is the
philosophical claim that the counterinductivist methodological rule is sound.
160 CHAPTER 6

Hence a plausible reconstruction of Feyerabend’s position is that he is

grounding his philosophical claims on his causal explanation, and his causal
explanation on his historical thesis. Now I believe that the philosophical claim
is a plausible conclusion from the causal explanation, in the sense that the
additional premises needed to complete the inference are all rather plausible.
It is also obvious that the philosophical claim does not follow from the
historical thesis alone. Let us then examine the step from the historical thesis
to the causal explanation, that is the following argument: Galileo practiced a
counterinductivist methodology; therefore, his success is due to this method¬
ology. This is a common type of historical inference, much studied by philo¬
sophers of the historical and social sciences, though rarely in a context like
the present one. The full details of the argument would be the following:
(1) Galileo was a successful scientist.
(2) There must be a cause for this success, since every event has a cause.
(3) Galileo practiced a counterinductivist methodology.
(4) His counterinductivist methodology is a possible cause for his success.
(Or, pedantically stated: the fact that (3) is a possible cause for the fact that
(!)•)
(5) There is no other feature of Galileo’s work which could have been the
cause of his success.
(6) Therefore, the counterinductivist methodology was the cause of his
success.
(1) and (2) raise no relevant problems. (3) is Feyerabend’s historical thesis,
which I am accepting here for the purpose of this discussion.32 (4), the
possible-cause claim, is something that I am inclined to accept as obvious
from the context in which the historical thesis (3) is discussed and supported;
in other words, Feyerabend’s discussions of (3) are designed to support (4)
as well. (6) is the causal explanation and does follow from (1) through (5).
The problem lies in (5), the denial of alternative causes. The existence and
history of the other interpretations shows that this denial, so necessary to
the argument, is indefensible. In fact, for each of the other interpretations,
one can formulate a parallel argument, whose only problem would be the
denial of alternatives. Thus I conclude that Feyerabend’s causal explanation
of Galileo’s success is not justified, even if his historical thesis were; and since
his historical argument in support of his philosophical claim depends on the
causal explanation, this claim is historically groundless.
Now, what has been said of the counterinductivist interpretation, and
what has been easy to do in this case because of Feyerabend’s explicitness,
could be said and done for the other interpretations. Each is subject to the
 CONCRETENESS AND JUDGMENT 161

same problem, and hence their philosophical relevance is problematic, though

their historical basis is unquestioned.
What is one to do? What one can do is to follow a combinatory or
synthetic approach. This is both an obvious line to try in such a situation, and
also in accordance with another traditional interpretation of Galileo, so far
unmentioned. Shapere’s book on Galileo,33 whatever faults it may possess,34
does have the merit of recognizing that a synthetic approach is the correct
one, since that is what the book attempts. It is true, however, that the book
fails in the attempt, and instead of a synthesis it ends up with what may be
called inconclusive eclecticism. I have discussed elsewhere35 the method¬
ological value of Shapere’s book (methodological in the sense of philosophical
methodology, that is at the level of the methodology for the history and
philosophy of science). There I did not mention the synthetic approach but
stressed another feature of Shapere’s method which is valuable, namely the
method of logical analysis which I emended and qualified for other purposes.
Be that as it may; I believe that we ought to accept the synthetic, multi¬
faceted approach and the logical-analytical approach suggested or inspired by
Shapere’s book, for there is no reason why synthesis should degenerate into
inconclusive eclecticism, or why logical analysis should degenerate into
scholasticism.36 However, for substance we must look elsewhere.
Substance is what one finds in Clavelin’s Natural Philosophy of Galileo,
perhaps the best recent example of the synthetic tradition,37 now under
consideration. His view is that

the reason, therefore, why no scientific problem was the same again as it had been before
Galileo tackled it lay largely in his redefinition of scientific intelligibility and in the
means by which he achieved it: only a new explanatory ideal and an unprecedented skill
in combining reason with observation could have changed natural philosophy in so
radical a way.38

The new explanatory ideal is the following:

To explain, according to Galileo, meant above all to proceed from a certain number of
principles and concepts and with the help of a model to an intelligible reproduction of
the phenomena under investigation. This definition highlights two characteristic traits of
Galileo’s rationalism, above all, the almost complete reduction of physical to rational
necessity: once the causal ideal of Aristotelian physics had been abandoned, explanation
had the sole task of establishing an implicative relationship between the facts, as derived
from a model, and the guiding principles of reason. At the same time, simplicity became
an important physical criterion: nature follows the simplest path, that is, the one which
permits, on the side of reason, the simplest deductions. However, explanation does not
merely tend to provide a rational reconstruction of phenomena; its true aim is to turn
162 CHAPTER 6

every physical problem into a mathematical one and thereby to [utilize] for its analysis
the existing mathematical science.39

Some comments and some analysis are needed here. Two basic features are
being attributed to Galileo, the first at the level of philosophy, or theory of
scientific practice, the second at the level of scientific practice or method.
Philosophically, Galileo formulated a new theory of scientific knowledge,
scientifically, he practiced a new skillful combination of thinking and sense-
experience. Clavelin’s account of Galileo’s scientific method or practice is
difficult to excel; it is the latest and best discussion of Galileo’s synthesis
of reason and observation. Occupying the second half40 of the chapter on
“Reason and Reality,” Clavelin begins by criticizing the Tannery-Koyr6
apriorist thesis about the role of experience in Galileo’s scientific practice;41
he then analyzes Galileo’s work on the cohesion of matter to illustrate his
consciousness of the limitations of mathematical reason and necessity of
experience;42 he concludes with an analysis of the simplifications made by
Galileo in his geometrization of the motion of heavy bodies to show his
awareness of the complexity of, and his judiciousness regarding, the relation¬
ship between reason and reality.
It is important to note that there is something very special about this
methodological feature described as “proper combination of reason and
observation.” It is not as specific as counterinductivism, Aristotelianism,
Platonism, empiricism, and apriorism; it is a somewhat formal feature, some¬
what empty of content, and yet it is not uninformative. It is very difficult to
make precise by way of definition; attempts to do so tend to destroy the
great explanatory power it possesses. Instead it is better to give content to
it by showing it to be inherent in various aspects of Galileo’s work, which is
what Clavelin does. At the same time, the concept has an openness which
allows one to apply it in new more complex ways, or to new unexamined
aspects of Galileo’s work; again, such an application is what Clavelin has
superbly carried out. I am convinced that it is this type of interpretation of
Galileo’s scientific practice or method that will stand the test of time.
Let us now look at Clavelin’s account of Galileo’s philosophy of science,
or theory of scientific practice. It occupies the first part of the chapter on
‘Reason and Reality’, and, as the second quotation above suggests, four
elements are involved. The first is a concept of explanation as rational re¬
construction of the phenomena;43 the second is an identification of phys¬
ical with mathematical intelligibility;44 the third is the principle of corre¬
spondence between rational and physical necessity;45 and the fourth is
 CONCRETENESS AND JUDGMENT 163

the principle of simplicity.46 It is not clear how, for Clavelin, these four
elements are related to each other and as a whole to Galileo’s scientific
practice. Regarding the latter question, the only answer I can find is that
Clavelin seems to think that Galileo’s philosophy of science represents an
‘ideal’ for his actual procedure. Regarding the former question, I think the
following would be a plausible reconstruction of Clavelin’s view of the inter¬
relationship of those four elements: intellectual reconstruction of the pheno¬
mena and mathematical analysis (in the sense of formulation of physical
problems in mathematical terms so as to make possible mathematical mani¬
pulations and inferences) would be part of what Galileo means by explana¬
tion, whereas the principles of simplicity and of correspondence between
rational and physical necessity would be a priori necessary conditions for the
possibility of explanation in the sense specified. However, I shall not pursue
such questions any further because, however interesting and complex
Clavelin’s view of Galileo’s philosophy of science is, I don’t think it goes far
enough. There are three main problems with it.47 First it does not do justice
to the wealth of Galileo’s remarks and reflections on scientific inquiry. In
other words, though Clavelin’s account is the best available, it is still an over¬
simplification of Galileo’s philosophy of science as shown by my earlier
analysis of the methodological content of the Dialogue Concerning the Two
Chief World Systems. Second, the account is still a premature systematization,
that is a systematization of Galileo’s philosophical remarks before making a
systematic attempt to identify and to state clearly all such remarks; this
means that such a philosophy of science is more of a theoretical elaboration
of Galileo’s own philosophy, than an account of it; in other words it is more
of a Galilean philosophy of science, rather than Galileo’s philosophy of
science. Third, in most of his analysis Clavelin takes Galileo’s philosophical
remarks out of context, being interested in relating those remarks to each
other rather than to the live, concrete scientific practice out of which they
arose.
What I have argued is that the philosophical content of Galileo’s work is
much richer and much more complex than anyone has suggested so far. It is,
however, possible to define the essential characteristics of Galileo’s philo¬
sophy of science: Galileo did not here have a philosophy of science, in the
sense of “philosophy” common among some philosophers of science. In the
theory of scientific method and knowledge, Galileo is not a systematic
thinker in the sense of having a relatively small number of basic principles,
systematically unified, and in terms of which all the facts to which they refer
can be understood. This does not mean that he is not a philosopher, but
164 CHAPTER 6

rather that he is a philosopher of a particular kind, namely an unsystematic

philosopher, very much like Socrates was in moral philosophy. What
characterizes such unsystematic (or, to use a more positive term, Crocean)
philosophy is a constant and sustained willingness to reflect on one’s activities
in order to understand and justify them, as the need arises in a concrete
situation, and a willingness to avoid inconsistencies, when these become
apparent. When such reflections center around concepts such as truth, know¬
ledge, method, nature, physical reality, mathematics, sense-experience, etc.,
we have Galileo’s case of Crocean philosophy; when the concepts are wisdom,
goodness, justice, piety, human life, etc., we have Socrates s case. In other
words, what characterizes Galileo’s philosophy of science is his skillful com¬
bination of scientific inquiry and reflection upon this inquiry, reflection in
the sense of remarks, judgments, explanations, justifications, etc., in short
his skillful synthesis of scientific practice and philosophical theorizing (about
it). Or equivalently, in order not to bias the description in favor of the
primacy of practice over reflection, we may speak of Galileo’s skillful applica¬
tion of philosophy of science to science, that is concrete putting into practice
of epistemological and methodological concepts. This feature of Galileo’s
philosophy of science is supported by all the evidence which does support
other interpretations and by all the interpretations I provided in my method¬
ological analysis of the Dialogue. It has all the advantages that the feature
described as “proper combination of reason and sense-experience” has at the
level of scientific practice or method. It is both informative and yet non¬
specific and able to take into account the wealth of philosophical content
that we find in Galileo. I believe it is the only feature rich enough to do
justice to the riches under consideration.

NOTES

1 These pages are those of G. Galilei, Opere, edited by Favaro, Vol. 7. In the rest of this
section, including Tables I-IV, all pages references are to this book.
2 R. H. Thomason, ed., Formal Philosophy, Selected Papers of Richard Montague. The
claim can be read in a Yale University Press advertisement on p. xiii of the APA Eastern
Division Program for 1972.
3 B. Croce, Teoria e storia della storiografia, published in Britain under the title Theory
and History of Historiography, and in the United States under the title History: Its
Theory and Practice, now reprinted by Russell & Russell.
4 Croce, History: Its Theory and Practice, p. 164.
5 B. Croce, Storia dell’eta barocca in Italia, pp. 60-64, especially p. 62.
6 F. Ueberweg, History of Philosophy, Vol. ii, Modern Philosophy, p. 28. In addition
to Ueberweg’s four lines, one finds in this book two pages on Galileo in a so-called
 CONCRETENESS AND JUDGMENT 165

‘Appendix II. Historical Sketch of Modern Philosophy in Italy’, written by Vincenzo

Botta of the University of Turin.
7 E. Br^hier, The History of Philosophy, The Seventeenth Century, pp. 10-11.
8 F. Copleston, History of Philosophy, Vol. iii, Late Medieval and Renaissance Philo¬
sophy, pp. 280-81, and 284-86.
9 W. T. Jones, A History of Western Philosophy, pp. 620-31.
10 D. Hume, Dialogues Concerning Natural Religion, pp. 136, 138, 150-51, 162-63,
214. Some of the connections are discussed by R. H. Hurlbutt III, Hume, Newton, and
the Design Argument, especially pp. 125—26, 140—46, 153-56. See also C. de Remusat,
Bacon, sa Vie, son Temps, et son Influence jusq’a nos Jours (2nd ed.; Paris, 1858), p.
396, who is reported to give Hume credit for being the first English writer to appreciate
Galileo as superior to Bacon; this is reported by C. W. Hendel, Studies in the Philosophy
of David Hume, p. 287, footnote 35. See also, the Introduction, above.
11 I. Kant, Critique of Pure Reason, ‘Preface to the Second Edition’, p. 20 (Bxii-xiii).
See also the Introduction above.
12 E. Husserl, The Crisis of European Sciences and Transcendental Phenomenology,
pp. 23-59.
13 J. Ortega y Gasset, En torno a Galileo. See also the Introduction above.
14 Quoted by V. Botta in his Appendix to Ueberweg’s History, p. 472, presumably from
Sir David Brewster, Lives of Copernicus and Galileo (Edinburgh Review, 1830) or his
Martyrs of Science (London, 1841); cf. Ueberweg, p. 473.
15 E. A. Burtt, Metaphysical Foundations of Modern Physical Science, pp. 103-104.
16 Ueberweg, History, p. 28. The date of Galileo’s death is printed by Ueberweg as
1641, whereas the correct date is 1642. I suppose this is another indication of his slight
interest in Galileo.
17 One history-of-philosophy book which, while not Crocean, comes close to doing this
sort of thing is K. Jaspers’s The Great Philosophers.
18 G. Vlastos, ‘Introduction: The Paradox of Socrates’, pp. 1—21.
19 Ibid., p. 20.
20 Ibid., pp. 19, 20.
21 Ibid., p. 20.
22 Ibid., p. 21.
23 Ibid., p. 20.
24 A. C. Crombie calls him a philosophical symbol in ‘Galileo Galilei, A Philosophical
Symbol’.
25 Paul K. Feyerabend, ‘Machamer on Galileo’, p. 297. Feyerabend’s counterinduction
should be distinguished from his anarchism. I believe the former is regarded by him as
merely a rule among other rules, which applies to the case of Galileo but may not apply
to others, whereas anarchism would be the formal super-rule (somewhat empty of
content). Thus Feyerabend’s anarchism may be regarded as an extreme formulation of
the concrete, Socratic-Crocean methodology of science which I am supporting here and
which emphasizes a move away from method and in the direction of judgment. The
present critique applies only to the conterinductivist methodology.
26 Ibid., p. 298.
27 W. A. Wallace, ‘Galileo and Reasoning Ex Supposition: The Methodology of the
Two New New Sciences', pp. 82-88. Further qualified support for Galileo’s Aristotelian-
ism is given in H. I. Brown, ‘Galileo, the Elements, and the Tides.
166 CHAPTER 6

28 Wallace, ‘Galileo and Reasoning Ex Supposition’,p. 95.

29 Howard Stein, ‘Maurice Clavelin on Galileo’s Natural Philosophy’, p. 397; see also
William Shea, Galileo’s Intellectual Revolution, pp. 150-55.
30 Thomas B. Settle, ‘An Experiment in the History of Science’; Stillman Drake,
‘Galileo’s Experimental Confirmation of Horizontal Inertia: Unpublished Manuscripts’;
and James MacLachlan, ‘The Test of an “Imaginary” Experiment of Galileo’s’. In
labeling these three historians ‘empiricists’ I am referring more to how they are perceived
in the philosophical community, rather than to an explicit and articulated empiricist
philosophy in their work, which could be described more fairly as merely aphilosophical.
31 R. Naylor, ‘Galileo: Real Experiment and Didactic Demonstration’, pp. 398-99;
idem, ‘Galileo and the Problem of Free Fall’, pp. 133-34; idem, ‘Galileo’s Simple
Pendulum’.
32 See Chapter 8, below, for a critical examination of this claim.
33 D. Shapere,Galileo.
34 See Michael S. Mahoney, ‘Galileo’s Thought’; Charles B. Schmitt, ‘Review of
Shapere’s Galileo’; and my ‘Philosophizing About Galileo’.
35 See my ‘Philosophizing About Galileo’.
36 C. B. Schmitt labels ‘scholastic in the bad sense’ certain features of the book.
37 Another good example is E. Agazzi’s ‘Fisica galileiana e fisica contemporanea’,
expecially pp. 46-47.
38 Maurice Clavelin, The Natural Philosophy of Galileo; p. 383.
39 Ibid., pp. 453-4.
40 Ibid., pp. 424—53.
41 Ibid., pp. 424-32.
42 Ibid., pp. 432-48.
43 Ibid., pp. 404-9.
44 Ibid., pp. 409-17.
45 Ibid., pp. 417—20.
46 Ibid., pp. 421-4.
47 For a detailed criticism, see below, Chapter 10.
 CHAPTER 7

THE PRIMACY OF REASONING:

THE LOGICAL CHARACTER OF

GALILEO’S METHODOLOGY

It has turned out (Chapter 2) that it is possible to reconstruct the whole

Dialogue from the point of view of the internal relationships among the pro¬
positions it contains. From this point of view the proposition that the earth
moves turns out to be the crucial one in the sense that the whole book is a
complex argument of which it is the conclusion. This is the logical point of
view, namely the point of view of reasoning or logical practice. It should also
be noticed that the methodological analysis of the Dialogue showed that the
topic of logic and reasoning was one of the most frequent ones; in fact, the
number of its discussions was exceeded only by the topic of experience, the
excess being merely one. In other words logic and reasoning are one of the
central methodological topics in the Dialogue. Because of these two ‘logical’
features of the Dialogue, it will be useful to inquire whether or to what
extent it is possible to interpret the other methodological topics in terms of
logic and reasoning, namely to what extent one can reduce the book’s
methodological discussions to discussions concerning logic and reasoning.
Let us first look at the specific topics1 discussed in those passages where
logic and reasoning are explicitly mentioned. In 59—712 we find a discussion
of the distinction between practice and theory in logic, or logical practice and
logical theory, or logic-in-use and reconstructed logic, or between reasoning
and logic in a narrower sense. In 290—9 we find a discussion of the distinc¬
tions between an argument and a rhetorical appeal, and between the form and
content of an argument. In 57—9 and 169—80 we find a discussion of some
relationships between reasoning and sense-experience; in 169-80 the relation¬
ship between reasoning and the Socratic method is discussed; the passage on
150-9 is a discussion of the value of reasoning; the one on 159-69 is an
illustration of the analysis and evaluation of arguments; and the passages on
139-50, 196-206, 399-406, and 462-70 are discussions of special types of
reasoning, namely probable arguments, hypothetical reasoning, ad hominem
arguments, and reasoning ex suppositione, respectively. Such explicit dis¬
cussions of logical topics are clear evidence of Galileo’s methodological self-
awareness about logic and reasoning. If one wanted to ‘systematize’ such
remarks, the following would be a possibility.
Reasoning is the subject matter of logic (i.e., of logical theory) in the sense

167
168 CHAPTER 7

that it is something whose structure and validity logic studies (59—71). The
basic structure of reasoning consists of the inferential interrelationships
among sentences or propositions, and the validity of reasoning (or logical
validity) refers to the propriety of such interrelationships (159—69 and
290-3). The construction of arguments is distinct but not separable from the
analysis of arguments, in the sense that arguments are never constructed in a
vacuum but rather in the context of counterarguments, and in the sense that
the analysis of counterarguments is the best practice and preparation for the
construction of one’s own positive arguments (150—9). The examination of
the consequences of more or less arbitrary assumptions is another valuable
practice, though one must be careful and realize that genuine hypothetical
reasoning is very difficult (196—206). One must also realize that reasoning
that proceeds from the unexamined assumptions of opponents has limited
methodological value, though great rhetorical effectiveness (399—406). More¬
over, so-called reasoning exsuppositione is rather tricky in that its validity
depends on whether the supposition in question is contextually free of
objections (462—70). Finally one of the things to be contrasted to reasoning
is sense experience; the following are two principles that should be kept in
mind: (1) whether or not sensory appearances correspond to reality is to be
decided by reasoning (57—9), and (2) whether or not there is a need to use
one’s senses is to be decided also by reasoning (180—8).
Experience. Let us now examine the discussions of other topics to see how
these other topics relate to reasoning. Let us begin with experience. The dis¬
cussion of conceptual frameworks in 38—57 relates to experience only
indirectly since it pertains primarily to conceptualization. Hence this is a
discussion of reasoning insofar as conceptualization is a type of reasoning,
which will be examined later. Here it should be noted that the discussion
emphasizes the harm that sense experience can cause.
In the next passage (57—9) the emphasis is on being critical of experience,
in the sense that we should ask ourselves whether what the senses reveal to us
corresponds to reality. An example referred to here, but not explicitly dis¬
cussed till later (159—69), is the vertical fall of bodies. Experience shows us
that bodies appear to fall vertically. Only reasoning can show whether or not
this implies that they actually do. The implication does not hold because it
can be easily shown that on a rotating earth, apparent vertical fall would
imply actually slanted fall.
The next discussion of experience (95 — 112) does show that for certain
purposes there is no substitute for it. It is tme that the same discussion shows
that what conclusions one draws from experience is a matter of reasoning.
 THE PRIMACY OF REASONING 169

However, this does not change the fact that the use of the senses is some¬
times essential. The proper use of this discussion is not to try to do away with
the observational element, but rather to use it by way of contrast in order to
get a better idea of what reasoning is and what it is not. In other words,
though Galileo regards reasoning as very important, he does not want to say
that it is the only activity in natural philosophy. To realize that other and
different activities exist, and that they are sometimes necessary, gives more
content to the emphasis on reasoning that Galileo practices and preaches. We
may say that this discussion of experience is part of an attempt at a definition
of reasoning, a definition by contrast.
The next discussion of experience (169—80) shows how only reasoning
can decide whether or not an experiment is called for. Again this presupposes
a distinction between the two, but it is evidence of Galileo’s emphasis on
reasoning rather than experience.
The passage on pages 206—9 discusses some of the ways in which one
should be critical of experimental results. The main connection with reason¬
ing is that only by reasoning can one decide whether a certain natural effect
is large enough to be detectable. This is another limitation of experience
vis-a-vis reasoning.
In the next passage (209—14) the over-all trend is toward emphasizing the
importance of experiments, especially in order to observe phenomena that
presumably couldn’t possibly happen according to certain lines of reasoning.
Here experience does perform an essential function, but even here it is
reasoning that has to choose the proper time or occasion for making the
experiment.
Concerning the emphasis on experience in the passage on physico-mathe-
matical synthesis (244-60), the following points should be noted. First, some
of the experience referred to (e.g., the times-squared law of fall) involves
observations which are mathematically described and analyzed, and hence is
experience filtered through a type of reasoning, namely mathematical reason¬
ing. Second, some of the experiences referred to (e.g., the body oscillating in
a tunneled earth) are thought-experiments rather than actual ones; but a
thought experiment is really a piece of hypothetical reasoning. Third, in the
synthesis of physical considerations and mathematical analysis, the element
of experience obviously enters in the physical part; but equally obviously it
cannot be experience itself, but rather reasoning, that determines the proper
combination - how, when, and exactly to what extent the two elements are
to be combined.
The discussion of the deception of the senses (272—81) is really a dis-
170 CHAPTER 7

cussion of reasoning since Galileo is claiming that the deception comes from
our being uncritical about our sense experience and from incorrectly reaching
conclusions about reality on the basis of sensory appearances; that is, the so-
called deception of the senses is deception of reasoning. Insofar as experience
is being mentioned, what we have here is again a contrast between it and
reasoning.
The discussion of simplicity, insight, and sense experience (349-68) is
one of the passages with the most definite emphasis on the importance of
experience, even though it is often misinterpreted as emphasizing ‘reason’.
Galileo is saying that he is enough of an empiricist that if it had not been for
the experiences possible with the telescope, he could never have accepted the
Copernican system. He is indeed awed by the insight of those who believed
it in the light of clear empirical counterevidence; he is not saying that it was
their superior reasoning that led them to what he now believes to be the
truth. Galileo’s point is that reasoning ought not to do that kind of violence
to sense experience, that is the kind of violence exemplified by Copernicus’
procedure. Hence it is clear that the discussion in this passage is about reason¬
ing, as much as it is about experience, which is all that I wish to claim. For
my examination of the extent to which the topic of logic and reasoning
prevails in the book is not predicated on any assumption that would equate
reasoning with “reason” in the sense of a priori reason. Though different
from sense experience, reasoning uses sense experience as much as needed to
avoid errors and arrive at the truth. In other words there is no reason to
equate reasoning with a priori reasoning; the latter does exist, but it is merely
a special case of the former; hence the present interpretation of Galileo, as a
logician in the sense of a conscious and skillful practitioner of reasoning and
of logical analysis, is not a ‘rationalist’ interpretation in the sense of portray¬
ing him as an apriorist.3
The discussion of the logic and methodology of theoretical explanation
(372—83) relates to experience insofar as it suggests that, in certain contexts,
sense experience has much more of a methodological than logical function.
That is, in the theoretical explanation of sense experiences, these cannot prop¬
erly provide the basis of reasoning to theoretical conclusions about reality, but
rather the basis of reasoning to conclusions about how to proceed in natural
investigation. In a sense, what we have here is another fact about reasoning, a
fact that can appear as a ‘limitation’ to those who may have been inclined to
overestimate its power. This contrast between logic and methodology is a
contrast between reasoning and acting in natural investigation, and hence it is
in a sense part of a contextual clarification of the concept of reasoning.
 THE PRIMACY OF REASONING 171

The last passage dealing with experience is a discussion of theories and

observations (409—14). Here reasoning is indirectly discussed in a way similar
to the passage just analyzed; acceptance and rejection of a hypothesis are
decisions to act in a certain way, namely to do and to decline from doing
further work on it, respectively.
Came, Explanation, and Understanding. The topic of causes, explanation,
and understanding is one of the most frequently discussed ones in the book
(eleven times). One of these passages (372—83) is identical with one already
examined under experience and hence will not be analyzed here. Let us look
at the other ten.
The discussion of the contrast between understanding and information and
of the Socratic method (180—8) is easily reduced to a discussion about
reasoning. For the understanding conveyed through the Socratic method
involves deriving what was previously not accepted from propositions that are
accepted, and this is precisely what reasoning is.
The discussion of criticizing and understanding (214—7) claims that
criticism is more effective when grounded on understanding. The connection
with reasoning is that Galileo seems to be claiming that if one wants to be
rational in one’s criticism, if one wants to reason about what he criticizes,
then he should make sure that he understands it. Hence we have here a plea
for reasoning.
The first discussion of causal investigation (237—44) may surprise those
who believe that Galileo did away with such investigations; actually there are
several such discussions. From the point of view of reasoning, causal
investigation is simply reasoning about causes, which makes it easy to see that
there is nothing in principle illegitimate about it. Hence the several discussions
of causal investigations can be easily understood from such a logical point of
view. The present passage also offers the occasion to note the ambiguity of
the phrase ‘reason for’, which can mean either ‘causal explanation for’ or
‘logical justification for’. Hence, reasoning, that is the search for logical
justifications, is to be distinguished from the search for causal explanations.
Here we then have another partial definition of reasoning by contrast.
The discussion of comprehensibility and truth (260—72) relates to reason¬
ing as follows. First, it is the discussion of a number of principles of reasoning
about comprehensibility and truth: that comprehensibility does not imply
truth, that incomprehensibility does not imply falsehood, that naming does
not imply comprehension, and that causal inexplicability does not imply
incomprehensibility. Second, though reasonable truth, i.e., truth to which
one can arrive by a process of reasoning, is a type of comprehensible truth,
172 CHAPTER 7

one cannot equate the two because of the possibility of intrinsic compre¬
hensibility. Hence, understanding is one of the things which is to be distin¬
guished from reasoning, though they are also related.
The discussion of inconceivability-claims and facts (281—8) is a discussion
of the following principle of reasoning: that it is wrong to argue that some¬
thing is not the case because there is no way to conceive how it could be
the case.
The discussion of the role, discovery, and justification of causes (431-6)
is, like the previous discussion of causal investigation, partly a discussion of
reasoning about causes and partly a discussion of the difference between
reasoning and (causal) explaining.
The passage on the choosing of the effect in a causal investigation (442—5)
is a discussion of a preliminary step to take in reasoning about causes. The
next passage (445—62) is a discussion of the following principle of reasoning:
that a proposed cause should be rejected if it does not allow an artificial
reproduction of the effect, and it should be accepted if such an artificial
reproduction is possible.
The discussion of confessed ignorance and occult qualities (470—1) can
be interpreted as involving the principle that it is sometimes better to refrain
from reasoning rather than arrive at a conclusion with little foundation.
The passage on the method of concomitant variations (471—84) is a dis¬
cussion of a principle of what some would call ‘inductive reasoning’. Or we
may say that it is a principle for the discovery (or conception) of causes.
Socratic Method and Unconscious Knowledge. The topic of Socratic
method and unconscious knowledge is discussed five times, and it is easily
reducible to logic and reasoning. The passage on cognitive awareness (112—7)
illustrates how the intellectual contents of the human mind are not always
open to introspection or known best to the person involved. When a second
person undertakes a Socratic cross-examination, the person involved is being
forced to reason about the rationale underlying his beliefs. In general,
Socratic cross-examination is one of the best instances of reasoning, of logic
at work. The second passage on this topic (169—80) has already been
examined in the context of the analysis of the discussions of experience; it
was found to be a paradigm example of a discussion of reasoning. The same
has been done for the third passage (180—8) in the context of our examina¬
tion of the passages dealing with understanding. The discussion of unconscious
knowledge (193—6) is almost a definition of a type of unconscious knowledge
as that to which one can arrive by reasoning. The last discussion of the topic
of Socratic method and unconscious knowledge (217—23) is also the most
 THE PRIMACY OF REASONING 173

explicit. Equally explicit is the emphasis on reasoning, for Galileo is saying

that knowledge is a kind of recollection in the sense that it is sometimes
acquired by reasoning based on facts ascertainable by reflection upon certain
chosen aspects of one’s experience.
Mathematics. Let us now examine the discussions about mathematics. The
first thing to notice about them is that mathematics is not at all the book’s
main methodological topic; it is discussed only five times. The first discussion
(33—8) is pretty clearly a plea for mathematics in the sense of mathematical
reasoning as opposed to number mysticism. In other words, mathematics is
simply a special case of what I mean by logic in this book, the case where the
reasoning deals with numbers, quantities, and geometrical figures. The dis¬
cussion of geometry and philosophy (223—9) is a discussion of the im¬
portance of geometrical reasoning in natural philosophy. The passage on
mathematics and physical reality (229—37) discusses the applicability of
mathematics; in so doing Galileo mentions another special feature of mathe¬
matical reasoning, namely its necessity; however, he is clear that this necessity
does not extend to physics, even mathematical physics, since the latter is the
apphcation of mathematics to nature, and every such application lacks
necessity. In other words, Galileo here seems to be discussing the nature and
status of mathematical reasoning: he is saying about its nature that it is
apodictic; and about its status he is saying that it is relevant to physics,
exactly how and why the application is possible, but that this application is
nonapodictic. Or we may say that Galileo is discussing apodictic reasoning,
claiming that it exists in mathematics, but that it has limitations since the
mere application of it involves nonapodictic reasoning.
The discussion of physico-mathematical synthesis (244—60) has been
examined before, as a discussion about experience. Here we may add that
such synthesis may be regarded as the apphcation of mathematical reasoning
to physical reality.
The passage on the role, discovery, and justification of causes (431-6) has
been examined before from the point of view of the topic of cause, explan¬
ation, and understanding. The references to mathematics in this passage are
not central, but they do deserve some discussion here. One of Galileo’s claims
is that practice with mathematical demonstrations makes one less willing to
accept a causal explanation as conclusively proved; this is easy to understand
now, since we have just seen that for Galileo mathematical demonstrations are
apodictic. Another claim that he makes here is that the direct experimental
confirmation of a cause can give a causal explanation a certainty comparable
to that of mathematics; here he seems to be saying that if in the domain of
174 CHAPTER 7

reasoning it is sometimes possible to have necessity, namely with mathe¬

matical reasoning, in the domain of sense experience it is possible to have a
comparable certainty, namely with the clear evidence of the senses.
Authority. The topic of authority is discussed five times, four as it relates
to Aristotle and once briefly as it relates to the Bible. Generally speaking,
Galileo’s plea against authority can be related to reasoning as follows: appeal¬
ing to authority is a good example of nonreasoning, but such nonreasoning,
unlike for example sense experience, is undesirable. It is interesting to note
that three of Galileo’s critiques proceed by focusing attention on Aristotle’s
practice, as opposed to his mere words, and then arguing that if we take this
practice seriously, then we need not take seriously certain of his claims.
The discussion of Aristotle’s authority as a logician (59—71) has already
been examined, since it is a direct and explicit discussion of logic and reason¬
ing. From the point of view of authority, Galileo is saying that insofar as one
is using authority in one’s reasoning, one has to make sure that the authority
is relevant, hence appealing to authority is no substitute for reasoning.
The next passage (71—82) discusses the role of authority in natural philo¬
sophy. Here Galileo gives another reason why appealing to authority is no
substitute for reasoning. If one is appealing to an authority, it is more
adequate to look not only at the authority’s conclusions but also at his
reasons and arguments. In so doing one will be examining the authority’s
manner of reasoning. Now, in some situations the facts may be such that, by
applying this manner of reasoning to them, one is led to a different con¬
clusion. In such a case there is a conflict between appealing to the authority’s
conclusion and appealing to his reasoning. Then one has to use his own judg¬
ment to decide which part of the authority to follow. Normally, the manner
of reasoning is a much more important aspect, and it should be followed. In
the present passage Galileo discusses this point by reference to Aristotle’s
claim that the heavens are inalterable and to his argument that this is so
because no changes have ever been observed in the heavens. In another
passage (346—9) Galileo discusses the same point by reference to the example
of Aristotle’s claim that the earth is the center of the world and his argument
that this is so because the earth is the center of celestial revolutions.
The passage on independent-mindedness (132-9) also shows that what
Galileo is really objecting to is the mindless appeal to authority, that is
appealing merely to the conclusions and not to the reasons or evidence.
Finally the passage on the Bible (383—5) merely dismisses it as a possible
authority in natural phenomena, since it does not contain reasoning about
natural phenomena and the claims it makes about them are unreasoned.
 THE PRIMACY OF REASONING 175

Method. There are five passages, interspersed throughout the book, which
in one way or another discuss the idea of method. They are the passages on
the role of authority in natural philosophy (71-82); physico-mathematical
synthesis (244-60); procedure vs results, and Aristotle’s authority (346-9);
logic vs methodology in theoretical explanation (372-83); and searching vs
results: the supremacy of method (484-9). All of these passages except the
last one have already been examined insofar as they contain discussion of
various other philosophical topics; when grouped together, they suggest that
Galileo is practicing and preaching the idea that the method of arriving at a
result is much more important than the result itself. This is something that
could be called methodologism; it is also a species of formalism, insofar as
the method/result distinction reminds me of the form/content distinction. It
is also a species of Socraticism, insofar as the maieutic character of the
Socratic method represents also an emphasis on method as opposed to end-
result.
How does such methodologism relate to the logicism that seems to be the
essential feature of the Dialogued It would be wrong to equate the two in
general since methodologism could take forms different or incompatible with
logicism. However, it is clear that to emphasize reasoning is a type of method¬
ologism, since reasoning involves the interrelationships among facts, ideas, or
beliefs, and the proper relationships can exist even when the specific ideas
are changed. It is also clear that the content of Galileo’s methodologism, i.e.,
that the specific type of his methodologism is logicism, since in three out of
the five passages (71-82, 346-9, and 484-9) he is explicit that the method
he has in mind is reasoning; and one other passage (244—60) has already been
examined and reduced to reasoning. Admittedly, the remaining passage
(372-83), discussing the logic vs the methodology of the theoretical explan¬
ation, does use a different concept of method, relating to acting rather than
reasoning. But this represents no real problem since the discrepancy could
properly be used to take this passage out of this group, because after all it
was so grouped only from an initial impression that it had something in
common with the other four. Moreover, the passage has already been
examined, for its relation to reasoning, in the context of the topic of sense
experience.
Anthropocentrism. Teleological anthropocentrism is my label for the
thesis criticized by Galileo that the universe exists for the sake and benefit of
man. The topic is discussed on four different occasions. The first passage
(82-95) is easily reduced to a discussion of logic and reasoning. For what he
does in this passage is to begin by asking what is the rationale on which the
 176 CHAPTER 7

belief is based and what are the conclusions that might be drawn from it; he
criticizes the rationale insofar as it consists of the earth/heaven distinction,
and he criticizes the alleged implication that celestial bodies are perfectly
round. In a sense Galileo is here making a plea for objectivity by saying that,
no matter how much we like anthropocentrism, we cannot let our emotions
prevail over reason in this case. In other words, one activity that can be
contrasted to reasoning is wishful thinking, even where the wish would
extend to all mankind. So one of the things that reasoning is not, is feeling or
emotion; and with respect to some issues, such as the present one, reasoning
must prevail; and finally, even when feeling may be allowed to prevail, only
reasoning can decide this.
In the discussion of the powers of human understanding (124—31), we see
that another version of anthropocentrism is the thesis that the human mind
is a measure of what can occur in nature. Galileo’s criticism of this shows that
his emphasis on logic and reasoning, which may be called logicism, is not a
species of “rationalism” as ordinarily understood. For one type of rationalism
would make optimistic, extravagant, and unrealistic claims about the ability
of man to know about nature merely by the exercise of his mind; and such
rationalism impresses Galileo as arrogant anthropocentrism, to be rejected.
He makes a similar point in the discussion of the primacy of nature over man
(289—90): since nature is prior to man, a priori reasoning has a limited
function.
The passage on objectivity and the concept of size (385-99) is a discussion
of some of the ways in which reasoning is dependent on sense experience and
on the faculty of imagination; for example, our reasoning about sizes depends
on our perception of sizes and on our ability to imagine various sizes.
However, such dependence is not equivalent to a dependence on human
interest; reasoning may have to take into account human interests, but it
cannot be subservient to them; instead it must rule over them.
Conceptualization. The three passages on conceptualization deal with con¬
ceptual frameworks (38-57), inconceivability-claims and facts (281-8), and
the superficiality of abstract answers (293—8). Conceptualization was the
label under which these three passages were grouped together in the context
of the analysis of the book’s main methodological themes; the label was
meant to refer to conceptual thinking. From the point of view of reasoning,
we can say that the topic in these three passages is a priori reasoning. In his
discussion of conceptual frameworks Galileo is suggesting that on certain
occasions a priori reasoning is proper; in fact, on such occasions what is
wrong is to infect it with empirical considerations. The occasions are situations
 THE PRIMACY OF REASONING 177

when what is at issue is the basic concepts in terms of which one is to inter¬
pret empirical material. The question of what is regarded as ‘natural motion’ is
an example. In the other two passages Galileo criticizes certain types of a priori
reasoning: in one (281—8) we have an empirical argument designed to throw
some doubt on whether one can ground factual falsehood on conceptual
difficulty; the last passage (293-8) suggests that often a priori reasoning is
misapplied, in the sense that relevant factual considerations are neglected.
Open-mindedness. The first passage (150-9) is an explicit discussion of
reasoning, and of its essential connection with open-mindedness, and hence it
has been examined above. Here it should be pointed out that in that passage
Galileo stresses how open-mindedness leads to reasoning. In the passage on
open-mindedness, intellectual cowardice, and curiosity (425—31) Galileo em¬
phasizes the converse connection, namely how closed-mindedness leads to a
failure to reason by summarily dismissing other views; this is also discussed
by Galileo by suggesting how reasoning can lead to open-mindedness, namely
that by examinating the argument in support of a view one may become inter¬
ested or intrigued by the argument, and as a consequence by its conclusion.
Ignorance. The four passages grouped under the label of ignorance contain
discussions of the desirability of modesty in one’s claims to knowledge. In
none of these passages except the last one (470—1) is such modesty the main
topic of discussion; consequently, three out of these four discussions have
already been examined from other points of view. Nevertheless, this theme of
epistemological modesty, mentioned on four occasions, is further evidence
that Galileo’s emphasis on reasoning, which we may call his logicism, is not
equivalent to ‘rationalism’, as ordinarily understood. It is interesting to
extract Galileo’s reasons for such epistemological modesty: from the passage
on the powers of human understanding (124—31) we get that rationalistic
arrogance is a sign that one has never understood or tried to understand any
one topic fully; from the discussion of independent-mindedness and the
misuse of authority (132-9) we get that if one doesn’t claim to know every¬
thing, then he will be taken more seriously when he does claim to know
something; from the passage on open-mindedness, intellectual cowardice, and
curiosity we get that modesty makes one more open-minded and arrogance
more dogmatic and rigid; finally, from the explicit discussion on confessed
ignorance and occult qualities (470-1), we get that epistemological arrogance
is likely to be caused by a failure to use reasoning and by the exercise of idle
imagination, vanity, and verbosity.
Simplicity. Simplicity appears as one of the main topics of discussion in
two passages (139-50 and 349-68) and as the single central topic in another
178 CHAPTER 7

(416—25). The passage on simplicity and probability (139—50) has already

been examined since it contains an explicit discussion of reasoning, specifi¬
cally of probable arguments; it is worth repeating here that Galileo claims that
reasoning depending on the principle of simplicity is merely probable and not
apodictic. Both in this passage and in the one on simplicity, insight, and sense
experience (349—68) Galileo says that in the light of empirical counter¬
evidence this probability reduces to zero, in the sense that it is not a sound
principle of reasoning to infer even probable truth from simplicity when there
are clear and unanswerable empirical difficulties. Such reasoning is unsound
in the sense that it is a bad policy to follow regularly, though of course it has
unpredictable exceptions, like Copernicus. That is, one cannot but marvel at
their insight, but one can speak of insight only after sound reasoning can
justify their conclusion, for example after one has been able somehow to
remove the difficulties. So reasoning based on simplicity must be handled
with care for two reasons: at best it is merely probable, since nature tends to
act, but we can’t be certain that it always does act, by means of the simplest
operations (139—50); and one must watch out for unanswerable empirical
counterevidence, which can easily offset these degrees of probability. A third
problem with reasoning based on simplicity is due to the complexity of the
concept, on which we have an explicit discussion (413—25): it is not always
possible to agree on which idea is simpler, since different criteria of simplicity
are often used, two frequently differing criteria being an abstract-mathe¬
matical one and an ontological-physical one.
Astronomy. The four passages on the nature of astronomical investigations
are, from the point of view of reasoning, discussions of some problems and
limitations with quantitative reasoning and of the role of precision and
exactness in reasoning. The passage on quantitative data and qualitative
conclusions in astronomy (299—337) discusses the problem that sometimes
quantitative astronomical data are so inconsistent that no quantitative con¬
clusion can be drawn from them. The passage on the reliability of astro¬
nomical measurement (337-46) discusses the problem that sometimes the
data are based on unreliable measurements. The discussion of computational
vs philosophical astronomy (368—72) is a critique of quantitative-mathe¬
matical reasoning in astronomy, insofar as such reasoning can lead to the
neglect of qualitative-physical reasoning. The passage on the inexactness of
astronomical instruments (414—6) discusses how sometimes the effect to
be measured is so small that great precision of instruments is essential; but
even here the great precision often derives from the great sophistication,
and this sophistication is often the result of a new qualitative idea.
 THE PRIMACY OF REASONING 179

Criticism. Out of the four passages under the heading of criticism, one
(206-9) has already been examined, in the discussion of sense experience,
and also it raises no new point in the present context. The remaining three
relate to reasoning as follows. The discussion of different categories of error
(117—24) suggests that in criticism sometimes there are things other than
reasoning to be evaluated, one of these external considerations being whether
or not an erroneous idea is being defended at a time when a better one has
been published. The discussion of criticizing and understanding (214—7) is
saying that, in criticizing reasoning, sometimes one should strengthen it
before finding fault with it. The discussion of the value of criticism (406-9)
suggests that criticism can benefit reasoning, in the sense that reasoning can
be improved as a result of answering objections made to it.
In conclusion, we may say that all the methodological discussions in the
book can be interpreted as centering around the topic of logic and reasoning;
what is being discussed directly or indirectly are such things as the relation
between logic and reasoning; the nature of reasoning; types of reasoning; the
relation between reasoning and other things such as sense experience,
appealing to authority, explaining, and feeling; and a number of principles
of reasoning.

NOTES

1 In this chapter, the discussion is centered around the topics as systematized in Tables
III and IV of the last chapter.
2 The numbers are page references to G. Galilei, Opere, edited by Favaro, Vol. 7.
Similarly for the rest of the discussion in this chapter. See Tables III and IV of the last
chapter.
3 For a further discussion of the distinction between reasoning and a priori rationalism,
see my criticism of Koyrd in Chapter 9.
 CHAPTER 8

THE RATIONALITY OF SCIENCE AND

THE SCIENCE OF RATIONALITY:
CRITIQUE OF SUBJECTIVISM

CHANGE, PROGRESS, AND RATIONALITY IN SCIENCE

Is science rational? What does its rationality consist of? Does it consist, for
example, of the effectiveness of its method? Does it consist of something
peculiar to science and demarcating it from other cognitive activities, or does
it consist essentially of a special case of general cognitive rationality? Three
types of answers are possible to such questions. One may say that science is
rational insofar as it evolves and develops in a way that can be understood
and explained, or insofar as one can show that the changes that it undergoes
are changes for the better, or insofar as it is the result of individuals thinking
reasonably and acting with good reasons. The rationality of science is reduced
to the explicability of scientific change in the first approach, to the provability
of scientific progress in the second, and to the rationality of scientists in the
third.
These three approaches derive from the types of questions that it is
possible to ask about a given scientific episode. One may ask, how and why
the episode took place, whether and how its occurrence was a good thing
(i.e., a change for the better), and whether the episode was rational in the
sense that persons involved in it acted rationally. Hence we may also say that
the three approaches derive from three distinct aspects of science, namely,
change, progress, and rationality (in a restricted sense of ‘rationality’,
according to which only conscious, intelligent beings can act rationally or
irrationally). Since this restricted sense of ‘rationality’ is the primary one, and
since the concepts of change and of progress apply rather well and specifically
to the two other aspects of a scientific episode, it is best to restrict the term
‘rationality’ to its primary sense, as we shall do below. We may also say that
the three approaches constitute three different types of inquiry, which are
respectively explanation, evaluation, and rational reconstruction (in the sense
of reconstruction of reasons).
These three approaches are related but distinct. The relations are as
follows. Sometimes it is possible to explain the occurrence of an episode as
being the result of scientists acting rationally; and moreover to show that a
given scientific change was one made by rational agents would define a sense

180
 THE RATIONALITY OF SCIENCE 181

in which it was a change for the better. In other words, the rationality of a
scientific change can sometimes explain its occurrence and sometimes justify
it as being good. The three approaches are however different, for obviously
not every change need be a change for the better; moreover, not all progress
derives exclusively from the rationality of the agents involved, since it could
be defined as approach to a certain desirable goal, and this may occur even if
the agents happened to have acted somewhat irrationally. Hence there is
nothing wrong with a certain amount of division of labor and with restricting
oneself to just one type of inquiry. However, just as the pursuit of a single
one of these topics does not guarantee success, so the simultaneous investiga¬
tion of all three does not necessarily involve confusion. For what one must do
is the following: if one’s claims and conclusions are about change then one’s
evidence and reasons must pertain to change and not progress or rationality;
if one’s conclusions are about progress then one’s evidence must pertain to
progress and not to change or rationality; and similarly for rationality.
This tripartite distinction helps to make sense of the work of recent philo¬
sophers of science who are historically oriented.1 It is also the one that seems
to be adumbrated by Stephen Toulmin’s project on human understanding.
His first volume2 dealt with what he calls “the collective use and evolution of
concepts” and corresponds basically to a study of scientific change. The
second projected volume was supposed to deal with “the individual grasp and
development of concepts”3 and would correspond to a study of scientific
rationality (in the primary sense). The third projected volume will presumably
deal with “the rational adequacy and appraisal of concepts”4 and promises to
be a study of scientific progress. The distinction also corresponds to Karl
Popper’s talk of the three worlds.5 His first world corresponds to that of
scientific change, his second to that of rationality, and his third to that of
progress.
The distinction could also be interpreted as follows, the study of scientific
change corresponds to history of science, the study of scientific rationality
to the psychology of science, and the study of scientific progress to the
epistemology of science. I would have no objection to this terminology if it
were not used in an invidious and prejudicial sense. However, usually talk of
the history, psychology, and epistemology of science goes together with
excluding the former two from the domain of philosophy of science, which is
thereby equated with the third. However, there is no reason why the study of
scientists’ reasoning cannot be carried out in a philosophical manner, nor is
there any guarantee that the epistemology of science will not degenerate into
a nonphilosophical linguistics of science, for example. And as regards the
182 CHAPTER 8

history of science, I have argued elsewhere6 that the explanation of scientific

change involves special methodological problems, of a kind that requires
philosophical sensibility, though the philosophy in question is not a philo¬
sophy of the historical development of science, but rather an ability to
perceive and use conceptual distinctions concerning the given subject matter,
so that such concrete philosophizing may be no more common among
scholars who call themselves ‘philosophers’ than among those who call them¬
selves ‘historians’.
However, this raises the question of the connection between my tripartite
distinction and the one usually made between historically oriented and
logically oriented philosophy of science. My distinction is not a simple expan¬
sion of the usual dichotomy. In fact, I would analyze most logically oriented
philosophy of science as being a mixture of mostly studies of rationality, of
some studies of progress, and a little history (mostly of contemporary
science); whereas the historically oriented philosophy of science consists
mostly of studies of change, with some studies of progress, and a few studies
of rationality; so that the difference is one of emphasis. However, a few
historical philosophers have emphasized the problem of rationality, notably
Mary Hesse7 and Paul Feyerabend;8 while a few philosophers who began with
a logical orientation, e.g., Karl Popper9 and Toulmin,10 ended up dealing
mostly with scientific change and progress.

SCIENTIFIC RATIONALITY

The main concern of the present investigation is scientific rationality. Few

recent philosophers have been as involved in solving, or at least dissolving,
this problem as Feyerabend; thus, since his views are relatively well-known
and accessible in several languages,11 I shall formulate mine in the course of a
critical examination of his recent book Against Method.
One of the most attractive features of Feyerabend’s approach is what I
shall call its concreteness. This is not his word, but it is obviously what he has
in mind when, for example, he criticizes Carnap’s excessively abstract
approach (p. 183, n. 7).12 Moreover, Feyerabend’s use of historical evidence
is an expression of this concreteness, for he is not interested in history as
such, but insofar as it puts the philosopher in better contact with scientific
practice.
Second, I agree with the anthropological aspect of Feyerabend’s approach
(Ch. 17, esp. pp. 249-60, and p. 252). This partly corresponds to what
some would call the phenomenological method (in the sense of existential
 THE RATIONALITY OF SCIENCE 183

phenomenology), and partly to the historical method, in one sense of ‘histor¬

ical’, according to which one suspends one’s own point of view and takes the
point of view of the historical agents. To be sure, Feyerabend does not always
practice this anthropological method that he preaches. For example, he
portrays Galileo as a counterinductivist, which is behavior in accordance with
the rule that “advises us to introduce and elaborate hypotheses which are
inconsistent with well-established theories and/or well-established facts”
(p. 29). However, none of this evidence, even if it were otherwise acceptable,
tends to show that Galileo introduces hypotheses which he believes are incon¬
sistent with well-established theories and/or facts; in other words, Feyerabend
gives no evidence that Galileo agrees with Feyerabend and with the Aris¬
totelians in thinking that the Ptolemaic system was well established. For
someone following the anthropological method it is Galileo’s actual thought
and beliefs that should be reconstructed. However, in spite of such lapses,
Feyerabend’s ‘anthropological’ orientation is to be welcomed.
A third acceptable feature of Feyerabend’s approach is his normative,
critical orientation. He wants to Find out not only what scientific rationality
is but whether it is good or bad, and he wants to reform it insofar as it is
bad. This may seem to conflict with the anthropological approach insofar as
the latter could be interpreted as emphasizing description, rather than pre¬
scription. There is no conflict, however, because Feyerabend is critical
primarily of contemporary science anthropologically understood, and he
grounds this criticism on past science also anthropologically understood. In
other words, he uses certain parts (historical stages) of science against other
parts. To be sure, Feyerabend also tends to exaggerate his criticism. For
example, noting that a streak of intolerance has developed within modern
science, he bemoans the fact “while an American can now choose the religion
he likes, he is still not permitted to demand that his children learn magic
rather than science at school” (p. 299). However, there is no good reason to
follow him in going that far. Moreover, some of his criticism is grounded on
extrascientific factors, for example that parts of modern scientific education
“cannot be reconciled with a humanitarian attitude” (p. 20); but such
criticism, however intrinsically plausible, derives its relevance from the other
grounded on the history of science. Thus, Feyerabend’s destructive, anti-
scientific pronouncements should not be confused with the healthy, re¬
formist, critical stance.
Fourth, I agree with Feyerabend’s argument that scientific rationality does
not consist of normative principles stated in terms of the theory/observation
distinction. His argument would be that as long as we use such concepts, then
184 CHAPTER 8

we are bound to value theories which are well-grounded on facts, and hence
to be intolerant toward theories that conflict with facts or other well-
established theories. But such intolerance will make it difficult or impossible
for better theories to be formulated or new facts discovered because a prin¬
ciple which Feyerabend labels ‘counterinduction’ can be shown to be a very
effective method. Feyerabend supports the counterinductive mle with an
historical and a philosophical argument. The historical argument consists of
evidence that most of the great discoveries in the history of science have had
this feature of conflicting with previously well-established theories and/or
facts. His philosophical argument has two parts, one relating to theoretical
counterinduction, the other to factual counterinduction. First, given a very
general theory, some facts cannot be discovered except with the help of a
conflicting alternative theory; hence if we value new facts, and if we want to
be critical toward a well-established theory, we must consider alternatives; in
short, the improvement of a general theory, no matter how well established,
is best done by the external criticism of contrast to other theories, rather
than by internal criticism of comparison with experience. As for factual
counterinduction, historical investigation shows that there is not a single
interesting theory that agrees with all known facts in its domain; and epistem¬
ological analysis shows that every observational report presupposes some
theoretical assumption; therefore it is a mistake to take ‘facts’ too seriously,
so seriously as to prevent the consideration of theories that conflict with
them. From such arguments Feyerabend does not conclude that counter¬
induction ought to be the new mle for the conduct of scientific research, but
rather that even such an obvious rule as the inductive principle which
counterinduction opposes is no guarantee of success (pp. 32—33). However,
he does conclude that there can be no rules of any kind, whereas the only
thing that follows is that there are no rules of the kind considered by him,
namely formulated in terms of the theoretical/observational distinction. In
other words, Feyerabend’s arguments ‘against method’, as he puts it, are
really arguments that we ought to do away with the theory/observation
distinction (cf., e.g., p. 168). However, as he himself argues in his book,
theories are not rejected until there is an alternative to replace them; applying
this to the theory of the ‘theory/observation’ distinction, we note that he has
not provided an alternative way of studying scientific rationality, which
would replace the philosophy of science based on the theory/observation
distinction. The present investigation is in part an attempt to provide such
an alternative by using the premise/conclusion distinction of logicians, that
is by using the concepts of argument and reasoning of elementary logic.
 THE RATIONALITY OF SCIENCE 185

This brings us to a consideration of Feyerabend’s discussion of the role of

reason in science, for after all his book is full of the rhetoric of irrationalism.
The first point to be made here is that even if he were otherwise right, his
irrationalism would be wrong. That is, even if it were true that “essential
ingredients of modern science survived only because reason was frequently
overruled in their past” (p. 145, ital. in original; cf. p. 155), this would only
show that it is sometimes reasonable to overrule ‘reason’, which can therefore
be only what was thought to be, but is not really, reason. Thus, in this type
of argument, far from it being the case that Feyerabend can’t lose (as some
critics have pointed out in other contexts),13 he can’t win.
Moreover, Feyerabend frequently (e.g., pp. 153—4) fails to distinguish the
nonrational from the irrational, and just because a certain procedure is not
in accordance with reason, he concludes that it is in violation of reason. Just
as frequently, Feyerabend concludes that something is irrational just because
it goes against what his opponents (be they empiricists, inductivists, or
Popperians) would call ‘rationality’ (cf., e.g., pp. 179-80). The best example
of this is his analysis of Galileo, which will be criticized in detail below. Here
I merely wish to point out that Feyerabend is at pains to point out that he
thinks that Galileo’s procedure was highly desirable, though it supposedly
goes against widely accepted methodological principles, so that those who
accept them would have to regard it as irrational.
Let us look at some of Feyerabend’s specific points against ‘reason’. His
considerations can be divided into three arguments, which we may call the
insufficiency of reason argument, the argument from incommensurability,
and the historical argument. In the first Feyerabend argues as follows. Since
the teaching of small children is not exclusively a matter of argument, and
since there are processes like the mastery of a language which look like the
result of reason but are due partly to indoctrination and partly to natural
processes of growth, it follows that nonrational growth is possible both in
adults and in institutions. Therefore “even the most puritanical rationalist
will then be forced to stop reasoning and use propaganda and coercion”
(p. 25) whenever conditions are such that “forms of argumentation turn out
to be too weak” (p. 25) to accomplish his ends. Feyerabend’s argument may
be accepted with the following qualifications. There is nothing irrational in
sometimes stopping reasoning; there would be only if one assumed that
reasoning is the only proper activity human beings can engage in, an
assumption no rationalist needs to make. Moreover, there is no need to resort
to propaganda and coercion, for propaganda is a perversion of rhetoric, while
coercion is a misuse of noncognitive causes, and such perversions and misuses
186 CHAPTER 8

may or may not be used by the rationalist; the only thing that follows is that
rhetoric and noncognitive causes may have to be used; but such things are
nonrational rather than irrational, and they are susceptible of being handled
properly as well as improperly. So the conclusion to be reached is that
reasons and arguments are not the only causes that affect human actions and
thought; there are also rhetoric and noncognitive causes that operate.
Now at this point, Feyerabend would like to add that therefore it would
be very strange if arguments were the only thing operating in scientific ration¬
ality, so strange that even if that were the case, then one should rebel against
such unnatural restrictions and help to create a more human science in which
the other factors are not excised. To this we may reply by asking why should
every human activity be a microcosm of the entire human life? Why can’t
there be some activity, e.g., science, where the only rules of the game involve
argumentation? As long as we remember that there are other things in fife,
why should this activity of reasoning which we have labeled “science” be
mixed with the others? I believe Feyerabend’s only plausible answer could be,
because as a matter of (historical) fact what we label ‘science’ does contain a
mixture of elements. Here we would leave Feyerabend’s general considera¬
tions and go to his historical argument, which therefore is more crucial to his
position than he makes it seem.
Before examining Feyerabend’s historical argument, however, let us look
at the one from incommensurability, which he thinks “creates problems for
all theories of rationality” (p. 214). Feyerabend calls two theories incom¬
mensurable (pp. 223, 228—9, 269) when the subject matter to which they
refer differs so radically that (1) the existence of the entities and processes
presupposed by one theory implies the nonexistence of the entities and pro¬
cesses presupposed by the other theory, and (2) it makes no sense to say that
one theory is a generalization of the other, or that they overlap. The main
step in the argument consists in trying to establish that there are scientific
changes where the theories involved are incommensurable, e.g., Aristotelian
and Galilean physics, classical and quantum physics, and Newtonian and
relativistic mechanics (pp. 224-5, 271, 276-7); given this incommen¬
surability, Feyerabend concludes that one cannot say that the new theory
which replaces the old is epistemologically better. However, if we distinguish
progress from rationality, then we realize that Feyerabend’s conclusion does
not imply that the agents involved in the transition did not act rationally.
In fact, while answering a number of philosophical objections against his
incommensurability claims (pp. 277—85), Feyerabend himself argues cogently
that, in spite of incommensurability, it is still possible to behave rationally;
 THE RATIONALITY OF SCIENCE 187

for example, a self-inconsistent observation report would disconfirm the

corresponding theory (p. 278); one could learn the meaning of the new
theory the way anthropologists learn the language of newly-discovered tribes
(pp. 278—82); and crucial experiments are still possible (pp. 282—3). So
Feyerabend has not shown that incommensurability creates problems for
theories of rationality, but rather that it creates problems for theories of
progress. Moreover, since, besides being distinct, rationality and progress are
related, the possibility of rationality opens up a minimal possibility of pro¬
gress, namely that changes involving incommensurable theories are changes
for the better in the sense that they result from the behavior of rational
agents.
But how shall we test for the actuality of these possibilities still left open
after Feyerabend’s arguments? The answer, whose suggestion can be found
in Feyerabend himself, is that the test is to be conducted by anthropological
field work: “Let us commence field work in this domain also, and let us
study the language of new theories not in the definition-factories of the
double language model, but in the company of those metaphysicians,
physicists, playwrights, courtesans, who have constructed new world views!
(p. 282). When the event under consideration is relatively far into the past,
like the emergence of modern science in the 17th century, the only kind of
field work possible is the analysis of historical records. However, these
historical records must be sufficiently rich, and the analysis must be con¬
ducted with an appropriate ‘anthropological spirit. For the emergence of
modern science a sufficiently rich record can be found in Galileo’s Dialogue,
parts of which have therefore been examined by Feyerabend, though, alas,
without a sufficiently ‘anthropological’ attitude, as I shall soon show.
This brings us to his historical argument against reason. Here he tries to
show that “essential ingredients of modem science survived only because
reason was frequently overruled in the past" (p.145). When we look at the
details of the argument (e.g., pp. 145-61, 179-80, 196-200) we discover,
as mentioned earlier, that the reason which was overruled was usually what
empiricists, inductivists, Popperian critical rationalists, or Imre Lakatos think
is reason. In other cases, there was no overriding of reason but rather a com¬
bination of it with nonrational factors; nor does Feyerabend show that such
nonrational factors were improper from their own point of view. All of the
strengths and all of the limitations of Feyerabend’s position are present in
his analysis of Galileo, and so it will be valuable to examine this in detail.
188 CHAPTER 8

GALILEO

The longest case study on which Feyerabend grounds many of his philoso¬
phical conclusions concerns Galileo. The account consists of four elements:
an interpretation of Galileo’s work in terms of a definite procedure which he
allegedly follows; a description of this procedure in terms of such concepts
as theory, observation, propaganda, appeals to emotion and prejudices, etc;
a historical explanation of the fact of Galileo’s success as resulting from the
procedures he used; and a favorable evaluation of Galileo insofar as the pro¬
cedure he follows is ‘perfectly reasonable’ (p. 143).
One need not have mentioned the first item in this list were it not for the
fact that all of Feyerabend’s talk against method may easily be interpreted as
showing that the concept of method is a useless one in the philosophy of
science. It is clear, however, from Feyerabend’s ‘actual procedure’ (his
method, as it were) that he is only arguing against “the idea of a method that
contains firm, unchanging, and absolutely binding principles for conducting
the business of science” (p. 23), i.e., “the idea that science can, and should be
run according to fixed and universal rules” (p. 295). In a few places (e.g.,
pp. 145, and 163) he even uses the forbidden word ‘method’ to refer to
Galileo’s procedure. So Feyerabend’s point is really that no general theory of
method is defensible, which says that all scientists in all situations use a
certain definite method; instead different scientists use different methods on
different occasions. It is important to note, however, that Feyerab^nd’s
account presupposes that the concept of method is a very useful one in
understanding what a scientist does; a method becomes the effective cause
(in a nondeterministic sense) in the historical explanation of a given scientific
success. I do not think that such method-explanations of success are objec¬
tionable on general grounds; I think that such accounts provide historical
understanding and can even be used to ground philosophical theses.14 How¬
ever, in this case the following questions must be raised.
First, it may be that Feyerabend is not being radical enough in ascribing
a certain definite method to Galileo; that is, it may be that not even the same
scientist follows the same method to any significant extent in his activities.
Perhaps there is ‘anarchy’ within the work of a given scientist. What I am
saying is that, for Feyerabend, science is an ‘anarchical’ enterprise, but
individual scientists are not.15 There is no logical inconsistency here; but
there is a tension within Feyerabend’s account since he tends to transfer the
characteristics from science-the-institution to the individual scientist. I believe
there is textual evidence for this tendency in Feyerabend, but let me illustrate
 THE RATIONALITY OF SCIENCE 189

the problem as follows. Feyerabend presupposes that one could not under¬
stand Galileo’s work unless one found some order in it, enough to speak of a
method or procedure. But if one does this with an individual scientist, why
shouldn’t he do the same with science as a whole? Of course, one is not
logically bound to conclude that there is at least as much order in science as
in a scientist, but if he chooses to order a given scientist’s activities by
reference to a definite method, why should he not do enough work so as to
order science as a whole into a method? Feyerabend might answer, because
this can be done in the case of a given scientist, e.g., Galileo, but not for the
whole of science. What follows from this is that one has to be very careful
about whether or not there is method within a scientist’s activities. To some
extent this problem reduces to that of describing the method in a sufficiently
complex and sophisticated way. However, if this complexity was too great
then we wouldn’t have a method. In these investigations I am partly testing
this assumption that Feyerabend seems to make.
Second, once we realize that Feyerabend depicts an anarchical science con¬
sisting of methodical scientists, the possibility arises for a methodical science
consisting of anarchical scientists. Given these two abstract alternatives, it is
not clear which Feyerabend would choose. Perhaps he would opt for a third
possibility of an anarchical science of anarchical scientists; though this may
be his wish, it would go against his actual account in Against Method.
Third, Feyerabend’s method-explanation of Galileo’s success never takes
seriously the question of the connection between the procedures and the
success. That is, even if Feyerabend is right about the methods he attributes
to Galileo, there is no argument that it was because of these methods that he
succeeded. There is no question that Feyerabend is committed to such a
causal claim (e.g., p. 112), which therefore requires careful examination. One
way to conduct this examination is to ask whether the situations in which
Galileo proceeded a la Feyerabend involved activities which would be
regarded as successful. For example, he examines mostly Galileo’s attempts to
prove the motion of the earth; however, few people, be they 17th century
contemporaries of Galileo, or present-day scholars, would agree that Galileo
succeeded in his attempts. Feyerabend would be the first to claim that
Galileo failed here. But if he failed when he was proceeding as Feyerabend
claims, then this is, if anything, evidence that such Galilean procedures led to
failure, not that they led to success.
Prescinding now from such problems, let us see how Galileo’s method is
described. Feyerabend deserves great credit for enlarging Galileo’s method
beyond the epistemological factors usually considered and for including
190 CHAPTER 8

rhetorical and artistic components. The epistemological factors considered

by Feyerabend are the following. (1) Galileo introduces and elaborates
hypotheses (e.g., motion of the earth, relativity of motion, celestial reli¬
ability of telescopes) which are inconsistent with well-established theories
(e.g., geocentrism, operative character of motion, optics) and with well-
established facts (e.g., vertical fall, apparent size of Venus and Mars); this
procedure is labeled ‘counterinduction (pp. 29, 77, 99—101). (2) Galileo
lets inadequate views mutually support each other, e.g., (a) “an inadequate
view, the Copernican theory, is supported by another inadequate view,
the idea of the nonoperative character of shared motion” (p. 89), (b) he
lets one refuted wiew — Copernicanism — support another refuted view
— the idea that telescopic phenomena are faithful images of the sky (pp.
141_2), and (c) “Galileo changed his view about the ‘neutral’ motions -
he made them permanent and ‘natural’ — in order to make them compa¬
tible with the rotation of the earth” (p. 96); Feyerabend gives no special
name to this aspect of Galileo’s procedure, which I shall label ‘dialectical’,
to follow some terminology suggested elsewhere.16 (3) Ad hocness: Galileo
uncritically accepts any and all ideas and observations which support
Copernicanism (e.g., p. 90, 93-8). (4) Galileo drastically reduced the con¬
tent of dynamics by replacing the Aristotelian comprehensive theory of
change, including locomotion, qualitative change, and generation and corrup¬
tion by a theory dealing merely with the locomotion of matter (pp. 90-
100, 160-1); this is what Feyerabend labels the ‘backward step’ (pp. 176,
153, 113).
The rhetorical aspects of Galileo’s method are allegedly the following:
deceptive tactics (pp. 70, 81, 87, 160); utterances which are arguments in
appearances only (p. 81); propaganda (pp. 81,90, 157, 160); psychological
tricks (pp. 81, 88, 154); persuasion by confusion (p. 84); clever techniques
of persuasion (pp. 141, 143); misleading insinuations (p. 160); distortion
(p. 160); appeal to emotion and prejudice (p. 154); jokes (p. 154); non
sequiturs(p. 154).
The artistic, aesthetic, literary factors include “style, elegance of expres¬
sion, simplicity of expression, tension of plot and narrative” (p. 157), “a
sense of humor, an elasticity . . . and an awareness of the valuable weaknesses
of human thinking” (p. 161).
The epistemological factors emphasized by Feyerabend are ones that go
against almost every principle held by orthodox philosophers of science; the
rhetorical factors seem to violate the ideas of the basic honesty and decency of
science, widely held by scientists and laymen alike; the aesthetic factors con-
 THE RATIONALITY OF SCIENCE 191

tradict the alleged gap between the arts and the sciences. Indeed Feyerabend
delights in being contrary. His contrariness reaches its highest pitch when he
argues that such Galilean procedure is ‘perfectly reasonable’ in itself (pp. 143,
145—61) and fruitful in other fields (pp. 163—4). In other words, Feyer-
abend's evaluation of Galileo’s method is the orthodox one, it is his descriptive
interpretation of the features of that method that is unorthodox; and since
his evaluation is partly grounded on his interpretation, that means that his
reasons for his evaluation are also unorthodox. Feyerabend does not always
effectively distinguish his description and his evaluation, though he often
seems to or tries (pp. 143, 156). Thus we often find him arguing that because
it can be shown (by Feyerabend himself) that a certain view was refuted or
inadequate, and because therefore Galileo should have seen that it was
inadequate, therefore it follows that Galileo regarded it as inadequate, and
hence that he was acting counterinductively; or that because it can be shown
(by Feyerabend himself) that a certain argument is incorrect, and because
therefore Galileo should have known that it was incorrect, therefore it
follows that he knew it to be incorrect, and hence that his using it amounted
to deceptive trickery. We might say that Feyerabend’s alleged descriptive
interpretations are really evaluations, and hence his whole case is historically
unfounded; and since, as we argued earlier, his two other arguments (from the
insufficency of reason, and from incommensurability) depended on the
historical argument, we might conclude that Feyerabend’s views do not merit
any further consideration.
This, however, would be a mistake. We need not deny our inclination to
think that Feyerabend is doubly in the wrong, namely that his historical
interpretations are descriptively wrong, and that //they were right, then that
would show that Galileo was wrong (irrational). The fact is, or rather my
suspicion is, that Feyerabend turns out to be right in spite of himself That is,
his account of Galileo is not really irrationalistic, but pseudo-irrationalistic,
and in reality Galileo proceeds rationally for Feyerabend; however, our idea
of scientific rationality must be expanded first to allow aesthetic and
rhetorical factors, and second epistemological practices proscribed by
orthodox philosophies of science; moreover the aesthetic and rhetorical
factors are by themselves merely alogical, and they must be judged by their
own criteria; finally, whatever unorthodox epistemological practices Galileo
may engage in, the real test of their rationality or propriety is their corres¬
pondence to basic and elementary forms of reasoning and argumentation,
rather than to philosophically articulated theories of scientific rationality,
most of which presuppose the theory/observation distinction, for example.
192 CHAPTER 8

Keeping this in mind, I now proceed to test my suspicion. I shall do so by

examining Feyerabend’s account of the tower argument.

THE TOWER ARGUMENT

The tower argument was one of the classical objections to the earth s rotation.
Galileo states and criticizes it in the following passage:

SALV. . . Aristotle says, then, that a most certain proof of the earth’s being motionless
is that things projected perpendicularly upward are seen to return by the same line to
the same place from which they were thrown, even though the movement is extremely
high. This, he argues, could not happen if the earth moved, since in the time during
which the projectile is moving upward and then downward it is separated from the earth,
and the place from which the projectile began its motion would go a long way toward
the east, thanks to the revolving of the earth, and the falling projectile would strike the
earth that distance away from the place in question. Thus we can accommodate here the
argument of the cannon ball as well as the other argument, used by Aristotle and
Ptolemy, of seeing heavy bodies falling from great heights along a straight line perpendic¬
ular to the surface of the earth. Now, in order to begin to untie these knots, I ask
Simplicio by what means he would prove that freely falling bodies go along straight and
perpendicular lines directed toward the center, should anyone refuse to grant this to
Aristotle and Ptolemy.
SIMP. By means of the senses, which assure us that the tower is straight and perpendic¬
ular, and which show us that a falling stone goes along grazing it, without deviating a
hairsbreadth to one side or the other, and strikes at the foot of the tower exactly under
the place from which it was dropped.
SALV. But if it happened that the earth rotated, and consequently carried along the
tower, and if the falling stone were seen to graze the side of the tower just the same,
what would its motion then have to be?
SIMP. In that case one would have to say ‘its motions’, for there would be one with
which it went from top to bottom, and another one needed for following the path of the
tower.
SALV. The motion would then be a compound of two motions; the one with which it
measures the tower, and the other with which it follows it. From this compounding it
would follow that the rock would no longer describe that simple straight perpendicular
line, but a slanting one, and perhaps not straight.
SIMP. I don’t know about its not being straight, but I understand well enough that it
would have to be slanting, and different from the straight perpendicular line it would
describe with the earth motionless.
SALV. Hence just from seeing the falling stone graze the tower, you could not say for
sure that it described a straight and perpendicular line, unless you first assumed the earth
to stand still.
SIMP. Exactly so; for if the earth were moving, the motion of the stone would be
slanting and not perpendicular.
SALV. Then here, clear and evident, is the paralogism of Aristotle and of Ptolemy,
discovered by you yourself. They take as known that which is intended to be proved.
 THE RATIONALITY OF SCIENCE 193

SIMP. In what way? It looks to me like a syllogism in proper form, and not a petitio
principii.
SALV. In this way: Does he not, in his proof, take the conclusion as unknown?
SIMP. Unknown, for otherwise it would be superfluous to prove it.
SALV. And the middle term; does he not require that to be known?
SIMP. Of course; otherwise it would be an attempt to prove ignotum per aeque ignotum.
SALV. Our conclusion, which is unknown and is to be proved; is this not the motion¬
lessness of the earth?
SIMP. That is what it is.
SALV. Is not the middle term, which must be known, the straight and perpendicular
fall of the stone?
SIMP. That is the middle term.
SALV. But wasn’t it concluded a little while ago that we could not have any knowledge
of this fall being straight and perpendicular unless it was first known that the earth stood
still? Therefore in your syllogism, the certainty of the middle term is drawn from the
uncertainty of the conclusion. Thus you see how, and how badly, it is a paralogism.17

The argument which according to Galileo’s spokesman, Salviati, begs the

question is the one contained in Salviati’s first speech in this passage; it may
be reconstructed as follows:

0) Bodies fall vertically.

(2) If the earth rotated, bodies could not fall vertically,
because

(3) If the earth rotated, then while a body was falling the place on
the earth directly below it would be carried along toward the
east and the body would land to the west of where it was original¬
ly dropped.
Therefore,

(4) The earth does not rotate.

Galileo has no objection to the (formal) validity of the last step in this
argument (the one from (1) and (2) to (4)); however, he questions its sound¬
ness in terms of whether premise (1) is justified. The Aristotelian spokesman,
Simplicio, gives the following justification:

(1) Bodies fall vertically

because

(5) Bodies are seen to fall vertically.

Now, with respect to this argument Galileo accepts the premise but questions
its (formal) validity. This validity depends on whether or not apparent vertical
194 CHAPTER 8

fall implies actual vertical fall. This implication needs justification because it
would not hold on a rotating earth. In fact, if the earth were rotating and
bodies were seen to fall vertically, then in actuality they would be following a
path slanted to the earth’s surface; that is, on a rotating earth, apparent
vertical fall would not imply actual vertical fall. Now, how could one possibly
justify the implication? The only relevant argument would seem to be the
following one:
(6) If the earth does not rotate, then apparent vertical fall implies
actual vertical fall.
(7) The earth does not rotate.

Therefore,
(8) Apparent vertical fall implies actual vertical fall.

Premise (6) is indeed true, and the argument is (formally) valid. Unfortunately
premise (7) is identical with conclusion (4) of the original argument, which is
being examined. In short, the vertical fall argument, from (1) and (2) to (4),
is such that one of its premises (1) is being justified on the basis of the same
proposition (7) [=(4)] it has for conclusion.
The important issues can be put into focus in terms of the following argu¬
ment, which combines the three separate segments just discussed, and where
some obvious symbolic abbreviations have been made and the proposition
numbers correspond:

(6) If not-R, then S implies V.

(7) Not-R.
(8) S implies V.
(5) S.
(1) ••• v.
(2) If R, then not-V.
(4) /. Not-R.

Logically speaking, this is a very interesting deduction; in particular, its three

steps are valid. Rhetorically speaking, however, the identity of propositions
(7) and (4) makes it worthless: someone who does not adhere to (4), will not
be any more favorably inclined toward it after tliis argument, which requires
such adherence already at the very beginning.
In Feyerabend’s account four main theses are involved. First (pp. 70-75),
the Aristotelian tower argument presupposes naive realism with respect to
motion (p. 75); and it is important to notice that this Aristotelian presupposi-
 THE RATIONALITY OF SCIENCE 195

tion takes the form of what Feyerabend calls a ‘natural interpretation’; that
is, “one does not First distinguish the apparent motion from the real motion
and then connect the two by a correspondence rule. One rather describes,
perceives, acts towards motion as if it were already the real thing” (p. 75).
Second (pp. 75—8), Galileo discovered this Aristotelian natural interpretation
counterinductively. Third (pp. 78—81), Galileo tests and examines the
Aristotelian identification of real and apparent motion by introducing a
different interpretation which “restores the senses to their position as instru¬
ments of exploration, but only with respect to the reality of relative motion.
Motion ‘among things which share it in common’ is ‘non-operative’, that is,
‘it remains insensible, imperceptible, and without any effect whatever’”
(p. 78). Fourth (pp. 81—92), Galileo uses, and needs to use, propaganda,
psychological tricks, and deception in elaborating this new concept of the
relativity of motion.
It is a great insight for Feyerabend to have arrived at the first of these
conclusions, though it must be qualified to restrict it to vertical fall, so as to
say that the tower argument presupposes, as a ‘natural interpretation’, the
identification of real and apparent vertical fall. Feyerabend himself goes
through the motions of making the qualification (p. 75), but he then forgets
its importance in the rest of his discussion. At any rate, it is possible to give a
textual and logical proof of the qualified conclusion, which will be done later
in another context.18 However, that same analysis also shows that Feyerabend
is completely wrong in his second thesis; Galileo is simply not aware of this
Aristotelian presupposition of the tower argument, so he neither discovers it
counterinductively, nor does he discover it at all. We can discover it by logical
manipulations, but these are merely suggested by the text, nowhere contained
in it.19 Feyerabend formulates his second thesis because he approaches the
philosophy of science with the spectacles of the theory/observation distinc¬
tion; so he wants to reconstruct some theory which the Aristotelians held,
and the one that Galileo introduced. If Feyerabend had approached the tower
argument as an argument, which is the way Galileo does, then he would have
been inclined to interpret Galileo’s discovery as one about an argument,
namely that the tower argument is a petitio principii. Though Galileo could
have proceeded to inquire about general theories and natural interpretations,
the fact is that he does not do so; and he does not do so because he seems to
be interested in arguments. Thirdly, Galileo introduces his concept of the
relativity of motion not to test a presupposition of the tower argument, but
to answer the objection to the earth’s motion from the deception of the
senses. Feyerabend thinks as he does because he fails to distinguish two
196 CHAPTER 8

different arguments against the earth’s motion, namely the one from vertical
fall and the one from the deception of the senses. The logic of the two argu¬
ments is different. Galileo correctly criticizes the first as being circular, the
second as being groundless,20 in the sense that the earth’s motion would not
involve any deception of the senses; and there would be no deception of the
senses because of the relativity of motion; in other words, it s not a case
where one does not see something he ought to see (the real slanted path on
a rotating earth); there is no reason why one ought to see such a slanted path
because our eyes can only detect motion relative to them, that our eyes are
so built can be proved by experiences on moving systems such as ships. Such
being the logic of the argument from the deception of the senses, and
Galileo’s use therein of the relativity of motion, Galileo needs neither pro¬
paganda, nor tricks, nor deceptions, which are figments of Feyerabend s
imagination, as we will see below. He so imagines because he thinks wrongly
that Galileo is using the relativity of motion in discussing the tower argument.
Feyerabend does not realize that Galileo’s analysis of this argument is not
designed to produce a positive, substantive conclusion about phenomena, but
rather a negative destructive criticism of the Aristotelian argument. For this
purpose Galileo does not need the relativity of motion. //Galileo were using
the relativity of motion in connection with the tower argument, and if he
were thereby trying to prove the earth’s motion rather than merely refuting
the argument, then Galileo would have to resort to propaganda, tricks, and
deception, since that task cannot be accomplished by legitimate means.
However, such a conditional is counterfactual.
At this point the question arises, whether Galileo is doing anything wrong
in the context of his discussion of the deception of the senses argument,
where he admittedly uses the relativity principle. In that context, Galileo
refutes the objection by arguing that there would be no deception of the
senses if we experienced apparent vertical fall on a rotating earth, because
only motion relative to the observer is perceivable. In this critique Galileo is
certainly subsuming vertical fall under what Feyerabend calls ‘Paradigm II’
(p. 87), namely the paradigm of motion in a moving system, like a boat. Is
he therefore being deceptive, etc. in so doing? Not at all, because this is the
same paradigm used in the statement of the objection, which is that on a
rotating earth our senses would be deceived in experiencing the appearance of
vertical fall while in actuality the fall was over a slanted path. In other words,
this different argument against the motion of the earth cannot even be made
without using Paradigm II. Hence, in answering it, it is quite proper to use the
same paradigm.
 THE RATIONALITY OF SCIENCE 197

And this brings us to Feyerabend’s fourth thesis. He is certainly right that

Galileo uses rhetorical alogical considerations which are interesting to
examine and which play an important function. However, their propriety is
a very complicated affair. In order to begin to understand this matter, it is
essential to have in front of us the text of the dialogue immediately following
the above quotation:

SAGR. On behalf of Simplicio I should like, if possible, to defend Aristotle, or at least to

be better persuaded as to the force of your deduction. You say that seeing the stone
graze the tower is not enough to assure us that the motion of the rock is perpendicular
(and this is the middle term of the syllogism) unless one assumes the earth to stand still
(which is the conclusion to be proved). For if the tower moved along with the earth and
the rock grazed it, the motion of the rock would be slanting, and not perpendicular.
But I reply that if the tower were moving, it would be impossible for the rock to fall
grazing it; therefore, from the scraping fall is inferred the stability of the earth.
SIMP. So it is. For to expect the rock to go grazing the tower if that were carried along
by the earth would be requiring the rock to have two natural motions; that is, a straight
one toward the center, and a circular one about the center, which is impossible.
SALV. So Aristotle’s defense consists in its being impossible, or at least in his having
considered it impossible, that the rock might move with a motion mixed of straight and
circular. For if he had not held it to be impossible that the stone might move both
toward and around the center at the same time, he would have understood how it could
happen that the falling rock might go grazing the tower whether that was moving or was
standing still, and consequently he would have been able to perceive that this grazing
could imply nothing as to the motion or rest of the earth.
Nevertheless this does not excuse Aristotle, not only because if he did have this idea
he ought to have said so, it being such an important point in the argument, but also, and
more so, because it cannot be said either that such an effect is impossible or that
Aristotle considered it impossible. The former cannot be said because, as I shall shortly
prove to you, this is not only possible, but necessary; and the latter cannot be said
either, because Aristotle himself admits that fire moves naturally upward in a straight
line and also turns in the diurnal motion which is imparted by the sky to all the elements
of fire and to the greater part of the air. Therefore if he saw no impossibility in the
mixing of straight-upward with circular motion, as communicated to fire and to the air
up as far as the moon’s orbit, no more should he deem this impossible with regard to the
rock’s straight-downward motion and the circular motion natural to the entire globe of
the earth, of which the rock is a part.
SIMP. It does not look that way to me at all. If the element of fire goes around together
with the air, this is a very easy and even a necessary thing for a particle of fire, which,
rising high from the earth, receives that very motion in passing through the moving air,
being so tenuous and light a body and so easily moved. But it is quite incredible that a
very heavy rock or a cannon ball which is dropping without restraint should let itself be
budged by the air or by anything else. Besides which, there is the very appropriate
experiment of the stone dropped from the top of the mast of a ship, which falls to the
foot of the mast when the ship is standing still, but falls as far from the same point when
198 CHAPTER 8

the ship is sailing as the ship is perceived to have advanced during the time of the fall,
this being several yards when the ship’s course is rapid.
SALV. There is a considerable difference between the matter of the ship and that of the
earth under the assumption that the diurnal motion belongs to the terrestrial globe.
For... 21

It was certainly a clever move for Galileo to begin his critique of the tower
argument (in the earlier quotation) by interpreting it to refer to actual
vertical fall. Feyerabend is right in saying that the Aristotelian argument
simply interchanges apparent and actual fall without distinguishing them; so
Galileo could have begun with either version of the argument. The fault of
the argument from apparent vertical fall (Sagredo’s first speech in the passage
just quoted) is that it depends on a premise which is as much in need of proof
as the conclusion at issue. This fault is less serious than the circularity of the
argument from actual vertical fall. It is indeed more effective to start with the
more serious criticism, and so Galileo’s first rhetorical move is a judicious one.
Having focused on the actual vertical fall version, Galileo had in front of
him a formally valid instance of denying the consequent {modus tollens). Of
the two premises of this argument, propositions (1) and (2) above, Galileo
could have questioned either one. If he had questioned the conditional pro¬
position, the circularity could not have been exhibited as easily since the
justification of this conditional premise is circular only insofar as it depends
ultimately on the nonconditional premise of the original modus tollens,
namely the proposition (1) that bodies really fall vertically.22 It is certainly
rhetorically effective, but certainly not improper, to exhibit the failure of an
argument in the easiest possible way.
Third, when Galileo comes around to discuss the apparent vertical fall
version of the argument, this is made to look as a revised version of the
original argument, and hence to some extent as ad hoc. This seems to be a
merely rhetorical fault, rather than a logical one. However, the argument is
also shown to be dependent on a premise which is as much in need of proof
as its conclusion. This feature, though not purely logical, is not purely
rhetorical either. For though questions of the comparative knowability of
propositions are context-dependent, the context is an epistemological one,
rather than one dependent merely on who is speaking to whom.
Fourth, the way that the discussion changes from apparent vertical fall to
the ship experiment involves rhetorical considerations. Someone might say,
following Feyerabend’s evaluation of Galileo’s rhetoric, that besides intro¬
ducing the argument from apparent vertical fall in a prejudicial manner, he
drops the topic in such a way as to make it look as if Simplicio was giving up
 THE RATIONALITY OF SCIENCE 199

on the argument and relying on a new one.23 This rhetorical evaluation would
depend on interpreting Galileo’s criticism of the ship/earth analogy,24 which
follows the passage just quoted, as part of his criticism of the ship experi¬
ment argument against the earth’s motion.25 This interpretation in turn
would involve attributing to Galileo another deception, namely that he would
be trying to show that a rock dropped from the top of the mast on a moving
ship would land at the foot of the mast, and similarly on a rotating earth,
while he does not believe in the analogy. However, it is itself a prejudicial
interpretation to regard the passage where Galileo criticizes the ship/earth
analogy as a logical part of his discussion of the ship experiment argument.
No sound reason can be given for this interpretation, whereas evidence can
be given to show that Galileo wants his criticism of the analogy to be part
of his criticism of the apparent vertical fall argument. The evidence consists,
first, of the logical fact that such a connection exists; for the alleged results
of the ship experiment can be cited as evidence that bodies refuse to move
simultaneously toward and around the center, and hence as support for
the corresponding premise in the apparent vertical fall argument, namely
the proposition that if the earth rotated, bodies could not be seen to fall
vertically. Moreover, there is textual evidence that this is what Galileo has in
mind. For example, the first time that the ship experiment is mentioned26 is
in the context of an argument where the impossibility of mixed motion is
given as the explanation for the alleged deflection on the moving ship.
Finally, the way the switch is made from the criticism of the analogy to the
criticism of the alleged experimental results27 suggests the same thing:
Salviati agrees that ‘up to this point’ the ship experiment as an argument
purely from analogy has not been considered. In other words, in this passage
Galileo is saying that though the ship experiment has just been criticized
insofar as it is the basis of an explanatory argument against mixed motion, it
is still standing as a pure analogy argument against the earth’s motion. I
believe Galileo is right here, and that a pure analogy argument has some force
even though the mechanism of the analogy is unknown, i.e., even though we
do not know why the analogy holds, as long as we are prepared to claim that
it holds. At any rate, the conclusion is inescapable that interpretations of the
rhetorical situation must be grounded on interpretations of the logical situa¬
tion; and since it emerged earlier that logical analysis was crucial for the
understanding of scientific rationality per se, this means that a systematic
logical analysis of a crucial work like Galileo’s Dialogue is very important for
for the philosophy of science.
In summary, scientific rationality is distinct from, but related to, scientific
200 CHAPTER 8

progress and change. Feyerabend’s approach can be easily misunderstood as

being merely negative but suggests two very important insights. First, the
anarchism of “anything goes” (which ought to be distinguished from the
specific counterinductive rule) may be understood as a manner of speaking
about judgment, and when so interpreted it receives the support of the
evidence of what I earlier called Galileo’s synthetic or “dialectical method¬
ology. Second, the propagandist-manipulative interpretation of scientific
rationality may be taken as being itself a rhetorical exaggeration of the truth
that rhetorical persuasion has an important role. Underlying both elements
stands the logical dimension of science (logical in the sense of the theory and
practice of reasoning), as a deeper level to which judgment is reducible, and
on which rhetoric needs to be grounded.

NOTES

1 Cf. I. Lakatos and A. Musgrave (eds.), Criticism and the Growth of Knowledge and my
Essay-review of this book.
2 S. Toulmin, Human Understanding, Vol. I: The Collective Use and Evolution of
Comcepts.
3 Ibid., p. ii.
4 Ibid.
5 K. Popper, Objective Knowledge: An Evolutionary Approach, pp. 106-90, esp.
pp. 106 and 153.
6 History of Science as Explanation.
7 M. Hesse, The Structure of Scientific Inference.
8 P. K. Feyerabend, ‘Explanation, Reduction, and Empiricism’; idem, ‘Problems of
Empiricism, Part II’; idem, ‘Consolations for the Specialist’; idem, Against Method.
9 K. Popper, The Logic of Scientific Discovery. This book is a translation of his Logik
der Forschung.
10 S. Toulmin, The Philosophy of Science.
11 P. K. Feyerabend, I probemi dell’ empirismo; idem, Contro il metodo; idem,
Einfuhriing in die Naturphilosophie; idem, Ausgewdhlte Aufsatze; idem, Wider den
Methodenzwangtheorie.
12 Hereafter references to Against Method will be made in parenthesis in the text.
13 E. Gellner, ‘Beyond Truth and Falsehood’.
14 Cf. my History of Science as Explanation, pp. 223-8.
15 Feyerabend has objected in private correspondence that he is not ascribing a single
definite method to Galileo since “he does different things in the case of dynamics and
in the case of optics (in the first case he changes the grammar of dynamical terms, in the
second case he changes sensations by substituting the telescope for the eye”. Though it
is true that Feyerabend examines Galileo’s work in these two contexts, and that in the
sense just described different procedures are operative, it is also true that both are sub¬
sumed under Feyerabend’s concept of counterinduction, the counterinductive idea in
 THE RATIONALITY OF SCIENCE 201

the former case being the earth’s motion, in the latter, the celestial reliability of the
telescope. Hence, Galileo’s activities are being ordered in a nonanarchical fashion.
16 A. Funkenstein, The Dialectical Preparation for Scientific Revolutions’, and my
‘Dialectical Aspects of the Copernican Revolution: Conceptual Elucidations and
Historiographical Problems’.
17 G. Galilei, Dialogue Concerning the Two Chief World Systems, tr. Drake, pp. 139—
140.
18 See below Chapter 12.
19 In my reconstruction (Chapter 12), twenty five steps are needed to arrive at a con¬
clusion of this form: R if and only if not-(S if and only if V). Feyerabend’s insight is to
realize that this proposition is implied by Galileo’s critique, but his error also occurs at
this point. Being under the wrong impression that Galileo is here not merely trying to
refute the tower argument, but also trying to prove the earth’s motion, and confusing
the tower and the deception of senses arguments, Feyerabend has Galileo support the
right side of this biconditional, namely not-(S if only if V), on the basis of the relativity
principle. Feyerabend’s reconstruction is the following: apparent and actual vertical fall
are distinct on moving systems like a ship, since only relative motion is visible in such
systems; hence apparent and actual fall with respect to the earth are distinct. Feyerabend
correctly points out that this amounts to subsuming vertical motion under ‘Paradigm II:
Motion of objects in boats, coaches, and other moving systems’ (p. 87) instead of under
‘Paradigm I: Motion of compact objects in stable surroundings of great spatial extension
- deer observed by the hunter’ (p. 87). He is also right in saying that the gap between
these two paradigms is unbridgeable, in the sense that there is no noncircular way of
choosing the former for the purpose of applying it to vertical fall. He is wrong, however,
in attributing such an argument to Galileo, who wisely avoided it.
20 For statements and critiques of the argument on the deception of the senses see
Galilei, Dialogue, tr. Drake, pp. 167—71 and 233-47, and Galilei, edited by Favaro,
Opere 7, 197-200, and 272-81.
21 Galilei, Dialogue, tr. Drake, pp. 140-42.
22 See my ‘Galileo as a Logician’, pp. 144—45. See also Chapters 12 and 14 below.
23 Galilei, Dialogue, tr. Drake, p. 141; idem, Opere 7, 167.
24 Dialogue, tr. Drake, pp. 141-43; Opere 7, 167-69.
25 Dialogue, pp. 143-49; Opere 7, 169-75.
26 Dialogue, p. 126; Opere 7, 151-52.
27 Dialogue, p. 143; Opere 7, 169.
 CHAPTER 9

THE HISTORY OF SCIENCE AND

THE SCIENCE OF HISTORY:
CRITIQUE OF APRIORISM

In emphasizing the role of reasoning in scientific rationality in the last

chapter, our discussion was carried out in a context where the reasoning was
taking the form of reason as opposed to unreason and was being distinguished
(but not opposed to or separated from) method and rhetoric, as other ele¬
ments of scientific rationality. The first danger that such emphasis carries is
that of being misinterpreted as advocacy of a priori rationalism, so the most
urgent thing to do now is to distinguish reasoning from rationalism. In
accordance with our concrete approach, this distinction will be elaborated
not by an abstract general definition of these two concepts, but by reference
to views which advocate apriorist rationalism while confusing the two
concepts, and to evidence materials which illustrate reasoning but not
apriorism. The materials are certain arguments in Galileo’s Dialogue, the views
are those of Alexandre Koyr6.
Few scholars in this century have had as great an impact on our under¬
standing of science as Alexandre Koyre. It is perhaps no accident that his
pioneering work on the subject is one dealing with Galileo. Entitled Etudes
galileennes the book has become both a model and a source of inspiration for
historians and philosophers alike. The influence on contemporary historians
of science is both explicitly acknowledged1 and bordering on the classical,
in the sense that his ideas and his approaches can now be freely borrowed and
adapted and have become internalized in their professional sensibilities.
As regards philosophical influence, the view propounded in Thomas S.
Kuhn’s Structure of Scientific Revolutions is the one which in the last decade
or two has caught the imagination, if not the intellect, of most scholars,
scientists, and laymen alike. It would perhaps be an exaggeration to repeat
the judgment I once heard from a philosopher, namely that, in Kuhn’s book,
what is new is not true, and what is true is not new, vis-a-vis Koyr^’s.
Though exaggerated the judgment is not unfounded, and it does reflect the
extent of Koyr^’s influence. At any rate one other philosopher, Joseph Agassi
in Towards an Historiography of Science, has focused on one of Koyre’s
central techniques — error analysis, has articulated it, synthesized it with Karl
Popper’s philosophy of science, provided it with novel content and illustra¬
tions, and advocated it as the most fruitful approach to the philosophy of

202
 THE HISTORY OF SCIENCE 203

science. Some of Feyerabend’s own critiques, examined in the last chapter,

remind one of Koyr^’s technique of error analysis, though of course, Feyer¬
abend’s central concern is to explore the limitations of reason by studying the
limitations of method and the power of rhetoric, whereas Koyrd’s concern
was to extol the power of reason by exploring the rationalistic character of
science, rationalist in the sense of a priori rationalism. This is all the more
curious since they both use similar approaches and evidence. To what extent
Koyr6’s rationalist apriorism is more adequate than Feyerabend’s pseudo-
irrationalistic anarchism will be examined presently.
I might pay my own homage to Koyr6 by interpreting his introduction of
methods of intellectual history into the historiography of science as a move
toward the technique of logical analysis, which is one whose powers I am
interested in exploring. However, to do this would be insincere lip service on
my part, for though it is true that Koyrd’s work is uniquely valuable as an
introduction to the use of logical analysis in the interpretation of science, yet
it needs such serious corrections, as it is shown below, that in their absence
I feel it is bound to lead to abuse and to aprioristic and rationalistic excesses.
At any rate, now that Koyre “is the master of us all”, to use words of
historian Charles C. Gillispie,2 the methodologically aware scholar, and most
of all one interested in Galileo, ought to ask himself whether uncritical
acceptance is any more justified here and now than it was at the time of
Galileo vis-a-vis that supreme “master of those who know”, Aristotle. To
combat one authority with others, one might say that, though Koyr6 may
be acknowledged as a founder of professional history of science in its present
form, yet a discipline which hesitates to forget its founders is lost, a dictum
coined by Alfred North Whitehead and popularized by sociologist Robert K.
Merton.3
To be sure there have been critics of Koyrd’s work. For example, Koyr^’s
erudition has been questioned by Eugenio Garin in 1957.4 It concerns the
character of Koyr^’s discussion of Galileo’s alleged Platonism. Garin argues
that Koyr^’s characterization of Platonism is abstract and unhistorical, and
hence when the problem is discussed in such terms it is insoluble. He exhibits
the inadequacies of Koyrd’s discussion of the actual historical situation of
Platonism (that is, of Platonism as a historical entity) by pointing out a
number of errors and confusions about relevant texts and persons. Garin
argues that Koyrd’s distinction between Platonism as mathematicism and
Platonism as mystical, magical numerology is a figment of Koyr^’s imagina¬
tion in the sense that it has no historical basis in texts such as Ficino’s works
or even in Clavius’s commentary to Euclid.5 Another example given by Garin
204 CHAPTER 9

is Koyr^’s error in identifying the ‘Lud. Buccaf.’ mentioned in Bonamici’s

De Motu as a certain Lodovico Buccafiga6 instead of Ludovico Boccadiferro;
the latter is a nonnegligible figure, was a professor in various Italian univer¬
sities including Bologna, and wrote several commentaries to Aristotle full of
Platonic as well as Aristotelian doctrines;7 moreover, he is mentioned in
Galileo’s own Iuvenilia (under the Latin name Buccaferrus).8 Another
example given by Garin is a reference to Crescas in Koyr^’s discussion of
Descartes’ notion (in Le Monde) of quantity of rest,9 when, as Garin states,
an educated person at the beginning of the seventeenth century could easily
read Crescas’s theses in the widely known Examen vanitatis doctrinae
gentium by Francesco Pico.10
If Koyre’s erudition can be so questioned, at the other end of the scholarly
spectrum, the logic of his reasoning has been faulted in a significant way. It
has been argued in my History of Science as Explanation11 that an analysis
of Koyr^’s central volume of Etudes galileennes shows that the conclusions he
himself draws are not supported by the evidence he himself gives.
Somewhere in between the extremes of these critiques lie two other
equally serious problems. One is Koyr6’s persistent confusion of the context
of scientific discovery and the context of scientific justification. There is
nothing wrong with focusing on one or the other of these contexts, as long as
one’s conclusions are supported by evidence from the appropriate context.
The error is to reach conclusions about one context on the basis of evidence
from the other context. Koyr^’s study of Galileo’s law of falling bodies
examines evidence from the context of justification since he examines the
various attempts found in Galileo’s (published) writings to give a theoretical
demonstration of it.12 On the other hand, Koyrd’s conclusions13 are about
the method which enabled Galileo to succeed, where success can only be
interpreted to pertain to his discovery of the law, since Koyr6 is at pains to
point out the inadequacies of Galileo’s various proofs of the law (and hence
presumably there was no success in the context of justification), and since
Koyr6 is contrasting that success to Descartes’s failure, which was clearly a
failure to arrive at the correct law of fall.14
It is interesting to point out that this type of criticism of Koyr6 has
emanated from two very different quarters. On the one hand, it has been
made in the terms just used in my work on historical method whose explicit
and primary aim is to work out a philosophy of the historiography of
science.15 On the other hand, a similar criticism is made in the domain of
pure Galileo scholarship, in Drake’s attempt to emphasize ‘Galileo Studies’ as
contrasted to the Koyr6-type of ‘Galilean Studies’.16 Drake’s distinction
 THE HISTORY OF SCIENCE 205

between the biographical context and the history-of-ideas context is similar

to that between discovery and justification, and his judicious work in the
former areas has led to some epoch-making results about Galileo’s discovery
of the law of fall.17
To this problem in Koyr^’s work which one may categorize as method¬
ological or historiographical, we may add one that pertains to scholarship as
such. Perhaps the single most striking feature that is apparent to the reader of
Koyr^’s works, or at least of the book for which he is most famous, is the use
of very long and very frequent quotations, interspersed with commentary.
I am somewhat embarassed to report my having discovered that in the trans¬
lation, citation, and arrangement of such quotations Koyr£ takes a number of
impermissible liberties. It is almost as if these texts were being quoted so as to
enable him to insinuate by distortion what could not be suggested by explicit
argument.
For example, on p. 278 of Etudes galileennes Koyr6 has a quotation for
which he gives the reference ‘Dialogo, II, p. 423’. The quoted passage occurs
neither on p. 423 (of the National Edition), nor in Day II, which at any rate
does not include that page; since it is Day III that includes that page, one may
look for the passage in Day III, but in vain. In fact, the passage is not from
the Dialogue at all, but from Jacopo Mazzoni, as the interested scholar can
discover by studying the rest of Koyr6’s footnote, which reads in part ‘cf.
Jacobi Mazzonii, . .., In Universam Platonis et Aristotelis Philosophiam
p. 187 sq. . . . ’ Thus one concludes that Koyr^’s footnote should have read
‘Jacobi Mazzonii, In Universam . . ., p. 187 sq.; cf. Dialogo, III, p. 423’
rather than the other way around ‘Dialogo, p. 423; cf. Jacobi Mazzonii
As it is, most readers will get the impression that Koyr6 is quoting from the
Dialogue, when in fact he is not.18
However, the influence of Koyr^’s work persists in spite of such demon¬
strated inadequacies in erudition, logic (reasoning), methodology (historio¬
graphy), and scholarship. So perhaps his appeal does not derive from the attrac¬
tion of his approach but from the attraction of his substantive thesis, namely
his apriorist-rationalist interpretation of science. For this reason, as well as
because Garin’s above criticism refers primarily to the first volume of Koyr6 s
book, and because Drake’s critique and mine above refer primarily to the second
volume, it will be good to examine in detail the third which contains a state¬
ment of that thesis classic in its clarity and simplicity. It will emerge that
Koyr6 confuses the activity of reasoning with the attitude of apriorist rational¬
ism, and that his evidence supports an interpretation of science from the point
of view of the former rather than one from the point of view of the latter.
206 CHAPTER 9

The third volume deals with the law of inertia and contains a chapter
entitled ‘The Dialogue on the Two Chief World Systems and the Anti-
Aristotelian Polemic’ (K205—38).19 It begins with an introductory discussion
(K205-11) which emphasizes the connection between physics and cosmology
in Galileo’s work and in the rise of modern science in general; then it contains
a brief discussion (K212—15) of the many aspects of the Dialogue: polemical,
pedagogical, philosophical, and autobiographical. This is followed by an
analysis of what Koyrd takes to be the central core of the book, namely the
physical arguments against the earth’s motion. His analysis consists of three
elements: a quotation of the main arguments (K215—19), an assessment of
Galileo’s criticism of these arguments (K219—20), and a justification of this
assessment (K220—38).
The passage quoted by Koyre is that in which, first, Simplicio quotes from
De Caelo Aristotle’s original four arguments, and, then, Salviati gives a
statement of several contemporary arguments inspired by the Aristotelian
ones (F150—53).20 The careful reader of Galileo’s Dialogue will be struck by
the last three lines of Koyrd’s quotation (K219) which read: “Moreover, the
same thing would happen in all cases where one would shoot a cannon: the
ball would pass above or below the mark according as one would shoot
toward the east or toward the west . . .These lines constitute an excessively
free translation21 for the initial segment of the last portion of Salviati’s
speech on pp. 151—53 of the Dialogue. This last portion reads:

And not only the shots along the meridians, but also those toward the east or toward
the west would not result right, the eastward ones resulting high, and the westward ones
low whenever the shooting were point-blank; for, since the path of the ball in both shots
would be along the tangent, namely along a line parallel to the horizon, and since if the
diurnal motion belongs to the earth, the horizon would be constantly falling in the east
and rising in the west (that’s why eastern stars appear to be rising and western ones to be
falling), therefore the eastern target would be falling under the shot, so that the shot
would result high, and the rising of the western target would render the westward shot
low. In this way one could not shoot right in any direction; and since our experience is
different, we are forced to say that the earth is motionless (F153, my literal translation).

The liberties that Koyrd takes are inadmissible because he thereby fails to see
or to inform his readers that Galileo is here reporting a distinct problem with
the earth’s rotation, deriving specifically and exclusively from the point-blank
nature of these gunshots. He is not merely repeating the east-west gunshot
objection, stated earlier in Salviati’s speech; in fact, the two problems require
different solutions and these solutions are given by Galileo in two different
passages (FI93-97, and F205-09 respectively). Nor is Galileo merely sum-
 THE HISTORY OF SCIENCE 207

marizing the arguments he has just stated, which Koyr£ suggests by em¬
phasizing the word ‘all’. To excise this passage the way Koyre does, and to
translate its beginning as he does, may be in accordance with his perception
of the repetitive nature of the Dialogue; he argues that this repetitiveness has
the important pedagogical and rhetorical function of familiarizing and
accustoming its readers to the new concept of motion (K234, 237). However,
if the alleged repetitiveness of the Dialogue is grounded on evidence like the
present one, then it is an oversimplification at best, or perhaps an invention.
Another reason why it is improper for Koyre to quote the passage the way
he does is that the point-blank objection and reply contain discussions of
rectilinear motion along the tangent to the point of firing. The context of
Koyr^’s discussion of the Dialogue is that of “Galileo and the Law of Inertia”,
namely, to what extent Galileo’s conservation of motion involves rectilinear
motion. Therefore, the point-blank objection would have been relevant,
regardless of whether or not it would disconfirm Koyr^’s interpretation.
Let us now examine his assessment of Galileo’s critiques. Here it is im¬
possible to do Koyr6 any greater injustice than the one his own words do
him. So let us quote this brief passage in full:

Let’s now pass to the criticism. It is at once very profound and very simple. Galileo tells
us that the arguments of the Aristotelians are nothing but paralogisms. They presuppose
what must be shown. And, no doubt, it is true. But an Aristotelian could very well not
accept the criticism, [which is] a consequence of the objection that Copernicus had
already addressed to him: Aristotle does not reason, as he pretends, by starting from the
facts, but on the contrary by starting from a theory. To this the Aristotelian could
respond with good reason:
(a) that it is impossible to reason otherwise;
(b) that Galileo does the same.
In fact, the Aristotelian reasoning presupposes a theory, or if you prefer, a particular
concept of motion, namely that of a process which affects the moving thing. It also
presupposes that sense perception permits us to apprehend directly physical reality, that
it is even the only means of apprehending it, and that, consequently, a physical theory
can never cast doubt upon the immediate data of perception.
Now, Galileo expressly denies this. He starts from directly opposite assumptions:
(a) that physical reality is not given to the senses, but on the contrary apprehended
by reason;
(b) that motion does not affect the moving thing, which remains indifferent to all
motion that animates it, and that motion affects only the relations between a moving
thing and one which does not move.
A paralogism from Galileo’s point of view, the Aristotelian reasoning is in itself
unobjectionable.
Nevertheless, dialectically speaking, Galileo no doubt has the right, at least within the
Dialogue, to designate the Aristotelian reasoning as a paralogism. For, before having
208 CHAPTER 9

stated the physical and mechanical proofs of the earth’s immobility, Galileo has already
laid down the double principle of the optical as well as mechanical relativity of motion
(K219-20).

This is at best an oversimplification, and probably a disservice to the very

rationalism that Koyre’s interpretations are meant to support.
First, it is simply not true that “Galileo tells us that the arguments of the
Aristotelians are nothing but paralogisms”. In Koyrd’s own quotation, we have
Aristotle’s four original arguments (from violent motion, from double circular
motion, from natural motion, and from vertical fall) plus five modem argu¬
ments (from the ship experiment, from vertical gunshots, from east-west gun¬
shots, from north-south gunshots, and from point-blank gunshots). Of these
nine arguments, Galileo claims that only three are paralogisms, namely the
argument from violent motion (FI59—62), the argument from double circular
motion (F162—4) and the argument from vertical fall (FI64—7). The prob¬
lems with the other arguments, according to Galileo, are as follows. The
argument from natural motion is the one to which most of the First Day is
devoted (F164); it presupposes an untenable concept of natural motion,
namely that straight and circular motion are two distinct instances of simple
motion, whereas Galileo argues that they are two different stages of natural
motion: straight motion can be acquired naturally but cannot naturally con¬
tinue forever, whereas circular motion can naturally continue forever but can¬
not be acquired naturally without straight motion (F38—62; see Chapter 2).
The ship experiment argument is simply based on a false premise, namely that
on a moving ship a rock dropped from the top of the mast lands away from
its foot (FI69—75). The east-west gunshot argument involves a failure of
hypothetical reasoning, namely a failure to take seriously the motion of the
earth, even when examining the consequences of such motion (FI93—7). The
vertical gunshots argument, though in one place it is incidentally described as
a paralogism just like the vertical fall argument (F200), in effect is shown to
suffer from a failure to take into account both the relativity and the composi¬
tion of motion, for this is what is actually discussed in the criticism (F197—
203). The north-south gunshot argument is criticized as failing to take into
account the conservation of motion (F203—5). Finally, the point-blank gun¬
shot argument, which Koyr£ confuses with the one from east-west gunshots,
is criticized as based on a phenomenon of such a small magnitude that it
could not be detected even if it existed (F205—9), namely that on a moving
earth the deviation from the horizontal would be of the order of a fraction of
an inch.
 THE HISTORY OF SCIENCE 209

Second, of the arguments that Galileo claims are paralogisms, it is not true
that the paralogism is always that “they presupposed what must be shown”.
Only the vertical fall argument begs the question, the other two are fallacies
of equivocation. The argument from violent motion misuses the ambiguity of
the proposition “the parts of the earth would also move circularly”, which
can mean either that these parts would move around their own centers or that
they would move around the earth’s center (F159—62); Aristotle’s second
argument allegedly commits the fallacy of equivocation because its conclusion
could mean either that the earth lacks the diurnal motion or that it lacks the
annual motion (FI62—4).22
Koyr^’s next error is to confuse begging the question (reasoning that pre¬
supposes what must be shown) with apriorism (reasoning based on theories
rather than facts). For the reason that he gives why it must be admitted that
the Aristotelian reasoning begs the question is that it starts from a theory
rather than from facts. Koyre’s confusion is a serious error. An argument
presupposes what must be shown when one of its premises depends on its
conclusion. Such reasoning is fallacious because an argument is an attempt to
show that since you accept the premises you must accept the conclusion; if
one of the premises depends on the conclusion, in the sense that it implicitly
assumes the conclusion, then the argument is assuming what it is trying to
prove, which is surely wrong. Nor would Simplicio or Aristotle deny this.
Nowhere in the Dialogue does Simplicio take lightly the accusation of begging
the question, which after all belongs to Aristotle’s own list of fallacies.23
On the other hand, to reason from a theory may or may not be correct,
depending on the adequacy of the theory. But even if the presupposed theory
is false or inadequate there is no logical error, no paralogism, unless the
theory from which you are reasoning is the same as the theory (or alleged
facts) to which you are reasoning; in this case you’are involved in circular
reasoning and are begging the question. In other words, not every instance of
reasoning from a theory is an instance of presupposing what is being shown.
Nor is every instance of begging the question an instance of apriori reasoning:
it may be that what you are trying to prove is a fact; and it may be that you
are proving it on the basis of other facts; then there would be no apriorism,
but if one of the facts on the basis of which you are proving the conclusion,
depends on the conclusion, then you are still begging the question. Circular
reasoning does not become legitimate merely because it goes on within a
domain of facts. In conclusion, then, apriorism is neither a sufficient nor a
necessary condition for circular reasoning.
Let us illustrate such abstractions with discussions from the Dialogue. The
210 CHAPTER 9

Aristotelian argument claimed by Galileo to be a paralogism in the sense of

begging the question is the following (FI64—6):

The earth cannot rotate because if it did bodies could not fall vertically; but they do
since they can be seen to graze the edge of a tower when dropped from it. (My
paraphrase.)

The final portion of this argument grounds the impossibility of the earth’s
rotation on the vertical fall of bodies, so that one premise of that final step
is the proposition that “bodies fall vertically”. Another portion of the argu¬
ment bases this vertical fall of bodies on the apparent grazing; hence it is
being assumed that this apparent grazing implies vertical fall. Now, Galileo
argues, this implication can be questioned because it would not hold on a
rotating earth (since if the earth rotated and bodies were seen to fall grazing
the tower, then their actual path would not be vertical but slanted). Since the
implication can be questioned, it is fair to ask for a justification of it: how do
you know that the apparent grazing implies vertical fall? One abstractly
possible justification would be the following: if the earth stands still then the
implication holds; the earth does stand still; therefore the implication holds.
There being no other means whereby an Aristotelian could justify the implic¬
ation, he would have to use this abstractly possible argument. In so doing,
though it is indeed true that the implication holds if the earth stands still,
he is also assuming that the earth stands still, which is the final conclusion
he wants to reach. Thus the Aristotelian argument from vertical fall begs the
question because its premise that bodies fall vertically presupposes (i.e.,
would have to be justified by the proposition) that the earth stands still.
Feeling the force of Salviati’s criticism, Simplicio with Sagredo’s help
states a new argument which can serve here as a good example of reasoning
starting from a theory (F166—7):

If the earth rotated then bodies could not be seen to fall grazing the edge of a tower,
since on a rotating earth this apparent grazing would imply that bodies would have two
natural motions, toward and around the center, which is impossible. But bodies are seen
to fall grazing the edge of a tower. Therefore, the earth can’t rotate. (My paraphrase.)

The final portion of this argument grounds the impossibility of the earth’s
rotation directly on the apparent grazing, which is self-evident. But it also
grounds the impossibility of rotation on the claim that if the earth rotated the
apparent grazing would not occur. In another portion of the argument, this
claim is grounded on two premises, a self-evident one and a theory. It is self-
evident that on a rotating earth apparent grazing would entail simultaneous
 THE HISTORY OF SCIENCE 211

motions toward and around the center of the earth. But it is merely a
“theory” to say that it is impossible for material bodies to have two
spontaneous motions toward and around the center. This proposition may be
merely a theory, but it is not the same as the conclusion of the argument.
Hence this argument does not beg the question, though it is an instance of
what Koyre would call a priori reasoning, reasoning from a theory.
Now, since Koyre is conflating the paralogism of presupposing what is
being shown with the problem of a priori reasoning or reasoning from a
theory, and since he claims that all the Aristotelian arguments are paralogisms,
it follows that what he means is probably that these arguments are all reason¬
ing from a theory. Let us ask whether he is right in so claiming. Unfortunately
not, if for no other reason than because the ship experiment argument, which
is quoted by him, is clearly not a piece of a priori reasoning. Koyre thinks
that it is probably because he thinks that Galileo’s criticism of it is a priori;
however, even if this were so, it would not make the original ship experiment
argument an a priori one. The simple truth is that someone may, on the basis
of a theory, criticize an argument which is not itself based on a theory.
Another reason Koyre might give for regarding the ship argument as a priori
is to say that the experiment had never been made (K225). However, this may
be a piece of a priori reasoning by Koyr6 himself; in fact Chiaramonti claims
in his 1633 book answering Galileo’s Dialogue that the experiment had been
made by a certain Giovanni Cotunio of the University of Padua.24 A final
reason Koyr6 might give is his belief that “it is impossible to reason other¬
wise” than from a theory; however, he gives no justification for this allegation,
unless it be the claim that the Aristotelians as well as Galileo did in fact
reason from a theory; hence such an argument would beg the question.
Let us continue to place qualifications on Koyr£’s claims to see if we can
find some truth in them. Though not all the Aristotelian arguments involve
reasoning from a theory, some of them do. Concerning these, can we agree
with Koyre that they are unobjectionable because it is impossible to reason
otherwise and because Galileo does the same? It is true that if it were im¬
possible to reason otherwise than from a theory, then an argument could not
be faulted for doing so. However, it is obvious that reasoning from a theory is
merely one type, and Koyr6’s claim that it is impossible to reason otherwise
is an extravagant exaggeration at best. In fact, he gives no justification of this
claim. Or perhaps his analysis of the Aristotelian and of Galileo’s reasoning
could be interpreted as a supporting argument, namely, that it is impossible
to reason otherwise than from a theory because both the Aristotelians and
Galileo did so. It would not be a very serious criticism of this argument to
212 CHAPTER 9

object that a generalization is being reached from two cases; its real problem
is that these cases are not typical. In fact, the relevant Aristotelian reasoning
consists of those arguments against the earth’s motion which happen to in¬
volve reasoning from a theory; hence, these arguments would be prejudicially
chosen to ground Koyr^’s generalization. On the other hand, the Galilean
reasoning under consideration is that whereby he criticizes these and other
Aristotelian arguments. Now, arguments criticizing a priori arguments are not
typical since, whatever a priori element they might contain, it could be the
result of the a priori arguments being criticized. Regarding Galileo’s argu¬
ments that criticize Aristotelian reasoning that does not start from a theory,
even if they did presuppose a priori elements, as Koyre tries to show, this
very feature would make them atypical, since they would then be a priori
criticism of empirical arguments. Since the question here is whether Koyr6’s
alleged impossibility of reasoning other than from a theory has been reached
by generalizing typical cases, we may dismiss Koyrd’s generalization.
Let us now examine Koyrd’s argument that the Aristotelian arguments
which start from a theory are unobjectionable because Galileo is doing the
same. Here it is important to see that, even if Galileo were reasoning from a
theory, even if Koyr^’s account had shown this, his conclusion still would not
follow: the Aristotelians could not justify their position with such a tu
quoque. There are two reasons for this. First, as Koyrd himself points out, an
important element of the theory from which the Aristotelians argued was
their emphasis on empiricism: that sense experience is the only means of
apprehending reality; hence the Aristotelians could not consistently have
admitted that their arguments against the motion of the earth were really
reasoning from a theory. Second, much of Galileo’s criticism consists of an
analysis of the a priori elements of the Aristotelian reasoning, together with
the argument that if one replaces these a priori elements with other, more
plausible theories one cannot then draw the conclusion that the earth must
stand still. Of course Koyr6 denies that these other theories, from which
Galileo reasons, were more plausible than those from which the Aristotelians
started; but even if Koyrd is right about the equal plausibility of the respective
presupposed theories, the situation is not otherwise symmetric. For, whereas
the Aristotelians were trying to prove the impossibility of the earth’s motion,
Galileo in his criticism of these arguments is merely trying to prove that the
Aristotelian reasoning is incorrect, not that the earth moves. In his criticism,
Galileo never argues that, given his new concept of motion, it follows that the
earth moves, but rather than given the new concept it follows the motionless¬
ness of the earth is not proved by the Aristotelian arguments. In other words,
 THE HISTORY OF SCIENCE 213

the crucial difference is that Galileo’s conclusion is that the Aristotelian

reasoning is incorrect, whereas the Aristotelian conclusion is that the earth
stands still; the Aristotelian point is that the earth stands still, Galileo’s is that
the Aristotelians haven’t proved their point. That is why it is wrong for the
Aristotelians to argue from a theory, but not for Galileo. Galileo argues from
a theory in the context of showing that from a different theory one could not
reach the conclusion the Aristotelians reach in their argument; since their
argument presupposes a theory, they haven’t proved their conclusion. For
example, consider the above mentioned argument given as a good example of
Aristotelian reasoning from a theory, the argument that based the impossibility
of the earth’s motion on the falling body’s apparent grazing of the edge of a
building, and ultimately, on denying the possibility of natural motion both
toward and around the center. Given a theory that allows this possibility, one
could not conclude the motionlessness of the earth from the apparent
grazing. Galileo reminds us very frequently that his criticism is merely dis¬
proving the Aristotelian arguments, not proving the earth’s motion, so much
so that it would be tedious to collect the references.
Though Galileo’s alleged reasoning based on a theory cannot justify the
Aristotelian procedure, Koyrd’s claim is intrinsically interesting, and it is
worth examining. In the passage quoted above, Koyrd claims that Galileo s
reasoning presupposes the principle that physical reality is apprehended by
reason, not by the senses, and the principles of relativity and of conservation
of motion. In the rest of his account (K220-38) Koyr6 then supports these
claims by analyzing the passages where Galileo answers the Aristotelian
arguments; special attention is given to the answer to the ship experiment
argument which is the only passage used by Koyrd as showing Galileo s pre¬
supposition of aprioristic rationalism. It should be noted first that, even if
Koyre is otherwise correct in these interpretations, he would not be justified
in claiming that the way in which Galileo presupposes the principle of
apriorism is the same as that in which he presupposes the principles of
relativity and of conservation, namely that all these principles are theories
from which Galileo starts in his reasoning, so that his reasoning is no better
grounded than that of the Aristotelians, who start from different theories.
The fact that the principle of apriorism is an epistemological one concerning
the nature of knowledge, and that the other principles are physical ones
concerning the nature of motion, is perhaps the least of their differences. The
more significant difference is that the physical principles are or could be used
as premises in certain parts of Galileo’s counterarguments, whereas the
epistemological principle is not so used. For example, in the answer to the
214 CHAPTER 9

ship experiment it is clear that Galileo is or could be arguing as follows: given

the principle of conservation, the rock will retain its horizontal motion even
after it is dropped; and given the principle of relativity, the simultaneous
downward motion will not constitute an interfering disturbance to this
horizontal motion, therefore, on the moving ship the rock will end up at the
foot of the mast. On the other hand, how would Galileo be committed to
apriorism, assuming that he were? If Galileo can argue, by reasoning of the
type just sketched, that the rock will land at the foot of the mast even on a
moving ship, then it follows that one has to accept such a result on a moving
ship unless one can show that there is something wrong with Galileo’s argu¬
ment. Now, to find something wrong with this argument one has to fault
either some steps in reasoning or some premises being assumed in it; since
presumably there is nothing wrong with the steps in reasoning, the only thing
one could question is the assumed premises; but, if Koyre is right, these
assumed premises are the principles of relativity and of conservation which
cannot be faulted empirically but only by opposing to them another concept
of motion such as the Aristotelian one; it follows that the only objection one
could have to Galileo’s reasoning is to produce the Aristotelian argument
concluding that the rock must fall behind on a moving ship. Now, this may be
in accordance with Koyre’s desire of showing that both Galileo and the
Aristotelians are reasoning from a theory, but it is not in accordance with his
claim that Galileo is assuming apriorism as a principle from which he reasons
to arrive at his physical conclusion, for it also follows from the above con¬
siderations that the apriorist principle is implied by Galileo’s allegedly a priori
argument. In other words, the apriorist principle is a consequence rather than
an assumed premise of Galileo’s ship experiment counterargument; the
principle follows from the argument, rather than the other way around;
Galileo does not need the epistemological principle to justify his physical
argument, but rather his physical argument (the fact that such an argument
can be given) justifies the epistemological principle. In short, if Galileo can
answer the ship experiment argument in the a priori way that Koyrb thinks,
then his answer is supporting, not assuming, apriorism. Moreover, and con¬
clusively for the present issue, if it is a fact that the Galilean and Aristotelian
arguments are a priori, then apriorism is itself a fact and not a theory, for it
would then be the consequence of a fact.
To explain further this difference between an assumption and an implica¬
tion, and thus to reinforce this criticism of Koyr6, I will call attention to an
Aristotelian argument whose answer does involve Galileo in an epistemological
assumption about the relationship between sense experience and reason
 THE HISTORY OF SCIENCE 215

similar (though not identical) with the apriorist rationalism here attributed to
him by Koyr6. The argument is that from the deception of the senses; it is
explicitly discussed by Galileo in the Second Day, at the beginning of his
critique of Chiaramonti’s book on the three new stars (F272—81); it is also
implicitly discussed at the beginning of the First Day (F56—62) and in the
discussion of the gunshot arguments (FI97—200). However, it is not one of
the arguments quoted or mentioned by Koyr6, though it would have provided
him with a proper illustration of how an argument against the earth’s motion
and Galileo’s answer to it assume an epistemological principle. The argument
is that if the earth rotated then our senses would be deceived insofar as (a)
we do not feel any motion, and (b) falling bodies would be appearing to move
vertically but would in reality move in a slanted path. Obviously, in order to
be able to conclude from this (by modus tollens) that the earth does not
rotate, one needs another premise which would be a denial of the ‘then’
clause of the conditional premise; we could say that our senses cannot be
deceived, or to use Koyre’s words, that “sense perception permits us to
apprehend physical reality directly” (K220). Galileo answers (F57—9, F166,
F272—80) that our senses tell us merely that bodies appear to fall vertically
(vision) and that the earth feels to be at rest (internal sense); to think from
this that bodies really fall vertically and that the earth is really at rest is to
draw a conclusion that involves reasoning; this reasoning happens to be
incorrect since we cannot say without qualification that apparent vertical fall
implies actual vertical fall (it does so only on a motionless earth), and since it
is not true that we can feel all motion (we can only feel changes of motion);
therefore, if the earth rotated it would not be our senses that would be
deceived, but rather our reason, or at least the reason of those who would
draw the unwarranted conclusion from the sense data; therefore the argument
from the deception of the senses is incorrect. But in any case, it is by reason¬
ing, not by the mere senses, that we can apprehend reality, or in Koyr^’s
words, “physical reality is not given to the senses, but on the contrary appre¬
hended by reason” (K220). From this ‘theory’ Galileo is refuting the
Aristotelian argument, which was based on the opposite theory. However,
I believe that, besides providing a good illustration for Koyr6’s point, this
argument provides a good illustration of something he does not want to
accept, namely that the two ‘theories’ are not on a par; at the very least
Galileo’s is the more plausible one; actually, in the sense in which they
relate to the present argument, Galileo’s ‘theory’ is true, whereas the Aristote¬
lian one is false. Hence, though Galileo’s critique of this argument illustrates
Koyr^’s point, it does not support it; though it confirms the letter of his
216 CHAPTER 9

thesis, it does not confirm its spirit, since he attaches an apriorist meaning to
Galileo’s epistemological assumption, whereas in the present context it has a
critical meaning. Galileo is being both a critical empiricist and a reasoning-
oriented rationalist.
The same type of non-apriorist rationalism is presupposed by Galileo,
though by way of implication rather than assumption, in the answer to the
ship experiment argument examined by Koyr6. Galileo is not a rationalist in
the sense of an apriorist, but in the sense that he likes to use reasoning and
arguments as much as possible, we might say a rationalist in the sense of a
logician, a logician-in-action. The difference is that the logician will not limit
himself to a priori reasoning, reasoning from a theory; some reasoning is
reasoning based on facts, where by starting from facts one attempts to justify
a conclusion; if the conclusion is a theory then we would have reasoning to a
theory; or the conclusion may be itself factual, if only facts are used as
premises and if all the inferences are strong. To be more specific, the thesis
implied by the present passage (F169—80) could be formulated as follows:
experiments are sometimes unnecessary to ascertain the results of a test, for
sometimes it can be argued on the basis of known or more easily ascertainable
facts, what these results must be. In fact, the passage can be reconstructed so
as to become an illustration of this thesis, as was done in Chapter 5.
In his analysis Koyr6 neglects to take into account the very important fact
that, though Galileo is justifying the conclusion of his own counterargument
by reference to the principle of conservation, this is not being merely
assumed but empirical evidence is given in its support. The principle of
conservation corresponds to proposition (1) in the above reconstruction,
which is justified by means of (a) and (b). (See Chapter 5.) Of course Koyr6
may say that the reasoning from (a) and (b) to (1) is not completely valid
from a logical point of view, or perhaps that other principles are being
assumed in this subargument. However, to the former it could be replied that
it is one thing for a proposition to be incompletely supported and it is
another for it to be completely unsupported; only the latter would be a
‘theory’ in Koyr6’s sense. To the second alternative one could reply that the
other assumed principles would probably themselves be supportable when
questioned, and though other principles might be assumed in these other
arguments, perhaps they too could be supported if required, or if not perhaps
one would be dealing with highly abstract, general, and universal metaphysical
principles, whose assumption would present no practical or physical problem.
Thus the impact of Koyr^’s point would end up being something to the effect
that whenever one is reasoning one is justifying certain propositions on the
 THE HISTORY OF SCIENCE 217

basis of others which for the time being are not questioned; that is, in reason¬
ing one can’t prove everything simultaneously. But these statements about
reasoning, far from presenting problems, constitute very elementary facts
about the nature of reasoning; they amount to saying that reasoning is a
step-by-step process. The only real question one could raise is whether such
an activity is effective. To this the facts of life and of history provide the
obvious answer.
Concerning the principle of relativity, the situation is more complicated.
Koyr6 would argue as follows: though this principle is not directly present
in the argument, the argument is assuming that “there is no way in which the
horizontal impressed virtue could be disturbed by the vertically downward
tendency due to weight”, which is a form of the principle of composition;
now, the composition of motion is a consequence of its relativity (K222—3).
So presumably Galileo’s answer is based on the principle of relativity in the
sense that it would have to be used to justify one of the premises in his argu¬
ment. However, Koyr^’s interpretation is not a faithful reconstruction of the
argument actually given by Galileo. It is indeed true that if one assumes the
principle of relativity of motion, and if one interprets it in such a way as to
imply the composition of motion, then one could arrive at Galileo’s result for
the moving ship (having also used the principle of conservation). However,
in the passage when Galileo gives his answer (F169—80), the relevant step
(FI75) is much more concrete: the vertical fall does not represent a disturb¬
ance to the horizontal motion because the body is indifferent to horizontal
motion, and because the cause of the horizontal motion (the impressed
‘virtue’) is distinct from the cause of the vertical fall (gravity). Of course,
Koyr6 could question the soundness of these inferences, though they are not
questioned by Simplicio, who merely questions the truth of the two premises,
that the body is indifferent to horizontal motion, and that the cause of this
motion is an impressed ‘virtue’. But even if the soundness of the inferences is
questionable, Koyre’s substitute argument is not unquestionable either, for
though it may contain no problem of the soundness of the inferences, it
contains the problem that the crucial premises are just being assumed. I am
inclined to believe that the inferences in Galileo’s actual answer are context¬
ually sound. It follows that this answer is not assuming, but proving
(supporting) the principle of relativity. This can be seen even more clearly
from Galileo’s answers to the vertical fall argument (FI64—6) and to the east-
west gunshot argument (FI93—7), as we will see below.
Koyr^’s evidence that the principle of relativity is being presupposed in
the sense of assumed consists of (1) the fact that Galileo obviously holds the
218 CHAPTER 9

principle, (2) the fact that by postulating the principle one could answer the
objections to the earth’s motion being considered by Koyrd, and (3) the
alleged fact that Galileo gives several statements of the principle before he
gives his answers to the objections (K220, n.3; K221, n.l; K222, n.l; K237).
(1) is indeed true, but by itself it does not support Koyre’s conclusion, since
it might be, as I think is the case, that Galileo holds the principle because it is
supported by his critiques, rather than because he needs to assume it in order
to make those critiques. (2) is irrelevant, even if true, for the fact that the
logic of the situation could be rearranged differently, does not change the
facts of the actual logic of the situation, as we have just seen for the case of
the ship experiment.
As for Galileo’s alleged previous references, they are misinterpreted by
Koyr£ (K220, n.3: F57, F101, F139, F141; K221, n.l: F139 ff.;K237). The
first reference alleged by Koyre occurs at the beginning of the First Day
where Simplicio objects to Salviati’s principle of circular motion, by saying
that it conflicts with the clear evidence of the senses that bodies are seen to
fall in a straight line (F57). Salviati answers, among other things, that it is
questionable, as he promises to show later, whether falling bodies really
follow a straight path. However, there is no statement of the principle of
relativity; there is only a reference to a future discussion, hence even if that
discussion were the proper one, the present passage could not be regarded as
evidence of any antecedent commitment to relativity. At any rate the subse¬
quent passages being referred to can only be those containing the critique of
the argument from vertical fall (F164-66) and the one containing the sugges¬
tion that the real path of a falling body is circular (FI88—93); and in these
passages the principle of relativity is not in sight.
Koyre’s second reference is to p. 101 where I find the general topic to be
the roughness of the lunar surface. More specifically, on that page we find
the ending of the discussion of the experiment comparing reflections onto a
wall from a flat and from a spherical mirror, and the beginning of the discus¬
sion where Salviati attempts to explain the just-observed experimental facts
in terms of eye irradiation and of the microscopic irregularities of the
reflecting surface. Perhaps, Koyr6 meant a different page.
Koyre’s other references (F139, F141, F139 ff., and F141 ff.) are to the
beginning of the Second Day, after the preliminary discussion criticizing the
slavish followers of Aristotle’s authority. The passage (FI39-43) begins with
Salviati saying:

Then let the beginning of our reflections be the consideration that whatever motion
 THE HISTORY OF SCIENCE 219

comes to be attributed to the earth must necessarily remain imperceptible to us and as if

nonexistent, so long as we look only at terrestrial objects; for as inhabitants of the earth,
we consequently participate in the same motion. But on the other hand it is indeed just
as necessary that it display itself very generally in all other visible bodies and objects
which, being separated from the earth, do not take part in this movement. So the true
method of investigating whether any motion can be attributed to the earth, and if so
what it may be, is to observe and consider whether bodies separated from the earth
exhibit some appearance of motion which belongs equally to all.25

Here Koyr6 not only takes this passage out of context, but he manages to
perpetrate an equivocation with the Italian word principio. This can mean
either ‘beginning’ or ‘principle’. The text suggests ‘beginning’ as the meaning,
and both Drake and Salusbury-Santillana26 translate it this way. Koyrd
interprets the meaning as being, “Let the principle of our reflections be . . .
More importantly, the context of the present passage is as follows. The dis¬
cussion mentioned by Koyr6 is part of a passage containing several arguments
favorable to the earth’s rotation (FI39—50); this is followed by another
passage containing statements of the arguments against (FI50—9); and then
come the critiques of these arguments (FI 59—244). The favorable arguments
are explicitly labeled by Galileo as being merely probable (because based on
the principle of simplicity); as requiring the removal of the apparently con¬
clusive counterevidence, since a single conclusive objection would overcome
all probable arguments (F148); and as being based on a principle of relativity
of motion which both parties agree is good old Peripatetic doctrine, and
perhaps even older than Aristotle (FI42). Therefore, the principle being
explicitly laid down here must be that of optical, not mechanical, relativity.
The reasons for this are the presentation of it as Aristotelian doctrine, just
mentioned, and the explicit requirement that it remains to be shown that
there is no conclusive counterevidence (F148—50). For Galileo is very clear
that on the basis of the kind of relativity mentioned here, because of the
principle of simplicity, it only follows- that, other things being equal, the
earth’s rotation is more probable. At this stage Simplicio believes that the
other things are not equal, since he thinks there are conclusive arguments
against the earth’s motion; also at this stage Salviati is confident that he will
be able to show that the other things are equal, though it is obvious that he
has not yet done so. The other things that have to be equal are all the
terrestrial phenomena usually mentioned as proof of the earth’s rest: these
phenomena have to be equally explicable by the motion and by the rest of
the earth. So Salviati knows that his task is to show this; but this is what is
usually called the principle of mechanical relativity, which is very different
220 CHAPTER 9

from that of optical relativity, and which everyone including Koyrd agrees to
be one of Galileo’s great innovations. But this mechanical relativity is not the
principle already stated in this passage but the one that remains to be shown.
In Salviati’s words:

All inconveniences will be removed as you propound them. Up to this point, only the
first and most general reasons have been mentioned which render it not entirely im¬
probable that the daily rotation belongs to the earth rather than to the rest of the
universe. Nor do I set these forth to you as inviolable laws, but merely as plausible
reasons. For I understand very well that one single experiment or conclusive proof to the
contrary would suffice to overthrow both these and a great many other probable argu¬
ments. So there is no need to stop here; rather let us proceed ahead and hear what
Simplicio answers, and what greater probabilities or firmer arguments he adduces on the
other side . . .
Finding out whether both positions satisfy us equally well will be included in the
detailed examination of the appearances which they have to satisfy. For we have argued
ex hypothesi up to now, and will continue to argue so, assuming that both positions are
equally adapted to the fulfillment of all the appearances.27

In fact, when the time comes to answer the objections, Galileo does not
answer them in the way portrayed by Koyr6, namely by appealing to the
principle of mechanical relativity, on the basis of which the answers would be
trivial, and then by trying to familiarize and accustom Simplicio to this new
concept (K222—3, K231—6). Instead Galileo engages in a detailed critical
analysis showing that the evidence adduced by the objections cannot be used
to decide whether or not we are on a moving earth. Let us take the vertical
fall argument (FI64—6). Of course, given the principle of mechanical
relativity, “if the rock and the tower share the same motion of the earth, then
this motion will be for them as nonexistent, and everything will happen as if
it did not really exist, that is, as if the earth were motionless”, as Koyr6 puts
it (K223). And, of course, “that’s exactly what the Aristotelian cannot
admit”, as Koyr6 also says (K223). But that is why Galileo attempts to prove
his point without previously assuming the principle of mechanical relativity.
As we saw above, Galileo argues, by very concrete and basically correct con¬
siderations, that the argument from vertical fall begs the question, because we
can’t know that bodies really fall vertically unless we know that the earth
stands still. Though Galileo here does not argue explicitly for the principle of
mechanical relativity, it is obvious what the next few steps of the argument
would be. Given Galileo’s detailed critique of the objection from vertical fall,
it follows that the phenomenon of vertical fall, that is fall which is really
perpendicular to the earth’s surface, cannot be used to prove the motionless¬
ness of the earth. But since we can’t know that vertical fall is a fact unless
 THE HISTORY OF SCIENCE 221

we know that the earth stands still, it follows that we can’t know that vertical
fall is a fact unless we know that the earth moves; and from the latter it
follows that vertical fall can’t be used to prove the motion of the earth either.
Therefore, vertical fall cannot be used to prove either that the earth moves or
that it doesn’t. So here we have one crucial phenomenon that has been shown
to obey the principle of mechanical relativity.
However, the fact that this is true for one phenomenon doesn’t mean that
it will be true for other phenomena; and so Galileo criticizes the east-west-
gunshot argument and ends up showing that the range of east-west gunshots
also obeys the principle. Thus Galileo is not being repetitious for the sake of
familiarizing Simplicio with a new concept by recourse to experience, as
Koyre seems to think. Moreover, there is no real need for him to make it
look as if it were Simplicio who was requiring the recourse to experience.
Koyre so claims when he says that “the reader contemporary to Galileo . . .
through Simplicio’s mouth, once again asks for a recourse to experience:
T should like, he tells us, to find some means of making an experiment con¬
cerning these projectiles . . . ’ ” (K234). A footnote refers us to p. 194 of the
Dialogue. When this is consulted one finds that the speaker is Salviati, not
Simplicio!28 Finally, the specific reasons, not based on a previous assumption
of mechanical relativity, why the range would be the same on a moving earth,
involve a comparison to shooting arrows from a moving cart (FI93—7), and a
comparison between this and shooting them by throwing them with different
speeds from a motionless cart. Galileo argues that the arrows can be made to
move through correspondingly different distances either by shooting with
equal force from a moving cart or by imparting them different initial speeds
from a motionless one, for the equal forces from a moving cart generate more
speed in the arrow in the direction of the cart and less in the opposite
direction than the same forces do on a motionless cart; the different traveled
distances, in turn, generate equal ranges from a moving cart. Galileo’s argu¬
ment hinges on the analogy between the earth and the cart, and on the
difference in effective force produced by applying the same force in different
directions on a moving cart. It follows, that the range of east-west gunshots is
a phenomenon that would occur in the same way whether the earth moves or
not. To that extent the principle of mechanical relativity is thereby justified.
In conclusion, we may say that Koyrd’s account of the logic of the Arist¬
otelian objections to the earth’s motion and of Galileo’s counterarguments is
mistaken both in its fundamentals and in its details. The basic problems seem
to be superficiality in logical analysis, oversimplification, injudicious ex¬
aggerations, and questionable manipulation of the text by means of excessive
222 CHAPTER 9

quotations, of taking passages out of context, and of not infrequent scholarly

carelessness. Nevertheless, Koyrd does deserve the credit for having called
attention to the logical structure and validity of Galileo’s arguments and to
his rationalism, even though he misunderstands the former as circular and
misinterprets the latter as apriorism.
Finally, it would be unhistorical to deny that the study of the history of
science made great progress with Koyre; to turn the clock backwards is
simply unthinkable. Nevertheless, even historical sensibility is not an absolute
methodological requirement; indeed scientific method (as conceived here,
Chapters 5, 6, and 8) tells us that the historical understanding of the Koyrd
case would presently be rather injudicious. For at a time when the technique
of error analysis which he pioneered is undergoing a ‘cancerous growth’,29
and when the apriorist rationalism which he defended is acquiring the status
of a dogma, the ‘scientific’ thing to do is to be critical of Koyrd. The fact that
the logical analysis advocated here may be seen to stem from his technique of
error analysis, and that our emphasis on reasoning may be taken to resemble
his emphasis on rationalism, ought to serve as a warning of how far our
‘science of history’ is from the scientism of history or any other mechanical
rule or oversimplified panacea.

NOTES

1 Professors Marshall Clagett and I. Bernard Cohen have dedicated to him their major
works, and Professor Charles C. Gillispie states that “he is the master of us all” (The
Edge of Objectivity, p. 523).
2 The Edge of Objectivity, p. 523.
3 Quoted from Whitehead’s The Organization of Thought in Merton’s On Theoretical
Sociology, p. 1, as the epigraph to Chapter I.
4 ‘Chi legga di A. Koyrd . . . pp. 406-408.
5 Ibid., pp. 406-407.
6 A. Koyrd, titudes galMennes, p. 44.
7 Garin, pp. 408-409.
8 Galilei, edited by Favaro, Opere 1, 134, 167, 172.
9 Koyrd, p. 340, n. 1.
10 Garin, p. 408.
11 M. A. Finocchiaro, History of Science as Explanation, pp. 86-116.
12 Koyr6, pp. 83-107 and 136-155.
13 Ibid., pp. 155-158.
14 Ibid., pp. 107-136.
15 History of Science as Explanation, especially pp. 234-238.
16 S. Drake, Galileo Studies, p. 14.
17
S. Drake, ‘Galileo’s Discovery of the Law of Free Fall’.
 THE HISTORY OF SCIENCE 223

18 This reference has been corrected in the English translation of Etudes galiUennes
just published under the title Galileo Studies (p. 233, n. 285).
Hereafter references to Koyr^’s Etudes galiUennes will be made in the text by
prefixing page numbers with ‘K’, short for ‘Koyr<E.

2°f r ',Game[’ 6dited by FaVar°’ °pere 7’ 150~53- Hereafter, references to this volume
of Galileo s Opere, which contains the Dialogue, will be made in the text by prefixing
page numbers with ‘F’, short for ‘Favaro’, the main editor of the ‘National Edition’ of
the Opere.
The English translation of Koyr^’s book does not read this way for this passage since
instead of translating Koyrd’s words, it quotes from Drake’s translation of the Dialogue.
Cf. Koyr6, Galileo Studies, p. 162.
22 See Chapters 2 and 16.
23 Topics, 162b34; Prior Analytics, 65al0, 64b33; cf. C. L. Hamblin, Fallacies pp
50-58.
24 S. Chiaramonti, Difesa al suo Anti-Ticone e Libro delle Tre Stelle Nuove, p. 339;
cf. G. Barenghi, Considerazioni sopra il Dialogo, p. 183.
25 G. Galilei, Dialogue Concerning the Two Chief World Systems, tr Drake n 114
cf. K139-40.
26 G. Galilei, Dialogue on the Great World Systems, tr. Salusbury-Santillana, p. 127.
Galilei, Dialogue, tr. Drake, pp. 122 and 124.
The English translation of Koyr^’s book corrects his mistake; cf. Koyr6, Galileo
Studies, p. 172, and p. 226, n. 181. What effect this correction has on the Galilean-
Aristotelian drama which constitutes the tenor of Koyr^’s text, the reader can iudge
for himself.
29 t
I o give an ironical twist to this phrase which Agassi used to characterize the pre-
Koyrd period in the historiography of science; see J. Agassi, Towards an Historiography
of Science, pp. 33-40.
 CHAPTER 10

THE ERUDITION OF LOGIC AND

THE LOGIC OF ERUDITION:
CRITIQUE OF GALILEO SCHOLARSHIP

The critiques of the last several chapters, undertaken after the analysis of
various aspects of Galileo’s Dialogue of the earlier chapters, have been doubly
instructive. In part, we have gained new or deeper insights into his work; but
we have also derived useful theoretical and methodological lessons concerning
scientific rationality, the history and the nature of philosophy, and the
historiography of science. Let us continue our two-faceted critiques by
exploiting the opportunities offered by classic Galileo scholarship from the
point of view of the logic of erudition.
Historians have seldom been good logicians, nor have logicians often
excelled in historical sensibility. Is this fact a historical accident or a logical
necessity? One way to explore this problem would be to undertake a method¬
ological examination of the history of logic; this could be done partly by
analyzing works by historians of logic with an eye toward the formulation of
methodological lessons, and partly by a philosophical, more or less speculative
reflection on the nature of logic, history, the history of logic, and the historio¬
graphy of logic. One useful distinction here would be between logical
theorizing and logical practice. Almost all works in the history of logic deal,
of course, with the development of logical theories. There is no reason, how¬
ever, to neglect unduly the history of logical practice; in fact, given the pre¬
ponderance of the former, now would seem an appropriate time to do some
work in the latter. This would give us a second way to explore the above
mentioned problem. But what is meant by logical practice? It is the more or
less self-conscious practice of reasoning combined with some explicit use of
(basic) logical concepts and terminology. Important works for the historian
of logical practice to consider would thus be Euclid’s Elements, Plato’s
Dialogues, Aquinas’s Summa Theologica, Machiavelli’s Discourses, Galileo’s
Dialogue Concerning the Two Chief World Systems, Hume’s Dialogues
Concerning Natural Religion, perhaps Marx’s Capital. It seems that the study
of such works would require both logical acumen and historical sensibility
and would therefore be very fruitful and challenging.
A third way to explore the problem would be to study the methodology
of the history of scientific thought, for though one may not want to equate
scientific thought with logic, few would deny that reasoning plays a central

224
 THE ERUDITION OF LOGIC 225

and effective role in science. I have carried out some such studies elsewhere.1
Fourth, one could study the logic of history, partly in the sense of the
nature of historical development, but primarily in the sense of the reasoning
of historians. Here one would discover whether or not the difficulties are
insurmountable.
I believe I can combine several of these approaches by a critical examina¬
tion of selected writings by historians of science dealing with Galileo’s
Dialogue. Most directly I will be dealing with the logic of historiography,
attempting to derive methodological lessons by analyzing the reasoning of
historians. Indirectly I will be dealing with the history and the historiography
of logical practice. Because of the place of Galileo and the Dialogue in the
history of science, this will be relevant to the historigraphy of science.
Finally, I will be dealing with the historiography of logical theorizing insofar
as some of the subjects of my discussion will be remarks by historians about a
number of logical concepts used by Galileo.
The choice for my analysis is Emil Strauss’s Notes to his 1891 German
translation of Galileo’s Dialogue2 because they are universally regarded as the
most informative and erudite ever compiled; Antonio Favaro’s text of the
Dialogue in the ‘National Edition’ of Galileo’s complete works, since this is
the standard and definitive critical edition;3 Stillman Drake’s translation of
the book since Drake is perhaps the greatest Galilean translator of all times
and the leading Galilean scholar of our time;4 and Maurice Clavelin’s Natural
Philosophy of Galileo5 because it is the most widely acclaimed recent work
on the topic.
Though no general solution to our problem will emerge, it will become
apparent that, when the subject of historical investigation is itself logic and
reasoning, even the erudition and scholarship of the best historians are not
enough; at least as important are logical skill and sensitivity.

ERUDITE COMMENTARY ON REASONING: STRAUSS

The erudition of Strauss’s notes is impressive; it has been recognized by

Drake, who explicitly expresses his indebtedness to him,6 and by Giorgio
de Santillana, whose notes to his revision of the Salusbury translation are full
of references to Strauss. Though these notes can still be read with profit and
perhaps still remain unsurpassed, the careful reader of the Dialogue can detect
a certain weakness in one specific area, namely logic and reasoning. It is
useful, then, to examine carefully Strauss’s comments on these topics.
226 CHAPTER 10

1. Let us begin with the note Strauss has to the passage where Galileo dis¬
cusses Aristotle’s authority as a logician (F59—60).7 Strauss comments:

The merits of Aristotle in logic are also emphasized elsewhere. His writings pertaining
to this part of philosophy are as a whole traditionally called Organon (instrument, tool).
[S502, n. 42]8

The other passage to which Strauss refers (FI57) need not concern us
here. This note contains two points: the interpretative claim that Galileo is
here emphasizing Aristotle’s logical merits, and the informative remark that
his logical treatises are entitled Organon.
These comments are beset by the following problems. Though it is true
that Simplicio is emphasizing Aristotle’s logical merits, the general tone of
the discussion, including Salviati’s speech, is critical of Aristotle’s authority
as a logician. Strauss’s misinterpretation is due to his failure to note the
distinction between logic-as-theory and logic-as-practice, which is contained
in Salviati’s remark that one “can be a great logician, but not too skillful in
knowing how to use logic” (F59—60); hence Strauss does not see that
Galileo’s emphasis on Aristotle’s merits in logic refers only to his merits as a
logical theorist, whereas regarding his skill as a logical practitioner Galileo is
emphasizing his deficiency, at least in reasoning about the topics at hand.
The second problem is that one wonders about the relevance of Strauss’s
remark concerning the title of Aristotle’s logical treatises. The only connec¬
tion seems to be purely verbal, in the sense that the passage contains some
remarks about organs, indeed almost a pun about the definition of logic as
the organ of philosophizing, and the difference between organ-making and
organ-playing, which suggests a difference between the making and the use
of the organ of philosophizing. But Strauss’s reference to Aristotle’s Organon
misses these subtleties, both the pun and the substance of Galileo’s point,
which is the distinction between logical theory and practice.

2. Galileo’s criticism of the contrariety argument includes a discussion of

the bar’s paradox (F66), in connection with the question of whether the
Peripatetics are obliged to argue that because heaven is inalterable, and the
earth alterable, heaven has a contrary and hence is alterable after all. Strauss
comments (S503, n. 50):

The well-known fallacy of the Cretan here repeated does not really belong to the class
which one calls ‘sorites’ in the narrow sense. The typical example of this is rather the
supposed argument of whether one can form a pile by adding one piece to a quantity of
things. I have translated the odd expression 'argomento cornu to’ by ‘pseudo-argument’.
 THE ERUDITION OF LOGIC 227

I am inclined to agree that the argument about the lying Cretan is not a
sorites, as Galileo calls it, and that it is strange to call it an ‘argomento
cornuto'. But why mention this and not the fact that Galileo’s version of the
paradox is incorrectly stated? This is so because Galileo’s statement is as
follows (F66):
(1) A Cretan said that all Cretans are liars.
(2) Therefore, he is lying in saying that all Cretans are liars.
(3) Therefore, all Cretans tell the truth.
(4) Therefore, he is telling the truth.
(5) Therefore, it is true that all Cretans are liars.
(6) Therefore, he is lying.
(7) Etc.
The problem with this is that (3) does not follow from (2). What does
follow is that not all Cretans are liars, which does not entail that he is lying.
What Galileo should have said is that the Cretan said that he was lying; then
it would follow, from his lying in saying that he is a liar, that he is telling the
truth.

3. At the beginning of the Second Day several of Aristotle’s arguments

against the motion of the earth are examined. In the course of his critique,
Galileo gives what look like verbatim quotations from Aristotle’s works. This
gives Strauss the opportunity for some very informative erudite commentary.
For example, when Galileo finds it useful to quote the Latin text of Aristotle’s
second argument (FI62—3), Galileo refers to it as ‘Text 97’. Strauss explains
(S520, n. 33):

Which Aristotelian text Galileo used can perhaps be determined from the reference
here given, namely ‘testo 97’, which I have translated as ‘Paragraph 97’. In our Aristotle
editions the paragraphs are commonly not numbered with consecutive numbers, but
rather their numbering begins anew with each chapter. Thus the place here cited is deter¬
mined as being in the first paragraph of the 14th chapter. In ‘Dido’s’ edition the
preceding 13 chapters comprise altogether 95 paragraphs, so that the arrangement of the
text used by Galileo does not agree exactly but only approximately with ours.

Such information is obviously of little relevance to the understanding of the

content of Galileo’s text; moreover, in this case it is even superfluous for the
identification of the corresponding passage in Aristotle since, in addition to
saying ‘testo 97’, Galileo also gives the reference “On the Heavens, second
book, Chapter 14”, which is sufficient.
In a previous note to the passage where Galileo first quotes Aristotle’s
228 CHAPTER 10

second argument in Italian (FI50), Strauss does have a word of explanation

(S515, n. 23):

By the apparent lagging behind is meant the fact that the planets require somewhat
longer than 24 hours to complete a revolution in the heavens; this longer period is
precisely the result of their particular movement (partly apparent, partly real).

This remark is certainly helpful, but it explains the easiest part of the argu¬
ment. How problematic the rest is can be glimpsed from de Santillana’s note
to this passage, which note, however, is itself very difficult to understand;9
its being almost unreadable is perhaps not surprising in the light of the fact
that it is an attempt to understand the meaning of Galileo’s passage by
reference to other non-Galilean texts, rather than by analyzing the other
remarks that Galileo makes about the argument where he criticizes it
(FI62-4). An interpretation of this whole passage will be given below
(Chapter 16).

4. Galileo’s criticism of the argument from vertical fall (F165—6) begins by

showing the sense in which the argument begs the question, or is a petitio
principii. Strauss comments that (S520, n. 35):

petitio principii is a well known logical technical term by which one designates an
invalid argument whose conclusion is substantiated by means of a premise dependent on
the conclusion. The Aristotelian syllogism, brought into canonical form, would read:
(A) If the earth moves, the body cannot fall vertically.
(B) But the body falls vertically.
(C) Therefore the earth does not move.
But in this case the second premise is questionable as long as the conclusion is question¬
able. - Here this second premise itself is called the ‘middle term’, whereas it is the
middle concept which is more usually referred to by this designation. - Ignotum per
aegue ignotum : the unknown by means of the equally unknown.

This is the full text of Strauss’s note 35 to the Second Day, of which I have
included the third part (which translates Galileo’s Latin term) in order to
point out that Strauss himself commits this fallacy in the second part of his
note, concerning ‘middle term’. The reason is that someone who doesn’t
already know what is the more usual meaning of ‘middle term’, will not know
it after he is told that it more usually refers to the ‘middle concept’, rather
than to a premise like (B) above, which admittedly is what Galileo is referring
to. Ordinarily the middle term is a term common to both premises of a
syllogism, and not present in its conclusion; for example, lY> is the middle
term in the following:
 THE ERUDITION OF LOGIC 229

All X are Y.
All Y are Z.
Therefore, all X are Z.

However, it is the first part of Strauss’s note that requires comment. His
definition of petitio principii is correct. His canonical formulation of the
argument is also correct, insofar as it goes. However, it is not clear what he
means in saying that (B) is questionable as long as (C) is. In accordance with
his own definition of petitio principii this should mean that (B) is dependent
on (C). But in what sense is (B) dependent on (C)? The answer is in the sense
that (B) is justified by assuming (C). To explain this answer Strauss should
have reconstructed Galileo’s reasoning, which is the following:10
(1) If the earth moves and the body grazes the tower, then the body is not
really falling vertically. (This is obvious on reflection.)
(2) Therefore, if the earth moves, then the grazing of the tower would not
imply real vertical fall.
(3) But, if the earth does not move, then the grazing of the tower does
imply real vertical fall. (This is obvious on reflection.)
(4) Therefore, the grazing of the tower implies real vertical fall if and only
if the earth does not move.
(5) But the argument is assuming that the grazing of the tower implies
real vertical fall, since (B) is justified on the basis of the grazing of the tower.
(6) Therefore, the argument is assuming that the earth does not move,
namely its conclusion (C).

5. One of the objections to the earth’s motion criticized by Galileo was

that a simple body like the earth could not have multiple natural motions
(F281-8). In the course of his criticism Galileo argues briefly that there is
no more incompatibility between the Copernican straight-downward motion
of heavy bodies and circular rotation of the earth than there is between the
Peripatetic straight-downward motion of heavy bodies and motionlessness of
the earth (F288). On this argument Strauss comments that (S542, n. 148):

the following consideration by Sagredo is somewhat sophistical, a special principle

implying rest is not necessary from the Peripatetic point of view. Galileo obviously does
not attach a very high scientific value to the whole discussion, as shown among other
things from the fact that he puts it in Sagredo’s mouth.

There are several problems with this remark. First, it is not clear why
Sagredo’s argument (F288) should be regarded as sophistical and not merely
invalid or incorrect. Second, it is questionable whether it is even incorrect.
230 CHAPTER 10

Sagredo’s point would be incorrect if he were asserting that the peripatetics

do in fact need two distinct principles, one for downward motion and the
other for rest for the whole earth; Strauss is right to point out that the rest of
the whole earth is a consequence of the motion of its parts toward the center
of the universe. However, Sagredo’s (and Galileo’s) point is that the mixing of
downward and circular motions by the Copernicans is less implausible than
the mixing of downward motion and rest by the Peripatetics; that just as the
two prima facie imcompatible states of motion and of rest do not really
conflict, there may not be any conflict either between the two motions,
straight and circular. It should also be noted that Salviati intervenes to say
that another reason why there may not be any conflict between terrestial
straight and circular motions, is that the apparently straight motion may be
really circular, which he had suggested in a previous digression (FI88—93).
The ‘fanciful’ character of this digression and its scientific error do not
invalidate its use in the present context, namely to suggest a plausible way to
resolve the straight-circular conflict.

6. At the beginning of the Third Day, after giving a sketch of the helio¬
centric system, Salviati complains about the existence of people who make
silly objections to the earth’s motion, for example, that it is impossible for
the heavy earth to go up above the sun. These people provide Galileo with the
pretext for the following incredibly subtle and humorous witticism:

One should not pay attention to such people, whose number is infinite, nor take into
account their fooleries; nor should one try to win the sympathies of men whose defini¬
tion contains only the genus but not the difference, in order to have them as companions
in very subtle and delicate discussions. [F355, my translation]

Strauss comments that (S551, n. 34)

according to orthodox logic every definition of a concept consists of the specification of

a ‘genus’ and a ‘specific difference’. The conventional definition of man, which at the
same time supplied one of the most common examples to illustrate the nature of
definition, correspondingly read ‘rational mortal animal’; if one then disregarded the
‘specific difference’ which is expressed by the term ‘rational’, then the mere genus
‘animal’ is left.

Here Strauss must be given the credit for getting almost right something
completely missed by other otherwise competent scholars. However, his
comment is not completely adequate. For if we take the definition of man to
be ‘rational mortal animal’, and if we delete from it the specific difference
‘rational’, then we are left not with ‘animal’ but with ‘mortal animal’; in this
 THE ERUDITION OF LOGIC 231

case the joke does not get off the ground since Galileo would be labeling these
people ‘mortal animals’. Instead Galileo must be thinking of the definition of
man as 'rational animal’, so that when we take out the specific difference
‘rational’ from this definition, we are left with the genus ‘animal’ and thus
with the clearly intended vituperative though unprosaic description of these
men as ‘animals’.

7. There is a passage in the Third Day where Galileo asks whether Tycho or
his followers have searched for any evidence from stellar appearances which
might confirm or disconfirm the earth’s motion (F399). Galileo answers that
they haven’t since they argue in an ad hominem manner, namely by deriving
anti-Copernican conclusions from Copernicus’s admission that there are no
changes in stellar appearances. This gives Strauss the occasion to comment
(S561, n. 70):

Ad hominem, ‘against the man’, a term in dialectics by which one understands an argu¬
ment which is not based on established or generally accepted premises but rather on
those held as correct by the opponent; such an argument has therefore of course no
scientific worth but can serve for the persuasion of an opponent.

The first thing to point out is that Strauss is, correctly, not equating
Galileo’s lad hominem argument’ with its modern meaning, which is the
fallacy committed when a case is argued not on its own merits but merely on
the basis of the motives or background of its supporters or opponents. Never¬
theless, there are several problems with this note. First Strauss is not explain¬
ing Galileo’s meaning of ad hominem argument because Galileo regards as
ad hominem the Tychonic argument based on the premise that there are no
annual changes in stellar appearances, which was a ‘generally accepted’
premise, rather than one merely held by Copernicans. In other words, it is not
true to say about the Tychonic argument being considered by Galileo that it
was “not based on established or generally accepted premises but rather on
those held as correct by the opponent”. Of course, it is indeed true that the
argument was based on premises which were ‘not established’, but that is
another matter, and that is precisely Galileo’s point, that the lack of changes
in stellar appearances was generally accepted, but not ‘established’ as a result
of search and research.
Second, Strauss claims that ‘of course’ such an argument has ‘no scientific
worth’. Of course this is more nearly correct of the type of argument that
Strauss is talking about. But is it true of the type that Galileo has in mind?
In this passage (F399) what Galileo finds wrong with Tycho and his followers
232 CHAPTER 10

is not that they argue ad hominem, but rather that they do so “more in
defense of another man than out of any great desire to get at the truth”,11
and hence that what they do “may suffice to refute the man, but certainly
not to clear up the fact”.12 This presupposes that ad hominem arguments can
be used to get at the truth, though they can also be used for merely rhetorical
or dialectical purposes.
All of this suggests that by ad hominem argument Galileo means an argu¬
ment which derives a conclusion unacceptable to an opponent from premises
accepted by him, but not necessarily by the arguer, or generally accepted, or
established. Hence such arguments (if otherwise correct) provide conditional
knowledge, namely that if one accepts the premises in question then the con¬
clusion undesirable to the opponent follows. In this passage Galileo correctly
points out that such arguments are improper to deal with the question of the
factual correctness of the premises; however there is no evidence to attribute
to him the erroneous claim that such arguments have no scientific value. On
the contrary, there is evidence from other works that supports my inter¬
pretation of Galileo’s conception of ad hominem arguments.13

EDITING A TEXT OF REASONING: FAVARO

The name Favaro certainly needs no introduction for historians of science

interested in Galileo. What I wish to do here is merely to call attention to a
few examples where the text of the National Edition could be improved.
Since these imperfections center around one definite area, namely logic, it is
of some importance to mention them.

1. One difference between the Favaro text and that of the 1632 edition is
the modernized spelling of the former. Usually no problem results, but the
following passage is an exception. On F57 Salviati has just completed his
argument that circular motion is the only type of natural motion, and that
straight motion is the simplest means of acquiring the natural state of rest at
the proper place or of circular motion. Simplicio’s reply to this is in fact an
argument designed to show that the natural motion of the earth (in the sense
of terrestrial globe) is straight, which is a direct refutation of the first part of
the conclusion just reached by Salviati; the 1632 edition (EP24)14 makes this
clear by using a capital ‘T’ in the spelling of ‘Terra’ (Earth). Here, as in most
other occurrences of this word, Favaro uses a lower case ‘t’; the net effect is
that one gets the impression that Simplicio’s reply is merely a claim that
Salviati’s conclusion conflicts with the sense-experience of straight-downward
 THE ERUDITION OF LOGIC 233

motion of terrestrial (heavy) bodies, and of straight-upward motion of light

bodies, which constitutes a misunderstanding of Simplicio’s own speech and,
more importantly, leads one to misconstrue Salviati’s natural motion argu¬
ment. I am not saying, ot course, that Favaro’s text necessarily gives such a
wrong impression, but only that it facilitates this misinterpretation, and
renders more difficult the correct one.

2. The logical significance of punctuation is implicitly admitted by Favaro

when he says, “We have however departed from the original edition (as we
did in preceding volumes) for what pertains to punctuation, which we have
rendered more rational” (F12-13). I would agree that on the whole Favaro’s
changes in punctuation are a great improvement of the 1632 edition. How¬
ever this is not always the case, as the following example shows. The passage
is such that it would be impossible to understand it from Favaro’s text,
whereas the original is less misleading by being more amorphous. The passage
is part of Galileo’s answer to the objection from birds, and the original
punctuation may be seen in the following literal, and line-by-line translation
from the 1632 edition (EP 180):

1 Your having more difficulty for this, than for the other
2 objections, seems to me, to depend on birds being animate,
3 and being able therefore to use force at will against the primary
4 motion innate in earthly things; in such a way precisely, that
5 we see them while they are alive flying even upwards, a motion
6 impossible to them as heavy bodies; whereas dead they cannot, but
7 fall downwards; and therefore you judge, that the reasons, which take
8 place for all sorts of projectiles mentioned above, cannot
9 take place for birds; and this is very true, and because it is true,
10 therefore we do not see Mr. Sagredo those projectiles doing, what
11 birds do; for if from the top of a tower you let
12 fall a dead bird, and a live one, the dead one will do the same, that
13 a stone does; that is it will follow first the general diurnal motion, and
14 then the downward motion, as a heavy body; but if the released bird is
15 alive, who will forbid it, the diurnal motion always staying with it,
16 from going by a beating of wings toward whatever part of the horizon
17 it likes most? and this new motion, being its own in particular, and
18 not shared by us, must become perceptible to us; and if it had
19 by its flight moved toward the west, who is to forbid it,
20 from returning to the tower by the same beating of wings?
21 Because finally to fly up toward the west,
234 CHAPTER 10

22 was nothing other, than to subtract from the diurnal motion, which has
for example ten degrees
23 of speed, a single degree, so that it was left with nine while
24 it was flying, and if it had alighted to earth, it would have regained
25 the ten common degrees; to which by flying toward the east, it
26 could have added one, and with the eleven return to the tower.
27 And in sum, if we will consider, and more intimately
28 contemplate the effects of the flying of birds, they do not
29 differ in anything from projectiles thrown in all directions,
30 except that these are moved by an external projector, and those
31 by an internal principle.

After many fruitless attempts to arrive at an accurate reconstruction, the

only interpretation that made sense became one consisting of two distinct
comments about the birds argument (see Chapter 16): a favorable one on
lines 1—9, and an unfavorable one on lines 9—26, with lines 27—31 con¬
stituting a remark extraneous to these two comments but summarizing the
entire discussion of the birds argument. Thus it was misleading for Favaro
to eliminate the period from line 26 (cf. F212.26),15 and to introduce one in
line 18 to replace the semicolon (cf. F212.19). Favaro makes the biggest
break in lines 1 — 18 on line 11 by replacing the semicolon by a colon; since
there is no way of combining the first 11 lines, he should not have left the
semicolon of line 9 (F212.10) but replaced it by a period. There is no
question that Galileo’s punctuation was odd, and his language in this passage
very colloquial; moreover, the question mark in line 20 and the capital letter
in line 21 are misleading. Nevertheless, Galileo’s original passage is easier than
Favaro’s, because Favaro gives the wrong clues.

3. The Favaro text introduces a number of corrections in the text of the

1632 edition (F10—13). Favaro states that most of these corrections are
grounded on (1) the “Errata” page of the 1632 edition (F10); (2) Galileo’s
own corrections handwritten in his copy of the 1632 edition (F10); (3) a
manuscript that exists for the discussion of the 1572 nova at the beginning
of the Third Day, on pp. F301.34-346.26 (F10); (4) the text of Galileo’s
Discorso del flusso e reflusso del mare, many passages of which he transcribed
verbatim in his composition of the Fourth Day (FI2, n. 1); and (5) the two
versions of the wording of the marginal postils, one alongside the text and the
other in a table at the end of the 1632 edition (FI 1 — 12). Besides these
groups of corrections, Favaro corrected a number of obvious typographical
errors of the 1632 edition which had not been noted by Galileo in his own
 THE ERUDITION OF LOGIC 235

C0Py (F10 11)- Favaro also states that he made a few other corrections on
the grounds that “the correction was necessary, otherwise the meaning would
have been seriously affected” (F10, n. 5) and that they fall into the pattern
of Galileo’s habit of exchanging one term for one opposite in meaning {Ibid.).
These corrections are Favaro’s replacement of Galileo’s retto (straight) by
circolare (circular) on p. 45, in the second postil; the replacement of Galileo’s
minori (minor) by maggiori (major) on F55.31; the replacement of Galileo’s
occidentali ... bassi, e alti . . . orientali (western ... low, and eastern . ..
high) by occidentali .. . alti, e bassi .. . orientali (western . .. high, and
eastern . . . low) on F208.27; the replacement of segante (secant) by tangente
(tangent) on F228.14; and the replacement of quoziente (quotient) by
prodotto (product) on F324.2.
To these I believe one should add at least the following two corrections,
though they were not made by Favaro. On p. F 193.31 I think that the word
should be annuale (annual) instead of diurno (diurnal). The sentence involved
may be translated literally as follows:

But this one, taken from bodies falling perpendicularly, I do not regard as one of the
strongest arguments for the immobility of the earth, and I don’t know what will happen
about gunshots, and especially about those against diurnal motion. [F193]

The speaker is Simplicio and the remark is made at the end of the criticism
of the objection from falling bodies and the beginning of the discussion of the
gunshot objections. I believe that the ‘those’ of the last clause refers to
‘arguments’ rather than to ‘gunshots’, and hence the last clause should be
‘and especially about those against annual motion’, so that the last sentence
would mean that Simplicio doesn’t know what will happen about the argu¬
ments from gunshots, and especially about the arguments against annual
motion. There are two reasons. First, logically speaking, if‘those’ referred to
gunshots and ‘diurnal’ is not corrected, then we would have an implicit state¬
ment of the opinion that the arguments from cannons shot in a direction
against the diurnal motion (i.e. gunshots in a westward direction) are the
strongest. However, in the list of geostatic arguments, there are none from
gunshots toward the west, though of course we have the argument from east-
west gunshots, that is the argument involving considerations about cannons
shot toward the west and toward the east. The logic of this argument is such
that no argument would remain if one considered only gunshots toward the
west, since the argument attempts to derive from the earth’s rotation the con¬
sequence that there would be a difference in the range of gunshots toward
the east and toward the west. Moreover, the east-west gunshots argument is
236 CHAPTER 10

nowhere in the book regarded as particularly strong, whereas one of the

arguments against the annual motion (the parallax argument) is so regarded,
so much so that Galileo in fact is unable to give a conclusive refutation of it,
but rather criticizes it by suggesting a research program to discover new
evidence that would refute it and confirm the earth’s annual motion (F164,
F406, and F407—16).
The second reason for my interpretation is linguistic. If the arguments
referred to in the last clause are arguments from cannon shots in a direction
against the diurnal motion, then there would be no need of the ‘and’ with
which the last clause begins. The use of this ‘and’ indicates that the structure
of the whole sentence is: But [(this . . . earth), and (I don’t . . . gunshots),
and (especially . . . motion)], where the third conjunct in this series intro¬
duces a relatively different claim, such as one about the special strength of
objections to the annual motion. However, if the term ‘diurnal’were correct,
then the structure of the sentence would be: But {[this . . . earth], and [I
don’t . . . about (gunshots, and especially . . . motion)]}, where the last ‘and’
would be superfluous.

4. My second example of overlooked correction involves Galileo’s exchange

of perche (because) and perd or percid (therefore) on F216.37—217.1. The
relevant passage may be translated literally as follows:

... all these are arguments for the truth of the conclusion, namely that whirling gives to
the moving body an impetus toward the circumference, if the motion is fast; and
because, if the earth turned on itself, the motion of the surface, especially near the
equator, being incomparably faster than those mentioned above, would extrude every¬
thing toward the sky. [F216.35-217.3]

The speaker is Salviati who in the sentences just preceding describes a number
of easily performable experiments supporting the generalization about the
extruding power of whirling. Here he is strengthening this objection to the
earth’s rotation, before criticizing it. It is obvious that in order to make sense
the ‘because’ must be replaced by ‘therefore’, a good example of Galileo’s
slips involving exchange of opposites; the slip of the pen is all the more under¬
standable here in the light of the similarity of spelling in Italian.

5. An example of a marginal postil uncorrected by Favaro is the second one

on F413: “If annual variations in the fixed stars were observed, the earth’s
motion would not be contradicted”. Since it is contextually obvious that if
annual stellar variations were observed, that would be a very strong (almost
 THE ERUDITION OF LOGIC 237

decisive) confirmation of the earth’s motion, logical intuition tells us that the
antecedent clause should be a negative one: “if no annual variations in the
fixed stars were observed . . . In fact, an examination of the text to which
the postil refers shows the need for such a correction, for the text has Salviati
affirm that, though if variations were seen that would prove the earth’s
motion, if variations were not seen that would not disprove it, since the
failure to see them might be due to immense stellar distances or to the failure
to look for them or to the failure to look carefully enough.

6. The passage on centrifugal force (F214—44) is generally considered as

one of the scientifically most important ones, hence it is important here to be
especially careful about what Galileo is saying and what he is not saying. On
F238.12—238.22 we have a passage which may be translated literally as
follows:

... no one will think that the cause of extrusion in the big wheel grows as the proportion
of the speed of its circumference to the speed of the circumference of the smaller wheel,
because this is most false, as for now a very easy experiment can show us somewhat
roughly: because with a stick one cubit long we could throw such a stone that we could
not with one six cubits long, even if the motion of the end of the long stick, namely of
the stone held therein, were more than twice as fast as the motion of the end of the
shorter stick; this would happen if the speed were such that in the time of a whole
rotation by the bigger stick, the smaller one would rotate three times.

The problem here is with the reasoning indicator ‘because’ (che) after the
colon. It makes little sense to our logical intuition. In fact, a check with the
1632 edition reveals that Galileo used a ‘che’ without the accent, which is no
reasoning indicator but rather means ‘that’, as in ‘the experiment that’.

THE TRANSLATION OF REASONING: DRAKE

English-speaking people are fortunate that there exists a scholar like Stillman
Drake whose efforts have made available to them in their own tongue almost
all of Galileo’s major works.16 Add to this the fact that Drake’s devotion has
enabled him to internalize the spirit of Galileo to an unprecedented degree,
and we get a truly unique situation. Not content with merely translating
Galileo, Drake has also formulated with increasing frequency, a number of
explanatory interpretations which cannot be easily excelled for their textual
accuracy or biographical well-foundedness, whatever their alleged faults from
other points of view.17 As if this were not enough, he has in recent years
undertaken the epoch-making task of deciphering and chronologically
238 CHAPTER 10

arranging Galileo’s unpublished and previously unanalyzed notes on motion,

and this effort has always led either to new insights into Galileo’s work or to
novel evidence for some old beliefs, or to new counter-evidence against recent
popular historical interpretations.18
Having said this much, I must add that in my own investigations I have not
always been satisfied with Drake’s translations.19 If these inadequacies were
those of an ordinary Galileo scholar, or if they pertained to random topics,
then it would not be worth discussing them explicitly. However, they do
center around one definite area, namely logic and reasoning; moreover, this
topic is relatively controversial since the practice of reasoning can be easily
confused with commitment to apriorist rationalism (as we saw in the case of
Koyre), and since the art of reasoning could be easily opposed to the art of
experimenting (though there is no opposition, since experiment is either an
activity distinct from reasoning or else it is observation guided by reasoning).
Before discussing a number of passages from the Dialogue to illustrate
these imperfections of translation, I want to give one example to illustrate
that Drake’s translation does indeed capture and convey the spirit of Galileo
very well, so that my criticism is perhaps as much of a comment on the
nature of translation in general as on the adequacy of Drake’s, or at least a
comment on the problems of translating a multi-faceted work like the
Dialogue. The example is a sentence from Galileo’s preface “To the Discern¬
ing Reader” in which Galileo is referring to his deceased friends Sagredo and
Salviati. Translated literally the sentence would read:

Now, since bitter death has deprived Venice and Florence of these two great lights while
they were in the brightest part of their years, I have wanted, as far as my meager abilities
can, to prolong the life of their reputation on these pages of mine by using them as
interlocutors in the present controversy. [F31 ]

Drake’s translation reads, “Now, since bitter death has deprived Venice and
Florence of those two great luminaries in the very meridian of their years, I
have . . . .” (D7).20 The important phrase here is “in the very meridian of
their years”, which is a more Galilean phrase than Galileo’s own “nel piu bel
sereno degli anni loro”. In the light of the astronomical dimension of the
book, Drake’s metaphor is very much of the type used by Galileo, who is
constantly finding appropriate metaphorical uses for terms which literally
apply to the natural phenomena he is discussing.21

1. Let us now pass to criticism. The very first sentence with which the
dialogue begins may be translated as follows:
 THE ERUDITION OF LOGIC 239

SALV. It was yesterday’s conclusion and plan that today we should discuss, as distinctly
and particularly as we can, of the natural reasons and their efficacy which by one side
and by the other have so far been produced by the proponents of the Aristotelian and
Ptolemaic position and by the followers of the Copernican system. [F33, my translation]

Instead of “the natural reasons and their efficacy”, Drake has “the character
and the efficacy of those laws of nature” (D9). The problem here is with the
‘laws of nature’, which is supposed to be a translation of the Italian ‘ragioni
naturali', which means natural reasons in the sense of physical arguments.
Drake’s rendition is misleading because it obscures the logical dimension of
the work, which is bound to be present in a discussion of arguments and their
strength. In other words, whereas the first sentence promises that the book
will be a work in applied logic (if we define logic as the study of the nature
and validity of reasoning), Drake’s translation makes it a work of physics (if
we define the latter as the study of the laws of nature).

2. In the course of the discussion of natural motion in the First Day,

Simplicio gives the following argument, as it is translated by Drake:

Who is there so blind as not to see that earthy and watery parts, as heavy things, move
naturally downward — that is to say toward the center of the universe, assigned by
nature itself as the end and terminus of straight motion deorsuml Who does not likewise
see fire and air move directly upward toward the arc of the moon’s orbit, as the natural
end of motion sursuml This being so obviously seen, and it being certain that eadem est
ratio totius et partium, why should he not call it a true and evident proposition that the
natural motion of the earth is straight motion ad medium, and that of fire, straight a
medio. [D32—3]

In a note, Drake tells us that the Latin eadem est ratio totius et partium
means that “the reasoning which applies to the whole applies also to the part”
(D471). However, this makes nonsense of Simplicio’s argument in a way
certainly not intended by Galileo. The argument is that since the natural
motion of the parts of the element earth is (visibly) straight down and that of
the parts of the element fire is (visibly) straight up, therefore the natural
motion of the whole earth is straight down and of fire straight up, in virtue of
the principle expressed in Latin. To be meaningful, the argument must be
appealing to the principle that the reasoning that applies to the parts applies
also to the whole, which is the reverse of Drake’s translation. In fact, the
Latin sentence means literally “the reasoning applying to the whole and to
the parts is the same”, which refers both to reasoning from parts to whole
and from whole to parts; hence, Drake’s translation is not incorrect from an
abstract point of view, but only in the context of the passage in which the
240 CHAPTER 10

Latin sentence occurs. However, in this context we have an inaccurate re¬

presentation of the logic expressed in the text.

3. Among the objections to the earth’s diurnal motion, several dealt with
gunshots. Though interrelated to one another, and as a group to the various
versions of the falling bodies argument, these gunshot objections should be
distinguished from each other if they are to be fully understood. However,
this is difficult to do from Drake’s translation where the same term ‘point-
blank’ is used both to refer correctly to shots which are ‘di punto in bianco'
(FI53.9, D127) or ‘di punto bianco’ (F205.33, D180), as well as to refer
incorrectly to shots which are ‘di volata' (FI52.20, D126). The latter type of
shot instead is one where the artillery piece is given a great elevation.22 The
confusion is compounded by the fact that the second time the latter phrase
occurs (F194.12) Drake drops it from his translation (D168).

4. In the critique of Aristotle’s objection from falling bodies, there is a

speech by Salviati which Drake translates as follows:

SALV. So Aristotle’s defense consists in its being impossible, or at least in his having
considered it impossible, that the rock might move with a motion mixed of straight and
circular. For if he had not held it to be impossible that the stone might move both
toward and around the center at the same time, he would have understood how it could
happen that the falling rock might go grazing the tower whether that was moving or was
standing still, and consequently he would have been able to perceive that this grazing
could imply nothing as to the motion or rest of the earth.
Nevertheless this does not excuse Aristotle, not only because if he did have this idea
he ought to have said so, it being such an important point in the argument, but also, and
more so, because it cannot be said either that such an effect is impossible or that
Aristotle considered it impossible. The former cannot be said because, as I shall shortly
prove to you, this is not only possible but necessary; and the latter cannot be said either,
because Aristotle himself admits that fire moves naturally upward in a straight line and
also turns in the diurnal motion which is imparted by the sky to all the element of fire
and to the greater part of the air. Therefore if he saw no impossibility in the mixing of
straight-upward with circular motion, as communicated to fire and to the air up as far
as the moon’s orbit, no more should he deem this impossible with regard to the rock’s
straight-downward motion and the circular motion natural to the entire globe of the
earth, of which the rock is a part. [ D141 ]

The logical impression being conveyed here is that the final and/or main con¬
clusion in this passage is the last sentence, “If he saw no impossibility in ... ,
no more should he deem this impossible . . . ”. Presumably this consequence
is being drawn from the preceding statements; perhaps the basis is the imme¬
diately preceding clause, i.e., that “Aristotle himself admits that ... to the
 THE ERUDITION OF LOGIC 241

greater part of the air”. But such an inference would make no sense, for it
would be like arguing that because Aristotle admits A, therefore if Aristotle
admits A then he should admit B. In fact, no other preceding sentence can
be used to ground the alleged conclusion; moreover, one will soon realize that
this alleged conclusion is intrinsically plausible in the context, and therefore
it would make more sense merely to assert it. If one checks the Italian text
(FI67) that is exactly what one finds; there is nothing in the last sentence
corresponding to ‘therefore if ... ’ but rather a phrase (se dunque) corres¬
ponding to ‘if then . . . ’. Moreover, the punctuation, which serves as a logical
indicator, is clearer in the Italian. In short, Drake’s last twelve lines are better
translated as follows: “The former cannot be said . . . necessary. The latter
cannot . . . greater part of air; if then he saw no impossibility ... of which the
rock is a part”. When so modified, the main conclusion in Drake’s second
paragraph is that “this does not excuse Aristotle”, i.e., that the defence of
Aristotle mentioned in the first paragraph does not really excuse him, for two
reasons. The first is that he ought to have said so, i.e., given explicitly that
defense; the second reason is that neither one of two things can be said. The
first of these things cannot be said because so will Salviati shortly prove; and
the second thing cannot be said because of an argument having the form;
Aristotle admits A, and if he admits A then he should admit B.
Drake’s inaccuracy is significantly misleading because it prevents appre¬
ciation of the geostatic objection being discussed and of Galileo’s critique of
it. In fact, Simplicio’s response is, “It does not look that way to me at all”
(D141), and then he says why; therefore it is important to know what ‘it’ is,
what Simplicio is denying. This must be the main or final conclusion in
Salviati’s argument, and if one gets it wrong (as one is bound to do from
Drake’s translation), then one does not understand what is going on; in
particular, Simplicio’s response would not make sense, and then one would
get the oversimplification that Aristotelianism made no sense, but that only
Galileo’s theories did; whereas it is more plausible to hold that the former
made some sense, but that Galileo’s madt greater sense.
Simplicio justifies his denial by saying, in Drake’s translation (which may
be accepted here):

If the element of fire goes around together with the air, this is a very easy and even a
necessary thing for a particle of fire, which, rising high from the earth, receives that very
motion in passing through the moving air, being so tenuous and light a body and so
easily moved. But it is quite incredible that a very heavy rock or a cannon ball which is
dropping without restraint should let itself be budged by the air or by anything else.
Besides which, there is the very appropriate experiment of the stone dropped from the
242 CHAPTER 10

top of the mast of a ship, which falls to the foot of the mast when the ship is standing
still, but falls as far from that same point when the ship is sailing as the ship is perceived
to have advanced during the time of the fall, this being several yards when the ship’s
course is rapid. [D141]

This would make no sense as a reply to Salviati’s thesis alleged in Drake’s

text, namely the proposition we have symbolized as “if Aristotle admits A,
then he should admit B”; for the last sentence about the ship experiment
sounds like Simplicio is just changing topic, while the first two sentences
seem to constitute a somewhat irrelevant counterargument of the form “if C
holds, then A follows but B does not”, where C is a statement of conditions
obtaining on a rotating earth. Such a counterargument does not lead to a
denial of Aristotle’s inconsistency in admitting A but not B, not only because
it says nothing about Aristotle, but also because the C is a proposition that he
obviously did not hold. However, in reality, such a counterargument does
plausibly lead to a denial of Salviati’s claim about the character of “Aristotle’s
defense”. Salviati had said that Aristotle’s defense seemed to consist in
holding the impossibility of mixed motion, toward and around the center.
Aristotle’s defense in turn was the following ‘defensive’ argument contained
in Sagredo and Simplicio’s speeches, just preceding Salviati’s:

SAGR. On behalf of Simplicio I should like, if possible, to defend Aristotle .... I reply
that if the tower were moving, it would be impossible for the rock to fall grazing it;
therefore from the scraping fall is inferred the stability of the earth.
SIMP. So it is. For to expect the rock to go grazing the tower if that were carried along
by the earth would be requiring the rock to have two natural motions; that is, a straight
one toward the center, and a circular one about the center, which is impossible.
[D140-141]

This may be called the apparent vertical fall argument, as distinct from the
actual vertical fall argument, which in the preceding dialogue has just been
shown to be circular. Salviati is showing how the apparent vertical fall argu¬
ment assumes the impossibility of mixed, down-and-around motion, and he is
arguing that this assumption is unjustified because Aristotle admits mixed up-
and-around motion, and if he admits this he should admit mixed, down-and-
around. Simplicio is implicitly agreeing with Salviati that the apparent vertical
fall argument assumes the impossibility of mixed, down-and-around motion,
and, explicitly, he is denying that this assumption is unjustified, for two
reasons: (1) that on a rotating earth the mixed, up-and-around motion of fire
would easily result, but the mixed, down-and-around motion of heavy bodies
would not, and (2) the ship experiment. If there is a problem with Simplicio’s
reply it is his failure in (1) to take seriously the idea (expressed in Salviati’s
 THE ERUDITION OF LOGIC 243

last sentence) that on a rotating earth heavy bodies would have a natural
tendency to follow the rotation, for Simplicio expects that the cannon ball
should be carried around forcefully by the rotating air. Regarding (2), Salviati
explicitly argues that there is a disanalogy between what might happen on a
ship and on a rotating earth (D141 —143).
To conclude, Drake’s above-mentioned inaccuracy is likely to lead one to
miss (a) the existence of the apparent vertical fall argument (as a distinct
objection to the earth’s motion), (b) the fact that one function of the ship
experiment is to be a part of this argument, and (c) Galileo’s argument for
the disanalogy between the ship and the earth.23

5. One of the objections to the earth’s motion was an argument based on

a ship experiment. In Drake’s translation, the argument says that

since when the ship stands still the rock falls to the foot of the mast, and when the ship
is in motion it falls apart from there, then conversely, from the falling of the rock at the
foot it is inferred that the ship stands still, and from its falling away it may be deduced
that the ship is moving. And since what happens on the ship must likewise happen on the
land, from the falling of the rock at the foot of the tower one necessarily infers the
immobility of the terrestrial globe. [D144; cf. F169-70]

Everything is acceptable here, except that the sentence “what happens on the
ship must likewise happen on the land” should read “what is true of the ship
must likewise happen of the Earth”.24 This is significant because it is clear
that the objection examined by Galileo is an argument from analogy, the two
analogues being the ship and the Earth; this fits very well with the critique
that follows, which amounts to the argument that, even if the analogy were
sound (cf. F167—9, D141—3) the objection is inconclusive because it is based
on the false premise that on a moving ship the rock falls away from the foot
of the mast (FI69—75, D143—9). On the other hand, as translated by Drake,
the objection seems to be a generalization from sea to land, and this may
mislead the reader into interpreting the ship experiment as a thought-experi¬
ment on a rotating earth, and Galileo’s counterargument as an attempted
proof of the earth’s motion to the effect that since the rock would fall at the
foot of the mast on a moving earth, it would also fall at the foot of the tower,
which is what happens, thereby confirming the earth’s motion.25

6. Between the end of the discussion of the falling bodies objection and the
beginning of the discussion of the gunshot arguments, there is a brief digres¬
sion about the flying of birds. Sagredo exclaims, “If only the flying of birds
gave me as much trouble as the difficulties caused by cannons and by all
 244 CHAPTER 10

other experiments mentioned above!” (FI93, my translation). In the rest of

his speech, and in Salviati’s response, the implication is that the objection
from birds is much stronger than all others. This emphasis tends to be lost in
Drake’s translation, where Sagredo’s exclamation reads, “If only the flying
of birds didn’t give me as much trouble as ... ” (D167— 8), which implies
that the flying of birds gives as much trouble as the other alleged counter¬
evidence.

7. In the discussion of the heliocentrism of planetary motions, Galileo

makes the following comment about certain people who give very silly argu¬
ments against the earth’s motion:

One need not take into account these people, whose number is infinite, nor notice their
fooleries; neither should we try to acquire as companions in very subtle and delicate
opinions, men in whose definition only the genus enters but the difference is missing.
[F355, my translation, my italics.)

Drake’s translation reads:

There is no need to bother about such men as these, whose name is legion, or to take
notice of their fooleries. Neither need we try to convert men who define by generalizing
and cannot make room for distinctions, just in order to have such fellows for our com¬
pany in very subtle and delicate doctrines. [D327, my italics]

The problematic phrase here is “uomini nella cui difinizione entra solo il
genere e manca la differenza”, which corresponds to the italicized expressions,
and which refers indirectly to the definition by genus and difference of man
as a rational animal, as we saw in our discussion of Strauss.

8. The next passage is Salviati’s disclaimer after the discussion of sunspots

in the Third Day. When translated literally the passage reads:

I will attribute to them neither the label of conclusive nor of inconclusive since, as 1 have
said other times, my intention was not to resolve anything about such a lofty matter, but
only to propose those natural and astronomical reasons which can be adduced by me for
the one and for the other position. I leave to others the determination, which ultimately
must not be ambiguous since, it being fitting for one of the two arrangements to be
necessarily true and for the other to be necessarily false, it is impossible that (staying
within the terms of human doctrines) the reasons adduced for the true side should not
manifest themselves as conclusive as the contrary ones vain and ineffective. [F383, my
translation]

In Drake’s translation, the last clause of the first sentence reads: “but merely
 THE ERUDITION OF LOGIC 245

to set forth those physical and astronomical reasons which the two sides can
give me to set forth” (D356). Moreover, he drops the qualification “neces¬
sarily in the first half of the second sentence, and the crucially important
comparison “as conclusive as... ” in the last clause:

I leave to others the decision, which ultimately should not be ambiguous, since one of
the arrangements must be true and the other false. Hence it is not possible within the
bounds of human learning that the reasons adopted by the right side should be anything
but clearly conclusive, and those opposed to them, vain and ineffective. [D356]

Finally, Drake misrepresents the structure of the claim in the second half of
the passage. Salviati is claiming primarily that ultimately the determination
must not be ambiguous; the reason he gives for this claim is that it is im¬
possible that the reasons adduced for the true side should not appear as con¬
clusive as the reasons for the false side appear inconclusive; and, in turn, the
reason for this impossibility is that it is fitting for one of the two arrange¬
ments to be necessarily true and the other necessarily false. In Drake’s trans¬
lation we find instead the respective truth and falsity of the two arrangements
as the direct reason for the ultimate unambiguousness of the decision, which
is less plausible, besides being inaccurate; while the impossibility in question
becomes the apparent final conclusion in the passage. In the light of the
book’s rhetorical and philosophical import, it is important to understand such
passages correctly. The present passage is particularly significant because it is
explicitly used in its mistranslated version by a well known scholar to reach
unwarranted interpretations.26

9. In his discussion of stellar dimensions one of the things that Galileo

shows is that the actual size of stars and their annual parallax required by
Copernicanism are much smaller than was believed by some of its opponents.
At one point he gives a rough calculation to show that the annual parallax
of a sixth magnitude star would be about the same as the solar parallax due
to the earth’s radius. Sagredo comments: “As a first step, this makes a big
drop” (F387, my translation). Salviati replies: “Indeed it does, since ...”
(F387, my translation); and then he continues his argument, arguing next
that the anti-Copemicans make such stars 10 million times bigger than
necessary. In Drake’s translation, Sagredo’s comment reads, “For a first step,
this is a bad fall” (D360), and Salviati’s response, “It is indeed wrong, since
. . . ” (Ibid.); this misrepresents a favorable judgment by Sagredo about
Salviati’s argument, as a negative remark.
246 CHAPTER 10

CONTEXT AND THE EPISTEMOLOGICAL CONTENT

OF REASONING: CLAVELIN

Maurice Clavelin’s Natural Philosophy of Galileo has been deservedly praised

ever since its original French edition in 1968. For example, Drake finds it
“a book that deserves the most careful reading in its entirety. Ultimately it
will prove to be as stimulating and as illuminating as the classic work of
Koyrd himself.”27 Comparing the relation between Koyre and Clavelin to
that between Galileo and Newton, Drake concludes that Clavelin has achieved
“a substantial advance. No longer, one may hope, will it be considered naive
or rash to disagree with Alexandre Koyre.”28 Howard Stein judges it to be
“by far the best account known to the reviewer of the oeuvre of Galileo; and
an examplar of a type of study in the history of thought that is, in the
reviewer’s opinion, both highly desirable and all too rare”.29 A. R. Hall calls
it “dominatingly the best study of its subject yet written”;30 characterized by
“wide range, depth, and subtlety .. . solidity and detail. . . Galilean scholars
will need to turn to this book again and again, and never without rich
rewards”.31 Michael Mahoney Finds that “if it is Galileo’s thought that one is
interested in, one can hardly find a better guide than Clavelin’s study”32 and
that it is “rewarding at every turn”.33 Even less favorable reviewers find it to
possess ‘obvious merits’34 and to be “a classical exposition de textes”.35
If, however, we examine carefully the evidential adequacy of Clavelin’s
interpretations dealing with logic and reasoning, we discover that they are
unsatisfactory, and that the root-cause of this deficiency is his practice of
interpreting passages out of context. This means that the practice is generally
widespread,36 since its problems have not been detected by his reviewers,
even those who wrote article-length studies. This in turn means that the
scholarship of logical practice needs fundamental reforms.
Clavelin’s interpretations are part of his account of Galileo’s philosophy of
science, which is alleged by him to have three elements: (1) a concept of
explanation as rational reconstruction of phenomena (C404-9);37 (2) an
identification of physical with mathematical intelligibility (C409-21); and
(3) the principle of simplicity (C421-4). Presumably (1) and (2) are part of
what Galileo means by explanation, whereas (3) is something like an a priori
necessary condition for the possibility of explanation in this sense. (1)
reminds us of the hypothetico-deductive model of explanation, (2) is the
mathematicism widely attributed to Galileo, and (3) is his supposedly well-
known commitment to simplicity.
The idea of explanation as rational reconstruction is interesting and
 THE ERUDITION OF LOGIC 247

intrinsically plausible, and in some sense it would be ‘nice’ if it was Galileo’s

idea. Unfortunately, the way in which Clavelin supports this point is in¬
adequate. He does so by referring to the discussion of the apparent motion of
sunspots in the Dialogue,38 Clavelin assumes that in this passage Galileo is
giving an explanation of this apparent motion (C404) and then proceeds to
determine the structure inherent in the explanation (C405-9). Galileo is
portrayed as first knowing the basic facts about the slanted and curved paths
of sunspots (C404), as being initially puzzled and ignorant of how and why
these paths result (C404), then as having put forth his explanatory hypothesis
that the sun’s axis of rotation is inclined to the ecliptic (C405—8), and finally
as having made and confirmed the prediction that the slant and the curvature
of the paths should alternate upwards and downwards at six-month intervals
(C408—9). In other words, Clavelin interprets this hypothesis as having the
function of (1) rendering comprehensible the previously known but not
understood fact that the path was sometimes slanted but straight and some¬
times curved but not slanted, and (2) predicting the annual period of the
direction of these paths. However, if one examines the relevant text in the
Dialogue, one finds that the hypothesis about the inclination of the sun’s
axis is presented by Galileo as having had the function of leading him to the
discovery of the details of the motion of sunspots, that is, the function not of
explaining known facts, but of leading him to previously unknown ones.
Galileo states that he originally thought the sun’s axis was perpendicular to
the ecliptic (F373, D345), but then, having accidentally observed only once
merely that a particular path was curved (F374, D346), he immediately got
the idea of axial inclination (F374, D347) which led him to observe further
and to test the reality of its consequences (F374—5, D347—8); then it so
happened that all the predictions were confirmed (F379, D352). I am not
saying that Galileo’s account in this passage is an accurate chronological
report; the chronological question is here irrelevant. What I am saying is that
Galileo is discussing logic of discovery and prediction, rather than logic of
explanation, as Clavelin thinks. So I conclude that the evidence presented by
Clavelin does not support the first element which he attributes to Galileo’s
philosophy of science, namely explanation as rational reconstruction with
predictive consequences, though this concept is intrinsically plausible and
though there may be other Galilean evidence in its favor. Finally, Clavelin is
to be commended for calling attention to the philosophical significance of
the passage, especially since a related nearby passage is used by him to
support another claim about Galileo’s philosophy of science, namely his
commitment to the principle of simplicity.
248 CHAPTER 10

The passage in question (F379—83, D352—6) is the one immediately

following the discussion of the details of the motion of sunspots and is used
by Clavelin to argue that, for Galileo, “the principle of simplicity became a
superior criterion of deciding between physical truth and falsehood” (C421).
Clavelin’s argument is that in this passage Galileo admits that a geostatic
explanation of the motion of sunspots is possible, but rejects it as being
inferior in simplicity since one would have to attribute to the sun two new
motions besides the annual and the diurnal, namely a monthly axial rotation
and an annual rotation of this axis of rotation. Now it is true that on one
occasion in this passage Galileo describes as ‘very simple’ (F382, D355) the
one motion attributed to the sun in the geokinetic theory; but he also
describes the geostatic explanation as ‘innovating’ (F382, D355) of solar
motions, and the geokinetic explanation as predictively fruitful (F379,
D352); and he calls attention to the inexplicability by the geostatic theory of
the annual period of the rotation of the sun’s axis, which it requires. More¬
over, he does not elaborate his simplicity remark in the way he does with the
others: the predictive fruitfulness of the geokinetic theory is supported by
the whole account of how he supposedly discovered the factual details
(F372—9, D345—52); the ‘innovating’ character of the geostatic explanation
is supported by the fact its statement emphasizes how the sun’s fourth
motion is somehow extra (F380—82, D353—5); and the geostatic inexplica¬
bility of the annual period of this fourth motion is a very decisive point in
favor of the greater explanatory power of the geokinetic explanation. Finally,
if we (the readers of this passage) try to elaborate on the simplicity remark,
the following would be the most plausible elaboration: using as clue Galileo’s
remark that the four ‘incongruous’ (F382, D355) motions attributed to the
sun in the geostatic explanation are reduced to “a single and very simple one”
(F382, D355) in the geokinetic theory, it seems that the latter is simpler
because in it there are two congruous motions by the earth and a single one
by the sun, whereas in the geostatic model there are four incongruous ones by
the sun. Now, since Galileo says nothing about the incongruity of these four
motions, we may conclude he is referring primarily to the opposite directions
of the sun’s annual and diurnal motions which he discusses elsewhere in the
Dialogue; it is not obvious that there is any additional incongruity for the
sun’s third and fourth motions, either between themselves, or between them
on the one hand and the annual and diurnal ones on the other. Hence the
comparison in terms of simplicity of the two explanations seems to involve
merely a comparison of the number of motions involved; that is, the
geokinetic explanation is simpler in the sense of requiring one motion less
 THE ERUDITION OF LOGIC 249

than the geostatic explanation. But if this is the meaning, then the point is
implicit in two of the other philosophically significant remarks, mentioned
above: to describe the geostatic explanation as ‘innovating’ of motions means
that it requires at least one extra motion; and to emphasize the geostatic
inexplicability of the annual period of the sun’s fourth motion calls attention,
among other things, to the fact that there is this extra motion in the geostatic
model. I conclude that Galileo’s simplicity remark in this passage is incidental,
because it is not elaborated in the text, and relatively superfluous, because its
information content is reducible to the other remarks made and elaborated
by Galileo. Hence it may be disregarded if we can give an interpretation of
the passage which takes into account the philosophical remarks that pre¬
dominate in it.
It is possible to give such an interpretation for the entire passage. It is the
following: Logically speaking, whenever certain observable facts are shown to
follow from a theory, this does not constitute a conclusive argument for the
truth of the theory, unless one shows that there is no other way of explaining
those facts (F379—80, D352—3); nevertheless, methodologically speaking,
given two different theories each of which explains the relevant facts, the
theory which has greater explanatory power and which is less ad hoc is the
better and more probable one. This makes sense of the scientific discussions
in the passage, as well as of all the philosophical remarks in it; for the whole
passage can be reconstructed as being an illustration of the philosophical
theses just formulated, which must therefore be attributed to Galileo him¬
self.39
The discussion about explanation and simplicity so far has shown
primarily that the passage under consideration does not support Clavelin’s
interpretation, but some other thesis about Galileo’s philosophy of science,
attributing to him a different philosphical claim; at the same time it may be
that Clavelin’s theses may be supported with other evidence. Next, in order to
justify further the incidental character of the simplicity remark in this
passage, and in order to refute another of Clavelin’s theses about simplicity
in Galileo, I shall discuss a passage where simplicity is a main topic. Clavelin’s
thesis is the following: “Nor did Galileo consider this appeal to the principle
of simplicity a mere expedient ... the greater simplicity of a theory must be
clear proof of its closer agreement with reality” (C423). The passage is the
one at the beginning of the Second Day in the Dialogue where Galileo states
a number of arguments favorable to the earth’s diurnal motion (FI39—50,
D266—83). The passage stresses the connection between simplicity and
probability, both by means of explicit philosophical remarks, and by the fact
250 CHAPTER 10

that the concrete scientific topics discussed can be reconstructed as illustra¬

tions of these philosophical theses. Note that this interpretation is of the
same type as the one given above for the passage on the motion of sunspots;
the similarity consists in the fact that the interpretation is formulated in such
a way as to do justice both to the philosophical and to the scientific content
of the passage, in such a way that we have an integration of both this double
aspect of the passage and also of all the main philosophical remarks made in
it. The reconstruction is the following: Arguments based on the principle of
simplicity are neither worthless nor conclusive, but rather probable; they are
not worthless because, in the absence of conclusive arguments, the simplest
idea is the most acceptable; they are not conclusive because a single piece of
counterevidence or conclusive counterargument is sufficient to refute an idea
based on simplicity; and they are probable because nature acts by means of
the fewest possible operations.40
Next let us look at the third element which Clavelin finds in Galileo’s
philosophy of science, his identification of physical with mathematical intel¬
ligibility. Clavelin’s evidence consists of “several arguments in the Dialogue,
all the more convincing in that they were not yet couched in the language of
geometry” (C412). The first is the “proof that if the earth did indeed spin
east [ward] on its axis every twenty-four hours, then a stone dropped from a
tower would not be deflected toward the west” (C412). This proof is recon¬
structed by Clavelin as follows: first Galileo shows that on a moving ship a
stone dropped from the top of the mast lands at the foot; next Galileo claims
that “the case of a tower carried along by the earth’s diurnal motion is in all
respects identical with that of the ship” (C413);hence he concludes that the
same thing will happen on the earth. Now, in this argument, according to
Clavelin, Galileo is identifying physical and mathematical intelligibility in the
sense that this physical argument and a mathematical proof “both involve the
reduction of a statement to one or several previously established propositions
with the help of successive identifications” (C413). This is a very weak sense
indeed; the interpretation is ingenious but has no basis in the text; perhaps
that is the reason why in his discussion (C412-13) Clavelin gives no references
to the Dialogue. I suggest, in fact, that we cannot find in this work the proof
constructed by Clavelin. The relevant passage (F164-75, D138-49) has three
parts and is a critique of three different versions of the falling bodies
objection to the earth’s motion. The first version says that the earth can’t
move because bodies fall vertically; Galileo shows that this argument is
circular because we can know that vertical fall is actual (as distinct from
merely apparent) if and only if we know that the earth stands still (D13 8-40).
 THE ERUDITION OF LOGIC 251

The second version grounds the earth’s immobility on the undisputed fact
that bodies appear to fall vertically; Galileo argues that this objection depends
on the impossibility of mixed motion, namely motion around and straight-
toward the center, which is invalidly justified by the alleged fact that bodies
dropped from the mast of a moving ship fall behind; he then argues that the
justification is invalid because the motion of the ship is violent, whereas the
earth s diurnal motion would be natural, and because the relevant portion of
the air does not move along with the ship but would do so near the earth’s
surface if the earth moved (F167—9, D140—43). The third version is an argu¬
ment from analogy, comparing a tower on the earth with a mast on a ship,
and inferring that just as the rock falls behind when dropped on the moving
ship, so it would fall behind on a rotating earth; Galileo criticizes this argu¬
ment by arguing that it is based on the false premise that the rock falls behind
when dropped on a moving ship (F169—75, D143—9).
It is another question, of course, whether the proof constructed by
Clavelin, though not present in the text, is suggested by it. However, even this
is doubtful because Galileo’s critique of the second version of the objection,
the argument from apparent vertical fall, amounts to a criticism of the
soundness of the analogy between the rotating earth and the moving ship.
This would prevent one from using in the proof the proposition (crucial in
Clavelin’s construction) that “the case of a tower carried along by the earth’s
diurnal motion is in all respects identical with that of the ship” (C413).
Once again, however, I wish to make my own criticism more than merely
negative. In fact, I can give a positive philosophical interpretation of the
passages that Clavelin is talking about. These interpretations, like the ones
I have given above, take into account both the various philosophical remarks
present in them and the concrete scientific discussions in the context of
which they occur. Following this approach, Galileo’s discussion of the ship
analogy argument (F169—80, D143—54) is an illustration of the following
thesis: Experiments are sometimes unnecessary to ascertain the results of a
test, for sometimes it can be argued on the basis of known or more easily
ascertainable facts, what these results must be. The details of Galileo’s discus¬
sion support this interpretation.41
The other relevant passage, namely Galileo’s critique of the earth-ship
analogy (F167—9, D141—3), is part of a longer passage (F159—69, D133—43)
which, when analyzed in the same way, can be shown to be illustrating the
value and usefulness of logical analysis.42
Let us now look at a second piece of evidence given by Clavelin to support
his claim that Galileo’s philosophy of science identifies physical and mathe-
252 CHAPTER 10

matical intelligibility. It is the passage (FI29—30, D103) at the end of the

First Day in the Dialogue where Galileo discusses the powers and limitations
of human understanding. Clavelin thinks it shows that Galileo “presented
mathematics as the most perfect knowledge to which man can aspire —
knowledge so perfect, in fact, that it can be compared intensively if not
extensively with divine understanding” (C414). However, Gavelin’s argument
looks plausible only because he neglects to quote the part of the passage
where Galileo indicates one important qualitative difference between divine
understanding and human understanding even for the case of mathematical
reason, namely God knows the infinitely many mathematical truths he knows
without using a step-by-step reasoning process. In other words, though it is
true that Galileo is making a philosophical point in this passage, this point is
much more complex than even Clavelin realizes. When the specific passage
mentioned by him is examined in context (F124—31, D98—105), and when
the examination is carried out, as before, by integrating all philosophical
remarks with one another and the philosophical theory with the scientific
practice, we get a thesis to the effect that the powers of human understanding
are likely to disappoint both the optimists and the pessimists.43
Clavelin ends his discussion of Galileo’s identification of physical and
mathematical intelligibility by stating that “Galileo reversed the traditional
roles of logic and mathematics” (C416). That is, Galileo gives logic the
function of analyzing given arguments and mathematics the function of
constructing new ones. Though the relevance of this alleged fact is not readily
apparent, let us disregard this problem and consider some of the evidence that
leads Clavelin to allege it as a fact. Clavelin quotes Galileo as saying that the
art of proof is acquired by the “reading of books filled with demonstrations,
and these are exclusively mathematical works, not logical ones” (C417,
quoted from F60). Once more, by taking the remark out of context, Clavelin
misses its main point, which is that the art of proof is acquired by reading
books which contain actual proofs, rather than books which contain theories
of proof. For in the context in which the remark is made (F59—60, D35),
Galileo is discussing how logical theory and logical practice are distinct; the
meaning of the phrase ‘logical books’ in the sentence quoted by Clavelin is
that of ‘books in the theory of reasoning’; and the meaning of the term
‘demonstrations’ is that of ‘logical practice’ or ‘actual reasoning’. Finally,
Galileo’s equation of logical practice exclusively with mathematics must be
regarded as merely a rhetorical excess by Galileo, since it does not tie in at all
with the rest of the passage in which it occurs. When this entire passage
(F59—71, D35—47) is analyzed by the same technique used before, we get
 THE ERUDITION OF LOGIC 253

the following results: The logical authority of Aristotle is not sacrosanct and
at any rate pertains primarily to the domain of logical theory rather than
logical practice, his authority as a logical practitioner cannot be grounded
on his undisputed authority as a logical theorist, but only on the soundness
of his concrete reasoning; the latter can only be tested by the logical analysis
of his actual arguments, which analysis one should therefore be free to carry
out; if you do this you will discover that his logical practice leaves much to be
desired.44
To conclude, our analysis of some concrete problems in the study of the
historical record of reasoning suggests that more than mere erudition and
scholarship is needed when the subject matter is the kind of logical practice
found in Galileo’s Dialogue; thus the logical theorist must become his own
scholar.

NOTES

1 In my History of Science as Explanation.

2 G. Galilei, Dialog Uber Die Bieden Hauptsachlichsten Weltsysteme, tr. Emil Strauss.
3 G. Galilei, Opere, edited by Favaro, Vol. 7.
4 G. Galilei, Dialogue Concerning the Two Chief World Systems, tr. S. Drake.
5 This is a translation of Clavelin’s La philosophic naturelle de Galitte.
6 Galilei, Dialogue, tr. Drake, p. xxvi.
Galilei, edited by Favaro, Opere 7, 59—60. Subsequent references to this book will be
made only in parenthesis in the text, by prefixing the page number(s) with an ‘F’, as a
reminder that Favaro’s edition is meant.
8 Galilei, Dialog, tr. Strauss, p. 502, note 42. Subsequent references to this book will be
made only in parenthesis in the text, by prefixing the page number(s) with an ‘S’, as a
reminder that Strauss’s edition is meant.
9 G. Galilei, Dialogue on the Great World Systems, tr. Salusbury-Santillana, p. 138, note
22.
10 For more details see Chapters 8 and 12.
11 Galilei, Dialogue, tr. Drake, p. 372.
12 Ibid.
13 Galilei, edited by Favaro, Opere 6, 316, 317, 319, and 321, translated in The Con¬
troversy on the Comets of 1618, edited and tr. S. Drake and C. D. O’Malley, pp. 276,
279, and 280; also Galilei, Opere 8, 105—6, and 5, 351—63. Cf. my “The Concept of Ad
Hominem Argument in Galileo and Locke’.
14 G. Galilei, Dialogo di Galileo Galilei Linceo . . . , p. 24. Subsequent references to this
book will be made only in parenthesis in the text, by prefixing the page numbers with
‘EP’, as a reminder that the editio princeps is meant.
15 This notation refers to line 26 on page F212; similarly for subsequent such references.
16 Besides the Dialogue and the Controversy on the Comets already cited, we have: Two
New Sciences, Discoveries and Opinions of Galileo, Galileo Against the Philosophers.
254 CHAPTER 10

Other Galilean works can be found in parts of the following volumes, translated by
Drake in conjunction with other scholars: On Motion and on Mechanics, and Mechanics
in Sixteenth Century Italy. Finally, his Galileo at Work contains translations of many
important letters, papers, and documents.
17 Most of his interpretations up to 1970 are contained in his Galileo Studies.
18 Besides the work already cited, see for example, ‘The Evolution of De Motu', ‘Galileo
and the First Mechanical Computing Device’, ‘Galileo’s “Platonic” Cosmogony and
Kepler’s Prodromus', ‘Impetus Theory and Quanta of Speed Before and After Galileo ,
and ‘The Uniform Speed Equivalent to a Uniformly Accelerated Motion from Rest’.
19 See my ‘Vires Acquirit Eundo: The Passage Where Galileo Renounces Space-Acceler¬
ation and Causal Investigation’, my ‘Galileo’s Space-Proportionality Argument: A Role
for Logic in Historiography’, and Drake’s ‘Velocity and Eudoxian Proportion Theory’.
20 Subsequent references to G. Galilei, Dialogue, tr. Drake, will be given only in
parenthesis in the text, by prefixing the page number(s) with a ‘D’, as a reminder that
Drake’s translation is meant.
21 See Chapter 3, dealing with the literary aspects of Galileo’s book. It should be
mentioned here that the purpose of Drake’s translation was to popularize the Dialogue,
and that it was with deliberation that he chose “reasonably easy reading in preference to
strict literalness, even at the price of taking certain liberties with the text” (D, p. XXV),
so that my criticism is aimed primarily at scholars who make uncritical use of Drake’s
translation. On the other hand I am not sure that the problems mentioned below are
above the layman’s head, and so I am inclined to think they could have been avoided
without falling into pedantry. It should also be noted that Santillana’s revision of
Salusbury’s translation is no substitute for Drake’s translation, primarily because the
Santillana edition is incomplete. See Appendix, note 3 for details.
22 Cf. the glossary in L. Sosio’s edition of the Dialogue: G. Galilei, Dialogo sopra i due
massimi sistemi, edited by Sosio, p. 574. See also the Vocabolario Illustrato della Lingua
Italiana 2, 1541. Santillana seems to commit the same oversight as Drake: cf. Galilei,
Dialogue, edited by Santillana, pp. 141, 183.
23 Additional support for my interpretation is the use of the subjunctive mood in
Simplicio’s reply (F167) to Salviati, within the sentence that I have structured as being
of the form “if C holds, then A follows but B does not” (quando Velemento del fuoco
vadia in giro insieme con I’aria . ..).
24 “quello che occorre della nave deve parimente accadere della Terra” (FI70).
25 Feyerabend is a notable example of such a misinterpretation; cf. his Against Method,
pp. 69-98, esp. pp. 83-87. Some of his critics disagree with the philosophical conse¬
quences to be drawn from his analysis but seem to share his misinterpretation, though
they ought to know better; for example, P. K. Machamer (‘Feyerabend and Galileo: The
Interaction of Theories and the Reinterpretation of Experience’, pp. 27-28), in
criticizing Feyerabend on this point, states the geostatic objection analyzed by him as
the argument that “if the earth were moving, then when a stone is dropped from the top
of a mast (on a ship) it would fall behind the mast (because the earth will have carried
the ship beyond the point where the stone was dropped)”. Feyerabend’s interpretation
also conflicts with Galileo’s explicit clarification at the end of his critique that “I have
not claimed to prove it yet, but only to show that nothing can be deduced from the
experiments offered by its adversaries as one argument for its motionlessness” (D154).
26 See E. McMullin, ‘Introduction: Galileo, Man of Science’, pp. 31 — 32.
 THE ERUDITION OF LOGIC 255

S. Drake, Review of La philosophic naturelle de Galitee p 277

28 Ibid.
” H. Stein, ‘Maurice Clavelin on Galileo’s Natural Philosophy’, p. 375.
A. R. Hall, Essay review of La philosophic naturelle de Galilee p 80
31 Ibid., p. 84.
32 M. S. Mahoney, ‘Galileo’s Thought’, p. 944.
33 Ibid., p. 945.
34 W. R. Shea, Review of La philosophic naturelle de Galilee, p. 125.
5 J. L. Heilbron, Review of La philosophic naturelle de Galilie, p. 342.
36 See, for example, W. Wisan, ‘Galileo’s Scientific Method: A Reexamination’; cf. my
review of the collection where this paper appears, New Perspectives on Galileo, edited by
R. E. Butts and J. C. Pitt. For the sake of perspective, it must be said that Wisan’s work
represents the best and latest study of the mathematical aspects of Galileo.
37 M. Clavelin, The Natural Philosophy of Galileo, pp. 404-409. Subsequent references
to this book will be given in the text in parenthesis, by prefixing the page number(s) by
, as a reminder that Clavelin’s book is meant.
Galilei, edited by Favaro, Opere 7, 372-79; idem, Dialogue, tr. Drake, pp. 345-52.
As before page references to these books will be denoted respectively by ‘F’ and by ‘D’.
For more details, see Chapter 5, section dealing with passage F372-83.
40 For more details, see Chapter 5, section dealing with passage FI39-50.
For more details, see Chapter 5, section dealing with passage F169-80.
For more details, see Chapter 5, section dealing with passage F159-69.
43 For more details, see Chapter 5, section dealing with passage F124-31.
44 For more details, see Chapter 5, section dealing with passage F59-71.
 CHAPTER 11

THE PSYCHOLOGY OF LOGIC AND

THE LOGIC OF PSYCHOLOGY:
CRITIQUE OF THE PSYCHOLOGY OF REASONING

The argument so far has been that Galileo’s Dialogue has classic significance
for science, and important implications for the philosophy of science, for the
historiography of science, and for the history and historiography of philo¬
sophy. It has also emerged that reasoning is the book’s central feature, both
at the level of scientific practice, and at the level of philosophical reflection.
It follows that reasoning, which of course is not to be confused with rational¬
ism, is an essential feature of scientifc rationality. However, though we have
thereby exhibited the structure of scientific rationality, we have not yet
examined the structure of this structure. That is, so far we have examined the
macro-structure of scientific rationality, but not its micro-structure. In order
to do the latter, we need to study the finer details of the reasoning in
Galileo’s book. There is, however, an additional independent reason why it is
desirable to undertake this study, namely that it provides materials for and
examples of a concrete approach to the theory of reasoning, and to the
foundations of a reformed science of logic. This I now proceed to show.
In recent years we have seen the emergence of new expressions of the old
critiques of formal logic. Some have questioned its generality, claiming that it
represents basically a theory of mathematical reasoning and completely
neglects the very wide and important area of so-called rhetorical or dialectical
reasoning which includes most of the arguments found in law, ethics, philo¬
sophy, the human sciences, and everyday life.1 Others have questioned its
empirical import, claiming that the empirical study of the psychology of
reasoning shows that formal logic is inaccurate as a description of human
reasoning, and misleading and useless for its theoretical explanation.2 Still
others have objected to its artificiality (or formalistic character), claiming that
it pays very little attention to informal argumentation in natural language,
the study of which was and ought to be its raison d’etre.3 There are some
who question its usefulness, arguing that the rigor and precision of its con¬
cepts present an intrinsic limitation for its practical application to ordinary
logical problems, which need to be encoded before and decoded after the
formal apparatus is used; and unfortunately the encoding step is usually as
debatable as the original problem was when discussed in its own terms.4
Finally there are those who object to the abstractness of formal logic, claiming

256
 THE PSYCHOLOGY OF LOGIC 257

that what is needed is not more theorizing but more analysis, not the con¬
struction of other abstract schemas, but the critical understanding of concrete
instances of reasoning.5 It is, of course, easy to defend logic from such
criticism by regarding it as a branch of pure mathematics; however, I know
of no philosophical logician who is willing to accept the full consequences
and practice logic as mathematicians do; the pretension always lingers that he
is dealing with arguments and reasoning rather than with purely abstract
entities (such as truth functions and sets).6 At any rate, just as there is a place
for logic practiced as mathematics (though it is up to mathematicians to
determine it), so there ought to be a place for the kind of general, empirical,
informal (natural), useful, and critical-concrete logic suggested by the above
mentioned criticism. However, qualifications and elaborations need to be
made to such critiques in order to get a better idea of what this type of
investigation should be like and a better understanding of the relation
between actual reasoning and formal logic. This is especially true since none
of the positive results of the scholars from whom such criticism has been
taken combine all the mentioned virtues.
Let us explore the sense in which a logic of reasoning could be empirical
by examining some recent work in the psychology of reasoning. Some of the
general questions to keep in mind are whether ‘empirical’ here ought to mean
experimental or historical, and whether introspective experience is an effec¬
tive empirical guide in this case.
The most extensive recent investigations into the psychology of reasoning
are those of Wason and Johnson-Laird.7 Three main conclusions seem to
emerge from their studies. First they argue that the propositional calculus
does not adequately characterize propositional reasoning because that
calculus is two-valued, truth-functional, and formal in the sense of ignoring
the specific content of propositions (p. 93). And they present evidence that
the content of a proposition has a significant effect on how people interpret
its logical form (pp. 54-85), that people do not attach a material conditional
meaning to the ‘if then’ (pp. 54-65 and 86-96), and that many statements
in everyday language presuppose a state of affairs which when unfulfilled
renders the statement neither true nor false but irrelevant or inapplicable
(pp. 90-1); as regards this last point they investigate in great detail the role
of negations and show that “a denial generally (not invariably) functions in
language to correct the preconception which it denies” (p. 30), so that “the
affirmative preconception has to be recovered before the meaning of the
negative can be grasped” (p. 39), and in turn “when a negative is itself denied
by an affirmative, it becomes difficult to keep track of the argument” (p. 53),
258 CHAPTER 11

that is, one finds people not reasoning in accordance with what one would
expect from formal logic.
Second, Wason and Johnson-Laird exhibit the empirical inadequacy of
predicate calculus by presenting evidence concerning the interpretation of
universal affirmative and of particular propositions. Concerning the former
there is a tendency to treat the subject term of ‘All S are P' as the whole
universe of discourse, which would justify its conversion to ‘All P are S1’; and
there is also a tendency to interpret ‘Some S are P' as equivalent to ‘Some S
are not P' (pp. 157—8).
These conclusions seem well founded, and the logician with a sensitivity
for actual reasoning will feel that they correspond to what he has known all
along. In fact, Wason and Johnson-Laird occasionally are explicit in expressing
their debt to a philosopher like Strawson.8 Strawson’s method of ordinary
language analysis may be regarded as the introspective psychology counter¬
part of Wason and Johnson-Laird’s experimental approach, hence as a type
of a fundamentally empirical approach, and a proper one at that.
Wason and Johnson-Laird’s most original work is perhaps that dealing with
the testing of general hypotheses. It is here that emerge the limitations of their
approach as well as a point in favor of formal logic. The main conclusion they
reach in these investigations is one about the effect of content on reasoning,
namely that the testing of a general hypothesis dealing with a real life situa¬
tion is in accordance with what one would expect from logic, whereas when
the hypothesis deals with abstract material the subjects’ performance diverges
from the supposedly logical one (p. 193). This conclusion will be shown
below to be based on an inadequate analysis of their own evidence; however,
it should be pointed out first that the authors seem to be right in their general
conclusion about the effects of content on the way people reason, namely
that the nature of the subject matter (concrete or abstract) affects logical
performance, sometimes facilitating it, sometimes hindering it.
Let us look at their experimental results. The first group of experiments
involves the testing of the hypothesis “If a card has a vowel on one side, then
it has an even number on the other side” (p. 173) by reference to four cards
whose visible, up sides show respectively the symbols E, K, 4, and 7; the
subjects are informed that each card has a letter on one of its sides and a
number on its other side, and their task is to name all and only those cards
which need to be turned over in order to determine whether the hypothesis
is true or false (p. 173); for a total of 128 subjects in the course of four
experiments the choices are as follows (p. 182):
 THE PSYCHOLOGY OF LOGIC 259

Cards 1 and 3: 59 subjects

Card 1: 42 subjects
Cards 1,3, and 4: 9 subjects
Cards 1 and 4: 5 subjects
Other combinations: 13 subjects

Since the right answer is ‘Cards 1 and 4’, this means that only 5 out of 128
subjects got the right answer.
The second experiment relevant to our discussion (pp. 190-1) involves
the testing of the hypothesis “Every time I go to Manchester I travel by
train”. The four cards show respectively the words Manchester, Leeds, Train,
and Car. The subjects are informed that each card has the name of a city on
one side and the name of a means of transportation on the other; moreover,
they are asked to imagine that each card represents a journey made by the
speaker and that the hypothesis represents a generalization about these
journeys. The task was once again to select all and only those cards that
needed to be turned over to check whether the generalization was true or
false. Ten out of sixteen subjects made the correct selection of cards 1 and 4
(as contrasted to two out of sixteen in a control group tested with abstract
material).
The third relevant experiment (pp. 191 — 192) involves both the (‘con¬
crete’) hypothesis “If a letter is sealed, then it has a 5d stamp on it” and
the (‘abstract’) hypothesis “If a letter has a D on one side, then it has a 5 on
the other side”. Instead of cards, two sets of four envelopes were used; one
(‘concrete’) set was arranged as follows: the back of a sealed envelope, the
back of an unsealed envelope, the front of an envelope with an address and a
5d stamp on it, and the front of an envelope with an address and a 4d stamp
on it; the other (‘abstract’) set of envelopes was arranged as follows: the front
of an envelope with a D written in the middle of it, the front of an envelope
with a C written in the middle of it, the back of an envelope with a 5 written
in the middle, and the back of an envelope with a 4 on it. There were two
tasks to be performed: testing of the abstract hypothesis by reference to the
abstract set of envelopes, and testing of the concrete hypothesis by reference
to the concrete set of envelopes. (For testing of the concrete hypothesis
subjects were told to imagine that they were Post Office workers engaged in
sorting letters.) There were two groups of subjects, and each group performed
both tasks in the opposite order. The results were that 21 out of 24 subjects
were correct in the concrete task, but only 2 out of 24 in the abstract task.
The difference in logical performance is certainly striking, but the real
260 CHAPTER 11

problem is to account for it. The explanation favored by the authors is the
following (pp. 182-187). First they interpret the logical structure of the
hypothesis being tested as that of the conditional If p, then q , p would
refer, respectively, to the propositions that a card has a vowel on one side, or
that I go to Manchester, or that a letter is sealed, or that a letter has a D on
one side; and q would refer, respectively, to the proposition that the card has
an even number on the other side, or that I travel by train, or that the letter
has a 5d stamp on it, or that the letter has a 5 on the other side. Second, in the
light of this interpretation, they refer to card 1 as the p card, card 2 as the
not-p card, card 3 as the q card, and card 4 as the not-p card. Third, their
reconstruction of the subjects’ reasoning is as follows:

It assumes that the subject will initially focus only on items mentioned in the rule. From
this hypothetical ‘list’, only cards which could verify the rule will be selected, as a
function of whether he assumes that the converse holds (selection of p and q), or does
not hold (selection of just p). Two levels of insight have been retained, but unlike those
in the preliminary model, they are no longer independent. Partial insight consists in
re ah zing that all cards should be tested, and that those which could verify, and those
which could falsify, should be selected, i.e., p, q and not-<7. Even if q had not been
selected initially, it will now be selected because it could verify the rule. Complete
insight consists in realising that only cards which could falsify should be selected, i.e. p
and not-g. [P. 185]

Fourth, the greater insight into the concrete task is attributed to the coin¬
cidence in it of a logical and of a causal analysis of the situation, the
assumption being that humans have a tendency toward causal analysis of
problems (p. 193).
It is unnecessary to examine in detail the arguments and evidence they give
in support of this explanation. One reason is that their reconstruction of the
subjects’ reasoning is somewhat unclear. For example, it is not clear how and
why the subjects with partial insight chose the q card (#3), whether it is
because they think that a p on the down side would verify the hypothesis
‘if p then q\ or whether they think that this discovery would verify its con¬
verse ‘if q then p' which they regard as equivalent to the hypothesis itself. If
the former then the question would arise how any subject could choose only
p, that is, why the subjects who choose only p do not regard that they would
have a verification if the third (q) card had p on the other side; if the latter
then how can the subjects who chose p, q, and not-<? fail to choose the second
card (not-p) which would yield a falsification if q were on the other side
(since this would falsify ‘if q then p’, and hence its converse allegedly
equivalent to it).
 THE PSYCHOLOGY OF LOGIC 261

However, the main reason for the inadequacy of the authors’ explanation
is their analysis of the hypotheses being tested as conditionals of the pro-
positional calculus, whereas they are universal generalizations of conditionals
involving quantifiers. This initial oversimplification in turn leads them to
attribute equal complexity to the concrete as to the abstract propositions,
whereas the latter are much more complex than the former. I now turn to
showing this.
The hypothesis in the first group of experiments reads: “If a card has a
vowel on one side, then it has an even number on the other side”. This should
be symbolized as follows:

(x) (y) (z) (Cx & Pxy & Syx &y^z & Szx —> Qxz)

where

Cx: x is a card;
Pxy: x has a vowel on side y;
Syx: y is a side of card x;
Qxz: x has an even number on side z

The generalization refers to four cards and eight sides. Let us use the following
individual constants and denote the given cards and their sides:

ci : card #1, with an ‘E’ on its up side;

c2 : card #2, with a‘K’on its up side;
c3 : card #3, with a‘4’on its up side;
c4 : card #4, with a ‘7’ on its up side;
Si i: ‘up’ side of card 1, showing ‘E’;
s2i: ‘down’ side of card 1;
s12 : ‘up’ side of card 2, showing ‘K’,
s22 : ‘down’side of card 2;
Si3: ‘up’side of card 3, showing‘4’;
s23: ‘down’side of card 3;
s14: ‘up’side of card 4, showing‘7’;
s24 : ‘down’ side of card 4.

The generalization is true if and only if it is true in each of the eight cases
to which it refers, that is if and only if all of the following sentences are
true.
262 CHAPTER 11

0\ Cct &Pcxsn & Ss a C\ & s 11 ^S 21 & Ss21c i ~^QciS2i ;

o2 Cc2 & Pc2sx2 & Ss 12 c2 & s 12^22 & Ss22c2 * Qc2S22 ,

03 Cc3 &Pc3S 13 & Ssx3c3 & s 13^s23 & Ss23c3 > Qc3S23 ,

04 Cc4 & Pc4sx4 & Ss i4 c4 & ^ 14 ~Ps 24 & Ss24c4 * Qc4S24 ,

o5 Cc! &Pcxs2l & Ss21 Cl & ^ 21 ^PS 11 & Ssncx —► Qc\ hi;
Oe Cc2 & Pc2s22 & Ss 22 c2 & S22^S12 & Ss \2c2 * Qc2S\2,

On Cc3 & Pc3S23 & Ss23 Cs & S23^S 13 & Ssx3c3 —* Qc2Sl3\
On Cc4 & Pc 4 S 24 & Ss 24 c4 & S24^S14 & Ss x4c4 -+Qc4sx4.

In each case the first, third, fourth, and fifth conjuncts of the antecedent are
true (because the symbols have been so chosen as to insure their truth). So
the truth value of each sentence depends on the truth values of the second
conjunct of the antecedent (which asserts that a given card has a vowel on a
given side), and on the truth value of the consequent (which asserts that the
corresponding card has an even number on the other side).
Now, for Oi, Pcxsu is true since card 1 does have a vowel on its up side,
but the truth value of Qcxs2l is unknown since we don’t know what number
is on the down side of card 1. Hence we need to turn over card 1 in order to
determine the truth value of ox.
For o2, Pc2sn is false since card 2 has a consonant on its up side. Hence,
without looking at the consequent of o2, we can conclude that o2 is true.
For o3, Pc3s13 is false since card 3 does not have a vowel on its up side;
hence o3 is true.
For o4, Pc4s14 is false since card 4 does not have a vowel on its up side
either, hence o4 is true.
For o5, Pcxs2l is false since we have been informed that every card has a
letter on one side and a number on the other, and hence the down side of
card 1 does not have a vowel; hence as is automatically true, by falsity of
antecedent.
For cr6, the same applies as for o5; so o6 is true.
For a7, its consequent Qc3sl3 is true because card 3 has an even number
on its up side; hence, regardless of the truth value of Pc3s23, a7 is true.
For ct8, its consequent Qc4sx4 is false because card 4 does not have an
even number on its up side; hence a8 would be false \iPc4s24 were true;but
we do not know the truth value of Pc4s24 since we don’t know what is the
letter on the down side of card 4; hence we need to turn over card 4 to deter¬
mine the truth value of a8.
To summarize, the generalization is demonstrably true in every relevant
case except the first and the eighth, which involve cards 1 and 4 respectively.
 THE PSYCHOLOGY OF LOGIC 263

These are all and the only cards that need to be turned over for an exhaustive
test.
This analysis shows that the answer which Wason and Johnson-Laird
regard as correct is indeed one in accordance with a formal-logical inter¬
pretation of the hypothesis in question. The analysis can be simplified some¬
what by considering a more abstract symbolization of the hypothesis; such a
simplified analysis will be closer to the one given by Wason and Johnson-
Laird, without, however, diluting away the quantificational aspect of the
problem, as they do. The hypothesis in question could also be symbolized as
follows:

O) 0) [Dxy —* (Px —* Qy)],

where

Dxy. x mdy are different sides of the same card;

Px: side x has a vowel on it;
Qy: side y has an even number on it.

If the individual constants denoting the sides of the cards are the same as
before (snm), then we would have again eight cases corresponding to the
following ordered pairs as abstract entities, where the truth-values have been
written under the relevant portions of the formulas:

(1) <s„,s21>: Dsns21->(ftn ->Qs2j);

T 1 Til

(2) <s12>522>: Ds\2s22 (PS 12 y Qs22)',

T T F T 1

(3) <SI3,S23>: Ds13s23—>(Ps 13—>Qs23);

T T F T F

(4) <s14, s24>: Ds14s24 —*(Psi4 —>2s24);

T T F T F

(5) <s21,Sn>: Ds2\Sn—*(Ps2i—yQsn)\

T T F T F

(6) <522.si2>: Ds22S\2 * (Ps 22

T T F T F

(7) <s23,s13>: 7)s23S13 * (Ps23 ^2^13);

T T ITT
264 CHAPTER 11

(8) <S24. Sl4 >: DS24S14-24->Qsh)-

T 9 ? ? f

Once again all the truth values are determinable except for the case of the
ordered pair of the up and down sides of the first card and the case of the
ordered pair of the down and up sides of the fourth card; so that cards 1 and
4 need to be turned over. From the point of view of formal logic it is
immaterial whether the denotations of the predicate symbols D, P, and Q are
those just given, which correspond to the first group of experiments, or
whether the following interpretations hold, which would correspond to the
‘abstract’ case in the third group of experiments mentioned above:

Dxy: x andy are sides of the same envelope;

Px: side x has a Z);
Qy: sideyhasa5.

Hence the structure of the ‘abstract’ hypothesis of the third group of experi¬
ments is identical to that of the hypothesis in the first group, and we would
expect analogous experimental results.
However the concrete hypotheses involve simpler formulas. In a second
group of experiments the generalization was:
“Every time I go to Manchester, I travel by train”.
This can be symbolized as follows:

(x) (Mx —* Tx),

where
Mx: x is a journey to Manchester;
Tx: x is a journey by train.

This is the more obvious and natural interpretation, even though a more com¬
plex one could be given, namely

(x) 00 [Jxy —* (Mx —* Ty)\,

where

Jxy : x is the destination in a journey where the means of transportation is

y\
Mx: the destination of journey x is Manchester;
Ty: the means of transportation of journey y is train.

In the simpler interpretation the universe of discourse is the set of journeys,

whereas in the more complex interpretation the universe of discourse is the
 THE PSYCHOLOGY OF LOGIC 265

union of the set of journey destinations and the set of journey means of
transportation. Because of the simpler interpretation, we may ignore the
more complex one.
Could the structure of the generalization in the abstract experiments be
simplified further than it was done above? Our simplified analysis resulted in
a formula which was a doubly universally quantified generalization of a con¬
ditional whose antecedent was an atomic formula with a binary predicate
and whose consequent was itself a conditional consisting of two atomic
monadic formulas:

00 0) [Dxy —> (.Px-* Qy)\.

Could we reduce this to something like

00 (Vx —* Ex),
where
Vx: x is a card with a vowel on it;
Ex : x is a card with an even number on it?

This would not have the intended meaning because this sentence would be
true for cards which had an even number and a vowel on one side and nothing
on the other, whereas the generalization being tested would be false in this
case. I do not think that the structure of this generalization can be simplified
any further. We do need a universe of discourse of sides of cards, and we need
to consider such sides in ordered pairs.
Let us now look at the concrete generalization in the third set of experi¬
ments: “If a letter is sealed, then it has a 5d stamp on it”, which can be
symbolized as

(x)(Sx->Fx),
where
Sx: x is a sealed letter;
Ex: x is a letter with a 5d stamp on it.

The universe of discourse is simply the set of letters.

Thus both of the concrete experiments involve universal generalization
with a single quantifier and with monadic predicates only:

(x) (Px-* Qx).

In either experiment we have only four cases, in each of which only one
individual constant is involved:
266 CHAPTER 11

JOURNEY EXPERIMENT

j\ : journey described by the first card, namely to Manchester;

j2 ; journey described by the second card, namely to Leeds;
/3; journey described by the third card, namely by train;
/4 : journey described by the fourth card, namely by car.

The four sentences involved are the following, where the truth values have
been written under the appropriate portion of the sentence:

(1) pj\ —»Q]\ ; (2) Ph —► Qi2; (3) m —► Qiz ; (4) Pj4 —> Qi4 •
Til F T 1 ITT 1 1 F

POST OFFICE EXPERIMENT

J?!: first letter, sealed;

£2 '• second letter, unsealed;
£3: third letter, properly stamped;
£4 : fourth letter, improperly stamped.

(l) i>£x —>Q^ ;(2)F£2 —*Q£2 ;(3)^3 —>2^ ;(4)P£4 —>Qh-

Til F T 1 ITT 1 1 F

In both experiments items 1 and 4 need to be turned over.

The differences in complexity between the concrete and the abstract
hypotheses may be summarized as follows.8 In the concrete case, the universe
of discourse is a set referring to a single relatively natural class of objects; in
the abstract case the universe is a set of relatively abstract entities, like sides
of cards. Second, in the concrete case only monadic predicates occur, in the
abstract there occur binary as well as monadic predicates. Finally, in the
concrete case we have only one universal quantifier; in the abstract case we
have two. Because of these differences, the possibility emerges that the great
difference in logical performance by the experimental subjects could be
accounted thereby. This possibility becomes likely in view of the existence of
certain disturbing factors, which are mentioned for other purposes by Wason
and Johnson-Laird, but which can be shown to have just the required effect.
These features are the natural (‘non-material’) interpretation of the ‘if-then’
and the relative ambiguity of the hypothesis in the abstract experiments.
Let us turn to the latter first. Wason and Johnson-Laird present consider¬
able evidence to show that a widespread and persistent feature of the abstract
experiments was “the failure to appreciate that the cards are reversible, i.e.
 THE PSYCHOLOGY OF LOGIC 267

the significance of a card is the same regardless of which part of it is exposed”

(p. 195). They call attention to this fact in their discussion of what they
regard as a more speculative and less plausible explanation of the poor per¬
formance, according to which the novelty and abstractness of the task
induces a temporary regression to less mature forms of cognitive functioning,
such as failing to appreciate the importance of a reversible operation (pp.
193—201). However, we need not worry about the authors’ reasons for
mentioning the fact. Instead we may use it as evidence that the abstract hy¬
pothesis as initially stated, without qualification, comment, or context, could
mean either one of two things:

(HI) If a card has a vowel on one of the presently visible sides, then it
has an even number on the other side.
(H2) If a card has a vowel on either one of its two sides, then it has an
even number on the other side.

The difference between these two hypotheses is their scope as generalizations:

the first refers to four entities, the second to eight. Earlier, in analyzing the
hypothesis in the experiment I interpreted it to mean (H2). However, the
overwhelming evidence from the experimental subjects is that they took it to
mean (HI). Let us examine the consequences of this (more restricted) inter¬
pretation. We could symbolize it as

(*) 0) [Uxy —* (Px —+ Qy)],

where Uxy. x is the up side of a card whose down side isy, and the other
symbols are interpreted as before. For this formula there are only four
relevant cases that would need to be checked, namely those corresponding to
the first four in the list above,referring respectively to <sn, s2i> , <$12, $22>,
<si3, s23>, <$i4, $24>. When the hypothesis is so interpreted, the only card
that needs to be turned over is the first one. This would explain the choice
of card 1 by a large percentage of the subjects in the abstract experiments
(42 out of 128, in one series).
In this manner, however, we could not explain the other responses. To
explain them we need to call attention to the fact that the natural inter¬
pretation of ‘if-then’ is not as a material conditional. The authors themselves
present evidence (pp. 61—65, 89—93) that the ‘if-then’ does not have the
truth table of formal logic. It seems clear, therefore, that in reconstructing
the subjects’ reasoning, we should use their nonformal truth table for ‘if-then’.
The authors themselves are aware that this should be done, however they
erroneously state without argument that the subjects’ responses are incon-
268 CHAPTER 11

sistent with their own interpretation of ‘if-then’ (p. 183). Let us see what
really emerges when we take this into account.
Wason and Johnson-Laird present evidence that one common interpretation
of the ‘if-then’ is in accordance with the following truth table, as contrasted
to the one used in formal logic:

p Q if p then q p —>q
(actual reasoning) (formal logic)
true true true true
true false false false
false true void true
false false void true

If we look at the list above, we see that cases 2 through 6 would be evaluated
as void or irrelevant, rather than true, because of the falsity of the ‘F
formula. The value of case 1 would be true or false depending on the truth
value of Qs2i ; hence card 1 would have to be turned over. The value of case 7
would be true or void (irrelevant), though we could be sure that it’s not false;
hence we would have to turn over card 3. The value of case 8 would be false
or void (irrelevant), though we would be sure that it can’t be true; hence card
4 would have to be turned over. What follows here is that, if we prescind
from the ambiguity of the hypothesis, and we interpret it in a ‘reversible’
manner (H2), then the response would be card 1, 3, and 4. A significant
group of subjects, in fact, made this response.
Another common interpretation of the ‘if-then’ is such that a conditional
is taken to imply its converse. Wason and Johnson-Laird have discovered that
this is especially true when the conditional is some sort of generalization
(p. 61) and when the material is abstract (p. 62). These conditions are pre¬
cisely those occurring in the hypothesis being tested in the abstract experi¬
ment. According to this interpretation, ‘if p then q’ would be an abbreviation
of ‘if p then q, and if q then p\ where the ‘if-then’ would obey the ‘defective’
truth table mentioned above. It is not easy, and perhaps unnecessary to
formalize such a concept of ‘if-then’. Nevertheless, I believe it can be applied
to the present situation as follows. The second and fourth card would be
ruled out as irrelevant, the second because it falsifies the antecedent of the
hypothesis, the fourth because it falsifies the antecedent of its converse. Card
1 needs to be turned over to determine the truth value of the hypothesis as
stated, card 3 in order to determine the truth value of its converse. These are
the choices made by the largest percentage of subjects (59 out of 128) in the
original group of experiments.
 THE PSYCHOLOGY OF LOGIC 269

To summarize this explanation of the experimental results using abstract

material, we may say that it is possible to give an empirically supported
rational reconstruction of the subjects’ reasoning, both those who get answers
in accordance with formal logic and those that don’t. In the case of one
common response (card 1), it is possible to give a reconstruction completely
in accordance with formal logic, based on an interpretation of the original
hypothesis different from the one intended by the experimenter, but at least
as plausible. In the case of another common response (cards 1,3, and 4), it is
possible to give a reconstruction which is formalistic in the sense of being in
the spirit of formal logic, but using a nonstandard truth table. In the case
of the most common response, a formalistic reconstruction is not easy or
perhaps not possible. The few subjects whose response is labeled ‘correct’ by
Wason and Johnson-Laird are merely those who happened by accident to
choose the meaning of the hypothesis intended by the experimenters and
who then reasoned in accordance with formal logic. This is the most that can
be said in favor of the approach espoused by the authors. For one finds it
simply incredible that the subjects could have actually been engaged in
reasoning as complex as that outlined above. In that case the root of the
problem would be the experimenter’s choice of the original abstract hypoth¬
esis, which is simply deceptively simple. It is really very complicated, from
the point of view of formal logic, and hence we may say it was a bad choice
on the part of the experimenters. Could a better choice have been made?
Since the hypothesis complexity derives from its abstractness, as we saw
above, it seems that the only better choice would have been a more concrete
hypothesis and situation, which they did use in another series of experiments.
But when an experiment is so ‘concretized’, how significant does it become?
Doesn’t it perhaps become trivial? Let us ask for example, what one can learn
from the concrete experiments mentioned above? One cannot deny that
when the concrete experiments are contrasted to the abstract ones they are
very interesting and significant; in fact, such a contrast leads Wason and
Johnson-Laird to conclusions about the importance of content in reasoning
and hence to question the empirical import of formal logic; or the contrast
may lead one, as it led us in the above critique, to conclude that the use of
abstract material in experimental situations is tricky since the very abstract¬
ness of the material makes it very complex, and what’s worse, gives it the
appearance of being simple, from the point of view of formal logic. Thus,
if one rejects the usefulness of experiments with abstract material, then the
question remains whether experiments with concrete material have any value
per se.
270 CHAPTER 11

Let us then examine these experiments by themselves. They certainly

show such things as that the subjects would be able to do post office work
involving such simple tasks as determining whether sealed letters have a 5d
stamp on them. But what does it show about their reasoning? In particular
what does it show about whether or not they reason in accordance to formal
logic? The authors themselves confess that the almost trivial simplicity of the
concrete task makes it extremely difficult to determine exactly what steps
the subjects are following in their reasoning (pp. 190—1, 192—3). Of course,
one could ask them to give their reasons, as the authors report that has been
done in the case of the responses to the abstract material. But to do this is
really to have them express arguments. This is certainly a step in the right
direction, but then it would have been better to ask the subjects to write out
arguments, rather than merely give answers to pointed questions. But if one is
going to have experimental subjects write out arguments, it would be better
to examine examples of already written out arguments, existing in various
kinds of literature, produced by people in more realistic and concrete circum¬
stances, when they had the time to be relatively explicit in the steps of their
reasoning. In short, these problems in Wason and Johnson-Laird’s account of
reasoning suggest that a better approach to the study of human reasoning is
the study of written works containing relatively explicit argumentation. Now,
one of the best examples of such work is Galileo’s Dialogue Concerning the
Two Chief World Systems, which is full of explicit statements and analyses of
arguments for and against the earth’s motion, a very concrete and realistic
issue at the beginning of the seventeenth century when the book was written.
It is in part in this spirit that our investigation into Galileo’s Dialogue should
be viewed.
We may conclude this chapter by saying that on the one hand formal logic
has more explanatory power than these psychologists believe since it turned
out to be possible to give full (or almost) formal-logical reconstructions of the
subjects’ reasoning in ways unnoticed by the authors. On the other hand,
when formal logic is combined with an experimental approach by devising
experiments in its terms, we seem to get the situation that formal reasoning
is either not formal or not reasoning, in the sense that we get either intuitive
and trivial reasoning or formal operations that are likely to be psychologically
unreal because of their complexity.

NOTES

l
C. Perelman and L. Olbrechts-Tyteca, The New Rhetoric; C. Perelman, The Idea of
 THE PSYCHOLOGY OF LOGIC 271

Justice and the Problem of Argument; idem, The New Rhetoric’; idem, ‘A Reply to
Henry W. Johnstone, Jr.’; S. Toulmin, The Uses of Argument.
2 P. C. Wason and P. N. Johnson-Laird, Psychology of Reasoning; and D. Osherson,
‘Models of Logical Thinking’, in Reasoning: Representation and Process in Children and
Adults, edited by R. J. Falmagne.
3 Y. Bar-Hillel, ‘A Neglected Recent Trend in Logic’; idem, ‘Comments’, in J. F. Staal
(ed.), ‘Formal Logic and Natural Languages (A Symposium)’; idem, ‘Argumentation in
Natural Languages’; idem, ‘Argumentation in Pragmatic Languages’.
4 M. Scriven, ‘Definitions, Explanations, and Theories’, p. 100; idem, Reasoning, p. xv.
5 S. Toulmin, The Uses of Argument, pp. 252-59.
6 See, for example, D. Kalish and R. Montague, Logic, p. 3; Benson Mates, Elementary
Logic, pp. 3—4;W. Salmon, Logic, pp. 1-17; I. Copi,Introduction to Logic, pp. 3-5.
7 Psychology of Reasoning; in this chapter, subsequent references to this book will be
made within the text by indicating the page number(s) in parenthesis.
8 It might be objected that the second concrete example is more similar to the abstract
one and rather unlike the first concrete one, since the second concrete example refers
to sides of envelopes which allow for permutations similar to the cards, whereas for the
journeys to Manchester of the first concrete example there corresponds no ‘opposite’.
This objection suggests that perhaps the structure of the post office problem is reflected
in the formula

Cx) (y) [Dxy-*■ (Px-¥Qy)),

where perhaps
Dxy. x andy are different sides of an envelope;
Px: side x is sealed;
Qy: sidey has a 5d stamp on it.

The meaning of this would be that if an envelope is sealed on one side then it has a 5d
stamp on the other side.
In reply I should begin by pointing out that the first concrete example (travel to
Manchester) can be given a more complex structure, in spite of the fact that it makes no
sense to speak of the ‘opposite side’ of a journey. I suggested such a more complex
structure in my discussion above. My reason for focusing on the simpler formula was the
mere possibility of the simpler interpretation, whereas I argued that a simpler formula
would have misinterpreted the abstract situation. In short, whereas in the two concrete
examples the simpler formulas do not do any injustice to the situation, in the abstract
case the simpler formula does. Thus the relevant similarity between the two concrete
cases remains, together with their common difference from the abstract example.
Ultimately, of course, this is due to the physical difference between the abstract and
concrete situations. On the one hand, in the post office example, the physical fact is
that an envelope can be sealed only one one side, and this reduces to four the number of
real possibilities covered by the generalization “if an envelope is sealed on one side then
it has a 5d stamp on the other side.” (Moreover, there is no problem in the original
hypothesis “if a letter is sealed, then it has a 5d stamp on it”, deriving from the
possibility that the stamp and the sealing might be on the same side.) Similarly, in the
travel example, let the generalization be expressed by saying: “for any given journey
accomplished by a given means of transportation, if the destination is Manchester then
the transportation is by train”. The apparent possibilities involve the following pairs:
272 CHAPTER 11

<Manchester, train>, <Manchester, car>, <Leeds, train>, <Leeds, car>, <train,

Manchester>, <cai, Manchester>, <train, Leeds>, and <car, Leeds>. Here the first four
entities are not really (physically) different from the last four, respectively.
This answer to the initial objection might now lead to another, namely that the
difference between the two concrete and the one abstract hypotheses is not one of
logical structure, but one of background assumptions, relating to the the physical facts
of the respective situations. This new objection is true in one sense, but not in the sense
required to make it valid. For I am not claiming that the difference between the two
types of hypotheses is merely one of syntactical structure, but rather that the syntactical
complexity of the relevant symbolic formulas is one way of discussing the logical
differences that has explanatory power; moreover, I argue below that another important
element in explaining the experimental differences is that the meaning of the abstract
hypothesis is contextually ambiguous in a way in which that of the concrete one is not.
 CHAPTER 12

THE RHETORIC OF LOGIC AND

THE LOGIC OF RHETORIC:
CRITIQUE OF THE NEW RHETORIC

If our critique of the psychology of reasoning is correct, works and texts

like the Galilean ones examined earlier provide more adequate data for a
theory of reasoning than that provided by the usual controlled experiments
done by psychologists. Moreover, since what we want is a scientific theory of
reasoning, there is no reason why we should not avail ourselves of the scien¬
tific methodology worked out earlier, when we discussed the methodological
content of Galileo’s science (Chapter 5), the essential characteristics of his
methodology (Chapter 6), and the problem of scientific rationality (Chapter
8). Though that methodology does not provide us with simple recipes and
rules to follow, it does provide us with a model to emulate and a basic idea to
follow. The idea is to isolate or define a number of relevant concepts or
categories, and while thinking in their terms to practically exercise our judg¬
ment on the concrete materials being studied. For a theory of reasoning, the
material is of course arguments, and the relevant categories are such potentially
conflicting polarities as: logic and history, logic and psychology, the concep¬
tual and the empirical, logic and rhetoric, the abstract and the concrete, the
formal and the contextual, the theoretical and the practical, the descriptive-
explanatory and the prescriptive-evaluative, the general and the particular.
Having already examined the first few of these, let us go on to logic and
rhetoric.

LOGIC AND RHETORIC

Formal logic is primarily the study of mathematical reasoning, or to be more

exact, of those aspects of reasoning which predominate and acquire their
clearest expression in mathematics. This is a fact that has always been ex¬
plicitly recognized by the best logicians.1 It is also something which is usually
presupposed in an implicit manner by those who claim to want to deal with
everyday reasoning.2 The fact has also provided the basis for a very cogent
though nonconstructive diatribe against formal logic.3 The creators and
followers of the so-called new rhetoric4 have followed a more constructive
approach by working out the details of an alternative theory of reasoning,
centering around those aspects which predominate in legal arguments.
It is not clear what the relation between formal logic and the new rhetoric

273
274 CHAPTER 12

is or ought to be. Perhaps the two are parallel studies of different particular
domains of reasoning, formal logic dealing with mathematical arguments, the
new rhetoric dealing with legal arguments (and/or everyday arguments, and/
or evaluative arguments). Perhaps the two are complementary studies of
distinct aspects of all reasoning, formal logic dealing with the techniques of
establishing the truth of propositions, the new rhetoric dealing with the
linguistic techniques of securing the acceptance of propositions (either
inducing or increasing adherence to them). A third possibility is that formal
logic is the special case of the new rhetoric where the only way of securing
acceptance is to establish the truth of the proposition.
In spite of this ambiguity of purpose one cannot but applaud the efforts
of Perelman’s ‘new rhetoric’ and of his associates and followers. They are
correct in emphasizing the danger that the exclusive reliance on the formal
logic of reasoning has the irrationalistic tendency of leaving most of the
human sciences and of human affairs in the realm of the arbitrary and
capricious. They are also correct in mentioning the humanistic value of the
emphasis by the new rhetoric on the noncompelling, nonnecessitating (but
not non-existing) force of argumentation, which leaves the door open for a
sound notion of ‘free will’. Third, very commendable is their desire to take
the study of rhetorical problems out of the hands of propaganda ministers,
Madison Avenue men, public relation experts, and preachers from the pulpit,
who are the only groups who have in recent times studied these problems.
Nevertheless an attempt must be made to resolve the ambiguity for the
following reasons. If we have two separate, parallel fields then in order to
decide whether to use formal logic or the new rhetoric for the understanding
of a concrete instance of reasoning, one would have to decide whether one is
faced with a demonstration (deduction) or with an informal argumentation.
In other words we would have to apply whatever concepts of demonstration
and of argumentation had been arrived at as a result of the two different
studies; ultimately such applications would reduce to determining whether
the reasoning under investigation was more similar to mathematical or to legal
reasoning. No doubt, in many cases there would be no problem with such a
determination; however, if rhetoric and logic are essentially points of view
which though distinct can be applied to the same material (as the second
alternative above suggests), then the determination could be expected to run
into trouble, because in this case the demonstration/argumentation question
would have to be decided in terms of the propriety of adopting one of the
two points of view (with the understanding that the other point of view
would also be possible), rather than in terms of which properties are present
 THE RHETORIC OF LOGIC 275

in the material. If the third of the above alternatives is the correct one, then
the question would have to be decided in a correspondingly different way.
Thus the application of the new rhetoric depends on how its relation to
formal logic is conceived since most actual reasoning contains a mixture of
elements of deduction and argumentation (and other things besides).
A culturally, scientifically, and historically important illustration of these
problems relates to Galileo’s Dialogue Concerning the Two Chief World
Systems. Was the book an attempt to establish the truth of the earth’s
motion, or an attempt to induce or increase the acceptability of this idea in
the minds of Galileo’s contemporaries? If the former, the book was a failure,
if the latter a success. And to which activity, demonstration or argumentation
(or both?) did the Church’s condemnation of Copernicanism of 1616 refer?
At any rate, does not this controversy show that the propriety of treating
something as the subject of demonstration or argumentation is historically
dependent, since what can only be treated rhetorically at one time may later
become the subject of demonstration? Moreover, since argumentation is
proper, indeed necessary, when something is incapable of proof, it would
seem that the theory of argumentation presupposes (at least in these cases)
the theory of proof. In other words, for mixed cases of reasoning, which are
the typical cases, rhetorical analysis can only come after logical analysis. Let
me explore this idea by discussing the rhetorically all-important concept of
begging the question.
The concept of begging the question (petitio principii) is a very important
one in the new rhetoric because it is the best example of a procedure which
is unobjectionable from the point of view of formal logic, but necessarily
faulty from the point of view of inducing or increasing assent.5 In view of
this importance, it is unfortunate that Perelman does not give a better
example of begging the question than the following one taken from Navarre,
who in turn was following Blass, involving a passage in Antiphon’s speech on
the murder of Herodes: “I would have you know that I am much more
deserving of your pity than of punishment. Punishment is indeed the due of
the guilty, while pity is the due of those who are the object of an unjust
accusation.”6 Perelman interpets this7 as an argument with the first sentence
as conclusion, the second sentence as major premise, and the missing minor
premise “I am the object of an unjust accusation”, which cannot be accepted
by the judges before deciding the case. This is rather inadequate because the
passage could also be interpreted as consisting of two different ways of
stating the same conclusion, stated one way and explicitly in the first
sentence, and stated with a different emphasis and implicitly in the second
276 CHAPTER 12

sentence; and there is no direct evidence in the passage (such as reasoning

indicators) to favor either interpretation. Moreover, even if we accept Perel-
man’s interpretation, it should be noted that the missing premise is not iden¬
tical with the conclusion; hence if Antiphon proceeds to give an argument
supporting that missing premise, and if this argument is independent of the
conclusion in Perelman’s reconstruction, then there would be no begging of
the question. In other words, the passage quoted by Perelman could be the
last step of a long argument which does not assume the proposition that “I
am much more deserving of your pity than of punishment”. Of course, in
order to determine this condition we would have to look at the rest of the
speech; so I am not concluding that Perelman’s interpretation is wrong, but
rather that it is not justified by the passage he presents, and hence that the
passage is not a good example of what he is trying to illustrate. Unfortunate¬
ly, this is a general problem characterizing Perelman’s most important book
(The New Rhetoric): though a great merit of this book is that the examples
and materials mentioned are both more numerous, substantive, and realistic
than the artificial trivialities found in most logic books, nevertheless the
passages are usually taken out of context and the propriety and strength of
their illustrative power depends on assumptions which are not made explicit
and which may or may not be accurate. It follows from this that Perelman’s
approach would be improved if his discussions were centered around different
portions of one and the same work. Of course it is not easy to find a single
book rich enough to provide illustrations for and ground discussion of all the
concepts in the ‘new rhetoric’. However, there exists such a book, and it is
Galileo’s Dialogue.
Besides being excessively abstract in the sense just specified, the new
rhetoric may be improperly abstract in the sense of suggesting that rhetorical
analysis is independent of logical analysis. If we again consider Perelman’s
example of begging the question, it seems obvious that the rhetorical judg¬
ment that a begging of the question is being perpetrated presupposes some
kind of logical analysis, either the one given by Perelman himself, or an
expanded one along the lines I suggested. So, in any case, in order to under¬
stand the general concept of begging the question, or in order to explain a
given instance of it, one needs beforehand to rely on concepts of formal logic,
such as premises implying a conclusion. In other words, the very definition of
begging the question is full of formal-logical ideas: an argument whose
(formal) validity requires it to have as a premise a proposition which is either
identical with its conclusion or supportable only by means of an argument
having that conclusion as a premise. This is not to deny that the importance
 THE RHETORIC OF LOGIC 277

of such arguments is a rhetorical one; but what this means is that only when
they are examined from a point of view different from formal logic, can one
see their ineffectiveness and why they are to be avoided. Thus, if the aim is
to induce or intensify adherence to a certain proposition, then one cannot
accomplish this by means of any process which at some stage before the last
assumes that the adherence has taken place.
In summary then, we may agree with Perelman that the concept of begging
the question is the place to start to begin to understand the differences
between formal logic and the new rhetoric, as well as the importance of the
latter. At the same time I feeHhat it is also the place to start to see the limita¬
tions of the new rhetoric, and how these limitations may be overcome; the
limitations center around its abstractness in the double sense of separateness
of rhetorical analysis from logical analysis and of the analyzed material
(reasoning) from its context. Far from being separable from logical analysis,
rhetorical analysis presupposes it though the two are (conceptually) distin¬
guishable and distinct; moreover, it is precisely in rhetorical analysis that the
(linguistic) context of reasoning acquires paramount importance. A beautiful
illustration of these ideas can be given with an argument contained in
Galileo’s Dialogue, and the illustration will suggest the importance of this
book as material for theorists of reasoning.

A QUESTION-BEGGING ARGUMENT

The argument is Galileo’s tower argument, discussed in previous chapters for

its relevance to the understanding of scientific rationality and other matters.
Those discussions were relatively cursory since it was sufficient then to
suggest the broad outline of Galileo’s reasoning. Here, however, we shall need
a more detailed analysis since we are attempting to develop a rigorous,
empirical, historical approach to the study of reasoning, and so we need to
experience and discover what it is really like to do this sort of investigation.
The general context of Galileo’s situation is the question of whether or not
the earth rotates. The argument from vertical fall was one of the classical
objections to the earth’s motion. The portion of dialogue relevant to our
purposes here is the passage consisting of the quotations cited in Chapter 8.
Referring to that passage, we find that the first step of the original Aristotelian
argument can be reconstructed as follows:
(1) If the earth moved, the place of ejection of a body thrown
vertically upward would move along with it and the body would
fall some distance west of that place.
278 CHAPTER 12

(2) Therefore, if the earth moved, bodies would not fall vertically.
(3) But bodies do fall vertically.
(4) Therefore, the earth does not move.

Next Salviati asks Simplicio how he knows that bodies fall vertically.
Simplicio answers that we know it because they are seen to fall vertically.
This argument by Simplicio is the following:

(5) Bodies are seen to fall vertically.

(3) Therefore, bodies do fall vertically.

Now comes the crucial step. By what some would call a thought experi¬
ment, but is really a brilliant piece of hypothetical reasoning, Salviati gets
Simplicio to recognize that

(6) if the earth moved and bodies were seen to fall vertically, then
they would (actually) not be falling vertically.

From this Salviati, with Simplicio’s approval, infers that

(7) if one does not assume that the earth stands still, then one cannot
conclude that bodies would actually fall vertically from the mere
fact that they are seen to fall vertically.

This justification of (7) with (6) is repeated later in Sagredo’s speech. Hence
we must regard as a partial restatement of (6), the reason for (7) given by
Simplicio:

(8) if the earth moved, then bodies would (actually) not be fading
vertically.

In making Simplicio say this, Galileo means that if the earth moved, then
(since bodies are seen to fall vertically) they would actually not be falling
vertically. The following argument elucidates the sense in which (8) is a
partial restatement of (6) and must be what Galileo has in mind.
Letting

M = the earth moves

S = bodies are seen to fall vertically, and
V = bodies do actually fall vertically,

the reasoning could by symbolized as follows:

(6) If M and S, then not-V.

(6a) Therefore, if S and M, then not-V.
 THE RHETORIC OF LOGIC 279

(6b) Therefore, if S, then if M then not-V.

(5) S. (This is obvious, since bodies are seen to fall vertically.)
(8) Therefore, if M, then not-V.

Transformations of this sort also shed some light on what may have been
in Galileo’s mind in inferring (7) from (6). Using the same abbreviations:

(6) If M and S, then not-V.

(6.1) Therefore, if M, then if S then not-V.
(6.2) Therefore, if M, then not-(if S then V).8
(6.3) Therefore, if not-(not-M), then not-(if S then V).
(6.4) Therefore, if one does not assume not-M, then one cannot con¬
clude V from S.

Finally equating, ‘not-M’ with ‘the earth stands still’ and substituting for V
and S, (6.4) gives (7). The crucial and problematic step is that from (6.3) to
(6.4). For the moment, suffice it to say that (6.4) expresses in rhetorical
language what (6.3) says in logical language.
It is clear how from (7) Salviati reaches his next conclusion to the effect that

(9) the Aristotelian argument begs the question by assuming the very
same thing it is trying to prove.

In inferring (3) from (5), the Aristotelian argument is concluding that bodies
would actually fall vertically from the mere fact that they are seen to fall
vertically. In the light of (7), it follows that it is being assumed that the
earth stands still (doesn’t move), which is the argument’s conclusion (4).
What is not so clear is why Salviati does not give this simple explanation
to Simplicio when he asks for a justification of (8). The way Salviati expresses
himself is the following:

(10) The conclusion of the argument — that the earth doesn’t move —
ought to be unknown, for otherwise it would be superfluous to
give the argument.
(11) The middle term premise — that bodies fall vertically — ought to
be known, for otherwise one would be proving the unknown by
means of the equally unknown.
(12) (But it has already been argued that) if one does not first know
that the earth doesn’t move, one cannot know that bodies fall
vertically.
(13) Therefore, the certainty of the premise derives from the un¬
certainty of the conclusion.
280 CHAPTER 12

(14) Therefore, the argument is a petitio principii and seriously faulty.

The problem with this argument is (13); one might be inclined to be very
literal and claim that it neither implies (14) nor is implied by (10), (11), and
(12). But the real problem is its obscurity. In this context the word ‘certainty’
means ‘knowledge’, so it might seem that (13) means that knowledge of the
premise derives from lack of knowledge of the conclusion. Though this does
not make sense, it is a step toward the correct interpretation. For what seems
to be happening is that Galileo is carried away by the stylistic brilliance of
the expression, “the certainty of the premise derives from the uncertainty of
the conclusion”. The logic of the situation is such that the term ‘certainty’
cannot be performing a function analogous to that of the term ‘uncertainty’.
The sentence is best interpreted as saying that knowledge of the premise
derives from the conclusion, which is supposedly unknown. This follows
plausibly from (10), (11), and (12), and in turn we can see that, by the
definition of petitio principii, the argument is one because the premise
depends on the conclusion. And we can see that one fault with petitio prin¬
cipii lies in its contradictory epistemological presuppositions, in our case that
knowledge of one proposition derives from another unknown proposition.
After this explanation of the nature and seriousness of the paralogism in
the Aristotelian argument, Sagredo, with the help of Simplicio, objects to this
criticism, especially to the justification of the crucial (7), which is practically
synonymous with (12). The objection is basically that (6), the reason for (7),
hides a contradiction in its antecedent, and that a new argument for the
motionlessness of the earth can be formulated:

(15) If the earth moved, then it would be impossible that bodies

should be seen to fall vertically.
(5) But bodies are seen to fall vertically.
(4) Therefore, the earth doesn’t move.

And the justification for (15) is that

(16) if the earth moved, then if bodies were seen to fall vertically, they
would simultaneously have two motions — toward and around
the center; and

(17) it is impossible to have two such motions.

As Salviati implicitly admits, this revised argument is not open to the same
objection as the original one. In particular, (16) is an immediate consequence
of (6), indeed another way of expressing the same thought. What the
 THE RHETORIC OF LOGIC 281

Aristotelian is doing at this point is to turn the tables around. Whereas,

Galileo had used (6) to show the question-begging character of the original
Aristotelian argument, the Aristotelian is using (6) to formulate a revised
argument not beset by the same problem. Of course, his having done so is an
implicit admission that the original argument was indeed question-begging.
As Salviati notes, the worth of the revised argument depends wholly on
the alleged impossibility of motion being both toward and around the center
(17). He finds three things wrong with this premise:

(18) If Aristotle had (17) in mind when giving the original argument,
he should have explicitly said so (since it is such an important
part of the argument);
(19) the mixed motion mentioned in (17) is not only not impossible,
but necessary (something which Galileo promises to demonstrate
later);
(20) Aristotle himself would not accept (17), since he allows that
kind of mixed motion when he admits that fire and some of the
air move up by nature and rotate by participation.

(18) is obviously true, but it is not clear in what sense, if any, it makes the
revised Aristotelian argument faulty. This question will be pursued below.
In the case of (19), it is clear how the argument would be affected: one of
its premises would be false and it would be unsound (though formally valid).
It is also clear that (19) is physically true, in the sense that such mixed
motion is possible, though not necessarily ‘necessary’. However, the relevant
critical question would be whether (19) is true in the context of Galileo’s
discussion, that is, whether he does give the promised argument, and whether
it is contextually sound. He does fulfill the promise, at least in part, in his
argument dealing with the falling of bodies on a moving ship, and I believe
the argument is basically sound. But these matters will not be pursued here.
(20) is partly true, because Aristotle does allow mixed motion for fire
{Meteor., 344all). But whether it is completely true, i.e. whether Aristotle
would consequently allow mixed motion for falling bodies is questioned by
Simplicio on the ground that

(21) there is an important difference between the two situations:

whereas particles of fire are so light that they could easily be
carried along by the rotating air, falling rocks are so heavy that
they would not be.

This objection is misconceived first because the question is the possibility,

282 CHAPTER 12

not the likelihood of such mixed motion, and second because the Aristotelian
fire particles were supposed to be carried along not by the rotating air but by
the rotating lunar sphere, which compared with them was as tenuous, or even
more so, as air is compared to rocks and cannon balls. Thus Salviati’s third
criticism of the revised Aristotelian argument seems correct. Its effect is to
make this argument faulty in the sense of un-Aristotelian.
The statement of (21) is followed by a discussion of the ship experiment
as possible evidence for (17). This further discussion need not concern us
here.9

EVALUATIONS

So far the main concern has been with the reasoning as found in the text. The
evaluation of that reasoning involves primarily two questions. Is the revised
Aristotelian argument faulty, and if so how, given that it is a revision of the
original question-begging argument? And is the following argument valid and
if so how: if one does not assume that the earth stands still, then one cannot
conclude that bodies actually fall vertically from the mere fact that they are
seen to fall vertically, because if it is not true that the earth stands still, then
it’s not true that if bodies are seen to fall vertically they actually fall vertically.
Let us pursue this second question first. It was noted above that the con¬
clusion is simply a rhetorical formulation of the premise. When this is noted,
the first thing that comes to mind is that Galileo might be using two princi¬
ples relating the rhetorical and the logical situation:

Principle Logic Rhetoric

I If p, then q. If one assumes p, one can conclude q. I.e.,

one can conclude q from p.
II If p is not true, If one does not assume p, then one cannot
then q is not true. conclude q.

Then, using the earlier abbreviations, one could start with

(6.3) if not-M is not true, then it is not true that if S then V;

apply Principle II and get

(6.31) if one does not assume not-M, then one cannot conclude that if
S then V;
 THE RHETORIC OF LOGIC 283

from which, by applying Principle I, we get

(6.32) if one does not assume not-M, then one cannot conclude that one
can conclude V from S;

which means the same as

(6.33) if one does not assume not-M, then one cannot conclude V from
S;

which is the desired conclusion (7).

The problem with this interpretation is that whereas Principle I is intrin¬
sically plausible, II is not. Moreover, II can be shown to conflict with I. In
fact, ‘if p is not true, then q is not true’ is equivalent to ‘if q, then p’; and if
we apply I to the latter we get ‘one can conclude p from q\ which differs
from the result of applying II to the former, namely from ‘if one does not
assume p, then one cannot conclude q\ since this is equivalent to ‘one can
conclude q only if one assumes p\
Fortunately, there is a simpler explanation of Galileo’s argument, using
only Principle I. Start with

(6.2) If M, then it’s not true that if S then V.

Apply I to the main conditional to get

(6.21) If one assumes M, then one can conclude that it’s not true that if
S then V.

Now apply I to the conditional in the consequent of (6.21), to get

(6.22) If one assumes M, then one can conclude that it’s not true that
one can conclude V from S;

that is, combining the two occurrences of ‘conclude’,

(6.23) If one assumes M, then one cannot conclude V from S.

Now it is obvious in the context that

(6.24) If one does not assume not-M (that the earth stands still), then
one is assuming M (that the earth moves).

Hence we get

(6.25) If one does not assume not-M, then one cannot conclude V from
S;
284 CHAPTER 12

which is the desired conclusion (7). Thus the step from (6) to (7) in Galileo’s
reasoning seems valid.
Let us now consider the second evaluative problem, which is to determine
whether the revised Aristotelian argument is somewhat faulty because it is a
revision of the old one. The revised argument grounds the immobility of the
earth on the alleged impossibility of motion being both around and toward
the center; it is formally valid though unsound insofar as this premise is false
and unjustified. The question is, what does it matter if Aristotle did not have
the revised argument in mind when he gave the original one. Is it really im¬
proper to revise the original Aristotelian argument? The impropriety, if any,
is clearly not a ‘formal’ one, any more than question begging is. It has to do
with the way the argument originates. The Aristotelian (Sagredo acting as
one) thinks of it when he clearly sees that if one infers actual vertical fall
from seen vertical fall, as he is, then one is assuming that the earth stands still.
Now, this was implied by the crucial hypothetical that if the earth moves,
then if bodies are seen to fall vertically they do not actually fall vertically.
So we might say that the Aristotelian makes the switch when he realizes the
damaging consequences of that crucial hypothetical. What he does at this
point is to draw from it a consequence that suits him, namely

(16) if the earth moved, then if bodies were seen to fall vertically they
would be moving with ‘mixed’ motion.

How does this consequence suit him? In the sense that according to it, the
motion of the earth implies mixed motion for falling bodies, which he then
quickly denies. The propriety of the revision then depends on the propriety
of this denial, more specifically on whether or not the denial is arbitrary. Not
whether it is true (absolutely or contextually), which relates to the second
possible deficiency of the revised argument which Galileo discusses; and
not directly whether the denial is un-Aristotelian, which relates to its third
possible deficiency discussed above.
That denial would be arbitrary if it were arbitrarily introduced. And it
would be arbitrarily introduced if it were un-Aristotelian, which it is in one
sense. Hence the premise on which the argument depends seems arbitrary.
The concept that this deficiency suggests is that of an ad hoc argument. This
could be regarded as a kind of question-begging argument, one where a con¬
clusion is supported by means of a premise which is as much in need of
support as the conclusion. This kind of argument is referred to by Galileo in
the passage, and it is described as “the attempt to prove ignotum per aeque
ignotum”, i.e. to prove the unknown by means of the equally unknown. The
 THE RHETORIC OF LOGIC 285

reference is made when it is taken for granted that the original Aristotelian
argument is not an instance of this kind of question begging. Hence the
original argument must be question-begging in a different, and presumably
more serious, sense.

RECONSTRUCTIONS AND THEORETICAL CONSIDERATIONS

We shall now explore this other concept of question begging and simulta¬
neously seek a deeper understanding of the original Aristotelian argument and
of Galileo’s criticism of it. With the above discussion behind us, we shall now
feel more free to engage in abstractions and reconstructions. In addition to
the previous abbreviations, we shall use the following:

W = bodies ejected vertically upward land some distance west of

the place of ejection;
X = falling bodies move with mixed motion which is simulta¬
neously toward and around the center.

One version of the Aristotelian argument can be reconstructed as follows:10

(1—2221) If not-M and S, then V.

(1—222) .'. If not-M, then if S then V.
(1—221) Not-M.
(1-22) :. If S, then V.
(1-21) S.
(1-2) :. V.
(1-11) If M, then W.
(1—1) If M, then not-V.
(1—0) .'. Not-M; from (1—2) and (1—1).

The last four lines would constitute the initial segment of the argument.
(1—21) and (1—22) constitute the justification of V. Since (1—22) is not true
if M is, it needs a justification, which can only be provided by (1—221) and
(1—222). The ultimate premises on which the final conclusion, not-M, rests
are (1—2221), (1—221) (1—21), and (1 — 11). Unfortunately one of them,
(1—221), is identical with the conclusion. The argument is question-begging
in the sense of being circular, i.e., using the conclusion as one of the premises.
The role of the crucial consideration ‘If M and S, then not-V’ is to motivate
the justification of (1—22) and thus force the Aristotelian to use (1—222) and
hence not-M as grounds.
This argument, however, is not the same as the original one, given in
286 CHAPTER 12

Salviati’s first speech, and containing only the last four lines of this recon¬
struction. It is only by logical analysis that the speakers can construct this
deduction from the original segment; since its rhetorical uselessness follows
(rather obviously) from the structure of this deduction, that shows that the
rhetorical analysis is being grounded on the logical one. It follows that one
cannot carry out the former in abstraction from the latter (though, to repeat,
two distinct points are being made in the two points of view). Moreover, it is
also obvious that there is no way of identifying the begging of the question
by just looking at the initial segment of the argument ‘(1-1 !)> (1-1); but
(1—2); (1-0)’; that is, by just looking at Salviati’s first speech in the
Galilean text under consideration; one also needs the other parts of the
passage, which provide evidence that this segment is really part of the longer
one.
How much context need one provide in rhetorical analysis? After all, in
Galileo’s book the discussion does not end where our quotation ends. I think
the answer must be that one needs as much context as is needed to justify
the rhetorical point one is making. Though some passages may be too short to
allow any rhetorical analysis to get started, the present passage is obviously
sufficient since it provides a good illustration of a central concept of the new
rhetoric, the concept of petitio principii. Further insight into this concept can
be gained as follows.
Galileo does not ask the Aristotelian why (1-11) is supposed to be true
and exactly how it implies (1-1). Let us pursue these questions, beginning
with the latter. First the Aristotelian would probably say:

(1—12) If W, then not-V.

(1 — 11) If M, then W.
(1 — 1) If M, then not-V.

The answer to ‘why (1-12)?’ would probably be:

(1—122) If W, then not-S.

(1—121) If not-S, then not-V.
(1—12) If W, then not-V.

(1-122) would be obvious. But, why (1-121)? Probable answer:

(1-1211) If V, then S.
(1—121) If not-S, then not-V.

Why (1-1211)? Because:

 THE RHETORIC OF LOGIC 287

(1 — 12112) If not-M, then if V then S.

(1-12111) Not-M.
(1-1211) If V, then S.
Thus the justification of (1 — 12) would contain not-M as a premise. This
would indicate a second circularity in the original argument.
Now, is (1—11) true? The Aristotelian would say so because:
(1 — 112) If V, then if M then W.
(1-111) V.
(1-11) :. If M, then W.
The first premise here is obvious, but in order to justify the second he would
probably give the same justification as for (1-2), thus making the original
argument trebly circular.
It should be noted that both this third circularity and the first are the
result of the attempt to justify V empirically, by grounding it on S. But the
Aristotelian might resort to an a priori justification. He could argue, for
example, that bodies really fall vertically because it is their nature to move in
a “real’ straight line toward the center of the earth. This would be circular
insofar as ‘actual vertical fall’ here means “actual rectilinear motion toward
the center of the earth”. Insofar as there is a reference to the ‘nature’ of
bodies, we would have a disguised refusal to give a justification. Hence the
final conclusion would be resting on a proposition equally in need of support.
The Aristotelian could also argue that bodies actually fall vertically
because it is their nature to move in a ‘real’ straight line toward the center
of the universe, and this center coincides with the center of the earth. Then
Galileo would question this second premise. The Aristotelian could support it
a priori or a posteriori. The empirical argument, as Galileo shows elsewhere11
is itself circular; the a priori argument would presuppose that the earth stands
still and so would render circular the original objection to the motion of the
earth.
The revised Aristotelian argument is the following:
(2-2122) If M and S, then not-V.
(2—2121) If not-V, then X.
(2—212) If M and S, then X.
(2-211) Not-X.
(2-21) :. Not-(MandS).
(2—2) .'. If M, then not-S.
(2-1) S.
(2—0) .’. Not-M.
288 CHAPTER 12

The basic problem here is that (2-211) is as much in need of support as

(2-0). The argument begs the question not in the sense that its conclusion is
one of the ultimate premises, but in the sense that among its premises there is
a proposition which is no better known than the conclusion.
A rhetorical question would immediately arise already at the stage where
this new argument is given, namely, how this argument (2) is related to the
original (1). From a logical point of view they are different arguments;
however, rhetorically speaking, they are really two versions of the same
ambiguous argument. The ambiguity in this root argument derives from the
confusion of apparent vertical fall and actual vertical fall, which ambiguity
is present in Aristotle’s original statement of the argument and which is
retained even in Salviati’s statement in his first speech in the passage. So one
of the things that Galileo accomplishes in this section of his book is to
distinguish these two meanings of vertical fall. It also follows that what we
have here is also a beautiful illustration of what Perelman regards as “another
difference of paramount importance between [rhetorical] argument and
formal proof. The standard logical calculi are formulated in artificial
languages in which any one sign has one, and only one, meaning; in natural
languages the same word often has different meanings. 12 Though Galileo
is not using any formal calculus, he is using logical analysis to strip Aristotle’s
original ‘rhetorical’ argument of some of its ‘rhetoric’. That is, Galileo is
trying to turn the problem at hand from one where only rhetorical analysis
would be relevant to one suceptible to some logical analysis. Of course,
Galileo is not eliminating completely the rhetorical aspects of the problem,
nor would he want to; but he is using logic so that the rhetorical problem is
moved to a different point. This means that the line of separation between
rhetoric and logic shifts as knowledge grows; or to be more exact, more and
more arguments enter the domain of logic after crossing the borderline with
rhetoric; or, again, since I believe that the two are not separate domains but
distinct aspects, what happens is that there is a growth of problems of the
kind where the question of adherence becomes reduced to the question of
establishing truth.
Let us now consider some arguments which explore the consequences of
the most important propositions which are indisputable in this context:

(3—1) If M and S, then not-V.

(3—2) If not-M and S, then V.
(3—3) If M and V, then not-S.
(3—4) If not-M and V, then S.
 THE RHETORIC OF LOGIC 289

(3-5) S.
(3—6) If S and M, then not-V; from (3—1).
(3—7) If S, then if M then not-V; from (3—6).
(3—8) If M, then not-V; from (3—7) and (3—5).
(3—9) If V, then not-M; from (3—8).
(3—10) If S and not-M, then V; from (3—2).
(3—11) If S, then if not-M then V; from (3—10).
(3—12) If not-M, then V; from (3—11) and (3—5).
(3-13) V if and only if not-M; from (3-12) and (3-9).
(3—14) If M, then if S then not-V; from (3—1).
(3—15) If M, then not-(if S then V); from (3—14).13
(3—16) If (if S then V), then not-M; from (3—15).
(3—17) If not-M, then if S then V; from (3—2).
(3—18) (If S then V) if and only if not-M; from (3—17) and (3-16).
(3—19) If M, then if V then not-S; from (3—3);
(3—20) If M, then not-(if V then S); from (3—19).
(3—21) If (if V then S), then not-M; from (3—20).
(3—22) If not-M, then if V then S; from (3—4).
(3—23) (If V then S) if and only if not-M; from (3—22) and (3—21).
(3—24) Not-M if and only if (S if and only if V); from (3—23) and
(3-18).
(3—25) M if and only if not-(S if and only if V); from 3—24.

(3—1) is Galileo’s crucial consideration, his ‘thought-experiment’. (3—2),

(3—3), and (3—4) are obvious truths, which the Aristotelians themselves had
implicitly or explicitly accepted. (3—5) is an empirical fact.
(3—13) tells us that to say that falling bodies actually fall vertically is
equivalent to saying that the earth stands still. Hence it seems difficult to
avoid circular reasoning as long as one grounds the immobility of the earth
on actual vertical fall. Besides allowing Galileo to criticize the original
Aristotelian argument, (3—13) is useful to him as a step toward his ultimate
aim of justifying the motion of the earth. For he now knows that to do this
he has to deny that bodies actually fall vertically. His formulation of the
principle of inertia14 is a step toward the justification of that denial.
(3—18) is a somewhat weaker assertion than (3—13). It may be interpreted
as saying that to conclude actual vertical fall from visible vertical fall is
equivalent to assuming the immobility of the earth. Hence, as long as one
gives that kind of empirical justification of actual vertical fall when justifying
the stability of the earth, one is arguing in a circle. It should be noted that
290 CHAPTER 12

the step from (3—14) to (3—15) cannot be invalidated by the theory of

material conditionals since this theory is irrelevant here. Or, to be more exact,
one is assuming the principle that

if (if S then not-V) then not-(if S then V)

which is truth-functionally invalid, but rhetorically sound, and likely to be

linguistically true of the natural language ‘if-then’.15
(3—24) tells us that to say that the earth stands still is equivalent to
regarding S and V as equivalent. Hence, justifying not-M involves a failure
to distinguish S from V. And (3-25) tells us that to distinguish apparent
from actual vertical fall is to assert that the earth moves. Now, the first
thought that comes to mind is to use these facts to prove the motion of the
earth by arguing that since it is one thing for bodies to be seen to fall vertical¬
ly and another for them to actually fall vertically, it would follow that the
earth moves, by (3-25). Hence Galileo could be claimed to have, with his
criticism of the Aristotelian argument, not merely refuted the argument, but
also its conclusion, i.e. proved the motion of the earth. The fact that Galileo
did not claim so much should make us suspect that there is something wrong
with this suggestion.
The problem with the suggestion must be that the distinction between S
and V is not subsisting in a Platonic heaven, but is something dependent on
M, an empirical fact. The consequences of this are somewhat surprising. For it
would seem that the nonequivalence between S and V is a conceptual matter,
accessible by pure thought. But if this were so one would be able to prove the
motion of the earth in the manner indicated, which seems an invalid proce¬
dure. Hence the distinction between S and V is either not conceptual, or
some conceptual distinctions are empirical at least in the sense of having an
empirical origin. The last alternative seems the correct one. In other words,
the use and misuse to which (3-25) could be put suggests for us an
asymmetry between empirical and conceptual knowledge: empirical know¬
ledge can generate conceptual knowledge, but not vice versa.
In conclusion we may say that the approach to the study of reasoning
suggested by the new rhetoric promises to be a very fruitful one if we distin¬
guish properly between rhetorical and logical considerations, if we keep in
mind the primacy of the latter, and if proper attention is paid to the details
of actual reasoning, for example by the careful study of texts such as
Galileo’s Dialogue. It has also emerged that the concept of question-begging
is indeed the one with which to begin for a proper appreciation of the distinc¬
tions and of the relations between logical and rhetorical analysis. Finally, the
 THE RHETORIC OF LOGIC 291

logical which is primary here does not refer without qualification to orthodox
formal logic, which is viewed as a theory of reasoning, but to the activity of
reasoning itself.

NOTES

1 See, for example, A. Tarski, Introduction to Logic, pp. xiii-xiv.

2 See, for example, D. Kalish and R. Montague, Logic.
3 S. Toulmin, The Uses of Argument.
4 C. Perelman and L. Olbrechts-Tyteca, The New Rhetoric, C. Perelman, The Idea of
Justice and the Problem of Argument', idem, The New Rhetoric’; idem, ‘A Reply to
Henry W. Johnstone, Jr.’.
5 Perelman, ‘New Rhetoric’, p. 146; idem, New Rhetoric, pp. 110-14.
6 Perelman, ‘New Rhetoric’, p. 146; idem, New Rhetoric, p. 113.
7 Ibid.
8 This inference from (6.1) is problematic since if the ‘if-then’ in these sentence
schemata is interpreted as a ‘material conditional’ then (6.2) is not a logical consequence
of (6.1). The full resolution of this problem would ‘entail’ writing a book like A. R.
Anderson and N. D. Belnap’s Entailment, or like E. W. Adams’s The Logic of
Conditionals, which the reader is advised to study before making superficial objections.
Here the following comments will have to suffice. Partly I would want to say that I am
reconstructing Galileo’s reasoning, and if it is invalid, that is not the fault of my recon¬
struction; and if one objects that I ought to use the Principle of Charity, I can assure him
that I have already done so, and I challenge him to find a more accurate reconstruction.
Partly I would say that there is no good reason why the ‘material’ interpretation of the
‘if-then’ should be accepted here, that I have deliberately not symbolized it by means of
the ‘horseshoe’ symbol but retained the more vague natural language expression. Partly
I would say that perhaps the ‘if-then’ of these schemata should be interpreted as formal
implication; then ‘not-(if S then V)’ would mean that S does not formally imply V,
which does follow from ‘S formally implies not-V’, under the plausible assumption that
we are dealing with consistent sentences. Partly one could try to solve the problem by
adapting certain ideas from Adams’s Logic of Conditionals: that it is more proper to
conceive of conditionals as having probabilities rather than truth values (pp. 69—102);
that their probability is not the probability of being true but the ratio of the probability
of the conjunction of antecedent and consequent to the probability of the antecedent
(pp. 5ff.); and that inferences involving conditionals are best evaluated in terms of the
probabilistic soundness criterion that it should be impossible for the premises to be
probable while the conclusion is improbable (p. 1). Partly I would say that I am
advocating a practice-oriented logic or theory of reasoning, and thus it is logical prin¬
ciples that are to be tested by and derived from appropriately selected actual reasoning,
and not the other way around. Partly I would refer the reader to the comments I make
when this and similar problematic inferences arise again, e.g., at the end of this chapter,
and in Sections 4, 8, and 12 of Chapter 16.
9 For details on this, see Chapter 5 and Chapter 16, Section 10.
10 The numbering of the propositions in the two main arguments below follows the
numbering system introduced in Chapter 14. But nothing here depends on such labeling.
292 CHAPTER 12

11 G. Galilei, Dialogue, tr. Drake, pp. 34-37; cf. Chapter 16, Section 3 below.
12 Perelman, ‘New Rhetoric’, p. 144.
13 For the problematic character of this step and of the one in (3-20), see comments
below and Note 8 above.
14 Galilei, Dialogue, tr. Drake, p. 147.
15 See Note 8 above.
 CHAPTER 13

THE LOGIC OF SCIENCE AND THE SCIENCE OF LOGIC:

TOWARD A SCIENCE OF REASONING

We have seen that neither formal logic, nor the psychology of reasoning, nor
the new rhetoric, offer completely sound approaches to the science of reason¬
ing, though each has something valuable to contribute. Formal logic is
rigorous and systematic but formal, i.e., it is no empirical science; the
psychology of reasoning is sound insofar as it is empirical, but its experi¬
mental method has intrinsic limitations and needs to be supplemented, and
perhaps replaced, by an observational, historical, introspective approach; the
new rhetoric is valuable for the generality of its aim but needs to be made
more concrete, vis-a-vis both the materials of reasoning and logical analysis.
Let us continue our critical review of available approaches.

INDUCTIVE LOGIC

Inductive logic is defined by some as the study of logical probability, or

probable inference, or the degree of confirmation which the premises lend
to the conclusion of an argument, with formal validity as the special case
where the logical probability of the conclusion on the basis of the premises is
one. Others define it as the study of those types of arguments that are most
common in the empirical sciences, and hence from this point of view induc¬
tive logic is doubly relevant to our inquiry since it could offer us both a logic
of science and a science of logic, that is both a theory of scientific reasoning
and a scientific theory of reasoning. Because of this presumed generality, and
because of this emphasis on nonmathematical reasoning, inductive logic may
appear to provide the sort of thing we are looking for in our attempt to
develop a science of reasoning, and hence it deserves to be examined. More¬
over, the question arises of how inductive logic relates to the new rhetoric
since one of the central features of the argumentation studied by the latter is,
as Perelman himself states, that it is strong or weak, more or less convincing,
rather than valid or invalid, that is, rather than conclusive in an all-or-none
fashion, as is the case with formal proof.
Most recent work in inductive logic is too interdependent on formal logic,
the mathematical theory of probability, and statistics.1 Mostly for this
reason, it is neither sufficiently empirical nor sufficiently concrete. It does

293
294 CHAPTER 13

not deal significantly and realistically with actual scientific reasoning, which
is alleged to be its central subject matter. Rather it theorizes ad infinitum
about artificial and trivial arguments, such as simple induction by enumera¬
tion, which seldom occur in the actual life of science. Nevertheless, if one
examines the more elementary parts of the subject, it appears to contain the
seeds for remedying the insufficient generality of formal logic. Whether this
appearance corresponds to reality, is the question presently to be answered.
The fundamental problem haunting elementary inductive logic is nothing
less than whether there are such things as inductive arguments for it to study.
For one thing, there are the Popperian objections.3 Popper is convinced no
one has succeeded in solving Hume’s problem of the justification of induction,
which is the question of whether one is justified in reasoning from experienced
instances to other instances not yet experienced.4 Most philosophers would
probably agree that induction has not been justified.5 Popper then argues as
follows: since it is undeniable that humans learn from experience, and since
because of Hume’s arguments people could not learn inductively from
experience, it follows that induction is seldom if ever used in reasoning; that
is to say there is no such thing as inductive reasoning, and hence nothing for
inductive logic (as usually conceived) to study. Coupled with this destructive
criticism Popper makes the constructive suggestion that the important task is
to determine how people actually learn from experience. He then adds that in
science one finds the best and most explicit instances of learning from experi¬
ence, so it is these that deserve careful study. Now, though Popper and his
followers have practiced some of these preachings,6 their results have been
far from solid.7 A basic flaw is that, because of their underlying empiricism,
they have paid attention mostly to epistemological rather than logical
matters; insofar as they have examined the latter, they have been unduly
constricted by their falsificationism and fallibilism. In short, they have not
paid careful enough attention to the reasoning found in actual, historical
science. However, the general tenor of Popper’s approach to scientific
(‘inductive’) reasoning is acceptable: what is needed is an analysis, free of
inductivist preconceptions, of the actual forms of reasoning in science. In
conformity with such a program, the analysis of the arguments in a book like
Galileo’s Dialogue would seem to be very worthwhile.
Aside from this Popperian anti-inductivist criticism, unfortunately the
problem of the existence of inductive arguments can be seen by remaining
within the confines of Wesley Salmon’s Logic,8 and Salmon is certainly no
foe of induction. This book consists of two main parts dealing respectively
with deduction and with induction; the part on deduction is a standard
 THE LOGIC OF SCIENCE 295

elementary treatment of formal logic; the other part may be taken to deal
with inductive logic. This part contains an unresolved ambiguity, namely
whether inductive logic is the study of inductive arguments or the study of
inductive correctness. On the one hand, he discusses a number of types of
arguments which everyone would agree in regarding as instances of inductive
arguments, for example, induction by enumeration, statistical syllogism,
argument from authority, analogy arguments, and causal arguments;9 how¬
ever, he never formulates or makes use of a general definition of inductive
argument. What looks like such a definition is really a definition of correct
inductive argument, that is, an argument such that if all the premises are true
then the conclusion (1) is probably true but not necessarily true, and (2)
contains information not present even implicitly in the premises.10 However,
he never shows that any of the various types of admitted inductive arguments
are inductively correct in this sense, specifically that they satisfy condition
(1). Of course, he has intuitively plausible things to say about various faults
that may be present in inductive arguments and various ways of remedying
these faults, but there is no proof or even argument that when these inductive
arguments are in proper form, their conclusion is rendered probable by the
premises. Nor is this due to the elementary nature of the book, because
advanced texts and treatises do not do this either; that is, they do not show
that the probabilistic definition of inductive correctness properly describes
actual arguments, but rather merely that it can be made to describe various
formal structures, whose relation to actual arguments is extremely problem¬
atic. In short, if inductive logic deals with arguments definable as inductive,
then we don’t know that it deals with anything real because we have not even
been told when an argument is so definable, whereas, if inductive logic deals
with the inductive correctness of arguments, then we don’t know that any
real argument has the property as defined (ultimately because of Hume).
Moreover, it is a fact, admitted by Salmon himself,11 that any ‘inductive’
argument can be transformed into a ‘deductive’ one. Salmon dismisses this
problem by saying that the additional premises in such a transformation are
usually problematic as regards their acceptability, and that hence the trans¬
formation does not increase the overall argument soundness (which is a
function of the truth of premises and of the logical connection between them
and the conclusion). This argument, however, should be irrelevant to Salmon,
from his own point of view, according to which logic deals mainly with the
premise/conclusion connection, and not with the truth of premises. Hence,
from the latter point of view, the distinction between induction and deduc¬
tion evaporates.
296 CHAPTER 13

It is possible, however, to make a constructive elaboration of Salmon s

Logic. We could define inductive logic to be the study of inductive arguments,
and then define the latter to be arguments having an inductive form. An
inductive form would then be characterized by a semi-ostensive definition, in
terms of the various types of argument discussed by Salmon in his chapter on
induction, or appropriate modifications thereof. So an inductive argument
would be one having one of the following forms, or a form like them:

(1) Induction by enumeration: Z% of the observed members of F are G.

Z% of F are G.
(2) Statistical syllogism: Z%ofFareG.
x is F.
.'. x is G.
(3) Argument from authority: xassertsp.
P-
(4) Argument against the man: x assertsp.
Not p.
etc.
Now, the clause ‘or a form like them’ would make this definition open-ended
and would invite research to discover other forms like these. Some of these
are discussed by Salmon himself, namely analogy arguments and causal
arguments. But the form he gives for analogy arguments is unsatisfactory,
and for causal arguments he gives up the attempt to define them in terms of
a form. For the last type of ‘inductive’ argument discussed by Salmon,
hypothetico-deductive arguments in science, he provides a form, but this over¬
laps somewhat with the deductive form of ‘affirming the consequent’, so that
the question would arise why he discusses the inductive logic of what he
treated as a deductive type of argument in his discussion of conditional
arguments. The answer would be, because he believes that hypothetico-
deductive arguments (which affirm the consequent) are common in science.
But for the general logician interested in all types of human reasoning, this
would suggest that when a common type of argument is deductively invalid,
then the logician should construct or apply different, less stringent, more
realistic principles of evaluation. And this is one of the main points advocated
in the ‘new rhetoric’. At any rate, in ‘inductive logic’ as well as in the new
rhetoric, it becomes apparent that it is necessary to define and study common
types of actually occurring reasoning. To have at one’s disposal a collection of
interesting and important arguments is then of the utmost importance. Few
collections possess the realism of those which constitute Galileo’s Dialogue.
 THE LOGIC OF SCIENCE 297

The inductive correctness of such arguments could be investigated in an

analogous manner. Modifying properly the approach followed by Salmon, we
could say that the inductive correctness of arguments having the various
inductive forms depends on a number of identifiable, though not rigorously
formalizable, conditions. For example, inductions by enumeration depend
on whether the ‘observed F’s’ are sufficiently numerous and sufficiently
unbiased with respect to the total class of F’s; statistical syllogisms depend on
whether the percentage ‘Z’ is sufficiently high and whether the premises
embody all relevant evidence; arguments from authority depend on whether
x is a reliable authority concerning ‘p’, which in turn depends on a number of
particular conditions; arguments against the man depend on whether x is a
reliable anti-authority; etc. The identification of such conditions would
obviously have to be carried out in a quasi-empirical, concrete, historical
manner.

THE LINGUISTICS OF REASONING

At a memorable symposium12 Bar-Hillel lamented the fact that formal

logicians have devoted so little effort to the study of argumentation in natural
language. Calling the situation “one of the greatest scandals of human
existence”,13 he challenged “anybody here to show me a serious piece of
argumentation in natural languages that has been successfully evaluated as to
its validity with the help of formal logic”.14 In response to this challenge, the
participants produced a bibliographical list containing 66 items.15 An
examination of these items will easily convince one of the fairness of Bar-
Hillers words with which the printed version of the symposium discussion
concludes:

I asked for examples of application of formal logic to the evaluation of argumentation in

natural languages. I am first of all grateful to those of you who called my attention to
papers which I do not know; still all the examples - if I am not mistaken - were
exclusively taken either from philosophical discourse or from mathematical discourse.
Still, since all these discourses were held in natural languages, I will have to qualify some¬
what my original formulations, though only very slightly so.16

On another occasion Bar-Hillel called “one of the major philosophical prob¬

lems of our time”17 the following problem:

Is extant formal logic, deductive and inductive . . . sufficient for the formalization of all
arguments, of whatever degree of conclusiveness, that are used in ordinary language, or
at least in the sciences, or at least in the natural sciences (where by formalization I mean
298 CHAPTER 13

now “those preparatory operations in applied logic, whereby sentences of ordinary

language are fitted to logical forms by interpretation and paraphrase”), or is some
better-developed formal logic necessary for this purpose, or is formal logic altogether
insufficient for this purpose?18

In calling this a philosophical problem Bar-Hillel did not imply that it should
concern only technical philosophers; rather he was making a rhetorical point
since he was here addressing the XII International Congress of Philosophy
(1958). In fact, when speaking to logicians, as he was in the symposium
mentioned above, he would try to stress the fact that their neglect represented
something of a betrayal of the history of their discipline;19 whereas, when
speaking to linguists, he would try to point out the importance of this prob¬
lem to their work.20
I completely agree with Bar-Hillel about the urgency of the problem. Also
I accept his historical thesis that this lamentable situation results from the
fact that argumentation in natural languages developed historically into a
no-man’s-land between logic and linguistics.21 I also find very plausible his
analysis of the relevance of the problem to linguistics; he feels that semantics
is the central part of linguistics,22 and he claims that almost all linguists with
the qualified exception of Lyons23 and of Weinreich,24 have failed to see
“that meaning relations between linguistic entities are essentially deducibility
relations and, therefore, logical relations. They missed this fact partly because
a large number of the rules stating meaning relations can indeed be easily and
practically formulated in the form of entries in traditional monolingual
dictionaries (or lexica) .... Unfortunately the obvious efficiency of lexica
has blinded almost all the linguists who gave any thought to that issue into
believing that all of semantics is exhausted by lexicology, though a number of
them paid some attention to the fact that other rules were needed at least for
the purpose of determining how the meaning of longer linguistic entities is
to be composed out of the meanings of their component smaller entities
supposedly taken care of by the lexicon.”25 This has left “semantics, in an
atheoretical, sometimes even anti-theoretical, bloodless, and anemic state”.26
If Bar-Hillel’s critiques are nothing less than devastating, his constructive
suggestions are rather scarce. Nevertheless I believe they possess in promise
what they lack in accomplishment. First, I believe he is right in suggesting
that the chief obstacle to the understanding of argumentation in natural
languages is “the essential dependence of communication in such languages
on linguistic co-text (viz. the utterances, if any, that preceded the communi¬
cative act under scrutiny) and the extra-linguistic context (the general back¬
ground in which this communicative act was performed, the motives that
 THE LOGIC OF SCIENCE 299

brought it about, the cognitive and emotional background of the participants

in it, etc.)”.27 Second, “to attempt abstraction is always in order. But it is
never a priori guaranteed that by abstracting one will be in a better position
to solve a given set of problems, since for this purpose, sooner or later, the
abstractive step will have to be followed by a concretive one — abstraction by
concretion; otherwise, obviously, one will never come back to the original
problems”.28 Third, “we have to face the decisive choice: to apply formal
logic either to the sentences of natural languages ‘as is’, or to sentences in one
of the constructed language systems that ‘correspond’ (in a sense that has yet
to be clarified) to the sentences that occurred in the original argument”.29
Fourth, since “even the most developed methods known today for a direct
logical treatment of natural language sentences [the theories of Montague and
his associates30 ] are of limited use only . .. then there remains only the
‘indirect’ method”.31 Next, Bar-Hillel suggests that the indirect method
should be such that the original argument is reconstructed “in some
normalized, sterilized, regimented, systematized language to the sentences of
which it is then possible to apply the rules of formal logic”;32 from this,
“three formidable questions arise: (a) What is the exact nature of that
language system . .. ? (b) What is the exact nature of that formal logic that
will be applied to the sentences of that language system? (c) What is the
theory that guides us in performing this fateful transition from utterance of
natural languages to statements of the regimented language?”33 According to
Bar-Hillel too much work has been done in dealing with problems (a) and (b),
and practically none to deal with (c). Having taken us this far, Bar-Hillel
unfortunately stops.
Now, my primary concern is with (c). My suggestion is to lay the ground¬
work for such a theory by actually performing these ‘fateful transitions’ in a
sufficient number of cases and with respect to natural language argumentation
that has intrinsic and other interest. That is, I will be concerned first and fore¬
most with reconstructing the arguments m a book such as Galileo’s Dialogue',
then on the basis of such reconstructions one may construct some theory
relating the original material to the reconstructions.

THE PRACTICE OF REASONING

The problems about formal logic and the science of reasoning discussed so far
have all dealt with what may be called logical theory, in the sense in which
this may be contrasted with logical practice. That is, it has been argued that
the theory of reasoning needs to be empirical enough to take into account the
300 CHAPTER 13

actual reasoning processes that occur in the mind of homo sapiens, general
enough to study rhetorical and/or inductive arguments as well as the nature
of deduction, and concrete and realistic enough to take into account the fact
that almost all reasoning is expressed in natural languages. All of these
critiques are points to keep in mind when theorizing about or reflecting upon
reasoning; the next one is made from the point of view of someone interested
in the practical use of formal logic, that is to say, in using formal logic in
one’s own reasoning (about matters other than formal logic). I believe it is
Michael Scriven who has made this criticism most forcefully.34 He has
pointed out that there is overwhelming evidence from educational psychology
suggesting that the transfer of learning from one field to another is very little;
and when applied to formal logic this means that to learn formal logic may
make one proficient about solving problems in formal logic but has little
effect on one’s reasoning skill concerning everyday or other problems.35 To
this empirical argument, Scriven adds the following more conceptual one.
However adequate a theory of reasoning is, by the very fact of being a theory
it would consist of a number of systematically interrelated technical concepts
in terms of which one would try to solve various problems about reasoning.
Now arguments and reasoning problems are entities that can be found very
frequently in everyday life; if one wants to solve one of these problems by
the help of the theory of reasoning (formal logic), then one has first to
formulate the problems in the theoretical terminology, then apply the
theoretical principles and formal calculus of the theory to reach a theoretical
conclusion, and finally translate this theoretical conclusion into the terms of
the original problem so that it can be seen to be a solution to it as originally
presented, and not to some other problem. Unfortunately, Scriven adds, as a
matter of socio-historical fact, everyday reasoning problems are normally
such that the theoretical encoding and decoding steps are as problematic and
debatable as the original problem; hence the undeniable rigor, precision, and
reliability of the intra-theoretical (formal-logic) manipulations are of no
practical consequence.
The general cultural and pedagogical importance of these considerations
by Scriven is undeniable and difficult to exaggerate. It is not clear, however,
how his emphasis on the practice of reasoning can be made relevant to formal
logic. After all, the latter is the (theoretical) study of reasoning, and Scriven
himself has stressed the fact that the practical application of any theory does
not obey the same laws as the theory itself; for example, the theory may be
exact, but its practical application will necessarily be inexact; or the theory
may be part of a science, but its practical application will necessarily be an
 THE LOGIC OF SCIENCE 301

art. In other words, is Scriven saying merely that if one is interested in

practicing reasoning and improving reasoning skills, then the study of formal
logic should have a relatively low priority, and that if one is interested in the
practical application of formal logic, then one’s expectation better not be too
high? Or is Scriven also saying that if one is interested in the science of logic,
i.e., in theorizing about reasoning, then he should pay the proper attention to
the practice of reasoning? With regard to the former claims, Scriven’s point is
well made. But I believe he is also claiming the latter, in connection with
which a problem arises, namely, how is this latter critique distinct from the
earlier ones from the point of view of psychology, rhetoric, and linguistics?
For, after all, the psychologists are saying that formal logic does not pay the
proper attention to how humans actually reason, the adherents of the new
rhetoric are saying that formal logic tends to ignore the argumentative
practice of lawyers, jurists, moralists, preachers, students of the human
sciences, and laymen; and Bar-Hillel is saying that formal logic ignores the
practice of argumentation in natural languages. I believe Scriven’s point is
different, and I would describe its novelty to be the suggestion that the
theory of reasoning should consist of reasoning about reasoning'. That is, our
logic books should be full of arguments about arguments, rather than full of
theories, generalizations, formulations and systematizations of principles and
concepts about arguments which would require a subsequent application in
order to be of practical use. Certainly, that is the character of his latest book,
Reasoning. Thus the main element of the science of reasoning would be
primarily reasoning practice in the sense of analysis and evaluation of actual
arguments, and the theory of reasoning would have the secondary place of
formulating and systematizing principles for the critical analysis of argu¬
ments.
In summary, because the value of formal logic ultimately depends on its
contributions to the improvement of actual reasoning, and because the
theoretical character of formal logic makes its practical application to actual
reasoning problems highly suspect, the most valuable thing for the logician to
do is to engage in and produce arguments which analyze and evaluate other
arguments about ordinary matters. Contributions to logic so conceived would
be conclusions about arguments actually put forth as a matter of history,
based on their critical analysis. The contribution envisaged here concerns the
arguments in support of the earth’s motion put forth by Galileo in his
Dialogue. This inquiry will have a double logical relevance since even a super¬
ficial look at the book shows that most (though not all) of Galileo’s own
arguments are themselves about arguments, specifically about all the classical
302 CHAPTER 13

and contemporary objections to the earth’s motion. In other words, Galileo s

book consists mostly of a critical examination of all the available arguments
and evidence against the earth’s motion; the rest consists of the presentation
of favorable arguments, partly arising from considerations made in the course
of those critiques. What this means is that in the Dialogue Galileo primarily
acts as a logician in the sense under discussion of argument analyzer or
practioner of logic. But this means also that many of our own arguments will
be two steps removed from the subject matter of the earth’s motion, insofar
as they will be arguments about Galileo’s arguments about the anti-Copernican
arguments; nevertheless this will introduce variety and depth in our own
investigations. So, at any rate, these investigations promise to be logical in
at least two ways.
Actually there is a third way in which such an investigation would be of
relevance to logic. This involves considerations of the nature of science and
follow a line of reasoning suggested by Scriven himself.36 It is a fact that the
desire that logic be a science is very much on at least the lips of logicians.37
Agreeing with this goal, Scriven would argue38 that the proper model for
(‘empirical’) science ought not to be celestial mechanics, which is a very
untypical science on account of the isolated character of the solar system.
He would say that the proper (proper because realistic) model for science
should be something like engineering or medicine, where progress is measured
in terms of solutions to concrete practical human problems. According to this
model of science, for logic to be scientific would be for it to produce results
about actual reasoning problems, that is to produce justified conclusions
about arguments of the past or present. Now, to add my own twist to this
argument, I would say that since Galileo is regarded by almost everybody as
a model scientist, and since logicians want very much to be scientific, it will
be very important for them to examine his work to see to what extent it can
be used to justify their own techniques of formalization and abstraction.
They might discover that, or at least would be led to ponder whether, (a)
their anti-empirical orientation is a version of the empty scholasticism that he
fought, (b) formalization and abstraction are only sometimes appropriate,
and (c) it is important to discuss methodological and philosophical prob¬
lems (though one doesn’t have to go to the extreme of metaphysics). If
besides making such discoveries they would also discover that an essential
feature of his work was argument analysis, then perhaps they would be more
willing to truly scientificize logic since they could thereby also logicize
science.
 THE LOGIC OF SCIENCE 303

THE CRITICISM OF REASONING

Finally a word should be said about Stephen Toulmin’s approach to logic and
his critique of formal logic since his Uses of Argument is one of the earliest
statements of the recent literature critical of formal logic. Toulmin embodies
and expresses all of the critiques that I have discussed under the separate
headings: he has the psychologist’s concern with actual human reasoning, the
interest in legal reasoning and non-compelling argumentation exhibited by the
adherents to the new rhetoric, the kind of sensitivity to the natural languages
in which argumentation is usually expressed that Bar-Hillel recommends, and
the practical orientation that is so much at the heart of Scriven’s concern.
Moreover, he ends his critique by advocating a type of investigation that is
very close to the one I have been groping toward in my restatement and
justification of the above critiques. For example, according to Toulmin

logic conceived in this manner may have to become less of an a priori subject than it has
recently been . . . Accepting the need to begin by collecting for study the actual forms
of arguments current in any field, ... we shall use ray-tracing techniques because they
are used to make optical inferences, presumptive conclusions and ‘defeasibility’ as an
essential feature of many legal arguments, axiomatic systems because they reflect the
pattern of our arguments in geometry . . . But not only will logic have to become more
empirical; it will inevitably tend to become more historical. To think up new and better
methods of arguing in any field is to make a major advance, not just in logic, but in the
substantive field itself: great logical innovations are part and parcel of great scientific,
moral, political or legal innovations. In the natural sciences, for instance, men such as
Kepler, Newton, Lavoisier, Darwin and Freud have transformed not only our beliefs, but
also our way of arguing and our standards of relevance and proof.39

However, it should be noted that my endorsement of the various critiques

has been a qualified one in most cases. My qualification and criticism of such
critiques have attempted to define a sense in which logic is more important to
the study of reasoning than is commonly supposed; for example, I argued
that formal logic is more relevant to the kinds of experiments considered by
Wason and Johnson-Laird than they suppose, and that some of their
psychological models are inadequate because of an inadequate formal-logical
analysis of their own evidence; I also argued that the rhetorical analysis of
arguments presupposes their logical analysis, that only the actual performance
of the logical analyses required in the reconstruction of natural language
argumentation can provide the groundwork for a linguistics of reasoning, and
that logical practice in the sense of argument analysis is fundamental in the
improvement of reasoning skills. In short, I have been arguing that logic is
the key to the study of reasoning, though such logic may not be formal logic
304 CHAPTER 13

as usually conceived. I think it is necessary to make a similar point in connec¬

tion with Toulmin’s views.
Besides advocating a logic with the features mentioned in the above quota¬
tion, Toulmin wants to make logic into a comparative study in the sense that
“validity is an intra-field, not an inter-field notion. Arguments within any
field can be judged by standards appropriate within that field, and some will
fall short; but it must be expected that the standards will be field-dependent,
and that the merits to be demanded of an argument in one field will be found
to be absent (in the nature of things) from entirely meritorious arguments in
another”.40 This is the result of his conviction that “all the canons for the
criticism and assessment of arguments, I conclude, are in practice field-
dependent, while all our terms of assessment are field invariant in their
force”.41 Toulmin realizes42 that this leads to the problem of whether the
differences between the standards of different fields are irreducible, and
whether a general logic is possible. His answer seems to be that a general
account of the structure of arguments is possible; in fact, he provides one.43
However, a general theory of validity presumably is not possible, which
would explain his emphasis in the present book on description as opposed to
prescription. I think his answer is unsatisfactory for two reasons. One is that
if the layout of arguments is common to all fields (which is not to say that all
arguments have the same structure, but rather that the classification of argu¬
ments into several kinds does not derive from their structure), if then the
layout is general then the possibility would arise of interfield argumentation
about the canons of assessment. Of course, I would agree with Toulmin that
a responsible interfield criticism of standards cannot be too externalistic and
reconstructionist, if it is to have a good chance of success; familiarity with and
awareness of the standards internal to the field being criticized would be very
important. Nevertheless, it would not seem to be in principle impossible to
have general criteria, even if we accept the notion of fields of arguments.
A second objectionable feature of Toulmin’s account here is that it tends
to neglect the interdisciplinary nature of most creative reasoning. Toulmin’s
own book is a good example of this since it is an attempt to introduce
jurisprudential standards of assessment in traditional logical questions. A
classical example is, of course, Galileo whose supreme achievement was the
introduction of standards of mathematical reasoning into physical arguments,
as well as the introduction of standards of physical arguments into astro¬
nomical ones;44 to respect the traditional integrity of the three fields,
mathematics, physics, and astronomy, would have made it impossible for
modern science to emerge. I do not think Toulmin would deny this; and in
 THE LOGIC OF SCIENCE 305

the above quotation where he speaks of the historical character of logic, he

comes close to making the same point. Nevertheless, it is undeniable that his
book contains a strong tendency toward acritical or uncritical description.
Of course, to emphasize sympathetic understanding and description is certainly
a good rhetorical device in opposition to the irresponsible prescriptionism
which is the stock in trade of so many logicians and epistemologists. How¬
ever, a general emphasis on acritical comparison would not be good for logic.
The point is not to avoid criticism, but to avoid unfair criticism which fails to
understand the argument being criticized.
The way to avoid unfair criticism is to take account of the context and
problem situation in which the argument under consideration was given. The
comparative method that Toulmin wants to inject into logic may be taken
partly to contain this suggestion, and to that extent it is of course quite
acceptable. The contextual method not only is a very powerful tool in the
critical understanding of a given argument, but it contains an almost built-in
check against abuses and excesses of the kind to which a formal logician
would be inclined to succumb. For example, the propriety of generalizing the
desirability of Galileo’s introduction of mathematical standards into the
study of motion depends on whether or not the situation is sufficiently
similar to Galileo’s; hence it would not be proper to say today that it is
proper to mathematize the study of reasoning, though this may have been
proper a century ago; moreover, Galileo was not merely mathematizing
motion, but physicalizing mathematics by putting physical content into
mathematical arguments. Analogously, it would not be proper to say that
more mathematical modelling is needed in the science of psychology, since
the problems facing this science are and have always been of a different
nature than those of physics.45 Similarly, just as one may find proper
Toulmin’s attempt to introduce standards of jurisprudence into logical
matters, so one may also find appropriate Wason’s attempt to reform jurispru¬
dence by injecting a little more logical awareness into it.46
In conclusion, a general logic seems possible if we concentrate on the
critical understanding of actual reasoning, which concentration carries along
with it an empirical, historical, concrete or contextual, practical, and critical
orientation.

NOTES

1 See, for example, H. E. Kyburg, Jr., Probability and Inductive Logic, and its biblio¬
graphy on pp. 199-247. Noteworthy exceptions are M. Hesse’s Structure of Scientific
Inference and L. J. Cohen’s The Probable and The Provable.
306 CHAPTER 13

2 E. g., W. Salmon, Logic, pp. 13-17, and 81-117.

3 K. Popper, The Logic of Scientific Discovery, idem, Objective Knowledge, pp. 1-31;
and J. Agassi, ‘The Role of Corroboration in Popper’s Methodology’.
4 Popper, Objective Knowledge, p. 4.
5 See, for example, B. Skyrms, Choice and Chance, pp. 1-69.
6 J. Agassi, Towards an Historiography of Science', I. Lakatos and A. Musgrave (eds.),
Criticism and the Growth of Knowledge.
7 See my History of Science as Explanation and my Essay-review of Criticism and the
Growth of Knowledge.
8 Pp. 1-17, and 81-117.
9 Salmon, Logic, pp. 81-105.
10 Ibid., pp. 13-14.
11 Ibid., p. 17.
12 Y. Bar-Hillel, ‘The Role of Formal Logic in the Evaluation of Argumentation in
Natural Languages’, an intersectional symposium at the Third International Congress for
Logic, Methodology, and Philosophy of Science (Amsterdam, August 26, 1967), printed
in J. F. Staal (ed.), ‘Formal Logic and Natural Languages’.
13 Ibid., p. 256.
14 Ibid.
15 Ibid., pp. 281-83.
16 Ibid., p. 281.
17 Y. Bar-Hillel, Aspects of Language, p. 108.
18 Ibid., pp. 107-108.
19 Staal (ed.), ‘Formal Logic and Natural Languages’, pp. 256-57; Bar-Hillel, Aspects of
Language, pp. 202, and 206-207.
20 Aspects of Language, pp. 182-201.
21 Ibid., p. 182.
22 Ibid.
23 J. Lyons, Structural Semantics', cf. Bar-Hillel, Aspect_ ^f Language, p. 183, n. 1.
24 U. Weinreich, ‘Explorations in Semantic Theory’, in Current Trends in Linguistics,
edited by T. A. Sebeok, pp. 395-477.
25 Aspects of Language, p. 183.
26 Ibid., p. 182.
27 Ibid., pp. 207-208.
28 Ibid., pp. 208-209.
29 Ibid., p. 209.
30 Bar-Hillel’s intended reference is to Montague, whose papers have posthumously been
collected in Formal Philosophy, edited by R.H. Thomason. The reference is anachronistic
in the sense that nowadays one would have to examine the game-theoretical semantics
approach stemming from J. Hintikka’s work, which can be found in his The Semantics
of Questions and the Questions of Semantics, and in E. Saarinen, ed., Game-Theoretical
Semantics. Such an examination is beyond the scope of the present work.
31 Aspects of Language, p. 212.
32 Ibid., p. 213.
33 Ibid.
34 M. Scriven, ‘Definitions, Explanations, and Theories’, p. 100; idem, Reasoning,
pp. xiv-xv.
 THE LOGIC OF SCIENCE 307

35 Scriven, Reasoning, p. xiv.

36 M. Scriven, ‘The Frontiers of Psychology: Psychoanalysis and Parapsychology’; idem,
‘A Possible Distinction Between Traditional Scientific Disciplines and the Study of
Human Behavior’; idem, ‘Psychology Without a Paradigm’; idem, ‘Science: The Philo¬
sophy of Science’, International Encyclopedia of the Social Sciences (1968 ed.), 14:
83-92; idem, ‘A Study of Radical Behaviorism’; idem, ‘Views of Human Nature’.
37 See, for example, W. V. Quine, Methods of Logic, p. 1.
38 Scriven, Reasoning, p. xv.
39 S. Toulmin, The Uses of Argument, pp. 257-58.
40 Ibid., p. 255.
41 Ibid., p. 38.
42 Ibid., p. 39.
43 Ibid., pp. 94-145.
44 S. Drake, Galileo Studies, pp. 95-112.
45 Scriven, ‘Views of Human Nature’.
46 P. C. Wason and P. N. Johnson-Laird, Psychology of Reasoning, pp. 218-28; and
P. C. Wason, ‘The Drafting of Rules’.
 PART III

THEORY OF REASONING
 CHAPTER 14

PROPOSITIONAL STRUCTURE:
THE UNDERSTANDING OF REASONING

It is now time to formulate a number of basic principles of logical theorizing.

Logic is conceived here, of course, as the theory of reasoning. In this chapter
we shall be concerned with only one of its two main problems.

PROPOSITIONAL STRUCTURE

In the theory of reasoning there are two primary aims, the understanding of
reasoning, and the evaluation of it. Combining the two we might say that its
goal is the critical understanding of reasoning. Generally speaking, the phrase
‘the understanding’ of reasoning is meant to convey the same idea as the
following: the explanation, the analysis of reasoning. The phrase ‘the evalua¬
tion’ of reasoning is meant to convey the same idea as the following: the
assessment of, the criticism (favorable or unfavorable) of, judgment upon
reasoning. We may say that analysis and evaluation study different aspects of
reasoning. Analysis involves the study of the nature or structure of reasoning,
evaluation the study of its value or worth. In analysis one is concerned with
finding out what the reasoning is, in evaluation with finding out how good
(i.e., valid) it is. From a third point of view, the understanding of reasoning
may be said to involve the description of reasoning; while its evaluation
involves prescriptions for it.
Reasoning is the activity of the human mind consisting of the giving of
reasons for conclusions, or the reaching of conclusions on the basis of reasons.
More exactly it is the interrelating of our thoughts in such a way as to make
certain thoughts dependent on others.
Reasoning is linguistically expressed in what are called arguments. An
argument is a basic unit of reasoning in the sense that it is a piece of reasoning
sufficiently self-contained as to constitute by itself a more or less autonomous
instance of reasoning. The occurrence of reasoning is normally indicated, and
can always be explicitly indicated, by the use of what may be called reason¬
ing indicators. These are words like the following (or phrases synonymous
with such words): therefore, thus, so, hence, consequently, because, since,
for. Reasoning indicator words, however, are only hints, since it is possible to
express reasoning without them and since it is possible for them to have other

311
312 CHAPTER 14

meanings which do not indicate reasoning; but we will not consider such
complications at the moment.
Reasoning indicators serve to interconnect what may be called the proposi¬
tional components of an argument. A propositional component of an argu¬
ment is any part of an argument which is capable of being accepted or
rejected by itself. A proposition is any propositional component stated as a
complete sentence so that it can stand by itself.
For example, consider the following passage:

All the circularly moving bodies seem to lag behind and to move with more motions than
one, except the first sphere (that is, the primum mobile)', therefore the earth, moving
around its own center and being placed at the center, must be moved with two motions
and must fall behind; but if this were the case, the risings and settings of the fixed stars
would have to vary, which is not seen to happen; therefore the earth is not moved.1

The only reasoning indicator here is the word ‘therefore’ in both of its occur¬
rences. The propositional components are as follows:

(PCI) All the circularly moving bodies seem to lag behind and to move
with more motions than one, except the first sphere (that is, the
primum mobile).
(PC2) The earth, moving around its own center and being placed at the
center, must be moved with two motions and must fall behind.
(PC3) If this were the case, the risings and settings of the fixed stars
would have to vary.
(PC4) Which is not seen to happen.
(PC5) The earth is not moved.

The propositions would be the following:

(Pf) Same as (PCI).

(P2) Same as (PC2).
(P3) If the earth were moved with two motions and fell behind, the
risings and settings of the fixed stars would have to vary.
(P4) It is not seen to happen that the risings and settings of the fixed
stars vary.
(P5) Same as (PC5)

As a second example, consider the following passage:

The earth cannot move circularly, because such a motion would be a forced one and
therefore not perpetual. The reason that it would be forced was that if it were natural,
the earth’s parts would also naturally move in rotation, which is impossible because the
nature of these parts is to be moved downward in a straight line.2
 PROPOSITIONAL STRUCTURE 313

The reasoning indicators here are as follows: ‘because’ in both of its occur¬
rences, and the phrase “the reason that_was that_The proposi¬
tional components are as follows:

(PCI) The earth cannot move circularly.

(PC2) Such a motion would be a forced one.
(PC3) Not perpetual.
(PC4) If it were natural, the earth’s parts would also naturally move in
rotation.
(PC5) Which is impossible.
(PC6) The nature of these parts is to be moved downward in a straight
line.

These may be expressed as propositions:

(PI) Same as (PCI).

(P2) The earth’s circular motion would be a forced one.
(P3) The earth’s circular motion would not be perpetual.
(P4) If the earth’s circular motion were natural, its parts would also
naturally move in rotation.
(P5) It is impossible that the parts of the earth should move in rota¬
tion.
(P6) The nature of the earth’s parts is to be moved downward in a
straight line.

An argument may thus be conceived as a series of propositions some of

which are being based on others, where the interconnections are expressed by
means of reasoning indicators. The simplest possible argument contains two
propositions; such a simple argument can always be expressed in either one of
two standard forms which are logically equivalent: (1) A, therefore B; or (2)
B because A. In both (1) and (2), B is the conclusion and A is the reason or
premise. In other words, although both ‘therefore’ and ‘because’ are reason¬
ing indicators, they indicate different ways to express reasoning; the proposi¬
tion preceding the word ‘therefore’is the reason or premise, the one following
it is the conclusion; whereas the proposition preceding ‘because’ is the conclu¬
sion, and the one following it is the reason. The three-dots symbol V.’ is used
as an abbreviation for the word ‘therefore’ and a convenient way of referring
to simple arguments; so that ‘A,B’ is short for (1), or equivalently (2).
A complex argument is one which is made up of at least two subarguments
combined in such a way that the conclusion of one subargument is simultane¬
ously a reason of the other subargument. The simplest complex argument is
314 CHAPTER 14

one of the form: A because B, and B because C (or equivalently: C, therefore

B; therefore A). Here B is the reason of the subargument ‘A because B’ (or
‘B, therefore A’) and also the conclusion of the subargument ‘B because C’
(or ‘C, therefore B’). Every proposition in a complex argument falls into one
and only one of the following categories: intermediate proposition, final
reason, final conclusion. An intermediate proposition in a given complex
argument is a proposition which serves as the conclusion of one subargument
and as a reason of another subargument. A final reason in a given complex
argument is a proposition which is a reason of some subargument but not
the conclusion of any subargument. The final conclusion in a given complex
argument is a proposition which is the conclusion of some subargument but
not the reason of any subargument. In the example here, A is the final con¬
clusion, B is the one and only intermediate proposition, and C is the one and
only final reason.
The propositional structure of reasoning refers to the interrelationships
among its various elements, i.e. among its various subarguments and among
the various propositions. Such structure may be pictured in a structure
diagram which is constructed in accordance with the following rules:3
(1) Label each proposition in the argument by some number or letter.
(2) Represent each proposition by enclosing its number or letter in a
closed figure.
(3) The fact that a given proposition is a reason supporting another is
indicated by a solid line leading up from the first to the second.
(4) The diagram will have at the top a proposition which is supported by
one or more other propositions but which does not itself support any other
proposition; this proposition at the top is the final conclusion of the argu¬
ment being diagrammed.
(5) The diagram will have one or more propositions which support other
propositions but which are not themselves supported by anything else; such
propositions are the final reasons of the argument being diagrammed.
(6) Some structure diagrams may have propositions which both support,
and are supported by, other propositions; such propositions are the inter¬
mediate propositions of the argument being diagrammed. That is, they are
propositions which are logically placed between the final reasons and the final
conclusion of the argument and which are reasons from the point of view of
what they immediately support and conclusions from the point of view of
what they are immediately supported by.
These rules are to be understood so that the diagrams for the two argu¬
ments considered above would be the following:
 PROPOSITIONAL STRUCTURE 315

In the first example, P5 is the final conclusion; PI, P3, and P4 the final
reasons; and P2 is an intermediate proposition: it is the conclusion of the
subargument ‘PI, P2’ and one of the reasons of the subargument ‘P2, P3,
P4,P5’.
In the second example PI is the final conclusion; P4 and P6 are the final
reasons; and P3, P2, and P5 are intermediate propositions. P3 is the reason of
‘P3, PI’ and the conclusion of ‘P2,P3’; P2 is the reason of the latter, and
the conclusion of ‘P4, P5, P2’; and P5 is one of the reasons of the latter,
and the conclusion of ‘P6, P5’.
There is a standard way of labeling propositions in structure diagrams:
(A) Label the final conclusion ‘C’.
(B) Label all intermediate propositions and final reasons by an ‘R’ and use
numerals as follows to distinguish among the various reasons:
(1) Reasons which support the final conclusion are labelled Rl, R2, etc.
(2) Reasons which support the same intermediate reason, will have all
digits the same, save the last one (e.g. reasons supporting R12, will be labelled
R121, R122, R123, etc.).
The standard diagrams for the two arguments above would be, respectively,
the following:

Standard Diagram 1 Standard Diagram 2

316 CHAPTER 14

Of course, although some of the symbols in these two diagrams are the same,
they correspond to different propositions. The point of this standard labeling
is to give each proposition a label that indicates relatively easily and clearly its
logical function in the argument.
These ideas may be further illustrated by considering other passages where
Galileo states arguments very explicitly. For example:

since when the ship stands still the rock falls to the foot of the mast, and when the
ship is in motion it falls apart from there, then conversely, from the falling of the rock at
the foot it is inferred that the ship stands still, and from its falling away it may be
deduced that the ship is moving. And since what happens on the ship must likewise
happen on the land, from the falling of the rock at the foot of the tower one necessarily
infers the immobility of the terrestrial globe.4

The logical structure of this argument may be pictured as follows:

The various propositions are:

C: The terrestrial globe is immobile.

Rl: From the falling of the rock at the foot of the mast it is inferred
that the ship is standing still, and from its falling away it may be
deduced that the ship is moving.
R2: What happens on the ship must likewise happen on the land.
R3: On the land the rock falls at the foot of the tower.
R11. When the ship stands still the rock falls to the foot of the mast,
and when the ship is in motion it falls apart from there.

A fourth example is part of the argument that natural straight motion is

impossible:

straight motion being by nature infinite (because a straight line is infinite and
indeterminate), it is imposible that anything should have by nature the principle of
moving in a straight line; or, in other words, toward a place where it is impossible to
arrive, there being no finite end. For nature, as Aristotle well says himself, never under¬
takes to do that which cannot be done, nor endeavors to move whither it is impossible to
arrive.5
 PROPOSITIONAL STRUCTURE 317

The structure is:

The propositions are:

C: It is impossible that anything should have by nature the principle

of moving in a straight line.
R1: Straight motion is motion toward a place where it is impossible to
arrive.
R2: Nature never undertakes to do that which cannot be done, nor
endeavors to move whither it is impossible to arrive.
R11: Straight motion is by nature infinite.
Rill: A straight line is infinite and indeterminate.

The fifth example is:

Generation and corruption occur only where there are contraries; contraries exist only
among simple natural bodies, movable in contrary motions; contrary motions include
only those made in straight lines between opposite ends; of these there are but two,
namely, from the middle and toward the middle; and such motions belong to no natural
bodies except earth, fire, and the two other elements; therefore generation and corrup¬
tion exist only among the elements.6

Its structure is:

The propositions are:

C: Generation and corruption exist only among the elements.

R1: Generation and corruption occur only when there are contraries.
R2: Contraries exist only among simple natural bodies, movable in
contrary motions.
R3: Contrary motions include only those made in straight lines
between opposite ends.
318 CHAPTER 14

R4: Of motions made in straight lines between opposite ends, there

are but two, namely, from the middle and toward the middle.
R5: Motions from and toward the middle belong to no natural bodies
except earth, fire, and the two other elements.

The sixth and last example here is:

. . . because the third simple motion, namely, the circular, about the middle, has no
contrary (because the other two are contraries, and one thing has but one contrary)
therefore that natural body to which such motion belongs lacks a contrary and, having
no contrary, is ingenerable and incorruptible, etc., because where there are no contraries
there is no generation, corruption, etc. But such motion belongs to celestial bodies alone;
therefore only these are ingenerable, incorruptible, etc.7

Its structure is:

C: Only celestial bodies are ingenerable, incorruptible, etc.

R1: That natural body to which circular motion belongs is ingenerable
and incorruptible, etc.
R2: Circular motion belongs to celestial bodies alone.
R11: That natural body to which circular motion belongs lacks a
contrary.
R12: Where there are no contraries there is no generation, corruption,
etc.
Rill: The third simple motion, namely, the circular, about the middle,
has no contrary.
R1111: The other two simple motions are contraries, and one thing has
but one contrary.

THE RECONSTRUCTION OF ARGUMENTS

Some arguments are such that what they need is not so much analysis, but
 PROPOSITIONAL STRUCTURE 319

rather synthesis, as it were. For example, in Galileo’s Dialogue some argu¬

ments occur in the course of long, often poetical, and usually dramatically
intense passages, and they need to be reconstructed. In general, a reconstruc¬
tion of an argument is a restatement of it such that no logically extraneous
propositions are included and such that all logical interconnections among the
stated propositions are explicitly and clearly indicated, by means of reasoning
indicators.
For example, Galileo’s critique of the ship experiment, which takes about
four pages of text,8 may be reconstructed as follows: It is not true that when
the ship is in motion the rock falls away from the foot of the mast because
once the rock is dropped it will tend to move in two directions: (1) vertical,
or toward the center of the earth, and (2) horizontal, or around the center of
the earth along with the ship. The rock will tend to move in a horizontal
direction with the speed of the ship because the nature of bodies is such that,
when left to themselves, they continue in a state of rest, if they are at rest, or
in a state of uniform motion, if they are in motion. That the latter is
probably true may be seen as follows: balls sliding down a smooth inclined
plane move down the incline with a continuously increasing speed; and balls
climbing up a smooth inclined plane move with a continuously decreasing
speed; therefore, balls moving along a plane going neither up nor down move
with a speed that neither increases nor decreases, that is, a ball moving in a
horizontal direction will move in that direction with constant speed.
Galileo’s discussion of vertical fall9 has the following two elements.
(a) Reconstruction of the Aristotelian argument from vertical fall: The
earth does not move because bodies thrown vertically upward fall vertically
back to the same place from which they were thrown and this could not
happen if the earth were moving; for if the earth were moving, then the place
of ejection would move along with it while the projectile was going up and
down, and if that place moved then the body would fall some distance away
from it.
(b) Reconstruction of Galileo’s criticism: The Aristotelian argument is a
‘begging of the question’ fallacy because the way the Aristotelians would
support their premise that bodies fall vertically would be to say “because
they are seen to fall vertically”. Now, to say this assumes that apparent verti¬
cal fall implies actual vertical fall. And to show the latter you have to assume
that the earth is at rest because if the earth moves, then apparent vertical fall
does not imply actual vertical fall (it would imply actually slanted fall).
Another example is the Analytical Summary presented above (Chapter 2),
which is a reconstruction of the whole main argument of the book.
320 CHAPTER 14

THE INTERDEPENDENCE OF REASONS

Let us now examine a passage from the Dialogue which is noteworthy. (1)
because it is easily overlooked on account of its literary structure, which has
the character of a parenthetical discussion; (2) because it contains an intrinsi¬
cally interesting argument, indeed a distinct Aristotelian argument for celestial
inalterability, in addition to the one based on contrariety of motions and to
the one based on the alleged lack of observed changes; (3) and because it
contains arguments which can serve as good examples for some additional
ideas in logical analysis. The passage10 contains a discussion of what may be
called a teleological argument for celestial inalterability. To be more exact, it
can be reconstructed as a series of three things: (A) an Aristotelian teleologi¬
cal argument for celestial inalterability, (B) a Galilean critical counterargu¬
ment, and (C) an Aristotelian defensive reply to the counterargument. The
reconstructions are as follows.
(A) Reconstruction of the Aristotelian teleological argument for celestial
inalterability. There are two ways to show that there cannot be celestial
mutations. First, mutations would make celestial bodies imperfect since
celestial mutations would be superfluous. They would be superfluous because
celestial bodies need only light and motion to fulfill their purpose. And this
is so because their purpose is to serve man.
Second, celestial mutations would be vain and useless, and nature does
nothing in vain. They would be vain and useless because they would not
occur for the benefit of man, and they would not occur for the benefit of
celestial inhabitants (since there are no such creatures).
(B) Reconstruction of the Galilean counterargument. This argument is
deficient for several reasons. First it assumes that change makes things
imperfect, and this assumption is false because change seems to make things
more perfect. This is so because it is usually the less perfect things that are
devoid of change (a desert is less perfect than a garden, a dead animal is less
perfect than a living one), on the other hand, when things are both perfect
and devoid of change (for example, precious metals), they are precious
because they are scarce.
Second, it seems unlikely that the purpose of celestial bodies is to serve
the earth since this would mean that divine and eternal things would be
serving a base and transitory body.
Third, to say that there are no men on the celestial bodies is probably
correct, but from this it doesn’t necessarily follow that there are no changes
there, since (from this same proposition) it does not necessarily follow that
 PROPOSITIONAL STRUCTURE 321

there are no creatures there; in fact, to think that it does follow would be like
arguing by someone who had lived all his life in a forest, that there could not
be things like ships or fish. Now, if there could be creatures there, then
there would be changes for their sake.
Fourth, the conclusion seems wrong since some terrestrial changes derive
in part from celestial bodies and to cause a change without being changed
seems unlikely, so that celestial bodies too must be alterable.
Finally, to say that the purpose of celestial bodies is to serve man is
unjustified because the only argument you could give is the following: “For
the purpose of everything else is to serve man. For example, horses, herbs,
cereals, fruits, birds, beasts, and fishes exist for the comfort and nourishment
of men.” Now, the problem with this argument is that at best it shows only
that the purpose of all terrestrial things is to serve man; and from this we
cannot conclude that the purpose of celestial bodies is also to serve man
(especially for an Aristotelian who distinguishes sharply between the two
domains and believes that different principles apply).
(C) Reconstruction of the Aristotelian reply to the counterargument. This
criticism is invalid for two reasons. First, it is not true that the original
(Aristotelian) argument assumes that change makes things imperfect. In fact,
the claim that change makes things imperfect is nowhere explicitly stated in
the argument; nor is this claim implicitly employed as the reason why change
would make celestial bodies imperfect since the reason is given explicitly and
it is that celestial mutations would be superfluous. (In other words, the
assumption in question is not being made in the original argument because
there is no explicit or implicit step in it saying “mutations would make
celestial bodies imperfect because mutations make things imperfect”; instead
the step reads “mutations would make celestial bodies imperfect because
celestial mutations would be superfluous”.)
Second, though various parts of the earth are transitory and subject to
change, the earth as a whole is not transitory but as eternal as the celestial
bodies; thus if celestial bodies serve the earth, this means that certain eternal
bodies are serving another eternal body, and there is no absurdity in that.
Several comments should be noted about these reconstructions. First, the
Aristotelian argument has two apparently distinct versions in the text, and it
would be inaccurate to ignore either one; so I have kept both versions in my
reconstruction. Second, the first point in the Galilean counterargument is
indeed incorrect, as the first point in the Aristotelian reply correctly points
out. Third, the second Galilean criticism also has a cogent objection in the
second Aristotelian reply. Fourth, the third and last Galilean points are
322 CHAPTER 14

plausible ways of integrating the text discussions pertaining to those topics,

respectively; however, the reconstructed last point involves more interpreta¬
tion and more addition of ideas not present in the text. Fifth, the only way
to integrate the discussion about precious metals in the Galilean critique
seems to be as in the second part of the first point; it has prima facie plausi¬
bility, but merely prima facie, for Simplicio could add the following third
point to his reply: “Third, to say that precious metals are precious because
they are scarce is true only in the sense ‘'primarily because they are scarce ,
since their relative scarcity does contribute somewhat to their preciousness.
At any rate, the proposition is irrelevant; the relevant one would be: precious
metals are perfect because they are relatively scarce; but this latter proposi¬
tion is not true since their scarcity has nothing to do with their perfection.'"
But there is no textual justification for including this in the reconstruction.
These reconstructed arguments provide good examples for what I shall
call the interdependence and the independence of reasons. Two reasons which
immediately support the same proposition (and which would thus have the
same number of digits as subscripts in a diagram) are interdependent when
each depends on the other in order to support that proposition and each
alone is insufficient or incomplete to provide that support. When reasons are
interdependent the rules for standard labeling apply without change. When
two reasons are not interdependent, they are called independent. So two
reasons are independent of each other when each does not depend on the
other to support the conclusion based on them (whether or not each alone is
sufficient to provide that support, i.e., whether or not each is interdependent
with some other reason). In other words, two reasons may be independent of
 PROPOSITIONAL STRUCTURE 323

each other and still be interdependent with other reasons. The standard
labeling of independent reasons is as follows: the small letters ‘a’, ‘b’, ‘c’, etc.
should be placed after the label for the proposition they support, to distinguish
one set of reasons from another independent set; and then these small letters
should be carried for each proposition, as in the above diagram.
Applying these ideas, the diagram for the reconstructed argument A above
would be the one just drawn. The various propositions are:

C: There cannot be celestial mutations.

Ral: Mutations would make celestial bodies imperfect.
Rail: Celestial mutations would be superfluous.
Ral11: Celestial bodies need only light and motion to fulfill their pur¬
pose.
Ral111: The purpose of celestial bodies is to serve man.
Rbl: Celestial mutations would be vain and useless.
Rb2: Nature does nothing in vain.
Rbl 1: Celestial mutations would not occur for the benefit of man.
Rbl2: Celestial mutations would not occur for the benefit of celestial
inhabitants.
Rb 121: There are no celestial inhabitants.

Here, Ral and Rbl are independent, whereas Rbl and Rb2 are interdepen¬
dent, and so are Rbl 1 and Rbl2.
The structure diagram for B would be:

Some of the more important propositions are:

C: Argument (A) is deficient.

Ral: Argument (A) assumes that change makes things imperfect.
324 CHAPTER 14

Ra2: The assumption that change makes things imperfect, is false.

Ra21 lal: A desert is less perfect than a garden.
Ra211 b 1: A dead animal is less perfect than a living one.
Ra212: When things are both perfect and devoid of changes (for example,
precious metals), they are precious because they are scarce.
Rcl: To say that there are no men on the celestial bodies is probably
correct.
Rc2: From (Rcl) it does not necessarily follow that there are no
changes on celestial bodies.
Rc21: From the same (Rcl) it does not necessarily follow that there are
no creatures on celestial bodies.
Rc211: To think that from (Rcl) it does follow that there are no creatures
on celestial bodies, would be like arguing by someone who had
lived all his life in a forest, that there could not be things like
ships or Fish.
Rd 1: The conclusion of argument (A) seems wrong.
Rdl 1: Celestial bodies too must be alterable.
Rel: To say that the purpose of celestial bodies is to serve man is
unjustified.
Rell: The only argument you could give is the following: “For the
purpose .... nourishment of men”.

Note that the giving of examples to support a generalization (which is hap¬

pening in ‘Ra21 lal, Ra211bl, .'. Ra211’) normally generates independent
reasons. Note also that in a list of critical remarks, each critique is usually
independent of the others; here the Ra’s are independent of the Rb’s, etc.
Finally note that the Ra subargument of B deals with an assumption made
presumably in connection with Ral of A; that the Rb subargument of B
relates to proposition Ral 111 of A; that the Rc subargument ofB relates to
proposition Rbl21 of A; that the Rd subargument of B tries to directly
refute the final conclusion C of A; and that the Re subargument of B relates
to proposition Ral 111 of A.

The propositions and diagram for argument C are:

C: The criticism of argument B is invalid.

Ral: It is not true that argument A assumes that change makes things
imperfect.
Ral2: The claim that change makes things imperfect is not implicitly
employed (in argument A) as the reason why change would make
celestial bodies imperfect.
 PROPOSITIONAL STRUCTURE 325

Ral22: The reason is that celestial mutations would be superfluous.

Rbl If celestial bodies serve the earth, this means that certain eternal
bodies are serving another eternal body.
Rb2 There is no absurdity in ‘that’.

One advantage of this idea of the independence of reasons is that it allows

us to integrate into a unified whole separate arguments, if they have the same
final conclusion. For example, consider the following passage:

This principle being established then, it may be immediately concluded that if all integral
bodies in the world are by nature movable, it is impossible that their motions should be
straight, or anything else but circular; and the reason is very plain and obvious. For
whatever moves straight changes place and, continuing to move, goes ever farther from
its starting point and from every place through which it successively passes. If that were
the motion which naturally suited it, then at the beginning it was not in its proper place.
So then the parts of the world were not disposed in perfect order. But we are assuming
them to be perfectly in order; and in that case, it is impossible that it should be their
nature to change place, and consequently to move in a straight line.
Besides, straight motion being by nature infinite (because a straight line is infinite
and indeterminate), it is impossible that anything should have by nature the principle
of moving in a straight line; or, in other words, toward a place where it is impossible to
arrive, there being no finite end. For nature, as Aristotle well says himself, never under¬
takes to do that which cannot be done, nor endeavors to move whither it is impossible
to arrive.11

The second paragraph contains one argument whose structure was analyzed
above (section on ‘Propositional Structure’). The first paragraph contains the
following argument:
326 CHAPTER 14

The propositions are:

C: It is impossible that it should be the nature of integral bodies to

move in a straight line.
R1: It is impossible that it should be their nature to change place.
Rll: If the motion which naturally suited bodies were to go ever
farther from their starting point and from every place through
which they successively pass, then the parts of the world would
not have been disposed in perfect order.
R12: The parts of the world are in perfect order.
Rill: Whatever moves straight changes place and, continuing to move,
goes ever farther from its starting point and from every place
through which it necessarily passes.
R112: If the motion which naturally suited a body was to go ever
from its starting point and from every place through which it
successively passes, then since the beginning it was not in its
proper place.

Combining the two arguments we get:

 PROPOSITIONAL STRUCTURE 327

LATENT STRUCTURE

The Aristotelian teleological argument is also a good example to illustrate

certain ideas about what I shall call latent propositional structure. The
application of these ideas to that argument allow us to give substance to the
intuitive judgment that the two versions of the Aristotelian argument have
something in common, in spite of appearances. As I mentioned above, the
appearances are that we have two distinct versions, the only common proposi¬
tion being the conclusion. Probing their latent structure allows one to discern
other points of contact.
The latent propositional structure (or more simply, the latent structure)
of an argument consists of those propositions which are not explicitly stated
in the argument but are implicitly assumed or taken for granted by the giver
of the argument. In one case, a proposition is part of the latent structure, or
more simply a proposition is latent, when for a particular step of the argu¬
ment it is needed, in addition to the other explicit propositions involved in
that step, to fully justify that step. In another case, a proposition is latent
when it is one of the reasons being implicitly used to justify one of the final
reasons in the explicit argument. Because of their position in structure
diagrams, in the first case the propositions may be called horizontally latent,
in the second case vertically latent. In a structure diagram all latent structure
is drawn in dotted lines.
For example, consider the first version of the teleological argument (first
paragraph of reconstructed argument A above). Its explicit structure is
Diagram A. (Call explicit structure the interrelationship of the propositions
which are explicitly stated). If we include the latent structure, we would have
Diagram B.

Diagram A (cf) Diagram B (£)

(rT) (R2;
(Rn
(Rvi) (R12 :•
(Rm

(mvj)
(mn)
(riiii )r tRiii2 ’
328 CHAPTER 14

C: There cannot be celestial mutations.

Rl: Mutations would make celestial bodies imperfect.
R2: Celestial bodies are perfect.
Rl 1: Celestial mutations would be superfluous.
R12: Anything in which something superfluous happens is imperfect.
Rill: Celestial bodies need only light and motion to fulfil their purpose.
Rl 111: Their purpose is to serve man.
Rl 112: The only way in which celestial bodies can serve man is to move
and send light to the earth.
Rl 1111: The purpose of all terrestrial things is to serve man.
Rill 12: The purpose of celestial bodies is the same as the purpose of
terrestrial things.
Rill 121: The principle of anthropocentrism (e.g., “man is lord of the
universe, not just the earth”; or “man is the measure of all
things”.)

For a second example consider the second version of the teleological argu¬
ment (second paragraph of reconstructed argument A above). Its explicit
structure is Diagram C; if we include the latent structure, we get Diagram D.

C: There cannot be celestial mutations.

Rl: They would be vain and useless.
R2: Nature does nothing in vain.
Rl 1: They would not occur for the benefit of man.
R12: They would not occur for the benefit of celestial inhabitants.
 PROPOSITIONAL STRUCTURE 329

R13: Only things beneficial to man or other intelligent creatures are

not vain and useless.
Rill: The only way in which celestial bodies can serve man is to move
and send light to the earth.
R121: There are no such creatures.
R1211: There are no men there.
R1212: Man is the only inhabitant of the universe.
R12121: The principle of anthropocentrism (e.g. “man is lord of the
universe, not just the earth”, or man is the measure of all
things”.)
Note than in Diagram B, R2, R12, and R1112 are horizontally latent,
whereas R11111, Rill 12, and R111121 are vertically latent. In Diagram D,
R13 is horizontally latent, whereas Rill, R1211, R1212, and R12121 are
vertically latent. Note also that R1112 of Diagram B is the same proposition
as R111 of Diagram D, and that R111121 of Diagram B is the same proposi¬
tion as R12121 of Diagram D. What does this mean? These identities are the
“points of contact” that one’s logical judgment perceived earlier.
Let us apply the notion of latent structure to the argument from vertical
fall12 which was examined in detail in other contexts earlier. The original
Aristotelian argument is the following:14

C: Not-M.
Rl: If M, then not-V.
R2: V.
Rll: If M, then W.

Now, it we include the latent structure below R2, we would get:

R21: S.
R22: If S, then V.
R221: Not-M.
R222: If not-M, then if S then V.
R2221: If not-M and S, then V.

{R222V
330 CHAPTER 14

The circularity here derives from the identity, C = R221. If we include the
latent structure under Rl, we would get:

The propositions would be:

R12: If W, then not-V.

Rl21: If not-S, then not-V.
Rl 22: If W, then not-S.
R1211: If V, then S.
R12111: Not-M.
Rl2112: If not-M, then if V then S.
Rill: V.
R112: If V, then if M then W.

The second circularity derives from the identity, C = R12111. And since
Rill = R2, if the latent structure under R2 were repeated for Rl 11, C would
then reappear a third time as one of the propositions latent vertically down
from Rill.
To summarize, we have formulated a number of basic principles for the
logical analysis of reasoning, by using as illustrative examples various argu¬
ments and passages from Galileo’s Dialogue. The main result has been that for
the understanding of reasoning, which is a pre-condition of its evaluation, one
needs to identify its propositional structure, which refers to the inferential
interrelationships among its propositions. The concept of proposition here
developed is one derivative from that of reasoning, so that the identity of
 PROPOSITIONAL STRUCTURE 331

individual propositions is dependent on the role they play in the various steps
of an argument. The principles will be indispensable for carrying out the
analyses of Chapter 16.

NOTES

1 G. Galilei, Dialogue, tr. Drake, p. 137.

2 Ibid., p. 133.
3 These rules are adapted from R. B. Angell, Reasoning and Logic, pp. 369-93. The
same is true for most of the ideas in this chapter.
4 Galilei, Dialogue, tr. Drake, p. 144.
5 Ibid., p. 19.
6 Ibid., p. 39.
7 Ibid.
8 Ibid., pp. 144-48. For more details, see Chapter 16.
9 Ibid., pp. 139-41. For more details, see Chapters 8 and 12.
10 Ibid., pp. 58-62. See also, Chapter 16.
11 Ibid., p. 19.
12 Ibid., pp. 139-41.
13 In Chapters 8 and 12, besides earlier in this chapter.
14 The abbreviations here are those of Chapter 12.
 CHAPTER 15

ACTIVE INVOLVEMENT:
THE EVALUATION OF REASONING

Besides possessing propositional structure, reasoning can be evaluated as being

good or bad. The favorable evaluation of reasoning seems to arouse less
interest among both practical reasoners and logical theorists. It is not clear
why this is so: perhaps it is because the truth shines more brilliantly than
error and hence needs less display to prevail and flourish than error does in
order to be avoided. I am inclined to think that a fundamental reason is the
fact that contrasts are more instructive and significant, for unfavorable
evaluation is much more likely to involve considerations about good and bad
reasoning if for no other reason than that one’s own unfavorable evaluation,
in order to be adequate and well-grounded, must be itself an instance of good
reasoning. In short, in unfavorable evaluation one is (presumably) being
exposed to both good and bad reasoning, the bad reasoning being unfavorably
evaluated, and the good reasoning justifying the unfavorable evaluation;
hence, one gets a better intuitive feeling and theoretical understanding of
good and of bad reasoning, and of the difference between the two. Thus, the
criticism of reasoning, as we may call the unfavorable evaluation of reasoning,
is a topic very much worth studying.

FALLACIES

The only explicit attempt by the great majority of contemporary logicians to

deal with the criticism of reasoning is the accounts of fallacies found in
elementary logic textbooks. Before we proceed further, it will be good to
examine them.1 These accounts usually contain four elements: a general
definition of the concept of fallacy, a description of various practices which
are categorized as fallacies, a classification of fallacies into various groups, and
an illustration of the descriptions of fallacies with examples.
As concerns the concept of a fallacy, there is a tendency to regard any
logically incorrect argument as a fallacy. But then, either explicitly by giving
a definition of a more specific notion, or implicitly by limiting the discussion
to only certain kinds of logically incorrect arguments, the concept of a fallacy
effectively becomes that of a type of common but logically incorrect argu¬
ment.

332
 ACTIVE INVOLVEMENT 333

The things which are categorized as fallacies include such practices as

appeals to force, to pity, to authority, ad hominem arguments, begging the
question, dicto simpliciter, converse accident, equivocation, amphiboly,
composition, division, post hoc ergo propter hoc, hasty induction, affirming
the consequent and denying the antecedent.
The classification of fallacies usually divides them into four major groups.
One group is made up of fallacies variously called linguistic, verbal, semilogi-
cal, or of ambiguity; a second group of ones variously called psychological or
of irrelevance; a third group is made up of so-called deductive, logical, or
formal fallacies; a fourth group of so-called inductive or material ones.
The fourth element of such accounts, the examples of the various types of
fallacies, is usually rather meager. It consists mostly, if not exclusively, of
more or less artificially constructed examples for the purpose of illustrating
the various descriptions of fallacies. Examples of fallacies actually occuring in
the history of thought or in contemporary investigations and controversies
are rare.
What is wrong with such accounts of fallacies? One problem concerns the
paucity of actual examples, just mentioned. It is in fact puzzling that logic
textbooks shouldn’t be able to come up with more examples of fallacies
actually committed given that fallacies are supposed to be common errors
in reasoning. One gets the suspicion that logically incorrect arguments are not
that common in practice; that their existence may be largely restricted to
logic textbook examples and exercises.
If someone doubts the fact being referred to let him consult some text¬
books.2 Let him consult Wesley C. Salmon’s Logic and discover as I, out of
methodological duty, have just done once again before writing the next
phrase, that no actual example is given for any of the fallacies mentioned
above. Let him consult R. J. Kreyche’s Logic for Undergraduates and fail to
find a single actual example of the above mentioned allegedly common
fallacies. Let him consult Cohen and Nagel’s Introduction to Logic and Scien¬
tific Method and discover that, though their account of what they call ‘abuses
of scientific method’ may be interpreted as containing some actual examples,
their account of formal, verbal, and material fallacies does not. Let him
consult Learnside and Holther’s Fallacy, The Counterfeit of Argument and
discover that, though the examples are more numerous, better, and less arti¬
ficial than in more standard logic texts, it is not clear how many of them
would remain fallacious when put in the mind or mouth of actual persons in
an actual situation.
Let him consult Monroe Beardsley’s Thinking Straight, which is more
334 CHAPTER 15

practically oriented than most books. He will find that, though the exercises
often concern actual examples, these are usually either prejudicially edited or
inadequate illustrations of the various fallacies and argument forms. If these
exercise examples had been good illustrations, one would find the author
giving more actual examples in his textual discussion. Whereas one finds only
one actual example of the fallacies mentioned above. On p. 217 (3rd edition,
1966) one finds a passage which may have been somewhat edited, since no
reference of the source is given. Even so, the passage is not a very good
example of an ad hominem fallacy, since for this purpose the author is forced
to misinterpret the argument being advanced, which is actually an argument
from analogy. Be that as it may, the point here is that even in such a practi¬
cally oriented textbook as this one, one finds only one alleged example of a
fallacy actually committed.
Finally let him consult what is perhaps the most ambitious, popular, and
widely acclaimed introductory logic textbook, Copi’s Introduction to Logic.
He will find there in the author’s account of fallacies only three actual
examples. The historical question that Stalin is reported by Harry Hopkins to
have asked Churchill at Yalta, “And how many divisions did you say the Pope
had available for combat duty?”, is given as an example of the fallacy of
appeal to force (p. 74). The second actual example, which is supposed to
illustrate the fallacy of appeal to pity, is part of the plea to the jury made by
the famous lawyer Clarence Darrow at the trial of Thomas I. Kidd (p. 78).
The third one, and in Copi’s own words a “considerably more subtle example”
of the appeal to pity fallacy, is Socrates’ refusal in Plato’s Apology to make
an appeal to pity. This example is perhaps too subtle. And rather subtle one
will find practically all of the eighteen or so quotations included in the
exercises among a larger number of less subtle but more artificial examples.
The conclusion I wish to draw from such “consultations” is not that errors
in reasoning are probably not common in real life, but that there probably are
no common errors in reasoning. That is, logically incorrect arguments may be
common, but common types of logically incorrect arguments probably are
not.
The problem I wish to raise here is, do people actually commit fallacies as
usually understood? That is, do fallacies exist in practice? Or do they exist
only in the mind of the interpreter who is claiming that a fallacy is being
committed?
The next problem I wish to raise concerns the classification of fallacies. It
does not concern the question of the various groups into which fallacies
should be subdivided, which is a problem often mentioned in textbooks and
 ACTIVE INVOLVEMENT 335

which actually turns out to be just a question of what names to give to the
various groups. The problem is that a given alleged fallacy cannot be under¬
stood to be a fallacy unless it is classified as belonging to a certain group. In
fact, the various groups derive from the various reasons given practices are
fallacies. In other words, the inadequacy of the classification of fallacies can
only derive from the inadequacy of the justification of their fallaciousness.
Hence, since the arbitrariness of the classification is usually admitted, the
arbitrariness of the justification should also be admitted.
In other words, the real problem here is whether any given alleged fallacy
is really a fallacy and why, and not what the various kinds of fallacies are. In
fact, if a fallacy is defined as a type of common but logically incorrect argu¬
ment, the various types would have to be the following: (1) arguments
claiming to be deductively valid but which are not; (2) arguments claiming to
be inductively strong but which are actually inductively weak;(3) arguments
claiming to have some inductive strength but which have none.
The problem then, with this element of logic textbooks accounts of
fallacies, is the failure to recognize that the problem is not one of classifying
a disputed practice which can be shown to be a fallacy on other grounds, but
one of showing that it is a fallacy. For there is no way for an argument to be
a fallacy without falling into one of the three above mentioned classes. This
makes one suspect that many of the disputed practices usually regarded as
fallacies may be either not fallacies or not always fallacies.
To investigate this in more detail, let us examine the second element of
textbook accounts of fallacies, the description of various devices which I
wish to call by the neutral term of “disputed practices”. One problem with
these descriptions is that they are usually prejudicial in the sense that their
fallaciousness is built right into the description. There would be nothing
wrong with this were it not for the fact that they then become logician’s
fictions or at best practices seldom found in reality (actual life). There is a
pattern in these biased descriptions, and it is the following. If the disputed
practice is a type of inductive argument, namely one claiming that the con¬
clusion is only strongly, but not conclusively, supported by the premises,
then the practice will be described as a type of deductive argument, namely
one claiming that the conclusion is conclusively supported by the premises.
If the disputed practice is a type of what might be called a partial argument,
namely one claiming that the conclusion is only partly, but not too strongly
supported by the premises, then the practice will be described as a type of
allegedly inductive strong argument. One might think that the pattern runs
out of material here, but it can be extended as follows: if the disputed
336 CHAPTER 15

practice is a type of non-argument, namely not an attempt to support one

proposition with others, then it will be described as an argument claiming
that certain propositions provide at least some support for another (the
conclusion). Finally, if the disputed practice is an argument having as
conclusion a special type of proposition, then it will be described as an
argument having another conclusion; the pattern (or shall I say the fallacy?)
is that of exaggerating the strength of the connection claimed between
various assertions or of creating one where none is claimed.
The pattern can be illustrated by considering some of the disputed
practices which textbook writers find most abhorrent. One of these is the
so-called fallacy of affirming the consequent. I need not remind the reader
that this fallacy is defined as that committed when a proposition is inferred
from a conditional of which it is the antecedent and from the consequent of
the conditional. But though it may be that all the textbook examples of
arguments having the form “If P, then Q\ Q\ .’.P” are indeed fallacies, that
does not mean that actual arguments having this form are normally fallacious.
In order to declare such an argument fallacious the logician must interpret
it to be a deductive argument, namely an argument that claims to be formally
valid. In other words, when the argument given says “If P, then Q, Q, P”
the logician must interpret this to mean “P D Q; Q; therefore, as a necessary
consequence of them alone, P”.
But the argument could also mean: “Q; the fact that P would explain the
fact that Q; therefore, no other explanation of Q being available, we may
presume that P”. Under the second interpretation there is nothing logically
wrong with the argument. Hence it is not arguments having the form of
affirming the consequent that are fallacies, but deductive arguments having
that form. That is, to show that the actual argument is a fallacy, the logician
has to argue that it is deductive. This will usually be a difficult, if not insur¬
mountable, task, since most such arguments are inductive; and evidence for
this is the fact that the textbook writer usually does not even attempt to
show that the argument does claim to be formally valid.
Another alleged fallacy beloved of textbook writers is the post hoc ergo
propter hoc manner of reasoning. This is described in W. C. Salmon’s Logic
as “concluding that B was caused by A just because B followed A” (pp.
101—2) and in Copi’s Introduction to Logic as “the inference that one event
is the cause of another from the bare fact that the first occurs earlier than the
second” (p. 82). No justification is given why these interpretations are
preferable to the following: “concluding that B was caused by A partly
because B followed A” or “the inference that one event is the cause of
 ACTIVE INVOLVEMENT 337

another from the fact, among others, that the first occurs earher than the
second”. These latter interpretations should be preferred because they are
more accurate in the sense that they correspond more closely to a type of
reasoning in which people actually engage.
Included in a third group of fallacies are usually appeal to force and
appeal to pity. A typical description of the first is “appealing to force or the
threat of force to cause acceptance of a conclusion” and of the second
“appealing to pity for the sake of getting a conclusion accepted” (Copi).
These descriptions are prejudicial in their reference to a conclusion, since a
conclusion is by definition a proposition which is part of an argument and
which is being supported by other parts of the argument called premises.
Hence appealing to force or pity to cause acceptance of a conclusion means
giving an argument in which the conclusion is supported not by appealing
to evidence but by appealing to force or pity. Of course these arguments
are fallacies of irrelevance, but irrelevant are also those notions of appeal to
force and to pity to actual appeals to force and pity. These could non-
prejudicially but along the same lines be described as “appealing to force
or to pity to cause acceptance of a certain proposition or to cause certain
action”. When so described, they can be seen to be methods, among others,
of which giving an argument is one, in order to cause acceptance of a certain
proposition. And the appropriateness of the method depends on the context.
What does not depend on the context is the truth of the claim that it is
a category mistake to regard typical actual appeals to force and to pity
as fallacies; being non-arguments they cannot be logically incorrect argu¬
ments.
The conclusion to be drawn from the above discussion is that the concept
of a fallacy as a type of common but logically incorrect argument is a chimera
since the various disputed practices usually referred to as fallacies are either
not common or not logically incorrect or not arguments. The chimera is
probably the result of either hasty generalization or over-simplification or
formalistic prejudice. That is to say, it is the conclusion of arguments having
the form of a hasty generalization, or of ones whose premises neglect too
many relevant considerations, or ones based on a formalistic prejudice. By the
latter I mean valuing too highly the form that a given argument may have and
thinking that the only useful notion of form is one such that arguments
having a given form are either all correct or all incorrect; this then leads to
regarding the disputed practices as always logically incorrect, whereas some¬
times they are and sometimes they aren’t.
My discussion may be interpreted to support and to be supported by two
338 CHAPTER 15

traditional philosophical doctrines, that evil is unreal and that the real is
rational.
To establish a connection with the doctrine of the unreality of evil it
suffices to regard fallacies as logical sins, or erroneous reasoning as logical
evil. This in turn involves the idea that truth, logical correctness, validity,
and rationality are values like goodness, beauty, and utility; the most recent
historical appearances of this idea were perhaps in the philosophy of
Benedetto Croce and of John Dewey. My discussion would then be giving a
meaning to the unreality of logical evil and would be indicating the sense
and extent to which it is true: actually occurring logically incorrect argu¬
ments are not very common because, to find one, the logician usually has to
exaggerate the strength of the logical connection between premises and
conclusion being alleged by the argument giver.
The real which in this context the rationalist would claim to be rational
is the actual reasoning practiced by people in their various activities. The
rationalist is a priori skeptical about whether fallacies are as common as intro¬
ductory logic textbooks would want to make him believe. The realist on the
other hand, that is to say the one who has a sense of reality, feels (a
posteriori) that people are simply not as irrational as those textbooks would
lead one to believe. The rationalist will try to find ways of interpreting actual
arguments so that they are logically correct; the realist will try to find as
accurate a reconstruction of actual arguments as possible. In so doing the
realist discovers that actual arguments do not tend to be logically incorrect as
much as textbooks lead one to believe; that is to say he discovers that the
rationalist is right. Realism and rationalism in this case coincide.
Of course, a thoroughgoing rationalist may be inclined to go to the absurb
extreme of claiming that no actual argument is ever fallacious. Absurd if for
no other reason than that would mean that the usual logic textbook accounts
supporting their concept of a fallacy are logically correct. To be sure the
rationalist might in this quest try to find evidence that those accounts are not
arguments, and hence not logically incorrect for categorial reasons. He may
find rationality in them by categorizing them differently. I personally don’t
know what this category would be, but I doubt very much that the ration¬
ality involved would be pedagogic rationality. The realist in me prevails here
and parts company with the rationalist.

UNDERSTANDING VERSUS EVALUATION

If reasoning cannot be criticized as easily as logic textbooks would have it,

 ACTIVE INVOLVEMENT 339

that is not to say that it cannot be criticized at all. What one can do is to
study concrete instances of argument criticism and then derive whatever
theoretical lessons one can; this corresponds to the approach advocated
earlier (Chapter 13). What I plan to do here is to study Galileo’s critiques of
various Aristotelian arguments as found in the Dialogue. The basic principles
of analysis will be the framework of propositional structure discussed in the
last chapter; moreover, each of my discussions will be based on a specific
indicated passage from Galileo’s book and will have three parts: a reconstruc¬
tion of an argument widely accepted in Galileo’s time, which will be referred
to either as the object argument or as the Aristotelian argument; a reconstruc¬
tion of Galileo’s critique as an argument about the object argument, which
will be referred to either as the Galilean criticism or as the meta-argument; a
logical commentary on my part containing discussions of concepts pre¬
supposed in Galileo’s critique and evaluations of Galileo’s meta-arguments.
Most of these evaluations of mine will be implicit in my reconstructions of
Galileo’s critiques, whose relatively obvious plausibility usually will not re¬
quire any explicit discussion.
One central lesson that will be emerging is that it is both possible and
effective to evaluate arguments ‘actively’ in the sense that the inferential
interrelationships among the propositions involved are tested by reasoning at
the level of, and largely in terms of, the object argument and by checking
whether what follows from asserted premises is the conclusions drawn in the
object argument or other propositions. This lesson is supported both by the
procedure of Galileo, who usually reaches conclusions different from the
Aristotelian ones, and by my own procedure in which I will usually reach the
same conclusions as Galileo; the active character of my procedure lies
primarily in the intricacy of the reconstructions on which my agreement with
Galileo is grounded; moreover, accuracy for all of these reconstructions is
claimed.
To appreciate the theoretical significance of this last claim it should be
stressed that one of the most fundamental problems faced by the serious
theorist and practitioner of reasoning is that of understanding vs. evaluation.
The problem derives from the fact that, besides being characterizable as some¬
thing possessing propositional structure, as explained in the last chapter,
reasoning is something susceptible of being good or bad, namely of being
evaluated favorably or unfavorably. Moreover, the following principle would
seem uncontroversial: the unfavorable evaluation of a given argument is
worthless if it involves a misunderstanding of the argument; that is, de¬
monstrated understanding is a necessary condition for responsible negative
340 CHAPTER 15

criticism. Now, to demonstrate understanding, normally an accurate recon¬

struction of the argument is needed; then the question arises as to how
elaborate, charitable, and explicit one should be in reconstructing the argu¬
ment, since assumptions implicit in the original must be stated explicity in
the reconstructed argument.
Scriven’s solution to this problem is what he appropriately calls the
Principle of Charity3 , which he formulates as follows:

the assumptions you identify mustn’t be too strong or they will be an unfair reconstruc¬
tion of the argument, since they will be fairly easily refuted even though the argument
might still be perfectly sound. On the other hand, the assumptions mustn’t be too weak,
or they won’t connect the stated premises to the conclusion. (Or they won’t support an
independent part of the conclusion that has so far received no support.) The assumption
mustn’t be a triviality of definition or fact — since it then isn’t worth mentioning. Nor
can it be a mere assertion of the fact that the arguer thinks this is a sound argument,
since that’s not worth mentioning. It must be something new but on the other hand, it
must still be true and relevant.4

As Scriven goes on to add, the Principle of Charity makes pretty good sense
in the abstract; the difficulty is, however, to put it into practice in a judicious
manner.5 The way to learn this is not by explicit formal elaborations and
qualifications of the principle, but rather by applying it to concrete instances;
or at least, such concrete applications are a pre-condition for an eventual
more formal or exact statement of the principle.6
Another aspect of the problem is that an asymmetry seems to exist in the
relations between understanding and favorable criticism on the one hand and
between understanding and unfavorable criticism on the other. That is, the
requirement of understanding plays a less important role, if at all, in favorable
evaluation: if the original argument has been reconstructed in such a way that
the reconstructed argument is sound, valid, or has significant merit, then such
a reconstructed argument acquires for the logician an importance and value
which it retains even if it is an inaccurate reconstruction. The point is as
simple as it is subtle, and in recent literature it has not escaped the explicit
mention of a philosopher such as Richard Montague.7
Why should this asymmetry exist? I believe the only way of explaining it
is to accept the kind of ‘practical’ approach to logic being advocated here.
For to admit the asymmetry is like saying that a good reconstructed argu¬
ment may be regarded as a logical achievement (relatively independently of
its accuracy vis-a-vis the original argument), whereas a bad reconstructed
argument will not be counted as a logical achievement unless it is an
accurate reconstruction. That is, the good argument has intrinsic logical value,
 ACTIVE INVOLVEMENT 341

presumably because or in the sense that logic is itself the practice of reason¬
ing, and hence a good instance of reasoning is a logical contribution. But
where is the logical value of the bad but accurately reconstructed argument?
Presumably it lies in the argument that goes from the original to the recon¬
structed argument; the accuracy of the reconstruction is another way of
talking about the goodness of this high level argument. The logical accom¬
plishment may thus again be regarded as an instance of logical practice, or
reasoning, as the “practical” approach to logic advocates.
Practical logic so conceived also explains the desirability of something like
the Principle of Charity: adequate understanding is necessary because other¬
wise our reconstruction will be inaccurate, and an inaccurate reconstruction is
a case of a bad or unsound higher level argument going from the original to
the reconstructed one.
Thus the Principle of Charity is being used everywhere in my analysis. But
this is not to say that I shall inquire directly or to any appreciable extent into
whether Galileo himself is using such a principle, i.e., whether the arguments
criticized by him are accurate reconstructions. This would be beyond the
scope of my investigation. However, it is important to note that, in view of
what I have called above his ‘active’ method of evaluation, Galileo is implicitly
conforming to the Principle of Charity. In short, Galileo’s meta-arguments
have logical interest primarily because they constitute good criticism of the
object arguments (as I have reconstructed them), whereas my reconstruc¬
tions (of object arguments and of meta-arguments) have logical interest partly
(indirectly) because of the logical value of Galileo’s meta-arguments, and
partly (directly) insofar as my reconstructions are validly grounded, i.e., are
accurate. Moreover of course, in my commentary below I will be articulating
general concepts for the understanding and evaluation of reasoning, in accord¬
ance with the approach to logic being advocated here.

NOTES

1 Such accounts have been criticized by C. L. Hamblin, Fallacies, pp. 9—49. His
criticism overlaps with mine and has had some effect. For example, I. M. Copi’s fourth
edition of his Introduction to Logic was the first re-edition of this book to appear after
Hamblin’s; in it Copi says that “thanks largely to the useful criticisms by Professor
C. L. Hamblin in his book Fallacies, some corrections have been made in Chapter 3”
(p. viii). When we look at this chapter dealing with ‘Informal Fallacies’ (pp. 72-107),
we find, however, that the ‘corrections’ really amount to nothing more than cosmetic
changes; see, for example, pp. 73, 78, and 82.
2 The books mentioned in my text are, I believe, a representative cross-section of books
342 CHAPTER 15

that have been traditionally well-known, widely-used, and influential. More recently (in
the seventies) a trend has emerged toward more realistically oriented textbooks. The best
such books, however, de-emphasize the fallacy approach: for example, Michael Scriven’s
Reasoning, which has probably the best discussion of the evaluation of actual, everyday
arguments, avoids the concept of fallacy altogether; Robert J. Fogelin’s Understanding
Argument: An Introduction to Informal Logic, which has probably the best collection of
actual, classic selections of arguments, emphasizes analysis and understanding, rather
than criticism. On the other hand, the fallacy-oriented textbooks tend to define so many
types of fallacies that most of what they call ‘fallacies’ involve faults and problems other
than errors of reasoning. The result is an unacceptable amount of violence to the funda¬
mental distinction between an argument and a non-argument; this is bound not only to
cause confusion to the student, but it affects the authors’ own judgment of what is an
argument and what isn’t. For example, Howard Kahane, on p. 28 of Logic and Contem¬
porary Rhetoric, misinterprets an insult as an error of reasoning when he gives as an
example of “the fallacy of ad hominem argument” Spiro Agnew’s famous remark that
“a spirit of national masochism prevails, encouraged by an effete corps of impudent snobs
who characterize themselves as intellectuals”. (For an analysis of some actual and non-
fallacious ad hominem arguments, see my ‘The Concept of Ad Hominem Argument in
Galileo and Focke”.) Another example is the invitation by S. Morris Engel, on pp. 63-
64 of With Good Reason: An Introduction to Informal Fallacies, to treat as a ‘fallacy of
ambiguity’ an explanatory remark found in John Maynard Keynes’s Treatise on Money.
“It is enterprise which builds and improves the world’s possessions. If enterprise is afoot,
wealth accumulates whatever may be happening to Thrift; and if enterprise is asleep,
wealth decays whatever Thrift may be doing”. (For a significant and fallacious argument
exploiting ambiguity, see my ‘Galileo’s Space-Proportionality Argument: A Role for
Logic in Historiography’; for a significant but nonfallacious use of ambiguity in reasoning,
see my ‘Logic and Rhetoric in Lavoisier’s Sealed Note: Toward a Rhetoric of Science”.)
It is obvious that the cavalier way in which these admittedly realistically fallacy-oriented
authors dispose of highly controversial issues can only encourage superficiality.
3 M. Scriven, Reasoning, pp. 71-73, 76-77, and 75.
4 Ibid., pp. 173-174.
5 Ibid., p.174.
6 This fits very well with Scriven’s conception of logic as reasoning about reasoning, or
at least with my interpretation of Scriven’s idea. See above, Chapter 13.
7 See J. F. Staal (ed.), ‘Formal Logical and Natural Languages (A Symposium)’;
Montague’s remark can be found on p. 275.
 CHAPTER 16

GALILEO AS A LOGICIAN:
A MODEL AND A DATA BASIS

For several reasons it is now time to engage in a minute critical analysis of the
arguments put forth in Galileo’s book. In part, it is time to evaluate the argu¬
ments which in Chapter 2 were merely described; in part it is useful to reana¬
lyze the Dialogue in accordance with the results of Chapter 7, which suggest
the idea of Galileo as being first and foremost a methodologist of reasoning; in
part it is important to give content to the program outlined in Chapters 11,
12, and 13; in part it is important to put to work the technical framework
developed in Chapter 14; finally one needs examples of Galileo’s critiques of
arguments, in accordance with the suggestions in the last chapter.
It turns out that almost all passages in the Dialogue can be analyzed as
being explicitly arguments about arguments. The higher level arguments will
be called meta-arguments, while the lower level ones will be called object
arguments. Object arguments are usually arguments about natural phenomena
advanced either by Aristotle, or by his followers, or by others whom Galileo
is criticizing. Meta-arguments are arguments advanced by Galileo in criticism
of various features of the object arguments and are thus only secondarily
about natural phenomena. (However, some of Galileo s criticism is substan¬
tive and pertains to the factual truth of a premise or presupposition of an
object argument, so that some meta-arguments, or parts thereof, are directly
about natural phenomena.)
It is the extent of the context under consideration and the explanatory
power of the reconstruction which determine whether or not one of Galileo’s
physical arguments is regarded primarily as part of a meta-argument: the
governing consideration is whether or not a particular physical argument can
be reconstructed as having the function of serving to criticize some object
argument contained in the passage. For example, Galileo’s own arguments
about natural motion (F43—57),1 about the ship experiment (FI70—5), and
about centrifugal motion (F217—24, 238—44) are primarily meta-arguments
about corresponding Aristotelian or geostatic arguments. Conversely, some of
Galileo’s met a-arguments occur in a context such that they are primarily
designed to support some of his physical theses by criticizing counter-argu¬
ments; such are his arguments criticizing the geostatic explanation of tides
(Fourth Day) and of the motion of sunspots (F372-83), those criticizing the

343
344 CHAPTER 16

Peripatetic explanation of the telescopic appearance of the moon (F95—112)

and of the moon’s secondary light (FI 12—24), and the one criticizing the
Aristotelian notion of the heliocentrism of planetary motions (F346—68).
The analysis of all meta-arguments in the Dialogue would have been un¬
necessarily long and excessively tedious; hence the following criteria of
selection have been used. First, I have neglected meta-arguments which are
critiques of counter-arguments and whose immediate context makes them
part of a physical argument which is not itself a meta-argument; examples are
the second group of arguments in the preceding paragraph. Second, I have
neglected passages which are primarily expository or narrative in character:
e.g., the discussion of the earth-moon similarities (F87—95) and differences
(F124—32); the discussion of Aristotle’s authority (F132—9); the presenta¬
tion of the probable arguments for the earth’s diurnal motion (FI39—50);
and the preliminary presentation of the geostatic arguments (FI50—9). Third,
I have neglected physical arguments whose immediate context does not make
them part of a meta-argument, such as those concerning conservation and
composition of motion (FI80—93), and concerning retrograde motion
(F368—93). The fourth criterion I have used is that of manageable length, so
that by starting at the beginning of the book I stopped only when I thought
I had gone through a sufficiently large number of analyses, mindful of quality
and variety, besides quantity. This led me to stop with the centrifugal motion
argument (F214-44) which is approximately at the middle of the book
(pagewise). This place is a very proper one since the rest of the Second Day
and the second half of the Third Day (F383—442) have already been
explicitly analyzed as meta-arguments in the analytical summary of my
discussion of the book’s logical structure (Chapter 2); similarly, Galileo’s
discussion of the 1572 nova at the beginning of the Third Day (F299-346) is
obviously, and was explicitly analyzed as, a meta-argument about Chiara-
monti’s argument that the nova was sublunary.
Finally, it will be seen below that Galileo’s meta-arguments are very
suggestive and instructive logically and lend themselves very readily to the
formulation of principles and clarification of concepts at a first level of
logical theorizing; this I have done in the subsections labeled ‘Comments’. It
should also be noted that the discussion in each section below has two
aspects, one of logical theory and one of logical practice; this is reflected in
the title of each section which includes both the name of an actual argument
and one or more notions useful in the theory of reasoning. This two-sidedness
is analogous to the concreteness of the methodological content elaborated
earlier (Chapter 5), where methodological ideas were being illustrated with
 GALILEO AS A LOGICIAN 345

scientific examples. In the present discussion ideas in the theory of reasoning

are illustrated by actual reasoning.
Some clarifications of a technical nature are needed before we proceed
further. Each analysis below consists of three parts: an object argument, a
meta-argument, and comments. The object arguments and meta-arguments
are reconstructions of the corresponding passage, and all their propositions
have been numbered basically in accordance with the standard labeling dis¬
cussed earlier (Chapter 14): in order to distinguish the propositions of the
object argument and of the meta-argument in a given section, the letter ‘A’
precedes the numbers of all propositions of the former (as a reminder that
Aristotelian arguments are usually meant), and the letter ‘G’ precedes all the
numbers of all propositions of the latter (as a reminder that Galilean argu¬
ments are being referred to). The object argument will usually have only one
final conclusion, to which the single digit ‘1’ has been assigned, so that all its
supporting propositions can then be numbered in the standard way with
numerals whose first digit will be ‘1’. The meta-argument will usually have
more than one independent subargument, and hence the distinct final con¬
clusions of these have been assigned the labels Gl, G2, G3, G4, etc., (the ‘G’
being a reminder that Galileo’s own arguments are being referred to); all
supporting propositions for each such final conclusion, say Gn, are then
numbered in the standard way by adding digits to the Gn in accordance with
the rules. Such notation achieves two purposes simultaneously: it describes
unambiguously the logical-propositional structure of all object arguments and
meta-arguments to be discussed, and it labels uniquely each Aristotelian and
Galilean proposition in the passages under discussion.
Given such numbering the construction of the corresponding structure
diagrams is a merely mechanical task, and hence such diagrams have been
drawn only in those few cases which seemed particularly interesting. For
added convenience, the various labels in a diagram have not been enclosed in
a circle or closed figure.
On some occasions in my comments, reconstructed arguments are stated
for which it is relatively more debatable whether they are contained in the
text, though it is clear that parts are. In such case the label ‘FG’ is used as a
prefix before the standard numbering by numerals.
In stating the various object arguments and meta-arguments, the labels of
their constituent propositions are written in parenthesis just before the
propositions themselves. On the other hand, the statement of a meta¬
argument will usually also contain labels of object argument propositions
used to express various claims about those propositions; in such cases no
346 CHAPTER 16

parentheses have been used around the labels of these object argument
propositions, so as not to create visual confusion with the labels of the meta¬
argument propositions. Analogously, in my comments, I refer to propositions
of both object arguments and meta-arguments by using their labels without
parentheses.
All references are to Volume VII of Galileo’s Opere. Readers interested in
logical theory and the theory of reasoning may simply ignore these refer¬
ences; others are thereby provided with the means of engaging in that type of
logical practice which is the reconstruction of arguments from original texts.
Finally, the order of the sixteen sections that follow is merely that in
which the arguments appear in Galileo’s book. Since from the point of view
of the theory of reasoning this order is neither one of ascending nor one of
descending significance, it is advised that theorists of reasoning do not read
them in this order, and that they read them in an order best suited to their
special interests. For example, those interested primarily in formal logic
should read Sections 4, 8, 12, and 11 before any others. Those primarily
interested in informal logic should begin with Sections 15, 3, 9, and 6. Those
primarily interested in the logic of science should begin with Sections 2, 14,
16, and 10. Others can use the various section titles, which attempt to be
descriptive; a list of these titles may be found at the very beginning of the
next chapter.

1. CONCLUSIONS VS REASONS VS CAUSES, AND THE

THREE-DIMENSIONALITY ARGUMENT (F33-38.24)2

Perhaps it is a mere coincidence, but the fact is that the Dialogue begins with
an implicit discussion of the fundamental logical distinctions between a
proposition and its justification and between a proposition and its implica¬
tions. The argument under discussion is the following:
(Al) The world is perfect because (All) it has the three dimensions of
length, width, and depth and (A12) these are all the dimensions that exist;
and (All) the world has three dimensions because (Al 11) three is a very
special number (in that three is (Al 111) the number of parts that everything
has, namely beginning, middle, and end; (All 12) the number used in sacri¬
fices to the Gods; and (Al 113) the least number of things required before the
word ‘all’ can be applied to refer to them collectively.)
The relevant proposition here is the three-dimensionality of the world,
All. Galileo is here accepting this proposition, but neither its alleged implica¬
tion (All, :. Al), nor its alleged justification (All 11, Al 112, All 13;
 GALILEO AS A LOGICIAN 347

A111; All). In short, Galileo is agreeing that the world is perfect and
that it has all three dimensions, but denying that there is a connection
between the two propositions such as to ground pertection on three-dimen¬
sionality; and he is also denying that three-dimensionality is to be grounded
on the special properties of the number three. He justifies his denial of (A11,
Al) by suggesting an alternative justification of the perfection of the world,
namely that it is the work of God; and he justifies his denial of (Al 11, /. Al 1)
by arguing that the world is three-dimensional because exactly three mutually
perpendicular lines can be drawn through any given point.
If we call the Aristotelian argument here criticized by Galileo, the three-
dimensionality argument, then we may say that Galileo’s own argument is
one about the three-dimensionality argument. Galileo is arguing that (Gl) the
three-dimensionality argument is doubly invalid, because (G11) perfection is
improperly grounded on three-dimensionality, and because (G12) the latter is
improperly grounded on the special properties of the number three; and he
justifies proposition Gil by saying that (Gill) the cause of the worlds
perfection is God, and he justifies proposition G12 by saying that (G121) the
cause of the world’s three-dimensionality is the fact that three and only three
mutually perpendicular lines can be drawn through a given point.
How correct is Galileo’s own argument? I would agree with Galileo’s fmal
conclusion, Gl, and with the two reasons he gives, Gil and G12. That is, I
accept Gl, Galileo’s justification (Gil, G12, Gl), and also propositions
Gil and G12. Why do I accept G11? For Galileo’s reason or for some other
reason? Galileo’s justification of G11 could be elaborated as follows:

(Gl 12) The world is perfect.

(Gill) The cause of this perfection is God.

When so elaborated I would accept the justification, but not the premises as
such. My difficulty would be with accepting G112 and consequently with
accepting Gill. Do I have a justification of Gil? To be sure, I would say
that I don’t see that there is any connection between perfection and three-
dimensionality; but this would be merely another way of expressing Gil.
However, I believe I could use my rejection of G112 as my reason for Gil;
that is, I would say that the perfection of the world is improperly grounded
on its three-dimensionality because the world is not perfect.
What of G12? Here I would accept both Galileo’s premise G121, and his
justification (G121, /. G12), which one may want to elaborate by adding as a
latent proposition (G122) that the world is three-dimensional.
Thus my own argument (F) about Galileo’s argument (G) would be the
348 CHAPTER 16

following: All the inferential steps of Galileo’s argument (Gil, G12, Gl;
Gill, G112, Gil; and G121, G122, G12) are acceptable; moreover, all
of its propositions are acceptable except for G112 and Gill; these two
propositions together amount to saying that the world is perfect because God
created it.
Let us summarize as follows. Some of the important propositions being
discussed are:
(a) The world is perfect.
(b) The world is three-dimensional.
(c) The number three has special properties.
(d) The world is perfect because God created it.
(e) The world is three-dimensional because three is the exact number
of mutually perpendicular lines that can be drawn through a given
point.
The Aristotelian argument is:

(a)
I
(b)
I
(c) .
Galileo’s critique is:

not-(b,a) not-(c,b)

(a)\d) (b)^Xe).

My conclusions are:
(1) If a and d, then not-(£>,a).
(2) If b and e, then not-(c,b).
(3) b.
(4) e.
(5) .'. not-(c,.'. b).
(6) Not-a.
(7) Not-d.
(8) Not-a,not-(Z>,a).
Formalism aside, Galileo’s technique here is first of all one of partial agree¬
ment, with respect to (a) and (b). Then he gives causal explanations of these,
namely (d) and (e). Then he uses these explanations together with the facts
being explained to refute the alleged logical connection between (a) and (b),
and the logical ground for (b). The net result is the logical dissociation of
 GALILEO AS A LOGICIAN 349

propositions (a) and (b) and their causal association, respectively, with God
and with three-fold mutual perpendicularity.
Another logical lesson we may derive is that it is very important to
distinguish our conclusions from our reasons, and reasons from causes. Con¬
clusions and reasons are distinct because one may accept a conclusion, e.g.
(b), but not the alleged reason, e.g. (c); or one may accept a reason, e.g. (b),
but not the conclusion drawn from it, e.g. (a). Causes and reasons are distinct
in the sense that a cause is given in a context where some fact is agreed upon
but what produces it is being disputed, e.g. when Galileo says that the world
is perfect because God created it, or that the world is three-dimensional
because only three mutually perpendicular lines are possible; whereas a reason
is being given in a context where some conclusion is disputed and accepted
facts are appealed to in order to settle the dispute; for example, for Galileo
what is in question is neither the world’s perfection nor its three-dimensiona¬
lity but whether three-dimensionality implies perfection and is implied by the
special properties of the number three.

2. LOGICAL EVALUATION, AND THE NATURAL

MOTION ARGUMENT (F38.25-5 7.4)

The Aristotelian argument being criticized here is (F38.25—43.11):

(A 1231) There are two kinds of simple motions, straight and circular.
(A 123) .'.There are two kinds of natural motions, straight and circular.
(A122) But, straight natural motion is toward or away from the center
of the universe.
(A121) And, circular natural motion is around the center of the
universe.
(A 12) .'.Natural motions are toward, away from, and around the center
of the universe.
(All) But, for each natural motion there is a corresponding natural
element.
(Al) .'.Earth and fire are the elements whose natural motions are,
respectively, toward and away from the center of the universe;
and the heavenly bodies are the bodies whose natural motion is
circular around the center of the universe.

Galileo’s critique (F38.25—43.11) is as follows: (Gl) The basic premise of

the argument, A1231, may be accepted since (Gl 1) simple motions are those
along simple lines, (G12) simple lines are those whose parts are congruent to
350 CHAPTER 16

one another, and (G13) the example of the cylindrical helix may be dismissed.
However, (G2) none of the conclusions drawn from this basic premise follow
from it: (G21) proposition A123 does not follow because (G211) what
follows from A1231, i.e. from a consideration of only such simple motions,
is that there is only one type of natural motion, the circular, and that straight
motion is merely the simplest means of acquiring the natural state of rest at
the proper place or of circular motion; (G22) proposition A12 does not
follow because (G2211) straight natural motion would be natural motion
along any straight line, and (G2212) circular natural motion would be natural
motion alone any circular path, and hence (G221) both A122 and A121 are
false; (G23) proposition A1 would not follow from All and A12 because
(G231) all that would follow from them is the existence of some elements
whose natural motion is straight and of one whose natural motion is circular,
and (G232) some independent justification would be needed to equate these
elements with earth, fire, and the heavenly bodies.
It is important to notice that Galileo seems to be concentrating on criti¬
cizing the validity of the Aristotelian argument, that is, the various inferential
steps of the argument supporting Al, which is very appropriate since this
argument was supposed to be an a priori one. The criticism has the very
interesting form of claiming that one proposition does not follow from
another because what really follows from it is some other proposition dif¬
ferent or inconsistent with the first. This type of logical evaluation might be
labeled ‘active’, for the critic engages in reasoning at the same (object) level as
the argument being evaluated, rather than reasoning only at the metalevel, so
that he produces a counterargument. The evaluation is nevertheless logical
(and not merely scientific) since it centers around the interrelationships of
propositions. It should also be noted that proposition G211 is not being justi¬
fied within this argument, but rather that it will be justified in the next one
about natural circular motion; it may be regarded as one further conclusion
that may be drawn from the final one in the next argument. It is this logical
connection which gives the natural circular motion argument of the next
section scientific value in the sense of the science of logic, independently of
its scientific merits in the sense of substantive natural science; that is, the argu¬
ment to be reconstructed in the next section has without doubt the logical-
evaluative function of showing that the Aristotelian natural motion argument
is invalid, in particular that proposition A123 does not follow from A1231.

Galileo’s natural circular motion argument (F43.12-57.4)

(G122) Straight motion cannot be the natural state of integral bodies.
 GALILEO AS A LOGICIAN 351

(G121) Straight motion cannot naturally be perpetual.

(G12) .'.Straight motion cannot be natural.
(G112) But, circular motion can be the natural state of integral
bodies.
(Gill) And, circular motion can naturally be perpetual.
(G11) .'.Circular motion can be natural.
(Gl) .'.The only type of natural motion is circular. (I.e., all natural
motion is circular.)
And

(G23) Straight motion is a good means of restoring order out of

disorder.
(G22) Straight motion is a possible means of creating order.
(G21) Straight motion is a natural way of acquiring circular motion.
(G2) .'.Straight motion is the simplest means of acquiring the natural
state of rest at the proper place or of circular motion.

Now, (G122) straight motion cannot be the natural state of integral bodies
first because (G122al) if it were then such bodies would be changing place,
and (G122a2) if integral bodies change place that means that the universe is
not in perfect order, and second because (G122bl) if it were then such bodies
would have a tendency to move through an infinite distance, since (G122bl 1)
a straight line is in principle infinite.
And, (G121) straight motion cannot naturally be perpetual because
(G1211) when straight motion is violent it obviously cannot be perpetual,
and (G1212) when straight motion is nonviolent (i.e., spontaneous or
‘natural’) it can be shown that it cannot be perpetual. This is so because
(G12121) nonviolent straight motion must be accelerated, and (G12122)
accelerated nonviolent straight motion cannot be perpetual. The reason for
the latter is that (G121221) when a body moves with accelerated nonviolent
straight motion it must be approaching a place toward which it has a natural
inclination (for (G1212211) otherwise it would not be accelerating), and
(G121222) when it has reached this place the acceleration must stop. And the
reason for the former is that (G121211) nonviolent straight motion is the
simplest way for a body to move from one place to another toward which it
has a natural inclination, (G121212) in moving from one place to another
toward which it has a natural inclination a body will acquire speed continu¬
ously and gradually (since (G1212121) there would be no reason to acquire
one degree of speed rather than another), and (G121213) if a body acquires
speed continuously and gradually then its motion is accelerated.
352 CHAPTER 16

Also, (Gil2) circular motion can be the natural state of integral bodies
because (G1121) a natural state of circular motion preserves order.
Finally:

(Gill 1113) acceleration occurs naturally when a body approaches the

point toward which it is inclined to move;
(Gil 11112) retardation occurs naturally when a body moves further
away from the point toward which it is inclined to move;
(G1111111) circular motion around a point toward which a body is
inclined can be interpreted as motion where the body is
simultaneously approaching and moving further away
from that point;
(Gllllll) circular motion around a point toward which a body has
natural inclination would be subject to simultaneous
acceleration and retardation;
(Gil 111) such circular motion would be uniform;
(Gil 11) it could be perpetual;
(Gill) circular motion can naturally be perpetual.

Comments. The accuracy of this reconstruction may be justified as follows.

The argument in support of G122 can be found on F43.12—43.30; the
argument in support of G121 can be found partly on F56.20-56.25 (the part
containing proposition G12122), and partly on F44.17—45.10, for the part
containing G12121. The argument supporting G111 can be found on F56.9-
56.20. The argument supporting G112 can be found on F55.3—55.4 and
F56.1—56.9. Proposition G23 can be found on F43.30—44.8; proposition G22
on F44.8—45.10 and F53.13-55.1; and proposition G21 on F45.ll—53.6.
Though the structures are obvious from the numbering, it is useful to have
them visually available in the following figures:

G1 G122

Gil G12 G122al G122a2 G122bl

Gill G112 G121 G122 G122bl1

 GALILEO AS A LOGICIAN 353

Gill G121

Gllll G1211 G1212

Gimi G12121 G12122

Gllllll G121211 G121212 G121213 G121221 G121222

Glllllll G1111112 G1111113 G1212121 G1212211

Finally, it is clear that Galileo’s natural circular motion argument involves

a concept of natural motion different from that of the Aristotelian argument.
For the latter, natural motion is identified with simple motion; for the
former, natural motion is simple motion which can be a perpetual state.
There is another concept of natural motion lurking behind both arguments,
namely that of natural as nonviolent or spontaneous; this is not the contro¬
versial concept in this context since such a concept would imply that only
observation could tell us which types of motion were natural, whereas the
present discussion is attempting an a priori determination.
Is Galileo’s counter-argument sound? All the inferential steps seem to me
rather plausible, but beyond saying this there is no way of answering the
question, other than by comparison with Aristotle’s argument: A1231,
.’. A123. Whereas this inference seems rather arbitrary, Galileo’s counter¬
justification makes use of several Aristotelian ideas, and it is in this sense that
what really follows from A1213 is G1 and G2 instead.
It should be noted that G1 is not saying that all circular motion is natural,
nor that all motion is natural in the sense that all movable bodies have only
circular motion. But is G1 telling us anything about actual motion? It is
telling us that if an actual motion is circular, it may be natural, and if it is
noncircular, it definitely is not. Primarily G1 tells us something about what
can and cannot be rather than what is or is not. It will help Galileo to prove
that the earth can move, that it is not impossible for it to rotate and revolve.
But this is not to say that the earth does move. Similarly, straight motion is
merely unnatural, rather than unreal.

3. EQUIVOCATION, COMPOSITION, CIRCULARITY, AND THE

UNIVERSE-CENTER ARGUMENT (F5 7.5-6 2.27)

The object argument. (A 15) The parts of the element earth and of the element
354 CHAPTER 16

water, namely heavy bodies, move naturally straight downwards, for (A 151)
if one drops a heavy body from the top of a tower with straight and vertical
edges, the body can be seen to fall along those edges and to land at the foot
of the tower. (A14) Natural straight downward motion is motion toward the
center of the universe, for (A1411) the natural motion of heavy bodies is
contrary to that of light bodies, (A 1412) the natural motion of light bodies
is straight-upwards, (A1413) straight-upward motion is toward the circum¬
ference of the universe, and hence (A141) the natural motion of heavy bodies
is toward the center of the universe. (A13) The parts of the elements fire and
air can be seen to move naturally upwards. (A 12) Natural upward motion is
straight motion toward the inside of the lunar orb. (A11) Whatever applies to
the parts applies to the whole. Therefore, (Al) the natural motion of the
whole Earth is straight toward the center and of Fire straight away from the
center.
The meta-argument. (Gl) The argument is unsound in that its last step is
really an equivocation, for (Gill 11) the sense of ‘natural’ in which the
premises are true is that of ‘spontaneous’ or ‘nonviolent’, and hence (Gl 111)
the sense in which Al follows is that the spontaneous motion of the whole
earth would be straight toward the center (if it had been violently removed
from it), but (Gl 112) the sense in which Al is being asserted is that the po¬
tentially perpetual state of the whole Earth is straight motion, for (Gl 1121)
this is the sense in which Al conflicts with a claim of the argument just
presented (Galileo’s natural circular motion argument), namely its proposi¬
tion G122, so that (Gill*) if ‘natural’ in Al means ‘spontaneous’ then the
inference is valid but irrelevant, and (Gill**) if ‘natural’ in Al refers to a
potentially perpetual state, then Al is relevant but obviously cannot be
inferred from the premises in the sense in which they are being asserted to be
true; it follows that (Gil) the argument is either valid but irrelevant or
relevant but invalid.
Morever, (G2) the first premise, A15, is questionable insofar as it claims
that the downward motion is straight, for (G211) the actual straightness of
free fall cannot be inferred from its apparent (visible) straightness, and hence
(G21) proposition A15 does not follow from A151.
Third, (G3) premise All is absurd, for (G31) if premise A15 is interpreted
in a sense in which it is true, namely to mean that the spontaneous motion of
heavy bodies is toward the center of the earth, then these two premises would
imply that the whole earth would spontaneously move toward its own center
if it were violently removed from whatever place it is in, and (G32) this is
intrinsically impossible.
 GALILEO AS A LOGICIAN 355

Next, (G4) the second premise, A14, is questionable insofar as it refers to

the center of the universe, and the supporting argument begs the question in
two ways. (G41) First, (G4111) proposition A1413 is not obvious since
(G41 111) it speaks of the universe circumference; therefore, (G411) one may
ask for a justification; (G412) the justification would have to be on the basis
of the proposition that straight-upward motion is seen to be away from the
surface of the earth; (G413) if so, then the argument would have to be:
“(A141321) Straight-upward motion is seen to be away from the earth’s
surface; (A14132) straight-upward motion is toward any circumference
greater than and concentric with the earth’s; but (A14131) the universe
circumference is concentric with the earth’s; (A1413) straight-upward
motion is toward the universe circumference”; but (G414) proposition
A14131 which is here being assumed is essentially identical to the one it is
helping to prove, namely proposition A14. (G42) Second, (G4211) proposi¬
tion A141 does not follow from its stated reasons, A1411, A1412, A1413,
since (G421 111) in a spherical finite universe any straight motion whatever is
toward its circumference, but only those motions toward the circumference
which are along diameters of the universe have contrary motions which are
toward its center, and hence (G42111) the natural motion of heavy bodies
could be contrary to ‘toward the circumference’ and yet not toward the
center of the universe; therefore, (G421) if proposition A141 is to follow,
proposition A1413 must be interpreted to mean that upward motion is
toward the circumference and along a diameter of the universe; but (G422) if
this is the meaning of A1413 then it needs a justification (since (G4221) the
only thing we know is that upward motion is along diameters of the earth),
and (G423) if it needs a justification the only one that could be given would
be based on the proposition that the centers of the earth and of the universe
coincide, and (G424) if A1413 is justified by assuming such an identity then
the argument supporting A14 would beg the question (since (G4241) to
assert such an identity is essentially the same as to assert proposition A14).
Comments. The accuracy of these reconstructions may be justified as
follows. ‘A151, .'. A15’ can be found on F58.28—58.35. The subargument
supporting A14 can be found on F58.38—59.7. The immediate reasons
supporting A1 can be found on F57.10—57.19. Galileo’s G1 argument is on
F57.20—57.29 and F56.25—57.10. The G2 critique is on F57.29—57.34 and
F59.16—59.18. The G3 argument is on F57.34—58.4. The G41 subargument
is on F60.10—60.24. The G42 argument is on F58.5—58.8 and F60.25—
61.15.
The last step of the object argument as stated gives rise to a good example
356 CHAPTER 16

of what modern logicians call the fallacy of composition, which is an argu¬

ment where it is concluded that the whole has a given property just because
its parts have that property. Of course, in the argument as stated one of the
premises, All, explicitly asserts this; hence, technically speaking, the argu¬
ment avoids the fallacy, In practice, however, it makes little difference
whether A11 is explicitly stated or not, for the problem of whether it holds
in any given case must be solved on a case-by-case basis. Galileo s critique G3
may be taken to be a proof of the formal invalidity of arguments of this type;
Galileo does this by calling attention to a property which it is logically im¬
possible for a whole to have, though it is in fact true that the parts do have
it.
It is interesting to note that Galileo’s critique of the validity of the last
step of the Aristotelian argument, namely his G1 argument, is essentially of
the form: the conclusion does not follow from the premises because what
follows from them is some other proposition, and this is so because if we take
the premises to be true then we could validly infer a proposition about the
spontaneous (or ‘natural’ in the sense of spontaneous) motion of the whole
terrestrial globe, and this proposition is different from A1 because A1 as
asserted must be speaking of natural motion in the sense of a potentially
perpetual state of motion. Such an active evaluation of reasoning conforms to
Galileo’s previous critiques.
Galileo does not in this context prove G211, but he does so later on
FI 64—9. One way of stating that later criticism is that all that follows from
A151, namely from apparent straight fall, is that if the earth does not rotate,
then free fall is actually straight; the crucial consideration in the proof is
that if the earth rotates then free fall is actually slanted to the earth’s
surface.
The argument supporting G4211 seems to criticize the validity of‘A 1411,
A1412, A1413, A141’ by the counterexample method of showing that it is
possible for the premises to be true while the conclusion is false.
Finally, it may be useful to summarize the two ways in which the object
argument begs the question. They both involve the subargument supporting
A14: if one of its supporting reasons is to be acceptable it needs to be
justified on the basis of a proposition essentially identical to A14, and if one
of its supporting inferential steps is to be correct one needs to assume the
same essentially identical proposition. We may also say that the argument
begs the question in the sense that when its latent structure is constructed,
the argument becomes doubly circular by having A14 reappear twice as an
underlying reason for itself.
 GALILEO AS A LOGICIAN 357

4. CONTRADICTION, AD HOMINEM, EQUIVOCATION, AND

THE CONTRARIETY ARGUMENT (F62.28-71.33)

The object argument. (A 12)Change does not occur unless there is contrariety;
(A11) there is no contrariety among heavenly bodies, since (A111) heavenly
bodies have circular motion, (A112) no motion is contrary to circular
motion, and (A113) contraries have contrary motions; therefore, (Al)
heavenly bodies are unchangeable. Now, to show that no motion is contrary
to circular motion, consider that (A1121) there are three types of simple
motions, straight-toward, straight-away from, and around the center, that
(A1122) the two straight motions are contrary to each other, and that
(Al 123) one thing can only have one thing as contrary. [F62.31—63.17]
The meta-argument. (Gl) The Aristotelian argument supporting Al is
invalid since (Gl 1) if argument Al3 were otherwise correct the valid conclu¬
sion to be inferred would be that either the earth as well as the heavenly
bodies is unchangeable, or that the heavenly bodies as well as the earth are
changeable, or that there is no connection between change and the straight/
circular motion distinction. [F63.18—64.11]
Second, (G21) the argument’s main intended consequence is that the earth
stands still, but (G22) its main premise, A12, is more difficult to ascertain
than this consequence. For (G211) the argument’s final conclusion, Al,
would be used as follows: “since heavenly bodies are unchangeable, and since
the earth is changeable, the earth is not a heavenly body, and therefore the
earth does not have the annual motion”. And (G221) the earth is a very large
and accessible body, but (G222) the connection between change and con¬
trariety is impossible to fmd in most phenomena, e.g., in the apparently
spontaneous generation of some insects, in the different rates of change in
most living things, and in changes resulting from the transposition of parts.
[F64.ll-65.24]
Third, (G3) the contrariety argument is problematic since (G31) if one
accepts it then one would have to accept the following self-contradictory
argument: “(A42) heavenly bodies have contraries, since (A421) heavenly
bodies are unchangeable, (A422) terrestrial bodies are changeable, and(A423)
changeability and unchangeability are contraries; but (A41) all bodies which
have contraries are changeable; therefore, (A4) heavenly bodies are change¬
able”. One would have to accept this argument since (G311) propositions
A422 and A423 are uncontroversial; (G312) proposition A421 is the conclu¬
sion of the contrariety argument, Al; (G313) proposition A42 follows from
its immediate reasons; and (G314) if one accepts the contrariety argument
358 CHAPTER 16

then proposition A41 could be justified as follows: “(A41) all bodies which
have contraries are changeable, since (A411) change does not occur unless
there is contrariety and (A412) this means either that (a) change does not
occur to a body unless it has a contrary, or that (b) change does not occur in
a region unless there is contrariety within that region, and (A413) the latter
cannot be; for (A4131) proposition (b) would imply that there is no change
within the region of the elements earth and water and no change within the
region of the elements fire and air, and (A4132) if so then terrestrial bodies
would be unchangeable”. [F65.25—67.18]
Fourth, (G4) proposition A11 is questionable since (G41) Aristotle would
regard heavenly bodies as the denser parts of the heavens, (G42) if heavenly
bodies are regarded as the denser parts of the heavens then differences of
rarity and density exist in the heavens, and (G43) if differences of rarity and
density exist in the heavens then a change-producing contrariety exists in the
heavens; for (G4311) differences of rarity and density give rise to the light/
heavy contrariety of terrestrial bodies, (G4312) this contrariety gives rise to
the upward and downward spontaneous motions, and (G4313) these motions
are allegedly the source of terrestrial changes, and hence (G431) differences
of rarity and density may be regarded as the cause of terrestrial changes;
moreover, (G4321) the cause of terrestrial as well as celestial differences of
rarity and density is the quantitative difference of more or less matter in a
given space, (G4322) the cause of terrestrial differences of rarity and density
is not the qualitative difference of heat and cold (since (G43221) the density
of solid substances changes little when their degree of heat changes signif¬
icantly), and hence (G432) the cause of terrestrial differences of rarity and
density is the same as the cause of celestial differences of rarity and density.
[F67.19—69.24]
Fifth, (G5311) premise A12 would be supported on the basis of the
changes and of the up and down motions of terrestrial bodies; but (G5312)
the term ‘bodies’ in the phrase ‘terrestrial bodies’ refers to the parts of the
terrestrial globe; therefore, (G531) the meaning of the term ‘bodies’ relevant
to premise A12 is that of ‘parts of integral (whole) bodies’. Whereas, (G521)
the meaning of the term ‘bodies’ in premise All is that of ‘integral (whole)
bodies’ since (G5211) A11 is supported on the basis of claims about integral
(whole) bodies. But, (G51) if the conclusion in A1 is to follow from the
premises then the term ‘bodies’ must be used with the same meaning in both
premises; (G52) if ‘bodies’ means ‘parts of integral bodies’ then premise A11
is groundless; (G53) if ‘bodies’ means ‘whole bodies’ then premise A12 is
groundless; therefore, (G5) if the conclusion follows then the argument is
 GALILEO AS A LOGICIAN 359

groundless (i.e., either the conclusion does not follow or it is based on

groundless premises). [F69.25—71.9]
Comments. (1) Galileo gives no explicit argument to justify Gil. However,
the following argument is implicit in the text in F63.18-64.11, and so I will
still use the label ‘G’ and a numbering to refer the various propositions to its
final conclusion G11:
(Gill) There are really two sides to the contrariety argument, Al, which
may be reconstructed as follows: “(A21) change does not occur unless there
are contraries; (A22) there are no contraries unless there are contrary natural
motions; (A23) contrary natural motions are only those straight motions
made in contrary directions; (A24) there are only two natural straight
motions made in contrary directions, namely toward and away from the
universe center; (A25) natural straight motion toward the universe center
belongs to the elements earth and water, and motion away from the center
belongs to the elements fire and air; therefore, (A2) change does not occur
except among terrestrial bodies”. And “(A322) bodies whose natural motion
is circular have no contraries, since (A32211) the other two types of simple
motion are contrary to each other, and (A32212) one thing can only have
one contrary, and hence (A3221) there is no circular motion contrary to
natural circular motion; but, (A321) where there is no contrariety there is no
change; therefore, (A32) bodies whose natural motion is circular are
unchangeable; but, (A31) natural circular motion belongs only to heavenly
bodies; therefore, (A3) only heavenly bodies are unchangeable”.
Now, (G112) the ‘natural’ motions being mentioned in these arguments
ought to be conceived either as nonviolent, spontaneous motions or as
potentially perpetual motions, since (G1121) the Aristotelian concept of
natural motions as simple has already been criticized.
(G1131) If ‘natural notion’ is conceived as potentially perpetual motion
then there are no contrary natural motions, since (Al 1311) it has already
been argued that in this sense of ‘natural’ there is only one type of natural
motion, namely the circular; but (G1132) if there are no contrary natural
motions then proposition A31 ought to be replaced by the proposition that
natural circular motion belongs to all integral bodies, to the earth as well as to
the heavenly bodies; (G1133) if the latter proposition replaces A31 then the
conclusion in argument A3 would be that the earth as well as the heavenly
bodies are unchangeable; therefore, (G113) if ‘natural motion’ is conceived as
potentially perpetual motion then the valid conclusion is that the earth as
well as the heavenly bodies are unchangeable.
(G1141) If ‘natural motion’ is conceived as nonviolent, spontaneous
360 CHAPTER 16

motion then straight motions toward and away from the center of their
wholes belong, respectively, to the denser and rarer parts of all integral bodies,
heavenly as well as terrestrial; (G1142) if this last clause is true then it ought
to replace proposition A25 in the argument supporting A2; (G1143) if A25 is
so replaced then the valid conclusion is that heavenly bodies as well as the
earth are changeable; therefore, (G114) if natural motion is conceived as
nonviolent, spontaneous motion then the valid conclusion is that the heavenly
bodies as well as the earth are changeable.
Finally, (G115) if ‘natural motion’ is conceived in both ways then it
follows that there is no connection between change and the straight/circular
distinction; for, (G1151) propositions G113 and G114 imply that if‘natural
motion’ is conceived both ways then the earth and heavenly bodies are both
changeable and unchangeable, and (G1152) the source of this contradiction is
the proposition that change does not occur unless there are contrary motions,
and (G1153) if we deny this proposition we get that it is not true that change
does not occur unless there are contrary motions, and (G1154) this denial
means that it is not true that change occurs if there are straight motions and
does not if there are circular motions, and (G1155) this means that change is
unconnected with the difference between straight and circular motions; now,
proposition G1152 is true because (G11521) in the following argument the
contradiction is derived from the proposition in question together with other
acceptable propositions, and the proposition in question is derived from
propositions explicitly used in the arguments A1 and A2: “(113/123—2)
change does not occur unless there are contraries; (113/123—1) there are no
contraries unless there are contrary (natural) motions; therefore, (113/123)
change does not occur unless there are contrary (natural) motions; but (122)
natural circular motions are not contrary, and (121) the natural motion of all
integral bodies - earth as well as heavenly bodies - is circular; therefore,
(12) the earth as well as heavenly bodies are unchangeable; moreover, (112)
nonviolent motions towards and away from the center of a whole are contrary,
and (111) nonviolent motions toward and away from the center of their whole
are the natural tendencies of all the parts of all integral bodies; therefore, (11)
the heavenly bodies as well as the earth are changeable; therefore, (1) the
heavenly bodies and the earth are both changeable and unchangeable”.
(2) Galileo’s second criticism does not use any term to refer to the fault be¬
ing claimed there for the contrariety argument. The fault may be described as
grounding a conclusion on a premise which is more difficult to ascertain than
the proposition which the conclusion is designed to support. An appropriate
term might be ‘useless’. To be more exact, we might say that an argument is
 GALILEO AS A LOGICIAN 361

useless when it is contextually a subargument of a bigger argument such that

the final conclusion of the first argument appears as an intermediate proposi¬
tion of the bigger argument, and a final reason of the first argument is more
questionable than the final conclusion of the bigger argument. For example,
consider two arguments with the following structures:

1 2

11 12 21 22

121 211^^212

2121

Argument 1 is useless if in the context argument 2 is being presented; if

1 = 21, 11 = 211, 12 = 212, 121 = 2121; and if either 2121 or 211 is more
questionable than 2.
Is uselessness a logical or rhetorical fault? The fault clearly does not refer
to the connections among the propositions in the various subarguments of
either argument involved; hence the fault is not logical. The fault is that in
the bigger argument the final conclusion is easier to ascertain than one of its
final reasons; this means that there is at least one argument which has for
final conclusion either the final conclusion of the bigger argument or its
negation, and which is much better than any argument supporting the final
reason in question; if so, the so-called bigger argument (#2 above) should not
be given, but if it should not be given, and if in fact it was not given (since
only the smaller argument was given), then the smaller argument is faulty
insofar as it suggests the bigger one; so what is wrong with the original
(smaller) argument is this suggestion, and this seems to be primarily a rhetori¬
cal failure.
(3) In stating the contrariety argument, supporting Al, Galileo uses the
reasoning indicator ‘if therefore’ (se dunque, F35.62). I am not talking about
the hypothetical indicator ‘if. . .then_’ which connects two simpler
propositions into a single, but more complex conditional proposition; rather I
mean a connective which relates propositions into an argument, and which is
more fully expressed as ‘if therefore . .. then’ with a proposition before and
one after the ‘then’. Moreover, the ‘if therefore’ is always preceded by
another proposition. These purely syntactical considerations suffice to
distinguish ‘if therefore’ from ‘if-then’.
362 CHAPTER 16

The next thing to consider is whether ‘if therefore’ is equivalent to

‘therefore if. The latter indicates that a conditional proposition is being
stated as a conclusion from what precedes, whereas the former indicates that
the (not necessarily conditional) proposition following the ‘therefore’ is a
conclusion from what precedes. But in order to understand this fully we need
to determine what is the full meaning of ‘if therefore’. Let us consider
Galileo’s own text as a concrete example, stating the propositions in the same
order and using the same connectives (but paraphrasing the propositions as I
already have for greater clarity):
(1) all generation and all decay involve contraries, so that (2) change does
not occur unless there is contrariety; but (3) contraries have contrary mo¬
tions; if therefore (4) there is no contrariety among heavenly bodies, since
(5) no motion is contrary to circular motion, then (6) heavenly bodies are
unchangeable. Here, the reasoning indicators, in the order in which they
occur, are: so that, if therefore, since, and then. ‘So that’ is unambiguous and
indicates the subargument ‘(1), (2)’; ‘since’ is also unambiguous, and it
indicates ‘(5),.'. (4)’. Keeping in mind the latent proposition that (7) heavenly
bodies have circular motions, semantical intuition tells us that (4) is being
grounded on (3), (5), and (7) together. Semantical considerations also tell us
that (6) is being grounded on (4) and (2). This gives the argument:

This structure corresponds to the argument as previously reconstructed, if

we let 6 = Al, 2 = A12, 4 = All, 7 = Alll, 5 = A112, 3 = A113, and if we
add 1 = A121 to the previous argument where it had been neglected for the
sake of incisiveness.
For the present discussion proposition 5 is an unnecessary complication,
since it might have been written in place of (3), leaving the latter latent. Thus
the meaning of ‘p; if therefore q, then f is ‘p; :.q\ r\ This makes ‘if there¬
fore’ a triadic connective, preceded by at least one clause and followed by at
least two, indicating the statement of two subarguments, with the clause
immediately following having the role of an intermediate proposition, the last
clause the role of their final conclusion, and the preceding clause the role of
their final reason.
 GALILEO AS A LOGICIAN 363

Finally, it should be noted that ‘p; if therefore q, then f may also be

expressed as ‘p; if then q, therefore f and as ‘p; if then q, r\
(4) Galileo’s third critical conclusion, G3, has been stated as attributing a
‘problematic’ character to the contrariety argument. It would have been
somewhat inaccurate and premature to use any other less general concept
since the text merely asks, without answering, what this criticism proves, and
since it is not immediately apparent what the criticism in fact amounts to.
Let us then look at the matter more closely.
There are three things to keep in mind, the original argument Al, the
critical counterargument A4, and the proposition that heavenly bodies are
unchangeable, Al = A421. Let us use the following abbreviations:

TA : terrestrial bodies are changeable (alterable);

Tc: terrestrial bodies have contraries;
HA : heavenly bodies are changeable (alterable);
He: heavenly bodies have contraries;
X: only contrariety yields change.

Three acceptable claims implicit in Galileo’s critical argument G3 are:

1. Ta;
2. if not-HA and TA, then ; and
3. if He and X, then HA.

On the basis of these, we could make the following deductions, where the
source and justification of each line is noted in parenthesis after each line
number:

4(3). If He, then (if X then HA);

5(2-3). if not-HA and TA, then (if X then HA);
6(5). if not-HA and TA and X, then HA;
7(6). if Ta and X, then (if not-HA then HA);
8(—)• if (if not-HA then HA) then HA;
9(7—8). if Ta and X, then HA;
10(1-9). if X then HA;
11(10). not-(X and not-HA);
12(11). either not-X, or HA.

Line 8 is obvious upon reflection; it is, in fact, an instance of the so-called

Law of Clavius. Line 10 says that if one accepts the connection between
change and contrariety, he ought to accept heavenly change; hence, the
Aristotelians ought to accept heavenly change. In other words, they are
364 CHAPTER 16

inconsistent in accepting the connection together with heavenly incorrupti¬

bility (line 11), and they should reject one or the other (line 12). It should be
noted that the contradictory propositions in Galileo’s counter-argument G3,
namely propositions A4 = Ha and A421 = not-H^, do not render it invalid;
rather they invalidate, i.e., render false, the conjunction of two (Aristotelian)
propositions, A41 and A421. In other words, Galileo’s argument is a partial
(but valid) reductio ad absurdum of heavenly unchangeability, partial
because the reductio depends on the assertion of the connection between
change and contrariety; from the point of view of the conjunction of these
two ideas, it is a full reductio ad absurdum. Thus heavenly unchangeability is
not refuted in the abstract, but only within the Aristotelian framework.
Let us now see what follows about the original contrariety argument A1.
The argument corresponds to proposition:

13. if not-Hc and X, then not-Ha .

Putting this together with previous results we get:

14(10—13). if not-Hc and X, then Ha and not-Ha; 4 and

15(14). not-(X and not-Hc).

Line 15 tells us that the antecedent clauses of line 13 are inconsistent. This,
of course, does not affect the formal validity of the corresponding argument;
all it means is that its premises are contradictory. So there is a contradiction
hidden within the premises of the contrariety argument. However, because of
this contradiction, the argument is practically worthless and rhetorically
ineffective, for the premises could never be all true, and hence the situation
could not arise where one could say that since all the premises are true,
therefore so is the conclusion. In other words, from the point of view of the
acceptance of the conclusion (heavenly unchangeability), one cannot be
properly persuaded by being told that he ought to accept not-HA because he
accepts or ought to accept X and not-He, for in view of line 15, he ought not
to accept both of these, and hence if he does, he should be shown this in¬
consistency, rather than being invited to reason in accordance with line 13.
We might thus label the contrariety argument ‘self-contradictory’ and note
that this is more of a rhetorical than logical impropriety.
Next, the following proposition follows directly from line 10, since if HA
follows from X alone, it will follow from X plus other propositions:

16(10). if not-Hc and X, then Ha .

When we compare this with the contrariety argument expressed in line 13, we
 GALILEO AS A LOGICIAN 365

can see the arbitrariness of this argument. Once more, the contrariety argu¬
ment is not formally invalidated by line 16, but rhetorically speaking, and
from the point of view of practical reasoning, one might ask why one should
draw conclusion not-HA when its opposite could just as legitimately be
drawn. In the light of line 16, the contrariety argument might be labeled
arbitrary, and for this reason rhetorically ineffective.
Moreover, in the present situation ‘if-then’ has perhaps an inferential
meaning, rather than being a material conditional.5 If the ‘if-then’ is inter¬
preted inferentially, then line 16 means that not-Hc and X imply HA, and
hence that ‘not-(if not-He and X, then not-Ha)’; this would then invalidate
the contrariety argument. In other words, if, by line 16, one can really infer
Ha from not-He and X, rather than merely hypothetically and abstractly do
so, as one does in formal logic, then one cannot also really infer not-HA. Thus,
in the sense of actual reasoning and active evaluation previously noted, one
might say that the conclusion of the contrariety argument does not follow
because what follows is instead the denial of it.
Next, by the same technique as before, from line 10 one may also deduce

17(10). if not-Hc and X and Tc, then HA.

The interesting thing that this tells us is that the contrariety argument is
incomplete, another rhetorical, alogical fault. In the contrariety argument
one infers not-HA from two premises, not-Hc and X. Line 17 tells us that
if the Aristotelians had also considered another proposition acceptable to
them, namely Tc, as a premise in the context of the same argument, they
would have arrived at a different (opposite) conclusion.
Let us now consider the following deduction:

18(9-13). if Ta and X and not-Hc, then HA and not-HA;

19(18). not-(TA and X and not-Hc);
20(19). either not-TA, or not-X, or ncrt-(not-Hc).

The last line does not depend on the assertion of terrestrial changeability
Ta. It says that the Aristotelians ought to reject either terrestrial change, or
the connection between change and contrariety, or heavenly lack of con¬
trariety. This is analogous to the previously justified conclusion of Galileo’s
fust criticism. The main difference is that the alternatives in line 20 include
heavenly lack of contrariety rather than heavenly unchangeability.
Finally, a formal way of investigating the formal validity of the contrariety
argument is by putting line 13 in the antecedent of a conditional and seeing
what follows. Let us begin with the following obvious assertion:
366 CHAPTER 16

21(—). if line 13 and not-Hc and X, then not-HA;

then:

22(10—21). if line 13 and not-Hc and X, then Ha and not-HA;

23(22). not-(line 13 and not-Hc and X);
24(23). either not-line 13, or He, or not X;
25(24). if not-Hc and X, then not-line 13.

Line 24 says that one ought to reject either the contrariety argument, or
heavenly unchangeability, or the connection between change and contrariety.
Line 25 says that if one accepts the latter two ideas, then one should reject
the contrariety argument. Another interpretation is that if one claims consis¬
tency for not-Hc and X, then one has to reject the validity of the argument.
This corresponds to our previous remark that its validity derives from the
contradiction in its premises. Since the Aristotelians accept heavenly lack of
contrariety and the connection, it seems that they should reject the argu¬
ment; hence the argument should be invalid for them. Does it follow from
this that the argument is invalid per set Of course, it should not be forgotten
that line 25, like everything else in this discussion depends on lines 1,2, and
3. The Aristotelians might, of course, reject some of these, though it is not
clear on what grounds. Of the three, line 3 might seem most vulnerable since
it corresponds to an argument suggested by the contrariety argument itself, or
to be more exact, they seem to be both instances of the same manner of
reasoning. In fact, they are what I have previously called two sides of the
same argument; however the two sides are different enough so that it is
possible to accept line 3 and reject line 13; for, disregarding X, they are the
converse of each other; and if the connection between change and contrariety
is held in only one direction, we would accept that whenever there is con¬
trariety there is change, but not that whenever there is no contrariety there
is no change. This would receive additional support from empirical considera¬
tions. In this manner the contrariety argument may actually be invalidated.
Of course, it could still be written in a formally valid manner, with an
unjustified premise, i.e., with the connection between change and contrariety
stated more strongly that is warranted; in this case its problem would be this
premise, and I feel inclined to say that this would not change the ‘logic’ of
the situation, meaning the substantive issue. But if one accepts this manner of
speaking, one is no longer equating logic with formal logic.
(5) Simplicio objects to argument A4 used in Galileo’s criticism G3 by
comparing it to the liar’s paradox:
 GALILEO AS A LOGICIAN 367

The Cretan said, “I am lying”.

Either the Cretan was lying, or he was telling the truth.
If he was lying, then “I am lying” was a lie.
If “I am lying” was a lie, then he was telling the truth.
.'. If the Cretan was lying, then he was telling the truth.
But if he was telling the truth, then “I am lying” was true.
And, if “I am lying” was true, then he was lying.
.'. If the Cretan was telling the truth, then he was lying.

In this paradox, from each one of the alternative assumptions one can deduce
its opposite. However, in the argument in question, though one can deduce
heavenly changeability from heavenly unchangeability, one cannot deduce
unchangeability from changeability. Hence, whereas the liar’s paradox gives
rise to an infinite series of deductions, the present argument terminates once
heavenly changeability has been derived. The following makes this clear:

(A52K
(A5112nChange occurs if and only if there are contraries.
(A5122)7
(A5121) The elements earth and water are the contraries of the elements
fire and air (since (A51211) the natural motion of the former is
contrary to that of the latter).
(A512) Elemental bodies are changeable.
(A5111) But heavenly bodies have no contraries (since (A51111) their
natural motion is circular).
(A511) A Heavenly bodies are unchangeable.
(A51) Heavenly bodies are the contraries of elemental bodies (since
the former are changeable and the latter are unchangeable).
(A5) Heavenly bodies are changeable.

The structure of this argument is:

A5

A511 A512

A5111 A5112 A5121 A5122

A51111 A51211.
368 CHAPTER 16

(6) Galileo’s critical argument G4 is interesting because it is basically ad

hominem in the (seventeenth century) sense that it derives a conclusion, G4,
previously unaccepted by an opponent from propositions accepted by him
but not necessarily by the arguer; in the present case the main such proposi¬
tion is that heavenly bodies are the denser parts of the heavens. In saying that
argument G4 is ‘basically’ ad hominem I am implicityly calling attention to
the fact that, in the way in which it is expressed, the argument is not com¬
pletely ad hominem-, for the immediate premises for the conclusion (proposi¬
tions G41, G42, and G43) are such that Galileo would accept them, the first
because of its character of (historical) report, the last two because of their
conditional structure; this subargument could have been stated in such a way
as to involve directly the disputed proposition and then would have been
explicitly ad hominem'. “proposition All is questionable since heavenly
bodies are the denser parts of the heavens, and hence the contrariety of rarity
and density exists in the heavens”. This is the way that subargument ‘G431,
G432,.'. G43’ is stated.
Rhetorically speaking, that is, from the point of view of persuasion, ad
hominem arguments are very effective since the arguer thereby provides his
opponent with reasons for accepting the conclusion; though the ad hominem
argument would not constitute a reason for someone who did not accept the
problematic premise(s), the opponent in the case does accept it (them), and
hence if the argument is otherwise correct, he is induced to accept the
previously unaccepted conclusion.
Let us now examine in more detail the above suggestion that conditional
propositions provide a logically impeccable means of expressing ad hominem
arguments. In G4 Galileo is saying essentially that: “(G4l') if one believes
that heavenly bodies are the denser parts of the (mostly extremely rarified)
heavens, then he ought to believe that there is a change-producing heavenly
contrariety; therefore, (G4') Peripatetics ought to believe in heavenly con¬
trariety, since (G42') they believe that heavenly bodies are the denser parts of
the heavens”. So far the argument is logically correct, indeed formally valid.
The rest of the argument may be reconstructed as follows: “(G411') if one
believes that heavenly bodies are the denser parts of the heavens, then he
believes that the terrestrial contrariety of upward and downward motions
causes terrestrial changes (since (G4 111') both beliefs are part of the
Aristotelian system); (G4121’) but if the contrariety of upward and down¬
ward motions causes terrestrial changes, then differences of rarity and density
are the ultimate cause of these changes (since (G41211') differences of rarity
and density cause the contrariety of upward and downward motions of
 GALILEO AS A LOGICIAN 369

terrestrial bodies), and (G4122') if differences of rarity and density are the
ultimate cause of terrestrial changes, then differences of rarity and density
among heavenly bodies cause heavenly changes (since (G4122T) terrestrial
and celestial differences of rarity and density have the same nature, for
(G412211') they derive both from the quantitative difference of more or
less matter in a given volume); therefore, (G412') if the contrariety of upward
and downward motions causes terrestrial changes, then differences of rarity
and density among heavenly bodies cause heavenly changes; but (G413') if
differences of rarity and density among heavenly bodies cause heavenly
changes, then there is a change-producing heavenly contrariety; therefore
(G4l') if one believes that heavenly bodies are the denser parts of the
heavens, then he ought to believe that there is a change-producing heavenly
contrariety.”
This argument has the following structure:

G4121r G41221'

G412211'

All propositions here are conditional, except for G4(, G42 , G4111 , G41211,
G41221', G412211'; I believe they are all acceptable and that their inferential
connections are correct. This may be compared with G4, whose structure is:

G43221.
370 CHAPTER 16

Let us now see what can be said by way of generalization. Suppose we have
the ad hominem argument ‘R1 since R11; but R2;C’, where the final reasons
R11 and R2 are accepted by the opponent but not by the arguer. The latter,
however, does accept that if R11 then R1 and that if R1 and R2 then C. If
the ad hominem argument is inferentially correct, then these conditions will
be true; putting the two together we get ‘if Rll and R2 then C”. Since the
‘opponent’ accepts R11 and R2 but not C, this shows his beliefs are inconsis¬
tent, i.e., that he should either reject ‘not-C’ or one of R11 or R2. This shows
that the central value of ad hominem reasoning is logical.
(7) What does criticism G5 amount to? That is, what further conclusion
could be drawn from G5? One reformulation would be to say that either the
conclusion does not follow, or else one of the premises is groundless; that is,
that the argument is either logically invalid or based on an unjustified
premise. We could also say that if the premises are well grounded then the
conclusion does not follow; this is not exactly the same as saying that if the
premises are true then the conclusion is false, and hence that the argument is
formally invalid; it is however to say that even if the premises are accepted
the conclusion need not be, and hence that the argument is rhetorically
invalid, i.e., ineffective from the point of view of persuading someone about
the conclusion on the basis of the premises.
The critique is also a good illustration of the interplay between logic and
rhetoric. On the one hand, it might seem that the formal validity of the
contrariety argument depends on the nonambiguous use of a particular
term, which is not a formal-logical characteristic of the term; for one might
have considered G531, G521, and G51 together and drawn the conclusion
that the argument is logically invalid, and then logical invalidity would have
been a consequence of a non-formal fact. On the other hand, it should be
noted that the ambiguity of the term ‘bodies’ is established by logical
considerations in the sense of considerations pertaining to the understanding
of the reasoning involved in arriving at the two premises All and A12. Thus
the formal invalidity of the contrariety argument depends ultimately on
logical matters broadly conceived, and this suggests a broadening of the
notion of logic. Is this broadening one in the direction of including rhetorical
considerations? I do not think so since the ambiguity of the problematic term
is being grounded on nonrhetorical features of the contextual arguments
supporting the two problematic premises, that is, the argument ‘A111, A112,
A113, .'. All’ and the implicit argument for A12; each of these subarguments
is being assumed not to be ambiguous. Rather the broadening involves a
greater emphasis on logical context, in the sense that the logical validity of a
 GALILEO AS A LOGICIAN 371

given argument may depend on that of a bigger argument which includes

the subarguments needed in the context to support the premises of the
given argument; the fuller argument may be such that if we take its final
reasons to be true, the final conclusion does not follow because those final
reasons lead to intermediate propositions in which a term is used ambiguous¬
ly; or the full argument may be such that if we take the final conclusion to be
true, then at least one of the final reasons from which it follows is obviously
groundless.
Criticism G5 may be taken to charge the contrariety argument with the
fallacy of equivocation, and it helps us to conceive this fallacy as an argument
where a term is used ambiguously in such a way as to conceal the fact that if
the term is given one meaning then the conclusion obviously does not follow,
and if it is given the other meaning then one of its premises is obviously
groundless.
(8) The arguments in this section can serve to illustrate and justify some
claims about ‘unless’ (se non) and ‘only’ (solo). Many logic textbooks advise
their students to translate ‘p unless q' as the inclusive disjunction ‘p V q' or
equivalently as the material conditional l-q-» p’.6 This advice reflects both
linguistic insensitivity and inadequate understanding of actual reasoning. It
seems more accurate7 to interpret ‘p unless q’ to mean ‘p if and only if not-r?’
and ‘only A’s are B’ as ‘all A’s are B and all non-A’s are non-B’. When ‘unless’
and ‘only’ are so understood, then the intuitively valid inferences in argument
A2 and in part in argument A3 become also theoretically and reflectively
valid, whereas the orthodox logician’s interpretations would render them
invalid. In fact, these arguments should be formalized as follows, where ‘iff
is an abbreviation for ‘if and only if’:
(A21) No change iff no contraries;
(A22) no contraries iff no contrary motions;
(A23) contrary motions iff contrary directions;
(A24) contrary directions iff toward and away from center;
(A25) toward and away from center iff terrestrial bodies;
(A2) .'. no change iff no terrestrial bodies.

And
(A32) All bodies with natural circular motion are unchangeable;
(A31) all heavenly bodies have natural circular motion, and all non-
heavenly bodies lack natural circular motion;
(A3) all heavenly bodies are unchangeable, and all nonheavenly bodies
are changeable.
372 CHAPTER 16

There is no problem with this version of argument A2, where it should be

noted that Galileo thinks that A25 ought to be replaced by: (A25') toward
and away iff denser and rarer parts of integral bodies. However, in argument
A3, the second clause of proposition A3 does not follow, and the second
clause of A31 is superfluous. Galileo does not mention this problem, which
he must have regarded as trivial. It could be argued that since ‘nonheavenly’
in this context means ‘terrestrial’, since the second clause of A3 does follow
from A2, and since in the text the second argument, A3, is stated imme¬
diately after the first, therefore the second clause of A3 is being grounded on
proposition A2, and therefore both parts of A3 follow. However, this would
not explain the presence of ‘only’ in the statement of A31, i.e., the problem
of the superfluousness of the second clause of our A31 would remain; in fact,
this presence of ‘only’ indicates that the ‘only’ of A3 is being grounded
thereupon. Hence argument A3 would seem to be inescapably invalid (in
part). At any rate, Galileo’s critical modification of argument A3 would not
have this fault since he wants to replace A31 by the proposition that (A3l')
all integral bodies have natural circular motion, and thereby conclude merely
that all integral bodies are unchangeable.

5. THE COUNTER-EXAMPLE METHOD, AND THE

A POSTERIORI ARGUMENT (F7 1.34-83.ll)

The object argument. (A11) No changes have ever been observed to occur in
the heavenly bodies; (A 12) changes are observed routinely on the earth;
therefore, (Al) the heavenly bodies are unchangeable [F71.34-72.30]
The meta-argument. (Gil) If there were changes in the heavenly bodies,
most of them could not be observed from the earth, since (Gill) the
distances from the heavenly bodies to the earth are very great, and (G112) on
the earth changes can be observed only when they are relatively close to the
observer. And, (G121) if there were changes in the heavenly bodies large
enough to be observable from the earth, these large changes would not be
observed unless careful, systematic, exact, and continual observations were
made; but, (G122) no such observations have been made (by the argument’s
proponents); therefore, (G12) if there-were changes in the heavenly bodies
large enough to be observable from the earth, they might not have been
observed. Therefore, (Gl) the a posteriori argument (i.e., the object argu¬
ment) is logically invalid. [F72.31-74.11]
Second, (G2) the argument is invalid because (G21) no terrestrial changes
would be noticeable to an observer on the moon before some particular very
 GALILEO AS A LOGICIAN 373

large terrestrial change had occurred, and yet (G22) terrestrial bodies are
obviously changeable and would have been so even before that occurrence.
[F74.12—74.23]
Third, (G3) the argument as stated is a stronger version of the one that
may correspond more closely to Aristotle’s letter, namely: “(All') no one
has ever observed any generation or decay of heavenly bodies in the heavenly
region; therefore, (Al') the heavenly region is unchangeable”. For (G31) this
argument would have the additional fault of sanctioning the following clearly
invalid argument: “(Al l") no one has ever observed any generation or decay
of terrestrial globes in the terrestrial region; therefore, (Al") the terrestrial
region is unchangeable”. (G32) What is wrong with the literal version of the
Aristotelian argument is that it implicitly contrasts heavenly bodies to such
terrestrial bodies as cities rather than to terrestrial globes. [F74.23—75.8]
Fourth, (G4) premise All is false since (G41) changes have now been
observed both within heavenly bodies and in the heavens. For example,
(G411) comets constitute changes in the heavens, regardless of whether they
originate there or in the terrestrial region [F76.17, F77.1—77.12]; (G412)
the new stars of 1572 and of 1604 are observed generation and decay in the
heavens [F76.16-76.19, F82.13-82.36] ; and (G413) sunspots are changes
on the body of the sun since (G4131) they appear and disappear at random
in the solar disc, (G4132) their apparent motion across the solar disc is slower
at the edges and faster in the middle, and (G4133) their apparent size and
shape is narrower at the edges and wider in the middle [F79.1—79.25]. On
the other hand, (G414) the arguments against these recent observations are
worthless; for example, (G4141) the author of the Anti-Tycho uncritically
assumes that comets are phenomena to which parallax theory is applicable
[F77.7—77.8], (G4142) he uncritically rejects contrary observations not
confirming his thesis [F77.9-77.12], and (G4143) he is inconsistent in his
attitude toward comets and toward new stars_(since (G41431) he attempts to
locate the former in the terrestrial region but regards the latter as irrelevant
on the grounds that they do not represent changes within well-established
heavenly bodies) [F77.13-77.18, F82.13-82.36]; and (G4144) the inter¬
pretations of sunspots as planets circling the sun and obscuring parts of it
conflict with the evidence of their random appearance and disappearance, and
of their changing apparent speed and shape [77.19—80.9].
Finally, (G5) the argument’s conclusion Al is false; for the fact that
(G511) changes have now been observed in the heavens shows that (G51)
heavenly bodies are changeable [F75.9—76.1 1, F80.16—82.7]
Comments. (1) It is interesting to note the presence of the second premise,
374 CHAPTER 16

A12, in the a posteriori argument. It strengthens the argument by providing

evidence that heavenly changes should be observable. Of course, Galileo’s
critique G1 invalidates this by pointing out some relevant differences between
heavenly and terrestrial changes. Though invalid, the a posteriori argument
does attempt to appeal to the right kind of evidence; the problem is merely
that it does not go far enough. In other words, the argument does not depend
on the assumption that everything that is not observable does not exist, as it
would if proposition All were the only premise; rather it assumes some
proposition to the effect that if something is observable when occurring on
the earth then it will be observable when occurring elsewhere. The problem
with this is that it makes a difference where the observer is located.
Galileo’s critique seems to interpret the a posteriori argument as an
explanatory one, in the sense that it presents the conclusion about heavenly
unchangeability as the explanation for the difference of observation men¬
tioned in the premises. Galileo then suggests two other ways of explaining the
difference: it may be due to the great distance between the earth and the
heavenly bodies and/or to the lack of sufficiently careful observations of the
heavenly bodies. This does not invalidate the Aristotelian explanation, in the
sense of showing that it, i.e. conclusion Al, is false; rather, the alternative
explanations invalidate the argument, i.e., the inferential link between
premises and conclusion, since there is no reason to prefer the Aristotelian
explanation to the one suggested in Galileo’s critique. To conceive Galileo’s
critique in this manner also makes it clear that he is not accepting the
premises as such, but that he is merely saying that even if we accept them,
that does not force us to accept the conclusion. In fact, he will soon question
premise All.
(2) In his second criticism Galileo is using the method of counterexample.
He envisages a similar situation involving the observation of the earth from
the moon; in such a situation the corresponding premises of an a posteriori
argument for the unchangeability of the earth would be true, and yet we
know that the conclusion is false. This shows that it is possible for premises
Al 1 and A12 to be true and conclusion Al false, and hence that Al does not
follow from All and A12. The method of counterexample is, of course, the
technique favored by formal logicians to prove invalidity, and it is a relatively
‘active’ method of criticizing reasoning insofar as it involves the creation of a
counterexample.
It is interesting to compare the counterexample method to the one
mentioned earlier in which one argues that a conclusion does not follow
because what follows is some other proposition different from it. In the
 GALILEO AS A LOGICIAN 375

present argument, Galileo might have said that the most that would follow
from the premises is that there have been no changes in the heavenly bodies
so far (i.e., during the period of human observation). It seems clear that the
method of alternative conclusion, as the latter technique might be called, is
active insofar as one has to think of the conclusion that does follow; it is also
more constructive than the counterexample method since, besides showing
what is wrong with the given argument, it shows what could be right.
Let us look at this in more detail. Let a be the argument ‘Rl, R2, .'. C’.
The counterexample method begins by formulating another argument a
having the same logical form as a, but with obviously true premises and
obviously false conclusion. Then one argues as follows: since a and a have the
same logical form, if a is valid then a is also valid; but a is invalid, because it
has true premises and false conclusion; therefore, a is also invalid. In the
method of alternative conclusion one argues that if Rl and R2 are true then
C' would also be true (perhaps on the basis of other premises R3, R4,etc.),
where C' is such that if C' is true then C is not; if follows that if Rl and R2 are
true then C is false; that is, Rl and R2 imply not-C; given the consistency of
Rl and R2, which is usually obvious or unproblematic, it follows that Rl and
R2 do not imply C, i.e., that a is invalid. Though the method of alternative
conclusion depends on the consistency of the original premises and the truth
of the other ones used — R3, R4, etc. - this does not make it less effective
than the counterexample method, since the latter depends on the identity of
logical form between a and a', which is almost always problematic. Moreover,
the rhetorical effectiveness of the latter depends on the counterexample having
some substantive and topical similarity to the original argument; an abstract
counterexample will be unconvincing, and this will normally be an indication
of the unsoundness of the justification that the original argument has the
proposed form. In using the other method this problem does not arise, even
though the additional premises R3, R4, etc., may be unfamiliar, since they
are necessarily relevant from the point of view of substance and topicality.
(3) Galileo’s criticism G3 constitutes a negative evaluation of argument
A1 in spite of the fact that it claims that A1 is better than the more literal
version of Al. For here Galileo strengthens his previous two criticisms by
pointing out that there he has criticized a stronger version of the argument
and not a straw man. To be sure, the strengthening is operative only at the
rhetorical level, but it is rather effective at this level.
Regarding the literal version of Al, Galileo is somewhat explicitly using the
method of counterexample. However, here we have a good example of the
limitations of this method, for we see that the counterexample introduced in
376 CHAPTER 16

G31 tells us that something is wrong with the literal version, but it does not
tell us exactly what is wrong, or why it is wrong, or why it is wrong in the
sense of making us understand what went wrong with it, so that we might be
able to avoid such an error in future reasoning. Thus a limitation of the
counterexample method is perhaps its relative ineffectiveness in improving
the practice of reasoning.
It is G32 that begins to diagnose the trouble, which turns out to be part of
an equivocation problem. For what G32 means is that the latent proposition
in the argument ‘A11',A1' ’ is A12': we constantly observe generation and
decay of terrestrial bodies in the terrestrial region; now, this proposition is
needed in the argument since it together with All' creates a contrast which
is then accounted for by Al'; but the contrast is spurious since ‘bodies’ in
A11' means integral bodies, whereas ‘bodies’ in A12' means parts of integral
bodies; hence there is nothing to explain; hence, there is no reason to con¬
clude Al'; in short, Al' does not help to explain the alleged facts mentioned
in the premises All' and A12' because there is nothing to explain.
(4) The main interest of Galileo’s critique G4 is that, though it pertains to
(one of the premises of) the a posteriori argument, it is itself a positive argu¬
ment in support of a thesis, G41. Here we should note that this proposition is
supported not only directly with positive evidence (G411, G412, and G413),
but also indirectly by criticism of available counterevidence and counter¬
arguments, namely by the argument supporting G414. In other words,
argument G4 is not only a meta-argument but also an argument in support of
heavenly changes; nevertheless, in the course of this object-level reasoning,
Galileo finds it useful and necessary to engage in some evaluation of counter¬
arguments. This supports a claim about the evaluative nature of reasoning
per se, whereas the other passages so far support a claim about the ratio-
cinative nature of the evaluation of reasoning.
(5) Galileo’s criticism G5 indicates that in the present case the falsehood
of a premise falsifies the conclusion. This is obviously not true in general, but
in this case one contributing factor is the fact that both premise Al 1 and
conclusion Al are denials of existence claims; hence the denials of these
denials are positive and very informative propositions.
It would be wrong to fault Galileo’s argument ‘G511, :. G51’ by saying
that this manner of reasoning is inconsistent with his criticism of the validity
of the a posteriori argument, given in Gl and G2, and that this reasoning is
only correct from the point of view of the empiricism of the Aristotelians.
For in Gl and G2 Galileo did not criticize the presupposed empiricism of the
a posteriori argument, but rather interpreted the latter as an explanatory
 GALILEO AS A LOGICIAN 377

argument, drawing an arbitrary explanation for conclusion. Hence Galileo

could defend his G5 reasoning by criticizing explanations of G511 other than
G51.

6. CHARITY, AND THE TELEOLOGICAL ARGUMENT

(F83.12-87.14)

The object argument. (Al) Changes among terrestrial bodies enhance the
perfection of the earth; for example, (A2) living organisms are more perfect
than dead ones, and (A3) gardens more than deserts. But, (A4) heavenly
changes would render heavenly bodies imperfect, since (A5) heavenly changes
would be of no use or benefit to man, and hence (A6) they would be super¬
fluous; therefore, (A7) unchangeability would enhance the perfection of
heavenly bodies. Therefore, (A8) heavenly bodies are unchangeable. This is
also shown by the fact that, since (A6) heavenly changes would be super¬
fluous, and since (A9) nature does nothing in vain, (A10) there cannot be
any heavenly changes.
The meta-argument. (G11) The inference from A5 to A6 is incorrect since
(Gill) man should not be regarded as the sole creature for whose sake and
benefit the whole universe exists. Moreover, (G12) the inference from A6 and
A9 to A10 is incorrect since (G121) even if we accept A9 it cannot be inter¬
preted to mean that nature brings nothing into existence which is superfluous
from the point of view of human needs. Hence, (Gl) the teleological argu¬
ment is unsound.
Comments. The propositional structure of the teleological argument is:

A8

A5

Since A5 and A6 appear in two distinct subarguments, we could also have the
following diagram:
378 CHAPTER 16

It is interesting to note that the alleged superfluousness of heavenly changes,

A6, supports heavenly unchangeability, A8, in two ways: first by means of
perfection considerations (A4, Al, A7), and second by means of natural
teleology, A9. It is this double use of proposition A6 that would make
standard labeling problematic in this case; in fact, it would be impossible for
the second diagram above, and misleading for the first insofar as the double
appearance of propositions A5 and A6 would be missed by the different
labels they would receive.
It should be pointed out that the subargument from A2, A3, and A5 to
A8 is a way of integrating into the object argument what would otherwise
appear as incorrect criticism in the evaluative meta-argument; for this sub¬
argument makes it obvious that the Aristotelians are not assuming the
excessively broad generalization that change makes all things imperfect. To
attribute this excess would be a violation of what may be called the Principle
of Charity,8 and there is certainly textual evidence that Sagredo (i.e., one
part or one mood of Galileo) commits this violation.9 At the same time, since
it has been possible to reconstruct the teleological argument as above, and
since this reconstruction is still subject to Galileo’s main criticism, it would be
a violation of that principle on our part (vis-a-vis Galileo) not to reconstruct
the teleological argument as we have.
(2) One might wonder whether the inference from A7 to A8 should be
faulted for affirming the consequent. Here the reasoning would be that a
fuller reconstruction of this step would be: (A81) if heavenly bodies were
unchangeable then their perfection would be enhanced; but (A82) heavenly
bodies are (or ought to be regarded) as perfect as possible, since, e.g., (A821)
they are the work of God, or whatever; therefore, (A8) heavenly bodies are
unchangeable. Here A81 would correspond to A7. However, there is no
reason to prefer this reconstruction to the following: (A811) if heavenly
 GALILEO AS A LOGICIAN 379

bodies were prefect then they would be perfect by reason of unchangeability;

but, (A812) heavenly bodies are perfect (since (A8121) . . .); therefore,
(A81) heavenly bodies are perfect by reason of unchangeability; therefore,
(A8) heavenly bodies are unchangeable. Here there is one extra step, the
original A7 has been reconstructed differently, and the problematic step has
been replaced by a valid instance of affirming the antecedent. The best way
of choosing between these two reconstructions would be to look at the
subargument from A1 and A4 to A7. A1 seems to mean: one of the effects
of terrestrial changes is greater terrestrial perfection. A4: one of the effects
of heavenly changes would be greater heavenly imperfection, i.e., if heavenly
bodies were changeable then they would be less perfect. Since this last
proposition implies A811, the second one of the above reconstructions is to
be preferred; nor does this make A1 superfluous since it is its presence that
suggests this interpretation of A4.

7. EQUIVOCATION, AND THE VIOLENT MOTION ARGUMENT

(F159.29-162.14)

The object argument. (Al) The earth cannot move circularly, because (A111)
such motion would be violent, and hence (Al 1) it would not be perpetual; it
would be violent because (Al 111) if the earth’s circular motion were natural
then its parts would also move circularly by nature, but (All 12) it is im¬
possible for the earth parts to move circularly by nature since (Al 1121) the
natural motion of the earth parts is straight downwards.
The meta-argument. (Gl) The violent motion argument is unsound since
(Gil) the clause “the parts of the earth would also move circularly by
nature” can mean either that these parts would move around their own
centers or that they would move around the earth’s center. (G12) In the first
case proposition Al 111 would be obviously groundless, and the step from
All 121 to All 12 obviously invalid. (G13) In the second case proposition
Al 112 would be groundless since (G131) the reason supporting it, All 121,
would be groundless: for (Gl311) the concept of natural motion used in
All 121 is either the one suggested elsewhere in the argument,namely in the
step from Al 11 to All, or it is that of actual motion under undisturbed
conditions; (G1312) if we use the concept of motion suggested in the step
from Al 11 to Al 1 then no straight motion could be natural, since (G13121)
no straight motion can be perpetual, and (G13122) the step from Al 11 to
All suggests equating the natural with the perpetual; (G1313) if natural
motion is actual motion under undisturbed conditions then there is no reason
to believe the natural motion of the earth parts is straight downwards, since
380 CHAPTER 16

(G13131) the conditions under which they move straight downwards are
contrived artificially and seldom occur in nature.
Comments. Galileo’s critique shows that either Allll is groundless and
A1112 does not follow from A11121, or that A1112 is groundless. Since the
groundlessness of Allll can be conceived as invalidity of the step from
A1112 to A111, the first alternative means that A1112 neither is implied by
All 121 nor imples Alll. Moreover, the groundlessness of All 12 derives
partly from the validity of the step from Alll to All; this means that if the
step from Alll to A11 is valid then A1112 is false; that is, if A1112 is true
then the step from Alll to All is invalid; hence, the second alternative is
also a type of invalidity. Thus the violent motion argument is invalid in any
case.
It might be objected that, though the step from Alll to All may suggest
equating the natural with the perpetual, it presupposes merely that non¬
perpetuity is a necessary conditon for nonnatural (violent) motion, hence
that if motion is perpetual then it is natural, and hence that G13122 and
G1312 are questionable. In short, an Aristotelian can consistently accept
both propositions All 121 and the step from Alll to All; and hence both
this step and proposition All 12; so that Galileo would be violating the
Principle of Charity in attributing to Aristotle the equation of the natural
with the perpetual.
It is true that to make perpetuity a sufficient condition for natural motion
formally validates the step from Alll to All. However, the statement of
such a sufficient condition is a mere restatement of the validity of the infer¬
ence; hence it would be entirely appropriate to ask for a justification of such
a sufficient condition. The equation of the natural with the perpetual, or to
be more exact, with the potentially perpetual, would provide one, and it is
in this sense that the step from Alll to A11 may be taken to presuppose the
above-mentioned equation. This, of course, would accord with Galileo’s
argument in F53—57.

8. IMPLICATION VS SUPPORT, AND THE TWO

MOTIONS ARGUMENT (F162.15- 164.25)

The object argument. (A121) All bodies known to move circularly, except
the primum mobile, have two motions; hence, (A12) if the earth moves
circularly then it has two motions. But, (A112) if the earth has two motions
then there would be a variation in the rising and setting points of the fixed
stars, and (A111) no such variation is observed; therefore, (A11) the earth
 GALILEO AS A LOGICIAN 381

does not have two motions. Therefore (Al) the earth does not move circular¬
ly*
The meta-argument. (Gl) The circular motion mentioned in the conclu¬
sion can refer either to axial rotation around one’s own center or to orbital
revolution around some point outside oneself.
(G24) If circular motion means axial rotation then the second motion
mentioned in the argument is an axial rotation in the opposite direction
causing the primary rotation to lag behind. (G23) If so, then the argument
can be reconstructed as follows: “(A12') one terrestrial rotation would imply
a second since (A 121’) all rotating bodies except the primum mobile have
two opposite rotations; but, (Al 1') a second terrestrial rotation is impossible
since (A112') if it occurred there would be variations in stellar risings and
settings, (Al 11') which there are not; therefore, (Al') the earth cannot
rotate.” Now, (G2111212) if Al21' is true, then if the earth were the
primum mobile the first terrestrial rotation would not imply the second, since
(G21112121) A121 means that one rotation implies a second if and only if
the rotation does not belong to the primum mobile; but, (G2111211) if the
earth rotates then it is the primum mobile, since (G21112111) to conceive
the earth as having axial rotation is to conceive it as the primum mobile;
therefore, (G211121) if A121' is true, then if the earth rotates its first
rotation does not imply a second terrestrial rotation; therefore, (G21112) if
A121' is true, then the first terrestrial rotation would not imply a second one;
but, (G21111) if A12' follows from A12l', then if A121' is true then the first
terrestrial rotation would imply a second; therefore, (G2111) if A12' follows
from A121', then Al 21' implies a contradiction; therefore, (G211) if A12'
follows from A121', then A121' is false; therefore, (G21) either A12' does
not follow from Al21' or Al21' is false. Moreover, (G22111111) the two
motions argument may be interpreted as supporting the impossibility of a
(first) terrestrial rotation, Al', by the impossibility of a second terrestrial
rotation. All'; therefore, (G2211111) All' would be pointless if a first
rotation were impossible independently of the second; therefore, (G221111)
Al l' would be pointless if a first rotation were impossible without the second
being also impossible; i.e., (G22111) All' would be pointless if it were true
that the first rotation is impossible and the second is not impossible; there¬
fore, (G2211) it is not true that the first rotation is impossible and the second
not; therefore, (G221) if the first rotation is impossible then so is the second;
therefore, (G22) if the second rotation were not impossible, the first would
not be impossible either.
(G33) If circular motion refers to orbital revolution then the second
382 CHAPTER 16

circular motion is the daily axial rotation. (G32) If so the argument can be
reconstructed as follows: “(Al") the earth cannot have the (annual) orbital
revolution because (A12") if it did then it would also have the (daily) axial
rotation, and (Al 12") if it had orbital revolution and axial rotation then
there would be variations in stellar risings and settings, (Al 1l”) which there
are not, so that (All") it can’t have both revolution and rotation”. But,
(G31) Al 12" is false, since (G311) it is not stellar variations in risings and
settings that follow from the earth’s joint rotation and revolution, (G3111)
as it will be shown later. Therefore, (G3) if circular motion refers to orbital
revolution then premise Al 12 of the original argument is false.
Comments. It should be noticed that I have stated neither Galileo’s fmal
conclusion (from Gl, G21, G22, G23, G24, and G3) nor the conclusion,
G2,to be drawn from G21 ,G22, G23, and G24.
The first problem to solve is that of the relationship between Galileo’s G21
critique and the following one that readily comes to mind. The step from
A121' to A12' is formally invalid if one does not assume that (A122') the
earth is not the primum mobile', next, one would ask the Aristotelians to
justify this latent proposition; now, it is obvious that the question of whether
or not the earth is the primum mobile is equivalent to the question of
whether or not the earth rotates on its axis, which is the question the argu¬
ment is trying to settle; hence the argument begs the question.
Let us use the following abbreviations:

R = the earth rotates;

S = the earth has a second rotation;
P = the earth is the primum mobile.

The two motions argument (i.e., one version of it) is then:

(Al21') (if Rthen S) iff not-P;

(A12') if R then S;
(All') not-S, since (Al 11');
(Al') not-R.

The step from Al 21' to A12' assumes:

(A 122') not-P.
But: RiffP.

Hence to support not-P one needs the following argument:

(A1222') not-R;
 GALILEO AS A LOGICIAN 383

(A 1221') R iff P;
(A122') /. not-P.

Since A1222' - A1the full argument is circular:

Al'

All- An'

A121' A122'

A1221' A1222'.

Of course, an argument cannot be circular if it is formally invalid; so this

argument must be valid. Hence this critique seems to be inconsistent with
Galileo’s, which suggests invalidity, since G211 implies that if A121' is
true then A12' does not follow from A12l', which is to say simply that
A12' does not follow from A121'.
Let us see whether this impression is correct. G2111212 is saying that:

(FI) if [(if R then S) iff not-P] then [if P then not-(if R then S)],

which is right. G2111211 says that:

(F2) if R then P.

G211121 is equivalent to:

(F3) if [(if R then S) iff not-P] then [if R then not-(if R then S)],

which is a formal consequence of FI and F2. G21112 says:

(F4) if [(if R then S) iff not-P] then not-(if R then S),

which follows formally from F3 only if one adds:

(F3.1) if [if R then not-(if R then S)] then not-(if R then S),

which is not formally valid, since it is refuted by assigning the truth-value

falsehood to R. However, this formal invalidity depends on a truth-functional
interpretation of ‘if-then’, which there is no reason to accept here. We may
interpret ‘not-(if R then S)’ as: one cannot conclude S from R.10 Then (F3.1)
means that if given R one cannot conclude S from R, then one cannot
conclude S from R. Hence, F3.1 is correct. Similarly for the argument going
from G211 to the proposition that A12' does not follow from A121'.
384 CHAPTER 16

The difference between the two critiques is that the circularity critique is
saying that the ‘full’ argument, i.e., the two motions argument as stated plus its
latent structure, is circular and (hence formally) valid; whereas Galileo s G21
critique implies that the argument as stated is invalid insofar as A12' does not
follow from A12l'. The formal logician would agree with this, giving as a
reason that what follows is only that if not-P then (if R then S). Galileo is
implicitly saying this, but more. The reason for Galileo’s more complicated
criticism stems, I believe, from the fact that he wants his G21 criticism to
connect with his G22 critique. In fact, G21 amounts to saying that A12 is
groundless since either it does not follow from the stated reason A121 or if
it does follow then A12l' is false, so that there is no way of justifying A12'
here. But A12' amounts to saying that All' implies Al', hence to say that
A12' is groundless is to say that there is no reason to think that Al l' implies
Al'. But if so then an obvious possibility to consider is whether or not Al 1'
supports Al', in some sense different from implication. I believe that the G22
critique is an attempt to answer this question. When Galileo says that “All
would be pointless if ...” he may be interpreted as saying that “Al 1' would
not support Al' if.. The crucial consideration G2211111 amounts to
interpreting the step from All' to Al' as an explanatory argument, so that
the G22 critique may be interpreted as follows: if Al 1' supports Al' then the
fact that Al' has to be the alleged explanation for the fact that All'; if so
then Al' should imply Al 1'; therefore, if Al l' supports Al' then Al' implies
All'. In drawing conclusion G2211 Galileo is assuming that All is not
pointless, i.e., that it supports Al', so that in G22 he concludes that not-All'
should imply not-Al', i.e., that Al' should imply All'. However, if this
conclusion is to fit with the rest of Galileo’s criticism, his meaning must be
that the two motions argument must be assuming that Al' implies Al l', so
that, since no reason is given for this implication, no reason is being given
why Al 1' supports Al'. Putting this together with my interpretation of the
G21 criticism, we get that: (G2) there is no reason to think that All' either
implies or supports Al'.
I have not included this proposition in my reconstruction because there is
another way in which he may be interpreted to be combining what I have
called the two lines of criticism, G21 and G22. This other interpretation
emphasizes Galileo’s explicit assertion (FI64.6—164.8) that the two motions
argument (in its first version) is self-contradictory, and it reflects more
accurately the tone of the text at the point where Galileo seems to combine
his two lines of criticism; moreover, as it will be seen below the interpretation
has great logical elegance. However, what it gains in accuracy and elegance it
 GALILEO AS A LOGICIAN 385

loses in psychological plausibility, since several nontrivial steps are omitted

by Galileo; also it has a difficulty which I will discuss shortly. The interpreta¬
tion would correspond to the following deduction, using earlier abbreviations.

(1) If R then S. (This is A12'.)

(2) A121'.
(3) If S then R. (This is G22.)
(4) IfA12l',then [if P then not-(if R then S)]. (This is G2111212.)
(5) If R then P. (This is G2111211.)
(6) If S then P. (From 3 and 5.)
(7) If P then not-(if R then S). (From 2 and 4.)
(8) If S then not-(if R then S). (From 6 and 7.)
(9) If (if R then S) then not-S. (From 8.)
(10) If [if (if R then S) then not-S] then [if (if R then S) then not-R].
(Formally valid.)
(11) If (if R then S) then not-R. (From 9 and 10.)
(12) If R then not-(if R then S). (From 11.)
(13) If [if R then not-(if R then S)] then not-(if R then S). (Inferen-
tially though not truth-functionally valid.)
(14) Not-(if R then S). (From 12 and 13.)
(15) (If R then S) and not-(if R then S). (From 1 and 14.)

The gap is the text is due to the fact that we find only (3) on FI 63.25—
163.27; ‘(2), (4), /. (7)’ on F163.29-163.36; (5) on F163.36-164.2; and
(15) on F164.3ff.
The difficulty with this interpretation stems from (3). Here the G21 line
of criticism is made to depend on the G22 line, which is the opposite of the
first interpretation where the possibility that not-S might support not-R,
through (3), is raised after it has been shown that there is no reason to believe
that not-S implies not R. Moreover, the present interpretation depends on the
assertion of (3), which in the first interpretation is merely considered in order
to conclude that the two motions argument should prove (3). Here, therefore,
the question arises whether (3) is justified.
Galileo begins in G22111111 by saying that not-S is a reason for not-R
and ends concluding in G22 that if S then R. This seems a curious way of
reasoning, tantamount to deriving the converse from a conditional proposi¬
tion. The crucial consideration is G2211111, for G22 plausibly follows from
it (on the assumption, which Galileo presumably ought not to be granting,
that the reason is not pointless). Galileo seems to be saying that if not-R is
asserted because of not-S then in the absence of not-S (i.e., given S) one
386 CHAPTER 16

would not be prepared to assert not-R. In other words, though it is not true
that if the falsity of p implies the falsity of q then the truth of p implies the
truth of q, it is plausible that if the rejection of p leads to the rejection of q
then the nonrejection of p leads to the nonrejection of q. In short, Galileo s
point is perhaps more rhetorical than logical.
But now we get the following problem. If this type of rhetorical point is
correct here, perhaps it could always be made whenever the argument has the
modus ponens form (if p then q\p\ q), for if one accepts q because of p,
then if one did not accept p one would not accept q. This generalization may
be accepted, by declaring it harmless, and then the problem evaporates. How¬
ever, there remains the problem that Galileo would be assuming that not-S
supports not-R. This problem may be solved by saying that Galileo would
then have shown that if not-S supports not-R then the argument is self¬
contradictory, and that therefore if one denies that it is self-contradictory
one is committed to saying that the conclusion A1' is not supported by A11 .
But one reason for this lack of support might be the presence of an implica¬
tion, hence the criticism of the implication ‘if A11' then A’ ’ would seem to
be required.
Finally regarding criticism G3, Galileo distinguishes it sharply from the
rest, calling it a case of ‘false consequence’ by contrast to G2 which he calls a
case of ‘paralogism’ and ‘fallacy’. This seems to amount to saying that
whereas G2 charges an error of reasoning, G3 does not. How is this possible,
in view of the fact that what is false according to G3 is a ‘consequence’,
specifically A112", which is a conditional proposition and hence false
because its consequent clause does not follow from its antecedent clause? The
main difference is that, whereas the two motions argument gives a reason for
A12, it does not for A112; there can be no mistake of reasoning unless there
is reasoning, and no reasoning supports A112, which the argument takes as
obvious. Presumably it is true and relatively obvious if it means A112'; it is
only when it is contrued as A112" that the stellar variations do not follow.
Even so one must be careful here, since certain (annual) stellar variations
would indeed follow, as Galileo himself shows later (F406—416). So it would
have been better for him to say here, as he says later, that the main problem
about these stellar variations caused by the earth’s revolution is that of
determining exactly what the nature of these variations is, exactly enough so
as to put them to experimental test. Since in the present passage Galileo
does show appreciation of the two motions argument construed as supporting
Al", we may take this to be the needed qualification to his otherwise
simplistic-sounding claim about the falsity of Al 12".
 GALILEO AS A LOGICIAN 387

9. GROUNDLESSNESS, AND THE MIXED MOTION ARGUMENT

(F166.30-169.21)

The object argument. (A12) If the earth rotated then a rock dropped from the
top of a tower would not land at its foot since (A12al) on a moving ship a
rock dropped from the top of its mast lands away (stern) from its foot, and
since (A12bl) if the earth rotated and the rock landed at the foot of the
tower then the rock would have a mixture of two (natural) motions, to¬
ward and around the center, (A12b2) which mixture is impossible because
(A12b21) on a rotating earth the air would be incapable of carrying around
falling bodies like rocks (as shown by the fact that (A12b211) strong winds
are observably incapable of impressing motion on heavy rocks). Since (All)
the rock falls at the foot of the tower, it follows that (Al) the earth stands
still. [F166.30—167.34, and FI68.6—169.5]
The meta-argument. (Gl) Proposition A12al does not support A12
because (Gil) there is great disparity between the case of a moving ship and
that of a rotating earth: First, (Gl 111) the ship’s motion is not natural where¬
as the earth’s rotation would be, hence (Gill) on a moving ship the falling
rock has no inclination to follow its motion, whereas on a rotating earth all of
its parts would have the tendency to rotate; second, (G1121) on a rotating
earth the lower part of the earth’s atmosphere would rotate along (because
(G11211) this lower atmosphere is trapped between mountains, and because
(G11212) it contains many earth vapors and exhalations), whereas (G1122)
on a moving ship the air between the top and bottom of the mast does not
follow its motion, so that (Gl 12) the medium through which the rock falls is
moving on a rotating earth but is not for the case of a moving ship. [FI 67.35 —
168.35]
(G2) Proposition A12b211 does not support A12b21 because (G21)
A12b211 exemplifies the case of the air causing a body to start moving after
being at rest, whereas A12b21 exemplifies the case of the air maintaining or
not impeding a body’s motion, and (G22) there is great disparity between the
two cases, and (G23) a situation properly analogous to that of A12b21 would
be a rock dropped from the paws of an eagle while carried by the wind.
[F169.6—169.21]
(G3) Proposition A12b21 does not support A12b2 because (G311) on a
rotating earth falling bodies would have circular motion around the center, by
nature, and hence (G31) they would not need to be carried around by the air.
[F168.11-168.18]
Comments. The structure of the object argument may be visualized as:
388 CHAPTER 16

A1

All A12

A12al A12bl A12b2

A12b21

A12b211

Here A12al has been interpreted as an immediate reason for A12 because
Galileo’s criticism G1 amounts to a criticism of the analogy presupposed in
arguing directly from A12al to A12. Similarly G2 criticizes the analogy pre¬
supposed in the step from A12b211 to A12b21. It is interesting to note that
Galileo strengthens his G2 criticism by mentioning explicitly, in proposition
G23, a correct analogue for the situation mentioned in A12b21; in short, his
G2 criticism is that the presupposed analogy does not hold partly because it
is obvious that it doesn’t if we look at the alleged analogues from a certain
point of view, and partly because the analogy that does hold is some other
specific one.
It would be possible, however, to reformulate the object argument by
structuring it as follows (using the same labels as before):

A1

All A12

A12bl A12b2
i
I I
A12al A12b21

A12b211.

Here the step from A12al to A12b2 would be an explanatory argument, in

the sense that the impossibility of mixed (down and around) motion would
be presented as the explanation for the alleged facts of the ship experiment.
Evidence for the textual accuracy of this interpretation of the falling bodies
argument is a statement of it given by Galileo in a passage where he states in
a preliminary way all arguments against the earth’s rotation (F151 —2).
 GALILEO AS A LOGICIAN 389

Galileo’s criticism could then be reformulated as alternative-explanation

criticism, the alternative being that the rock might fall behind because of the
nonnatural character of the motion imparted to it by the ship and/or because
of air resistance. Therefore, from the evaluative point of view, this reformula¬
tion of the object argument is no better than the other. If the reformulated
argument has an advantage, it is from the point of view of structure, which is
such that all the steps above A12b2 are correct, so that one may conceive of
the reformulated argument as one from the impossibility of mixed motion,
A12b2, and then evaluate it as groundless since the subarguments supporting
this impossibility are wrong. It should be noted that this groundlessness
remains even if one emphasizes the fact that the impossibility is supposed to
pertain to natural motions, for in that case Galileo’s critique of the natural
motion argument would become operative; the effect of that critique here
would be that the two ‘natural’ motions being mixed are natural in different
senses, the straight-downwards in the sense of spontaneous inclination, the
circular rotation in the sense of potentially perpetual state.

10. ANALOGY, AND THE SHIP EXPERIMENT ARGUMENT

(F169.22-180.18)

The object argument. (A121) When a ship stands still a rock dropped from
the top of the mast falls at the foot of the mast; but (A131) when a ship is
moving a rock dropped from the top of the mast falls away from its foot;
therefore, (A13) if the rock falls at the foot then the ship is standing still, and
(A 12) if the rock falls away from the foot then the ship is moving. But (All)
what is true of the ship must be true of the earth. Therefore, (Al) if a rock
dropped from the top of a tower falls at its foot then the earth is standing
still.
The meta-argument. (Gl) The argument’s main premise, A131, is false.
For (Gl 14113) motion is conserved on noninclined surfaces whenever
accidental and external disturbances are removed, since (Gl 141131) on
inclined surfaces downward motion is accelerated and upward motion is
retarded [F171 —3]; and (Gl 14112) when a ship is moving its motion takes
place on a noninclined surface, if accidental and external disturbances are
removed, since (Gl 141121) the surface of a calm sea is horizontal; and
(Gl 14111) when a ship is moving the rock before being dropped has the same
motion as the ship; therefore, (Gl 1411) the horizontal motion the rock has
before being dropped is conserved, if accidental and external disturbances are
removed; therefore, (G1141) the rock will continue to move horizontally
390 CHAPTER 16

with the ship after it is dropped, if accidental and external disturbances are
removed; therefore, (G114) the rock will fall at the foot of the mast on a
moving ship, if accidental and external disturbances are removed. (G113)
There are two relevant disturbances, the air and the rock’s vertical fall; but
(G112) the disturbance of the air may be ignored, since (G1121) its effect is
negligible; and (G111) the rock’s motion of fall does not disturb its horizontal
motion, since (Gil 11) these two motions are not contrary, and since
(G1112) they derive from distinct causes; (G11121) the fall derives from
gravity, while (G11122) the horizontal motion derives from the ‘virtue’ im¬
pressed on the rock by the ship before the drop. The latter can be shown as
follows: (Gil 1221) the rock’s horizontal motion after the drop may be
regarded as projectile motion; now, (G111222) the cause of projectile motion
is either the impressed ‘virtue’ or the medium (air) through which the motion
takes place. But (G111223) the cause cannot be the medium: (G111223all)
moving air pushes light substances more easily than heavier ones, so that
(G111223al) if the air carried the projectile then a ball of cotton could be
thrown further than a rock; (G111223bl) lead pendulums oscillate much
longer than cotton ones; (G111223cl) if the force cannot be impressed by
the thrower directly to the projectile, then it cannot be impressed by the
thrower to the air; (G111223dl) arrows can be shot against the wind; and
(G111223el) arrows travel much less when shot sideways as compared to
normal. It follows that (Gil) the rock falls at the foot of the mast even when
the ship is moving.
Comments. The argument from G1141131 to G114113 is on F171—3; the
rest of the argument up to G114 is on FI74; the rest up to G111222 is on
F175; from there on the argument is on F175—180.18.
Galileo’s critique amounts to a counter-argument designed to support a
positive claim, Gil, about the ship experiment; since this claim is a denial of
a premise, A131,of the ship analogy argument, the latter is thereby criticized.
The logical interest of Galileo’s denial is twofold. First, the denial is not
grounded on a direct experimental test, but rather on relatively easily avail¬
able empirical facts (G1141131, Gil 1223a 11/bl/cl/dl/el), and then these
are inferentially connected with phenomena on the ship; the creation of
such interrelationships is logical practice. Second, because of the structure of
the object argument, the effect of Gil is that nothing follows about the
earth’s motion from the ship experiment: for if we replace A131 by Gil in
the ship analogy argument, then A13 no longer follows, since though A13 is
the contrapositive of A131 it is not the contrapositive of Gil; moreover,
though A12 would still follow from A121, now it would also follow (from
 GALILEO AS A LOGICIAN 391

Gil) that if the rock falls away from the foot, then the ship is standing still;
this together with A12 would yield that if the rock were to fall away from
the foot of the mast, then the ship is both moving and standing still, which is
to say that the rock cannot fall away from the mast’s foot, which merely rein¬
forces the conjunction of A121 and Gil. From A121 and G11, adding All,
it would follow that if the earth stands still the rock falls at the tower’s foot,
and that if the earth rotates the rock also falls at the tower’s foot, from
which, since the rock does fall at the tower’s foot, it doesn’t follow either
that the earth stands still or that it rotates.
Since these considerations are essentially in Galileo’s text, they strengthen
my reconstruction of the ship analogy argument given above, the main
problematic spot being the inferring of A12 and A13 from A121 and A131,
which in the text Galileo justifies by ‘the principle of the converse’ (per il
converso, FI70.2). It would be superficial to object that in the statement of
the argument Galileo is committing the fallacy of affirming the consequent,
since (according to the commonly accepted meaning of ‘converse’) the con¬
verse of ‘if p then q’ is ‘if q then p\ so that he would be inferring A12 from
A131 and A13 from A121. It would be no less superficial to say that, since
Galileo does not accept the ship analogy argument, though he cannot be
faulted for its statement, which is accepted by the Aristotelian spokesman,
he could be faulted for not calling attention to this error in the statement of
the argument. Such criticism would be groundless or circular since there is no
more reason to equate Galileo’s ‘converse’ with the one in contemporary logic
textbooks than with the ‘contrapositive’ found therein; second, if Galileo’s
‘converse’ is interpreted in the logic textbook sense, this would make non¬
sense of his argument that the ship experiment proves nothing about the
motion of the earth, for from Gil and A121 he would have concluded that if
the rock falls at the foot then the ship, and hence the earth, is both moving
and standing still, which is to say that the ship experiment proves too much,
rather than nothing.
Finally, it should be mentioned that, as reconstructed, the subargument to
G111223 is not as explicit as it could be. The analysis of its latent structure is
left as an exercise to the reader.

11. DEDUCTION, INDUCTION, RHETORIC, AND THE

EAST-WEST GUNSHOT ARGUMENT (F194.12-197.1)

The object argument. (All) If the earth rotated, gunshots would range
further toward the west than toward the east; for (A1111) on a rotating earth
392 CHAPTER 16

the projectile’s motion toward an eastward target would be in the same direc¬
tion as that of the gun due to the earth’s motion, whereas (All 12) the
projectile’s motion toward a westward target would be in the opposite direc¬
tion as that of the gun, and hence (All 1) on a rotating earth the distance
between the gun and the point hit will be greater in a westward than in an
eastward shot by an amount equal to the total distance traversed by the gun
during the time of flight of both balls. Since (A 12) gunshots toward the west
and toward the east are observed to range equally, it follows that (Al) the
earth does not rotate.
The meta-argument. (G1131222) If the earth’s rotation implies unequal
ranges for east-west gunshots, then arrow shots which range, e. g., 300 yards
from a motionless cart would range 200 yards in the direction of the cart and
400 yards in the opposite direction, from a cart moving at a speed equivalent
to 100 yards per time of each shot [F194.27—195.13]; (G1131221) if so,
then from a moving cart both shots would range 300 yards when those in the
direction of the cart are shot with a force equivalent to a range of 400-yards-
from-a-motionless-cart, and when those in the opposite direction are shot
with a force equivalent to a range of 200-yards-from-a-motionless-cart
[F195.13-195.30] ; therefore, (G113122) if the earth’s rotation implies
unequal ranges for east-west gunshots, then from a moving cart both shots
would range 300 yards when those in the direction of the cart are shot with
a force equivalent to a range of 400-yards-from-a-motionless-cart, and when
those in the opposite direction are shot with a force equivalent to a range of
200-yards-from-a-motionless-cart. But (G113121) the greater and lesser
amounts of shooting force give rise to correspondingly greater and lesser
speeds for the arrows. Therefore, (G11312) if the earth’s rotation implies
unequal ranges for east-west gunshots, then from a moving cart both shots
would range 300 yards when those in the same direction move with 4 degrees
of speed, those in the opposite direction move with two degrees of speed, and
the cart moves with one degree of speed [F195.31—196.12]. But (G11311)
if from a cart moving with one degree of speed arrows are shot with equal
strength in opposite directions (the strength being equivalent to three degrees
of speed from a motionless cart), then those shot in the same direction move
with four degrees of speed and those shot in the opposite direction move
with two degrees of speed [F196.13—196.26], since (G113111) the effective
speed of the shots in the opposite direction is the sum of the cart’s speed plus
the speed imparted to the arrow by the crossbow, whereas for the same
direction the effective speed is the difference between them. Therefore,
(G1131) if the earth’s rotation implies unequal range for east-west gunshots,
 GALILEO AS A LOGICIAN 393

then if from a cart moving with one degree of speed arrows are shot with
equal strength (equivalent to three degrees of speed from a motionless cart) in
opposite directions, then both shots would range 300 yards. Therefore,
(G113) if the earth’s rotation implies unequal ranges for east-west gunshots,
then arrows shot in opposite directions from a moving cart range equally
[F196.26—196.31]. But (G112) if so, then on a moving earth east-west
gunshots would range equally; and (Gill) if the latter, then the earth’s
rotation would not imply unequal ranges for east-west gunshots; therefore,
(Gil) if the earth’s rotation implies unequal ranges for east-west gunshots,
then it does not so imply. Therefore, (Gl) the earth’s rotation does not imply
unequal ranges for east-west gunshots. (I.e., A11 is false.)
Rather, (G214) if the earth rotates, then the eastward projectile is moving
with a greater (absolute) speed than the westward one, since (G2141) the
(absolute) speed of the eastward projectile is the sum of what it receives from
the gunpowder and from the moving earth, whereas the (absolute) speed of
the westward projectile is the difference between what it receives from the
gunpowder and from the earth; and (G213) if the eastward projectile moves
(absolutely) faster, then it travels a greater (absolute) distance than the west¬
ward one; and (G212) if so, then the distance (relative to the moving earth)
traveled by the eastward projectile is the same as that of the westward pro¬
jectile, since (G2121) for the eastward projectile the relative distance equals
its absolute distance less the distance traveled by the gun, and for the west¬
ward projectile the relative distance equals its absolute distance plus the
distance traveled by the gun; and (G211) if the relative distances are the
same, so are the ranges, since (G2111) on a rotating earth the observed range
of a shot would be the difference between the absolute distance traveled by
the projectile and that traveled by the gun. Therefore, (G21) if the earth
rotates, then the ranges of east-west gunshots would be equal. Therefore,
(G2) east-west gunshots would have equal ranges whether the earth rotates
or stands still.
(G31) Proposition A111 follows from All 11 and All 12 if and only if
one assumes that, All 13, the distance between the gun and the point hit
equals the sum of the distances traveled by the projectile and by the gun for
a westward shot, and their difference for an eastward shot; and that, A1114,
in this sum and this difference the distance traveled by the projectile on a
rotating earth equals the distance traveled on a motionless earth. But (G32)
this equation, in All 14, amounts to assuming that the earth stands still.
Therefore, (G3) in the step from All 11 and All 12 to All 1, the argument
begs the question. [F196.34—197.1]
394 CHAPTER 16

Comments. The G1 criticism amounts to showing that in the object argu¬

ment the step from A12 to A1 is invalid, i.e., that A1 does not follow from
A12. It also means that (the facts about) east-west gunshots do not imply
that the earth stands still and do not constitute a valid objection to the
earth’s motion. G1 as such does not mean that the earth’s rotation implies
equal ranges, but in the argument supporting G1 there are implicit proposi¬
tions which suggest this implication, which is made explicit in the G2 argu¬
ment (proposition G21). Because of G21, east-west gunshots do not even
support the earth standing still, since though those facts could be explained
by the earth standing still, they could also be explained by its rotation, so
that an explanatory argument from east-west gunshots to the earth standing
still would be invalid, because arbitrary. In short, the argument from east-
west gunshots is deductively invalid (Gl), inductively invalid (G2), and
rhetorically invalid (G3).

12. INFERENTIAL VS MATERIAL ‘IF’, AND THE

VERTICAL GUNSHOT ARGUMENT (F 197.1-202.16)

The object argument. (All) In vertically upwards gunshots the projectile

returns to the gun. But (A 12) if the earth were rotating, the projectile would
return to a place west of the gun, since (A12al) if the earth rotates then
during the flight of the projectile the gun has been carried a long distance
toward the east, and since (A12bl) if the earth rotated and the projectile
returned to the gun then the projectile would have moved transversally, and
(A12b2) if the projectile had moved transversally, we would observe its
transversal motion (because (A12b21) otherwise we would have a denial
of the senses), but (A12b3) we do not observe any transversal motion.
Therefore, (Al) the earth does not rotate. [F197.6—197.14, and 200.6—
200.13]
The meta-argument. (Gl) The vertical gunshot argument begs the question
since (Gil) in grounding A12 on A12al the argument assumes the final
conclusion, Al, that the earth does not rotate; for (Gill) the rest of this
subargument (its latent structure) would have to be the following: “(A12al)
if the earth rotates then during the flight of the projectile the gun has been
carried a long distance toward the east, and (A12a2) if the gun is carried a
long distance toward the east then the projectile returns to a place west of the
gun, because (A12a21) the projectile is not carried the same distance toward
the east”; and (G112) the only justification that could be given for A12a21 is
that: “(A12a211) the projectile is on a motionless earth, since (A12a2111)
 GALILEO AS A LOGICIAN 395

the earth stands still”; and (G113) proposition A12a2111 is identical to Al.
[F197.1-197.10, and 200.13-200.19]
Moreover, (G2) proposition A12 is false since (G21) if the earth were
rotating the projectile would indeed return to the gun; for (G21al 11) on a
rotating earth both the projectile and the gun would be moving eastward, and
hence (G21all) the projectile would not start its up and down motion from
rest but rather would have its up and down motion combined with its east¬
ward rotation; since (G21al2) its eastward rotation is neither eliminated nor
impeded by its up and down motion, it follows that (G21al) the projectile
remains always over the gun. Also, (G21bl) observation would show that on
a ship an arrow shot upwards returns to the bow whether or not the ship is
moving. [F200.19-200.29]
Third, (G3) proposition A12b2 is groundless and false, groundless because
(G3al) there would be no denial of the senses if the projectile moved trans-
versally and we did not observe its transversal motion (i. e., because its
supporting reason, A12b21, is false), and false because (G3bl) we could not
observe the transversal motion of the projectile if it was so moving. In fact,
(G3bll) on a rotating earth when a cannon is fired at AC, by the time its
ball reaches its mouth, it will have been carried to DE, the ball will have
traveled the transversal CD, but we would have observed it to go along the
cannon’s barrel, which is not transversal [F200.30—202.16]:
A D

s'
*
xT
/
/
s'
/
s'
S

a'
c E

and there is no denial of the senses here because (G3all) it is the nature of
motion to be such that shared motion is imperceptible, and (G3al2) in this
situation the horizontal rotatory motion is shared by us and by the observed
projectile. [F197.15—199.26]
Comments. If the G1 criticism is to be correct in alleging that the object
argument presupposes what it is trying to prove, then the subargument
causing this circularity ought to be a valid deduction, and if so A12 would
396 CHAPTER 16

follow from its final reasons A12al and A12a2111, so that if(asG2 argues)
A12 is false then either A12al or A12a2111 should be false. Since A12al
would seem obviously true, it follows that A12a2111 must be false, which
amounts to a proof of the earth’s motion. This is puzzling and problematic,
the problem being analogous to the one that arose in Chapter 12 in connec¬
tion with the circularity argument from actual vertical fall.
Let us therefore ask whether A12al is really true. Its apparent truth
derives, I believe, from an ambiguity: ‘east’ in this context can mean either
“east from the point of view of absolute space” or “east from the point of
view of the earth’s surface”. The gun would be carried toward the east only in
the former sense, not in the latter. Now, let us see what is the sense of ‘east’
and ‘west’ in A12a2. This would be the sense in which it would follow from
A12a2111, which is the sense of absolute space for ‘east’, and the sense of
terrestrial surface for ‘west’. It follows that A12al is indeed true; hence the
above mentioned paradoxical proof of the earth’s motion remains.
Of course, this proof would evaporate if one cannot assert the falsehood
of A12, which is asserted in proposition G2, so one might question the
correctness of its supporting argument. The step from G21 to G2 presupposes
that ‘if p then q’ implies ‘not-(if p then not-*?)’, which as before could be
questioned for being truth-functionally invalid, but which I would not want
to question; the other steps seem correct. G21al2 could of course be
questioned, but Galileo could give his previous justification of it, involving
a claim about two distinct causes being in operation; at any rate there would
be no trace of presupposing a claim that the earth moves (which trace might
explain the paradox). G21al 11 seems either obvious or dependent merely on
conservation of motion. So our puzzle remains.
Let us look once again, and more closely, at the deduction that generates
the paradox, using the following abbreviations:

R = the earth rotates;

S = the projectile returns to the same place (on the earth’s
surface);
Eg = the gun is carried to the east.

The question-begging inference analyzed in G1 reduces to:

(T) if A12al and A12a2111, then A12.

The crucial conclusion in G2 is:

(2') not-A12.
 GALILEO AS A LOGICIAN 397

These may be rephrased as:

(1) if (if R then Eg) and not-R, then (if R then not-S);
(2) not-(if R then not-S);

which seem to entail

(3) not-[(if R then Eg) and not-R],

by modus tollens, and since

(4) if R then Eg

is true, we get

(5) R.

I believe that the problem is with the ambiguous meaning of ‘if R then not-S’.
There are several reasons why in (2) its meaning is inferential, that is, ‘not-S
can be inferred from R’: first, in the context of criticism G2, Galileo wants
to deny the validity of the inference from S (=A11) to not-R (=A1), which
denial is what (2) asserts; second, (2) is justified by proposition G21 asserting
that ‘if R then S’, which presupposes that ‘if R then S’ and ‘if R then not-S’
are not both true, which is only inferentially and not truth-functionally
correct (given the self-consistency of R, obvious in this case); third, if (2) had
a truth-functional meaning, R would follow immediately in one step by the
definition of the material conditional, which is false if and only if the
antecedent is true and the consequent false. On the other hand, in (1) the
meaning of ‘if R then not-S’ is not inferential since the context, Gl, of this
conditional is one where it is being treated as a single proposition, in order
to show that its justification presupposes not-R; moreover, in (1) it is obvious
that it is the main ‘if then’ connective which is the inferential one, so that the
whole meaning of (1) is that “(if R then not-S) can be inferred from (if R
then Eg) and not-R”. This shows that the above paradox was the result of a
fallacy of equivocation.11
Finally, the G21b argument presupposes an analogy between the moving
earth and a moving ship, which Galileo criticized in the context of the ship
experiment argument; however, I do not think there is any real conflict
because the disanalogy is destructive only with respect to the alleged obser¬
vation of different phenomena on a moving and on a motionless ship; given
the same effects, the disanalogy reinforces the inference from the ship to the
earth since on a moving ship there would be more possible causes for the
arrows to be left behind.
398 CHAPTER 16

13. EXPLANATION VS INFERENCE, AND THE

HUNTER’S ARGUMENT (F203.34-205.29)

The object argument. (A14) Hunters can hit birds while they are flying;
(A 13) the explanation of this fact is not the hunter’s anticipation of the bird’s
motion when aiming at it, that is, the practice of aiming at the point where
the bird will be when the bullet gets there rather than aiming at the bird,
since (A 131) hunters do not aim at the target by anticipating it; rather, (A 12)
by moving their gun hunters follow the bird in its motion while they aim, and
(All) in going from the gun to the bird the bullet retains the gun’s motion
parallel to that of the bird; therefore, (Al) the explanation is the fact that
hunters aim by moving their gun to follow the bird.
The meta-argument. (Gil) The explanation cannot be the practice of
aiming by moving the gun to follow the bird, since (G1111) the gun’s motion
is very small compared to the parallel motion of the bird, and hence (Gill)
the bullet cannot keep up with the bird; therefore, (Gl) the explanation must
be the combination of several factors: (a) aiming by moving the gun, (b)
aiming by anticipation to a small extent, (c) the use for charge of a large
number of small pellets rather than a single large ball, and (d) the great speed
with which the shot moves toward the bird.
Comments. The object argument is not an explanatory argument in the
sense that its conclusion states an alleged explanation for some fact stated in
the premises. Rather the conclusion is saying that the fact mentioned in
premise A12 explains the fact mentioned in A14, so that the argument is one
that tries to establish an explanatory connection, taking both the expli-
candum, A14, and the explicans, A12, as premises. Since it is clear that the
net effect of the whole argument is to account for (i.e., to explain) A14, we
might regard the whole argument itself as constituting an explanation. We
may then define an explanation as an argument whose premises include both
the explicandum and the explicans, and whose conclusion asserts that the
explicans explains, accounts for, or renders comprehensible the explicandum.
The meta-argument refutes the object argument explanation by refuting
some intermediate proposition presupposed by the step from A14, A12, and
All to Al, namely that the bullet’s motion parallel to the bird and imparted
to it by the hunter’s moving of the gun barrel is equivalent to the bird’s
parallel motion (and thus sufficient to make the bullet keep up with the bird).
Gllll and Gill are denying these propositions. Thus we may say that the
justification of an explanation would be a justification of the correctness of
the argument constituting the explanation.
 GALILEO AS A LOGICIAN 399

The next important point pertains to the structure of the explicandum per
se. Here it was stated merely as a proposition, but it is clear that not every
proposition will generate an explicandum; in fact, in this case the use of the
modal qualifier ‘can’ inA14 constitutes an implicit reference to the possibility
that perhaps hunters should be unable, or have difficulty with, hitting moving
birds. In short, what constitutes an explicandum is really a factually asserted
proposition, plus an argument having its denial as conclusion. In our case this
argument would be: “(A221) when a hunter aims and shoots at a flying bird,
the bird usually has a component of motion perpendicular to the line of sight
at any given moment; therefore, (A22) if a hunter shoots at the bird, where
he sees it, he will miss; but (A21) hunters do aim at birds when they shoot;
therefore, (A2) it should be impossible to hit a flying bird”. The explanation
given in the object argument really amounts to criticizing the inference from
A221 to A22 by pointing out that all that follows from A221 is that if a
hunter shoots at the moving bird without moving his gun then he should miss;
whereas explanation A13 was inferring merely that if A21 is true then A2 is
true, but denying A21. The G1 meta-argument introduces additional refine¬
ments in the step from A221 to A22.

14. SCIENTIFIC, AD HOMINEM, LOGICAL INVALIDITY,

AND THE POINT-BLANK ARGUMENT (F205.30-209.1 3)

The object argument. (A 12) If the earth were rotating then point-blank gun¬
shots toward the east would hit high and toward the west low, since (A121)
point-blank gunshots move along the tangent, WE, to the point of firing, O,
and (A 122) the earth’s rotation would cause an eastern target, Te, to descend,
e.g. to Te', and a western target, T\g, to rise, e.g. to Tyj, relative to the
tangent, WE\ but (All) there are no up and down deviations in point-blank
gunshots; therefore, (Al) the earth does not rotate. [F205.32-206.7]

The meta-argument. Since (Gill) on a rotating earth the gun as well as

the target are moving with respect to the tangent, (G11) the gun and the
400 CHAPTER 16

target are not moving relative to one another, and hence (Gl) the projectile
will travel from one to the other in the same way as on a motionless earth.
(G211) If A12 is true then the up and down deviations would be imper¬
ceptibly small, because (G2111) it can be shown that they would be of the
order of an inch; therefore, (G21) we do not know whether or not point-
blank gunshots actually exhibit these deviations; therefore, (G2) proposition
A11 is groundless.
(G3) The lack of deviation, A11, does not imply that the earth stands still,
Al, because (G3112) if the earth rotated then artillerymen would have
learned to aim point-blank gunshots as would be required on a moving earth
to hit on target, and (G3111) if artillerymen had learned to aim on a moving
earth then there would be no deviation, and hence (G311) if the earth rotated
then there would be no deviation, and hence (G31) the earth’s rotation does
not imply a deviation. [F208.20—208.25]
Rather, (G4) one might say that there would be a deviation if the earth
stood still. For (G41122) if artillerymen had learned to aim on a moving
earth and if the shot follows a tangential path, then artillerymen would be
aiming higher to hit lower toward the west, and lower to hit higher toward
the east, but (G41121) if artillerymen were aiming higher to hit lower
(toward the west) and lower to hit higher (toward the east), then if the earth
stood still point-blank gunshots would hit high toward the west and low
toward the east, and hence (G4112) if artillerymen had learned to aim on a
moving earth and the shot followed a tangential path, then if the earth stood
still then point-blank gunshots would hit high toward the west and low
toward the east; but (G4111) if the earth were rotating then artillerymen
would have learned to aim on a rotating earth; therefore, (G411) if the earth
were rotating (and shots followed a tangential path) then the earth standing
still would imply a deviation; therefore, (G41) to show that the earth
standing still would not imply a deviation you have to show that the earth is
standing still (or else that the shots do not follow a tangential path).
[F208.25—208.28]
Comments. The effect of Gl is to invalidate the argument from All to Al
by discrediting A12. But how correct is this criticism? Certainly it would be
wrong to say that it presupposes ‘circular inertia’,12 for Galileo does not say
that the path of the ball also curves downward (for the eastward shot), but
rather that the gun is continuously being inclined (F206.9); hence the pro¬
jectile would have an additional component of velocity ‘downwards’ with
respect to the (motionless) tangent, and it will be conserved (given the prin¬
ciple of conservation of motion, justified earlier). Rather, the situation here
 GALILEO AS A LOGICIAN 401

is like that alleged for the shooting at flying birds in the hunter’s argument
(F202—5), with the difference that the parallel speeds of both gun barrel and
target are here essentially the same. I say ‘essentially’ because if the surface
distance between the gun and the eastern target is of the order of 90 degrees
then the ‘downward’ speed of the latter is approximately equal to the
rotational speed, and hence much greater than the ‘downward’ speed of the
gun, which, being at the point of tangency (by definition), is only a fraction
of the rotational speed; however, the equivalence is approximately correct for
relatively small distances between the gun and the target. Thus the situation is
really like that of the north-south gunshots, concerning which Galileo himself
makes the explicit qualification that the answer he gives applies only for
ordinary relatively short ranges (F205.10—205.13). This gives additional
significance to the quantitative considerations made in G2.
Galileo’s criticism G2 concentrates explicitly on what might be called the
groundlessness of the crucial premise All. However, one could pursue the
same line of reasoning as follows:
If there is a deviation however small, then point-blank gunshots do not in fact
hit correctly; hence, because of G211, if A12 is true then A11 is not; but A12
means that the earth’s rotation implies not-Al 1; it follows that if the earth’s
rotation implies not-Al 1, then not-Al 1; i.e., if A11 then the earth’s rotation
does not imply not-Al 1; i.e., if All then All does not imply that the earth
stands still; i.e., All does not imply Al. Thus G2 can also be interpreted as
a critique of the validity of the point-blank argument, and a more effective
one than Gl.
The justification of G2111 is given explicitly by Galileo and amounts to
the following:
(1) Assuming that to reach its target a cannon ball takes about as
long as it takes a pedestrian to walk two paces, the ball will take
about one second to reach its target.
(2a) The western horizon rises 15 seconds of arc in one second of
time, since it turns 360 degrees in 24 hours (assuming we are at
the equator).13
(2b) Therefore, a western target rises 15 seconds of arc in one second
of time.
(2c) Assuming the distance from the cannon to the target is 500
braccia, the 15 seconds of arc are measured along a circle with a
radius of 500 braccia.

What Galileo seems to have in mind here (2a, b, c) may be explained by

402 CHAPTER 16

referring to the above figure. The cannon is resting at 0\E is the eastern and
W the western horizon. On a stationary earth the cannon ball would travel in
one second from O to the target Tw. On a moving earth the western horizon
rises and the eastern one falls so that in one second WE would pivot around O
to a new position W'E', where angle WOW' is 15 seconds; the target is now at
7V'; disregarding the eastward component of the target’s motion (since this
argument is considering only deviations above and below the tangent), 07V'
is also 500 braccia.

(3a) The chord for one minute, on a radius of 100000 units, is 30

units.
(3b) Therefore, the chord for one second, on a radius of 100 000 units,
is (1/60) X 30 = 1/2 units.
(3c) That is, the chord for one second, on a radius of 200 000 units,
is 1 unit.
(3d) Therefore, the chord for 15 seconds, on a radius of 200 000, is
15 units.
(3e) Therefore, the chord for 15 seconds, on a radius of 500 braccia =
(500/200000) X 15 braccia = 15/400 braccia = approximately
4/100 braccia.
(3f) Therefore, the target rises approximately 4/100 braccia, which is
approximately one inch.

What Galileo seems to be doing here (3a-f) is to find the distance from Tw'
to the line OW. He is approximating this distance by the arc 77V', whose
center is O. The approximation is correct since the angle TOTw' is very small
(15 seconds). Galileo’s answer checks with the following calculation:
57T
12X36

: approx‘ 12^7 = = aPProx.

(4) Therefore, the deviation is about 1/25 braccia, or about one inch.

What Galileo does in (4) is to reason that, since the target ‘rises’ one inch, the
shot will be one inch ‘low’.
An interesting question to ask about this part of the critique is whether it
is ad hominem, in the seventeenth century meaning of this term, that is
whether the argument draws consequences of ideas accepted by one’s
opponents, but not necessarily acceptable or accepted by oneself. The idea in
question here would be that involved in thinking of the horizon as rising and
 GALILEO AS A LOGICIAN 403

falling, on a moving earth. It seems un-Galilean to think that the motion of

the earth would involve the kind of shift of the horizon suggested by our
figure. And yet this helps rendering the passage comprehensible. All this is
evidence that the argument is indeed ad hominem.
Galileo’s criticism G3 constitutes another reason for the invalidity of the
step from All to Al.
I believe that G4 amounts to an argument to the effect that if the point-
blank argument is valid, then it is incomplete insofar as there is no justifica¬
tion that the earth’s motionlessness would not imply a deviation; one is
entitled to this proof in view of G411, Al, and All.
It is interesting to contrast the three reasons Gl, G2, and G3, for the
invalidity of the object argument. Both Gl and G2 involve scientific sub¬
stantive considerations, whereas G3 is a rather rhetorical kind of criticism;
this rhetoric pertains to the implicit consideration that what would count as
deviation and as proper point-blank aim depends partly on whether or not the
earth rotates. The reasoning in Gl is rather external insofar as there is an
appeal to both composition and conservation of motion, which the pro¬
pounders of the object argument do not accept; whereas criticism G2 is
wholly internal to the object argument, and this easily explains the greater
effectiveness of G2.

15. THE LOGIC OF PUNCTUATION, AND THE BIRDS ARGUMENT

(F209.13-212.30)

The object argument. (A 121) When birds fly they are moving at will; there¬
fore, (A 12) if the earth were rotating then flying birds would have to keep up
with the terrestrial rotation through their own efforts; but (All) it is incre¬
dible that they would have enough energy to move so fast; therefore, (Al)
the earth cannot move. [F209.13—209.25]
The meta-argument. (Gil) If the earth were rotating then the lower
regions of the air would be rotating with it; (G12) if so, then flying birds
would keep up with the terrestrial rotation through the agency of the air;
therefore, (Gl) proposition A12 is not correct. [F209.26-210.20]
(G2) The birds argument seems stronger than the projectile arguments
because (G211) birds can move at will against the innate tendencies of
terrestrial bodies (for example, (G2111) a live bird can fly upwards whereas
a dead one will fall downwards like any heavy body), and hence (G21) the
reasoning that applies to projectiles does not apply to birds. [F212.3—212.10]
(G311) The fact that the reasoning that applies to projectiles does not
404 CHAPTER 16

apply to birds means that on a rotating earth we would not see birds do the
same thing as projectiles, first because (G311al) if a dead bird were dropped
from a tower it would like a stone continue to rotate and also start fading,
but (G311a2) if the bird were alive then it would continue to rotate and also
use its wings to go where it wanted, and (G31 la3) if the latter happened then
we would only see the motion due to its wings (since (G311a31) this would
be the only motion not shared by us), and second because (G311bl) on a
rotating earth the bird’s ability to move at will would amount to an ability
to add or subtract degrees of speed with respect to the fixed rotational speed;
hence (G31) there would be no observable difference between the flight of
birds on a rotating earth and on a motionless one; hence (G3) the greater
strength of the birds argument over the projectile arguments does not make it
correct. [F212.10—212.26]
Comments. The reconstruction of G2 and G3 was made very difficult by
the punctuation found in Favaro’s standard critical edition, which, as usual,
is different from that of the 1632 edition; though normally an improvement
over the original, in this case Favaro’s punctuation was not. The relevant
Favaro text can be seen in the following literal translation, which is also a
line-by-line translation, so that the line numbers noted here correspond to
those on p. 212 of the Favaro edition:

3 Your having more difficulty for this than for the other objections,
4 seems to me to depend on birds being animate, and being able
5 therefore to use force at will against the primary motion innate
6 in earthly things, in such a way precisely that we see them, while they
are
7 alive, flying even upwards, a motion impossible to them as heavy bodies,
whereas
8 dead they cannot but fall downwards; and therefore you judge
9 that the reasons that take place for all sorts of projectiles mentioned
10 above, cannot take place for birds; and this is very true, and
11 because it is true, therefore we do not see, Mr. Sagredo, those projectiles
doing
12 what birds do: for if from the top of a tower you let
13 fall a dead bird and a live one, the dead one will do the same that
14 a stone does, that is it will follow first the general diurnal motion, and
then
15 the downward motion, as a heavy body; but if the released bird is alive,
who
 GALILEO AS A LOGICIAN 405

16 will forbid it, the diurnal motion always staying with it, from going,
17 by a beating of wings, toward whatever part of the horizon it likes most?
and
18 this new motion, being its own in particular and not shared by us,
19 must become perceptible to us. And if it had by its flight moved toward
20 the west, who is to forbid it from returning to the tower
21 by the same beating of wings? Because, finally, to fly up
22 toward the west was nothing other than to subtract from the diurnal
motion, which has,
23 for example, ten degrees of speed, a single degree, so that it was left
with
24 nine, while it was flying; and if it had alighted to earth, it would have
regained
25 the ten common degrees, to which by flying toward the east it could
have
26 added one, and with the eleven return to the tower: and in sum,
27 if we well consider and more intimately contemplate the effects
28 of the flying of birds, they do not differ in anything from projectiles
thrown in
29 all directions, except that these are moved by an
30 external projector, and those by an internal principle.

The punctuation in this passage suggests that the main break comes on
line 19, because of the period on that line. This would suggest that lines
19-30 constitute either the second of two main arguments in the passage
(the first being presumably on lines 3-19), or else a self-contained subargu¬
ment of the main argument. Either alternative runs into the problem that
lines 26-30 do not connect with anything else in this passage. If we mentally
delete these lines, and if we follow our logical intuition telling us that lines
19-26 do not constitute one of two main arguments, then we are left with
the following subargument interpretation: (lines 19—21) on a rotating earth
birds would be able to fly since (lines 21—26) flying on a rotating earth
would involve being able to add and subtract degrees of speed with respect to
the fixed rotational speed. (This two-proposition argument was collapsed into
a single proposition, G311bl, above.) Now, how would this subargument
connect with the rest of the passage? Looking at lines 3-19, the colon strikes
one as the most important punctuation mark; it indicates either a break inter¬
mediate in force between a period and a semicolon, or else that an explana¬
tion or justification follows. Whichever is the case, since lines 12—19 are
406 CHAPTER 16

obviously reasons for something, the next question is what is the claim which
they support. The suggested choice is the proposition (lines 11—12) that we
do not see projectiles doing what birds do. Unfortunately, logical intuition
vetoes this interpretation, since such a conclusion (lines 11 — 12) is about what
we see, and the propositions on lines 12—19 are about what would happen on
a rotating earth, and hence lines 11—12 do not connect with lines 12—19.
The problem is apparently alleviated by the fact that the reasoning indicators
on lines 10—11 seem to provide the reason for the claim on lines 11—12. This
would lead one to think that the main break in the entire passage comes with
the colon on line 12, for after all lines 12—19 connect pretty well with lines
19—26, better than they connect with lines 3-12. The second part of the
passage (lines 12—26) would then involve the minor problem of finding a
proposition supported by the claims made therein. The first part (lines 3-12)
would involve the major problems that lines 11 — 12 do not connect with lines
9—11, since on the one hand the ‘this’ of fine 10 should refer to lines 9—10,
but on the other hand one cannot conclude that we do not see birds and
projectiles behave similarly (lines 11-12) from the fact that “the reasons
which take place for all sorts of projectiles mentioned above, cannot take
place for birds” (lines 9-10). The situation would now be that lines 11-12
connect neither with what follows nor with what precedes; because of this,
one might try an appropriate reformulation of its literal meaning, namely that
we would not see on a rotating earth those projectiles doing what birds would
do. This proposition does connect very well with lines 12-26; hence lines
11—26 would now make sense as a unit; hence a main break should have been
indicated on line 10 or 11, by a period perhaps; this is especially true since
lines 3-10 can be easily interpreted as a distinct argument, as it is done in G2
above. The only residual problem that this division of the passage at line 10
would generate stems from lines 10-11, which strongly suggest an inferential
connection; the problem is solved by interpreting lines 10-11 (“and this is
very true, and because it is true, therefore ... ”) as a way of indicating that
the immediately preceding proposition (lines 9-10) can be rephrased to mean
what lines 11-12 state (according to the nonliteral interpretation worked out
above).
The preceding account describes the manner in which I arrived at my
reconstructions G2 and G3. But it would probably have been impossible for
me to arrive at them it 1 had not had the benefit of the more amorphous
punctuation in the 1632 edition of the Dialogue. Among other differences,
the latter has a semicolon on line 12 instead of a colon, a semicolon on line
19 instead of a period, and a period on line 26 instead of a colon.14
 GALILEO AS A LOGICIAN 407

16. IGNORATIO ELENCHI, QUANTITATIVE INVALIDITY,

INCOMPLETENESS, AND THE CENTRIFUGAL FORCE ARGUMENT
(F214-244)

The object argument. (A 12) If the earth were rotating, objects on its surface
would be scattered away from it, toward the heavens, because (A121) rota¬
tion has the power of extruding objects lying on the rotating body; but (All)
objects on the earth’s surface are not observed to be scattered toward the
heavens; therefore, (Al) the earth does not rotate.
The meta-argument. (G122) Proposition A12 does not follow from A121
because (G1221) what follows is either that (a) if the earth had always been
rotating then there would (now) be no objects on its surface, or that (b) if
the earth were to begin rotating then objects on its surface would be scattered
toward the heavens; but (G121) proposition All connects only with (b)
above; therefore, (G12) what follows from A121 and All is that the earth
has not (recently) begun to rotate. But (Gil) propositions A121 and All are
the final reasons of the object argument. Therefore, (Gl) what the object
argument really proves is not its conclusion, Al, but rather the conclusion
that the earth has not (recently) begun to rotate. [F214.30-216.ll]
Second, (G2) the argument is quantitatively invalid because (G21) in order
for A12 to follow from A121 the centrifugal tendency due to the extruding
power has to be sufficiently great; but (G22) it is not because (G221) the
downward tendency due to weight could overcome the centrifugal tendency
due to rotation; for (G2211) the centrifugal tendency is measured by the
tangential displacement (that would result if the body were extruded), and
(G2212) the downward tendency is measured by the displacement along the
secant from the body to the earth’s surface, and (G2213) the geometry of the
situation is such that a body can remain on the earth’s surface by undergoing
a very small secant displacement while it undergoes a very large tangential
displacement, and (G2214) the secant speed, however small, is always
sufficient to compensate for the tangential speed, however large, since
(G22141) from the center of a circle, O, one can draw a secant, 01, inter¬
secting a tangent, 7T', so close to the point of tangency, T, as to yield an
arbitrarily small ratio between the portion of the secant outside the circle,
IC, and the portion of the tangent to the point of tangency, IT. [F217.4-
224.28]
Third, (G31111) though it is true that the cause of extrusion increases as
the speed when the radius is constant [F238.2-238.6], (G31112) when the
speeds are equal the cause of extrusion decreases as the radius increases
408 CHAPTER 16

T I

[F243.25—243.26], so that (G3111) the cause of extrusion increases directly

with the speed but inversely with the radius [F238.25—238.27]; thus (G311)
it may be that the cause of extrusion remains constant when the speed is
increased as much as the radius, namely when equal number of rotations are
made in equal times [F244.6-244.il]; hence, (G31) the earth’s rotation
would cause as much extrusion as a wheel which rotated once in twenty four
hours [F244.ll—244.15]; therefore, (G3) a rotating earth would not scatter
its surface bodies toward the heavens. In support of G31112 note that
(G311121) the cause of extrusion is equal to the force needed to prevent the
body from escaping along a tangent, that (G311122) this force is that re¬
quired to compensate for the tangential displacement if the body had been
extruded, and that (G311123) this compensation is smaller for greater circles

12
 GALILEO AS A LOGICIAN 409

since (G3111231) when the linear speeds are the same (e.g., along TjCx and
T2C2) the deviation from the tangent is measured by the distance from the
tangent to the circumference (12Cx and I2C2), and (G3111232) this distance
decreases as the circle becomes larger (I2C2 < GCh) [F242.11—243.26].
Comments. (1) In accordance with criticism Gl, the argument from
whirling may be regarded as a classic instance of ignoratio elenchi (or
‘irrelevant conclusion’), in Aristotle’s own sense of an argument alleged to
prove one thing, but which at best proves something else.15
(2) I have used the term ‘quantitative invalidity’ in my reconstruction of
the G2 critique. The suggestiveness of this term is obvious; its importance
derives from the fact that it is a very important special case of failure of
implication. It is the case when the reasons why one proposition does not
follow from another involve quantitative considerations. In other words, it
is true that (A12) on a rotating earth objects would be scattered toward the
heavens, but the amount of such scattering is very small and also smaller than
the amount by which they move toward the earth’s center due to the weight.
Similar considerations would apply for G3.
(3) As historians of science have not tired of pointing out, in G2 the step
from G22141 to G2214 is incorrect. However, it is one thing to say that it
is invalid, and it is another to understand what is wrong, to explain why it is
invalid. The theoretically astonishing thing is that in the same context it is
Galileo himself who provides us with the means for such an explanation. It is
as if he had deliberately perpetrated the error in order to be able to make a
point in the theory of reasoning, which is to say not that he is using rhetori¬
cally improper procedures, but rather that this passage has an essentially
logical component.
Why, then, is the step erroneous? It is not clear how, for example, one
would use the method of counter-example; one would have to find the form
of both propositions, and then produce a pair such that the one having the
form of G22141 is obviously true, and the one having the form of G2214
obviously false. This would involve reformulating the argument, and the
faithfulness of the reformulation would present a very difficult, if not insur¬
mountable problem.
Another approach would be to say that G22141 as it stands is indisputably
true, whereas G2214 is false, this falsehood being uncontroversial nowadays,
but also such that it could or should have been admitted by Galileo himself.
After all, in stating the argument he asserts as a truth which he accepts, that
“heavy bodies, whirled with speed around a fixed center, can acquire impetus
to move away from that center, even when they are such as to have a pro-
410 CHAPTER 16

pension to go there naturally” (F216.19—216.21). Such an explanation of

this argument’s invalidity would be using the essential idea of the counter¬
example method. However, in bypassing the troublesome notion of form, it
would fail to provide understanding of what error is being committed.
If we turn to Galileo’s own account, we find that the argument is followed
by a discussion of the applicability of mathematics to physical reality (F229—
237), in which Salviati makes many subtle and crucial methodological
points.16 One of these is that the real problem is to use a correct mathematical
model for the physical situation, i.e., to approximate the physical processes
by the right type of mathematical entity (F234). Applying this discussion to
the present problematic inference, we may begin by saying that proposition
G22141 is a mathematical truth, but that G2214 is a physical proposition; to
say this is perhaps not to make a claim about the ‘logical form’ of these
propositions, but it is to say something about their nature or character. It
would follow that only to the extent that the physical entities and processes
mentioned in proposition G2214 are correctly approximated by the abstract
entities of the other proposition, is the inference correct. Hence it is not,
since the various geometrical segments mentioned in the premise correspond
to physical displacements, not to the speeds associated with these displace¬
ments, which speeds are the physical entities mentioned in the conclusion.
(4) The manner in which Galileo begins to state his G3 criticism reminds
one of the following equivocation charge: (FG414) The step from A121 to
A12 presupposes two propositions: A122, that the cause of extrusion
increases as the rotational speed does, and, A123, that the speed of terrestrial
rotation would be very great. Now, (FG4132) it is true that the cause of
extrusion increases with the rotational speed when the radius is constant,
namely that it increases as the number of rotations per unit time; but
(FG4131) it is false that the cause of extrusion increases with the rotational
speed when the number of rotations in equal times is constant, namely that
it increases as the radius; thus, (FG413) ‘rotational speed’ may refer either to
angular speed or to linear speed. Now, (FG412) if angular speed is meant then
A122 would be true but A123 false, since (FG4121) the earth makes only
one rotation in twenty four hours. Whereas, (FG411) if linear speed is meant
then A123 would be true and A122 false. Therefore, (FG41) not all premises
of the subargument to A12 can be simultaneously true. Therefore, (FG4)
A12 is groundless.
One problem with this argument is that though FG414, FG4132, and
FG4131 are in the text (F237.29—238.15), the other propositions are not.
Moreover, the step to FG413 is questionable since it interprets the denial of
 GALILEO AS A LOGICIAN 411

an increase with the radius mentioned in FG4131 as a denial of an increase

with the linear speed, whereas it really amounts to a denial of an increase
purely with linear speed. Third, proposition FG4131 is not in fact true.
Hence it would be uncharitable to attribute this argument to the text. Never¬
theless it does make (in the abstract) an interesting equivocation.
(5) Criticism G3 amounts partly to a critique of quantitative invalidity for
the object argument, but it also suggests a charge of incompleteness, insofar
as proposition G3111 shows that it is wrong to consider speed as the only
variable. Galileo charges incompleteness almost in so many words (F237.37—
238.2); moreover, he has a long discussion (F238.13-242.il) designed to
show that rotation is not the only phenomenon where speed is not the only
effective variable: for example, the resistance to being moved and to being
stopped increases as the speed and the weight (F241.1—241.11).
(6) I believe that the subargument to G311123 avoids the problem of mis¬
applying a mathematical model besetting the subargument to G2214. Whereas
in the earlier argument the model falsely represents a static situation, in the
present case the model is properly dynamic.

NOTES

1 As before, this is an abbreviated reference to G. Galilei, edited by Favaro, Opere 7,

43-57. All unqualified references in this chapter refer to this book.
2 As in earlier chapters, numerals after the period refer to lines.
3 In such contexts, ‘argument x’ means ‘the argument supporting x’.
4 The intermediate steps in this inference are:
(13.1) [if not-Hc and X, then not-HA] and [if X then HA];
(13.2) if [(not-Hc and X) and X] then (not-HA and HA).
5 Cf. the discussions at the end of Chapter 12 and in Sections 8 and 12 of this chapter.
See also the remarks in Note 8 of Chapter 12.
6 B. Mates, Elementary Logic, p. 81; I. M. Copi, Introduction to Logic, p. 252; W. C.
Salmon, Logic, pp. 49-50; and D. Kalish and R. Montague, Logic, p. 92.
7 My claim here is one of comparative accuracy, for there are some contexts where the
conventional translations would be acceptable, and many others where neither would be
adequate.
8 See M. Scriven, Reasoning, pp. 71-77.
9 Cf. the discussion in Chapter 14, Section on ‘Interdependence of Reasons’
10 If this is equated with formal implication (in the conventional model-theoretical
sense), then we would need, as an underlying contextual assumption, to exclude from
our object language those nonmolecular sentences that are inconsistent, otherwise we
would have a counterexample to (F3.1) when R is inconsistent. Another way out of the
problem might be to give a probabilistic analysis of these conditionals in accordance with
E. W. Adams’s Logic of Conditionals. A further possibility would be to explore the
relevance of the various ‘relevance-entailment’ systems found in A. R. Anderson and
412 CHAPTER 16

N. D. Belnap, Jr.’s Entailment. Cf. our discussions above at the end of Chapter 12, and in
Sections 4 and 12 of this chapter. See also our remarks in Note 8 of Chapter 12.
11 I realized this while comparing the two deductions below, the first one of which is in
the object language and correct, and the second one of which is in the metalanguage and
incorrect. The notation used is from B. Mates, Elementary Logic.

I
{1} (1) P—K2 P
A (2) (-P&P)—-> -Q TH
{3} (3) -[(P—>Q)&(P—»-Q)] P
{3} (4) (P->Q)->-(P-*-0) 3, R
{1,3} (5) -(P-►-Q) 1,4, MP
A (6) -p—>(P-► -Q) 2, TH
{1,3} (7) P 5,6, TH
{3} (8) (P->Q)->? 7, C

II
(1) Assume: a |= 0
(2) Not-a and 0 (= not-0
(3) Assume: not both O |= 0 and a |= not-0
(4) Not-(CT (= not-0); from (1) and (3)
(5) Not-a 1= (a |= not-0); (?) from (2) (?)
(6) a ; from (4) and (5).

In II, step (5), corresponding to step (6) of I, is wrong since (5) means: “not-a |= ‘a->
not-</>’ is valid”, which is not true, a counterexample being a = P, (p = Q.
12 Cf. G. Galilei, Dialogue, ed. Salusbury-Santillana, p. 193, note 64.
13 Galileo claims to give maximum advantage to his opponents by assuming the location
to be at the equator. Even Drake (Dialogue, p. 477, footnote to p. 181) claims that there
is really no such advantage, but it seems to me that there is; for, whereas at the equator
the horizon can be imagined to rise and fall 360 degrees in 24 hours, at the poles there
would be merely an horizontal turning with no rising and falling, and at intermediate
latitudes we would have a mixture of horizontal turning and rising and falling, the latter
being less than at the equator. Drake also does not seem to be entirely right in claiming
that “the demonstration is quantitatively worthless” (ibid.), for 1 show below that
Galileo’s figures seem to check.
14 See the section on Favaro in Chapter 10 above.
15 Aristotle, On Sophistical Refutations, 167a21; cf. Hamblin, Fallacies, pp. 31-32.
16 See Chapter 5 above.
 CHAPTER 17

CRITICISM, COMPLEXITY, AND INVALIDITIES:

THEORETICAL CONSIDERATIONS

SOME DATA ANALYSIS

The results of the previous chapter may be regarded as a data basis in need of
further analysis. Let us begin by making an inventory of the logical topics
found in Galileo’s book, as seen from the section titles of the last chapter.
The following table will serve our purpose.

PAGES ARGUMENT LOGICAL TOPIC

33-38.24 Three-dimensionality Conclusion vs reasons vs causes
38.25-57.4 Natural motion Logical evaluation
57.5-62.27 Universe center Equivocation, composition, and circularity
62.28-71.33 Contrariety Contradiction, ad hominem, and
equivocation
71.34-83.li A posteriori Counter-example method
83.12-87.14 Teleological Charity
159.29-162.14 Violent motion Equivocation
162.15-164.25 Two motions Implication vs support
166.30-169.21 Mixed motion Groundlessness
169.22-180.18 Ship experiment Analogy
194.12-197.1 East-west gunshots Deduction, induction, and rhetoric
197.1-202.16 Vertical gunshot Inferential vs material ‘if
203.34-205.29 Hunter’s Explanation vs inference
205.30-209.13 Point-blank Scientific, ad hominem, and logical
invalidity
209.13-212.30 Birds Logic of punctuation
214-244 Centrifugal force Ignoratio elenchi, quantitative invalidity,
and incompleteness

The first impression is that the logical topics are as varied as the physical
ones.
Next, it is interesting to count the number of simple subarguments of the
main arguments. The first number we arrive at is 32, there being 16 sections
each containing one object argument and one meta-argument. Denoting by
n-A the object argument of the nth section, and by n-G the meta-argument of
the same section, these 32 arguments would consist of the following number
of simple subarguments:

413
414 CHAPTER 17

1 -A: 3 9-A: 5
1- G: 3 9-G: 5+1+2=8
2- A: 3 10-A: 3
2- G: 6+3+1+3+4+6=23 10- G: 16
3- A: 4 11 -A: 3
3- G: 5+2+1+10=18 11- G: 7+5+1=13
4- A: 3 12- A: 4
4- G: 1+2+2+4+5=14 12- G: 2+5+4=11
5- A: 1 13- A: 2
5- G: 3+1+1+5+2=12 13- G: 3
6- A: 6 14- A: 2
6- G: 3 14- G: 2+3+3+4=12
7- A: 4 15- A: 2
7- G: 5 15- G: l+3+5=9
8- A: 3 16- A: 2
8-G: 7+6+3=16 16-G: 3+4+6=13.

Thus we have at least 229 self-contained arguments whose features may be

made the topic of further investigations.
Next, by examining the content of the various sections of our data basis,
we find that many more individual logical topics are discussed or mentioned
than is apparent from the section titles. The following index shows this,
where the numbers without hyphen or preceding an hyphen refer to the
section numbers of the last chapter, the ‘G’ numbers refer to the various parts
of the meta-arguments, and the ‘C’ numbers in parenthesis refer to my
comments:

INDEX OF INDIVIDUAL LOGICAL TOPICS

Active evaluation of reasoning, 2, 3-G1, Causes vs reasons vs conclusions, 1

3-G2,4-G1, 5-G2(C2) Charity, principle of, 6
Ad hominem argument, 4-G4(C6) Circularity, 3-G4, 8-G2, 11-G3, 12-G1
Ad hominem invalidity, 14 Composition, fallacy of, 3-G3
Analogical reasoning, 10 Conclusions vs reasons vs causes, 1
Analogy, 10 Converse of a given proposition, 10
Argument, ad hominem, 4-G4(C6) Counter-example method, 3-G4,
Argument, explanatory, 5-G1 (Cl), 8-G2, 5-G2(C2), 5-G3(C3)
9-(C), 11-G2 Deduction, 11
Argument, self-contradictory, 4-G3, Deduction, induction, and rhetoric, 11
8-G2 Deductive invalidity, 11-G1
Begging the question, 3-G4, 8-G2, 11-G3, Equivocation, 3-G1, 4-G5,4-G5(C7),
12-G1 7, 11-(C), 16-G3(C4)
 CRITICISM, COMPLEXITY, AND INVALIDITIES 415

INDEX OF INDIVIDUAL LOGICAL TOPICS

Errors in reasoning, explanation vs Invahdity, scientific, 14

description of, 16-(C3) Irrelevant conclusion, 16-G1
Evaluative nature of reasoning, 5-G4(C4) Liar’s paradox, 4-G3(C4)
Explanation vs inference, 13 Logical evaluation, 2
Explanatory argument, 5-Gl(Cl), 8-G2, Logical invahdity, 14
9-(C), 11-G2 Method of alternative conclusion, 2, 3-G1,
Fallacy of composition, 3-G3 3-G2, 4-G1, 5-G2(C2)
Groundlessness, 7, 8-G2, 9, 10, 11-G3, Method of counterexample, 3-G4,
12-G2 5-G2(C2), 5-G3(C3)
‘If-then’, 8-G2, 12 Nonterminating argument, 4-G3(C5)
Tf-therefore’, 4-(C3) Paradox, Ear’s, 4-G3(C4)
Ignoratio elenchi, 16-G1 Punctuation, logic of, 15
Implication vs support, 8 Quantitative invalidity, 16-G2
Incompleteness, 16-G3(C5) Reasoning and evaluation, 5-G4(C4)
Induction, 11 Reasons vs causes vs conclusions, 1
Induction, deduction, and rhetoric, 11 Rhetoric, 11
Inductive invalidity, 11-G2 Rhetorical invahdity, 4-(C7), 11-G3
Inference vs explanation, 13 Rhetorical strengthening of arguments,
Infinitely progressing argument, 4-G3(C5) 5-G3(C3)
Invalidity, 1, 2, 3-G2, 4-G1, 5-G1, 5-G2, Rhetoric, induction, and deduction, 11
6-G1,7, 8-G2, 11-G1 Scientific invahdity, 14
Invalidity, ad hominem, 14 Self-contradictory argument, 4-G3, 8-G2
Invalidity, deductive, 11-G1 Support, 9-G1, 9-G2, 9-G3
Invalidity, inductive, 11-G2 Support vs imphcation, 8
Invalidity, logical, 14 ‘Unless’, 4-(C8)
Invalidity, quantitative, 16-G2 Uselessness, 4-G2(C2)
Invahdity, rhetorical, 4-C7, 11-G3 Truth-functions, 12

The next step is to formulate a number of main topics by which to group

the individual ones. The following index is a possible elaboration:

INDEX OF MAIN LOGICAL TOPICS

Active evaluation, 2, 3-G1, 3-G2, 4-G1, 5-G2(C2)

Ad hominem: ad hominem argument, 4-G4(C6); ad hominem invahdity, 14
Begging the question and circularity, 3-G4, 8-G2, 11-G3, 12-G1
Counter-example, method of, 3-G4, 5-G2(C2), 5-G3(C3)
Equivocation, 3-G1, 4-G5,4-G5(C7), 7, 11-(C), 16-G3(C4)
Evaluation, methodology of:- logical evaluation, 2; evaluative nature of reasoning,
5-G4(C4); principle of charity, 6; explanation of errors in reasoning, 16-(C3)
Evaluation, principles of: fallacy of composition, 3-G3; uselessness, 4-G2(C2); infinitely
progressing argument, 4-G3(C5); irrelevant conclusion, 16-G1; incompleteness,
16-G3(C5)
416 CHAPTER 17

INDEX OF MAIN LOGICAL TOPICS

Groundlessness, 7, 8-G2, 9, 10, 11-G3, 12-G2

Induction, 11; reasons vs causes vs conclusions, 1; explanatory argument, 5-Gl(Cl),
8-G2, 9-(C), 11-G2; implication vs support, 8; support, 8-G1, 9-G2, 9-G3; analogy,
10; inductive invalidity, 11-G2; inference vs explanation, 13
Invalidity, 1, 2, 3-G2, 4-G1, 5-G1, 5-G2, 6-G1, 7, 8-G2, 11-G1, 14; rhetorical, 4-(C7),
11-G3; deductive, 11-G1; inductive, 11-G2; logical, 14; scientific, 14; quantitative,
16-G2
Linguistic expression: ‘if therefore’, 4-(C3); ‘unless’, 4-(C8); if then , 8-G2, 12, converse
of a given proposition, 10; logic of punctuation, 15
Method of alternative conclusion. See Active evaluation
Rhetoric: rhetorical invalidity, 4-(C7), 11-G3; rhetorical strengthening of arguments
5-G3(C3); induction, deduction, and rhetoric, 11
Self-contradiction, 4-G3, 8-G2

The above topics can, in turn, be reduced to three, as follows:

Evaluation categories: Evaluation methods:

circularity active evaluation
equivocation ad hominem argument
fallacy of composition explanation of error in reasoning
groundlessness method of alternative conclusion
incompleteness method of counter-example
infinite progression principle of charity
invalidity Elements of reasoning:
irrelevant conclusion deduction
question-begging evaluation
self-contradiction explanation
uselessness induction
linguistic expression
persuasion.

The difference between categories and methods of evaluation is that the

former are various types of attributes with which one may evaluatively
characterize reasoning, whereas the latter are techniques that the logical
theorist may use for arriving at and justifying the attribution of the various
categories. Hence categories and methods are also related insofar as the use of
a given method yields an evaluation in terms of some corresponding category,
and conversely, the evaluation in terms of some given category requires the
use of some correspondingly appropriate method. It also follows that besides
examining the internal relationships among the various concepts in each list,
we need to examine the external relationships among concepts from different
lists.
 CRITICISM, COMPLEXITY, AND INVALIDITIES 417

For the moment it suffices to note that the list of evaluation categories
is merely an expansion of the notion of principles of evaluation used at the
previous stage of theorizing, in the Index of Main Logical Topics, for the
purpose of including therein a number of miscellaneous terms. Similarly the
list of evaluation methods is an expansion of the previous topic of method¬
ology of evaluation.
Note also,that it now seemed better to list question-begging and circularity
as separate topics. Furthermore, the notion of an ad hominem argument has
been treated as a method of evaluation rather than a fault of reasoning since,
as clarified earlier, what is meant here is not the ad hominem fallacy of
contemporary logic textbooks, but rather the technique of criticizing an
opponent by deriving from his own assumptions conclusions which are un¬
acceptable to him. Similarly, though the method of explanation of errors in
reasoning may remind one of the so-called genetic fallacy, an examination of
the context in which this term was conceived [16-(C3)] indicates that what
we have here is a method for criticizing an argument by giving an analysis of
it which allows us to understand what error is being committed.
It is immediately apparent that all evaluation categories listed above are
negative or unfavorable ones. In part this is the result of our previous choice
of a data basis, which was constructed from Galileo’s critiques of Aristotelian
arguments. This represents no limitation, however, since all terms in the list
of evaluation categories are merely one side of a coin the other side of which
is a positive term; in other words, there is no conceptual difference between
the question of whether or not an argument is, for example, invalid and the
question of whether or not it is valid. By negating each term in the list we get
the corresponding positive term; in reality, it is such a pair of contradictory
terms that defines each distinct category.
The elements of reasoning are constitutive aspects of it in the sense that
reasoning is normally a mixture of the various activities listed. How exactly
they interrelate will be examined below, but it should be mentioned now that
there may be some overlap (e.g. between induction and explanation), and
there may be related activities that have not been but should be put in the list
(e.g., justification, probably related to persuasion). The term linguistic
expression refers to the use of such words as if, therefore, because, which use
may be regarded as a minimal (perhaps ‘operational’) definition of reasoning.
Evaluation was put in the list, not only because our evidence is clear that the
evaluation of reasoning is itself a form of reasoning, but also to explore the
more interesting possibility that reasoning is a form of evaluation. This would
be so if and to the extent that all other listed elements could be reduced to it.
418 CHAPTER 17

Finally, this list of elements has the following prima facie relationships to
the categories and to the methods. There may be a correspondence between
evaluation methods and elements of reasoning; for example, deduction and
method of counter-example seem to go together. On the other hand, some
evaluation categories may subdivide into as many special cases as there are
elements of reasoning; for example, we have already seen that invalidity may
be of a deductive, inductive or rhetorical kind. Other evaluation categories
may instead correspond to a definite element of reasoning; for example, it
seems that the uselessness of an argument is essentially a category relating to
persuasion.

CRITICISM, COMPLEX STRUCTURE, AND OPENNESS

Before proceeding with any further analysis, a number of obvious but far-
reaching conclusions supported by our data ought to be discussed. The most
pervasive features for the classification of arguments pertain to the distinction
of simple vs complex, and of critical vs constructive. The former is meant
in the rather precise sense defined in Chapter 14, namely that a simple argu¬
ment is one whose propositional structure contains no intermediate pro¬
positions, whereas a complex argument contains at least one intermediate
proposition. The critical/constructive distinction partly corresponds to that
of meta-argument vs object argument of last chapter, and partly to concepts
and terminology used in Henry Johnstone’s theory of argumentation.1 The
pervasiveness of these two distinctions contrasts with traditional logical
theories according to which the basic distinctions are those between
deductive and inductive, truth-functional and quantificational, and predicate
and relational arguments.
A second overwhelming fact about our data is that arguments are normally
complex. Though it is true that complex arguments are in a sense a series of
simple ones, it is also true that simple arguments are elements of complex
ones. So one may expect great differences between an approach that takes
complex arguments as fundamental, and one that takes simple ones as funda¬
mental. The usual accounts follow the latter, and hence they are probably not
merely a simplification, but an over-simplification. These considerations
accord very well with Hamblin’s theory of argument.2
Our data also indicate that most arguments are critical, rather than con¬
structive. It is not clear whether this is merely the result of our selection of
data, for after all they were compiled as an attempt to follow a practical
approach to the evaluation of reasoning in general, and to its criticism in
 CRITICISM, COMPLEXITY, AND INVALIDITIES 419

particular.3 It must be remembered, however, that our data are not biased
with respect to Galileo s Dialogue, for in our earlier explanation of our prin¬
ciples of selection, it was shown that the whole book consists of critical argu¬
ments. The question remains, of course, how typical Galileo’s book is. But
this, too, has been decided, when it was argued4 that Galileo’s book probably
constitutes the best and richest collection of arguments available anywhere;
though this makes it untypical in the sense of uniquely valuable, it does not
make it untypical with respect to the generalization presently being con¬
sidered. If a qualification is needed for this generalization, it is rather the
remark that, to use Johnstone’s expression, “critical arguments are more
fundamental than constructive ones”.5 Again, the great stress on constructive
arguments found in standard logic books probably introduces a second
distorting oversimplification in the theory of reasoning.
Besides such empirical support for the primacy of complex and of critical
arguments, there are also the following fundamental reasons. The basic pur¬
pose of an argument is to make a proposition more acceptable than it would
be in the absence of the argument.6 This means that without the argument its
conclusion would be less acceptable. But how could the acceptability of a
proposition be less, when there is no supporting argument? Only if there
are present certain objections to the proposition; the argument is then an
attempt to remove such objections. We may then say that an argument is a
defense of its conclusion from actual or potential objections. No argument
would be needed if there were no need to make the conclusion more accept¬
able, i.e. if there were no previous objections. Thus the objections are prior,
and objections are nothing but critical arguments.
Next we might note, that complexity fits very well with critical character,
foj- our data show that the meta-arguments were much more complex than
the object arguments, and in the context of Galileo’s book the former are
critical and the latter constructive. It is probably the case that complexity is
a consequence of critical character, though it is not clear how this is to be
shown. But this brings us to the primacy of complexity vis-a-vis simplicity.
A simple argument may be viewed as a piece of reasoning where the
acceptability gap and the inferential gap between final premises and final
conclusion are minimal. But as this gap decreases, so does the need for argu¬
ment. Hence we may expect arguments to be given mostly when the gap is
considerable. But if the gap is considerable, several intermediate steps are
probably required to establish a connection between final premises and final
conclusions; and this is to say that the argument will be complex.
Now the following objection comes to mind. How can the logical theorist
420 CHAPTER 17

be missing anything of importance by taking short, simple gaps as funda¬

mental, in view of the fact that complex arguments owe their success to the
reduction of a bigger gap to a series of short and simple ones? If the inter¬
mediate steps are removed from a complex argument, so that only the final
reasons and final conclusions are left, then what remains is likely to be no
argument, because no connection will be visible, and hence the acceptability
of the conclusion will not be increased by the presence of the premises.
In order to answer this objection, we need to call attention to another
fundamental feature of reasoning, pervasive throughout our data, which I
shall call its openness. This refers to the fact that, under normal conditions,
much of the propositional structure of reasoning is latent (in the sense of
Chapter 14); in other words, there are always implicit propositions whose
formulation and statement would make its structure more explicit. One
might say that arguments are normally incomplete, but because of the
negative connotations of the notion of incompleteness, it is preferable to say
that arguments normally have an open structure. Or one might say that
almost all arguments are inductive, in one sense of this term, according to
which the conclusion states more than is contained in the premises; but
because of the confusions surrounding the concept of inductive argument,
and because I wish to reserve the term induction to one of the elements or
aspects of (normal) reasoning, I will here avoid this term.
Now, the understanding of reasoning requires that its structure be made
(more) explicit. But to make the structure more explicit complicates it (in
the precise sense relevant here) since, as has been noted by Gilbert Ryle and
by Johnstone in another context,7 when one argues from explicit premises,
the debate instantly moves a step back, namely to the acceptability of the
premises themselves. It follows that to understand some piece of actual
reasoning, the chances are that it needs to be reconstructed as a complex
argument; therefore, the simple steps making up the latter are theoretical
abstractions; hence if one considers them in isolation, one is missing the point
of the reasoning; but in standard logical theories, the preoccupation with
simple arguments constitutes such consideration in isolation.
Structural complexity is then partly a consequence of structural openness.
But it is easy to see that the latter derives from the critical nature of reason¬
ing. For if reasoning primarily involves the removal of actual or potential
objections, then an argument will contain explicitly only those assertions
which are needed for such a specific purpose; the rest will be left implicit, in
the sense that it can be brought out as a result of new counter-arguments, of
theoretical analysis, or of theoretical evaluation.
 CRITICISM, COMPLEXITY, AND INVALIDITIES 421

To summarize, three of the most important facts about reasoning are: (1)
critical arguments are prior to constructive ones, (2) arguments with a com¬
plex propositional structure are prior to arguments with simple structure, and
(3) arguments normally have an open structure, namely are always susceptible
to further enlargement and complication of their structure. Moreover, the
critical dimension of reasoning seems even more fundamental than its com¬
plexity and its openness.

THE PRIMACY OF NEGATIVE EVALUATION

Since the concept of a critical argument seems so important in the theory of

reasoning, let us pause for some general reflection on this and related matters.
So far, a critical argument has meant an argument which finds fault with, or
passes unfavorable judgment upon, some proposition or some other argu¬
ment; that is, an argument whose conclusion is a negative evaluation. We
suggested in an earlier chapter,8 and our empirical data basis established, that
the negative evaluation of reasoning promised to be methodologically very
fruitful since it involves both good reasoning (constituting the critical argu¬
ment itself) and bad reasoning (constituting the subject matter of the
evaluation). Furthermore, it emerged in the last section that negative evalua¬
tion has something like a constitutive function for reasoning; this is implied
by the priority of critical over constructive arguments; the implicit and
intuitively-made negative evaluations in the process of reasoning tell us where
not to go, which conclusions not to draw. It is as if negative evaluations and
prescriptions, and prohibitive rules were more accessible and effective than
positive ones both at the level of the practice of reasoning and at the level of
logical theory.
This status for negative evaluations corresponds to trends and results dis¬
cernible in other approaches to logic and other fields of scholarship and of
culture in general.
For example, in the context of art history and criticism, E. H. Gombrich
claims that whereas “I do not think that there are any wrong reasons for
liking a statue or a picture .. . there are wrong reasons for disliking a work of
art”.9 This means that there are objective standarrds for negative evaluation,
but not for positive evaluation. Hence, critical arguments presumably obey
certain principles of evaluation, whereas constructive arguments do not.
A second example comes from psychotherapy, where one of the most
interesting approaches is the so-called rational-emotive psychotherapy,
originated by Albert Ellis.10 Its central doctrine is that neurotic disturbances
422 CHAPTER 17

are the result of errors in reasoning, of which the victim is unaware and
which he is unable to overcome. The therapy consists of discussing with the
patients the problems of the latter, in order to determine the specific pattern
in accordance with which they reason about personal and emotional matters.
One of its doctrines is that though the pursuit of happiness is not system-
atizable, the avoidance of unhappiness is: “If we can’t tell you how to be
happy, can we tell you how not to be unhappy? Paradoxically, yes. Because
while human beings differ enormously in what brings them positive con¬
tentment, they are remarkably alike in what makes them miserable.”11 What
this means is that, the criticism of arguments whose conclusions trigger
negative emotions is feasible, but the construction of arguments triggering
positive emotions is not feasible.
A third manifestation of the same trend is the philosophy of science of
Popper and his followers. In the Popperian approaches, the essential feature
of a scientific theory is its falsifiability or testability rather than its provability
or confirmability;12 the essential feature of rationality lies in the critical
attitude, i.e. in being open to criticism, rather than in being right or in being
in possession of the truth;13 the most significant feature of the growth of
scientific knowledge is the occurrence of errors and the struggle for their
elimination;14 and the most important thing about a scientific theory or any
belief in general is what one does with it, rather than how one acquired it.
We need not share their reasons for such emphasis on criticism and refutation,
namely that universal generalizations possess the property of falsifiability
in a sense in which they do not possess that of provability or probable con¬
firmability; that the same holds for hypothetical explanatory theories; that
the critical attitude is self-referentially consistent in a way in which the
justificationist attitude is not; and that the stress on error redeems from
obscurity and neglect significant portions and episodes of the development of
actual science. Instead it suffices to reinterpret their main doctrine from our
present point of view. This can be done best by using one of the most recent
and clearest formulations, given by Noretta Koertge: “A principle P is held
rationally relative to the knowledge available at time t if and only if (a) P is
held open to criticism [and] (b) there are no known cogent criticisms of P at
time t.”ls If we speak of arguments, rather than principles, statements, or
propositions, then we can automatically take care of the fact that this theory
relativizes the rationality of P to the available knowledge, since an argument’s
conclusion is always (by definition) relative to its premises. Second, to speak
of arguments is a way of objectifying the psychologists reference to the
‘holding’ of a principle open to criticism. Thus the Popperian nonjustific-
 CRITICISM, COMPLEXITY, AND INVALIDITIES 423

ationist theory of rationality can be reformulated in argument-theoretical

terminology as the claim that a rational argument is one which has no known
cogent objections; in turn, a cogent objection is a (meta) argument about
some aspect of an (object) argument such that the meta-argument has no
known cogent objections; and so on.
I think it is obvious that such a definition makes sense of the considerable
evidence in our data basis, for the Aristotelian arguments are invalid
(‘irrational’) because and insofar as the Galilean critical meta-arguments are
unanswerable (‘cogent’), while the latter are correct for the same reason.
Such a definition of validity may appear circular since it defines the
validity of one argument in terms of the validity of another. But this objection
would not be exactly right, since what happens is that the validity of one
argument is defined in terms of the validity of a higher level argument; one
might say that validity! is defined in terms of validity2, which is defined in
terms of validity3, etc., where validity , is validity at the object level, validity2
is validity at the next higher meta-level, and so on. Formal circularity is thus
trivially avoided.
Moreover, and more importantly, in practice there is seldom any circularity,
for the fact is that when the discussion is transferred to the metalevel, it
becomes quite effective, and although the first meta-argument may not be
decisive, the next step would be to criticize the meta-argument. In other
words, the above negative definition of validity provides a very effective
framework for the methodology of reasoning and for the methodology of the
evaluation of reasoning. This is the sense in which it is claimed to be correct.
In such a framework, critical arguments are crucial and more basic than con¬
structive ones, since the search for the former, and the failure of the search,
give worth to constructive, justificatory arguments.
Next, the best-articulated and most-widely accepted concept of validity is
the model-theoretic one, according to which an argument is valid if and only
if there is no model which satisfies the premises but does not satisfy the con¬
clusion.16 This corresponds to the less precise ideas that a valid argument is
one for which it is impossible that its premises be true and its conclusion
false, or one such that there is no argument of the same logical form having
true premises and false conclusion. If we call counter-examples such models
and arguments mentioned in the definiens of these definitions, then we may
say that a valid argument is one for which no counter-example exists. There is
no question that this concept of validity has great explanatory power,
especially in mathematical reasoning and in mathematical logic. Moreover,
the search for counter-examples, together with proofs that none exist, is not
424 CHAPTER 17

merely a hit and miss affair, and there is great power in the method of
assuming the existence of a counter-example and either deriving a contradic¬
tion or constructing an appropriate satisfying model. However, it is easy to
overestimate both the precision and the power of the definition, for ultimately
the concept of a model is relatively open and subject to transformations, as
the development of model theory shows. Nevertheless, from the present point
of view, the model-theoretic definition of validity is very interesting since it
can be interpreted as a special case of criticism or negative evaluation of an
argument; a disproof by counter-example is an extreme, clear-cut type of
objection to the original argument, while proof of nonexistence of counter¬
example is an extreme clear-cut type of failure to find any objections.
Finally, one of the most widespread approaches to the teaching of intro¬
ductory logic is what may be called the fallacy-avoidance or critical thinking
approach. This is reflected in the titles and subtitles of a large segment of
contemporary textbooks.17 The underlying idea is to teach how to reason
validly, or how to improve one’s reasoning, by teaching how to recognize
instance of invalid or fallacious reasoning. This stress on criticism is often
abused and frequently superficial, as we saw in an earlier chapter, and its
practical justification lies in the preponderance of fallacious reasoning in the
contemporary American scene. Its theoretical justification, however, must be
in the critical-evaluative nature of reasoning itself, in the fact that reasoning is
correct when it lacks specifiable faults, and hence reasoning is implicitly the
evaluation of evidence and of connections among propositions, together with
the avoidance of fallacious conclusions.

EVALUATION AND INVALIDITIES

I have been arguing in support of the conclusion that, on the one hand, both
the understanding and the evaluation of reasoning are forms of reasoning, and
that, on the other hand, both understanding and reasoning are themselves
forms of evaluation. The ratiocinative character of understanding is supported
by the fact that the understanding of reasoning is essentially the reconstruc¬
tion of an inferential structure of propositions out of the actual linguistic
material expressing the reasoning under consideration, and the process of
reconstruction is conceived as higher level reasoning whose premises are
empirical propositions about this material and whose conclusion is the re¬
constructed argument. The ratiocinative character of evaluation is supported
both by the fact the evaluation of arguments is normally expressed as a meta¬
argument about the argument being evaluated, and by the fact that such
 CRITICISM, COMPLEXITY, AND INVALIDITIES 425

evaluation has often an active character, namely it involves reasoning at the

same level as the argument being evaluated and arriving at different conclu¬
sions from the ones drawn in it. The evaluative character of understanding is
supported by the central role played by the principle of charity, for to
attempt seriously to understand a piece of reasoning is implicitly to display
a favorable attitude toward it, and without a certain amount of such appre¬
ciation one would be unable even to detect the alleged connections whose
presence is necessary for reasoning. Finally, the evaluative character of
reasoning has been supported by the fact that so much of the reasoning in
our data basis consists of evaluations, by a relatively general argument
exhibiting the fundamental importance of the notion of a critical argument,
and by showing that this stress on criticism (negative evaluation) corresponds
to ideas accepted in a number of more or less related areas. We have not yet
shown the mutual relationship between reasoning and evaluation by a
detailed analysis of our data basis. To this we now turn.
Let us begin our more detailed analysis with evaluation categories. The
most striking fact about them is their number and variety, by contrast with
the mere (formal) invalidity and falsehood of premises of standard logical
textbooks. It may be, of course, that the multitude of evaluations that we
have found to inhere in actual reasoning and in logical practice can be
reduced to these two, though this is prima facie doubtful, for after all our list
already includes invalidity. Of course, what we have called invalidity is only
the intuitive analogue of formal invalidity and not equivalent to it. On the
other hand, it did turn out that invalidity was by far the most common
evaluation category, so it seems promising to examine it first.
Invalidity is basically the failure of one proposition to follow from others:
a simple argument or inferential step is invalid if and only if the conclusion
does not follow from the premise(s) or reason(s). There are several different
reasons for such failure. It may be that the conclusion does not follow be¬
cause a counter-example exists (which would yield the special case of formal
invalidity). Clearly, such criticism is relevant only when the propositional
structure of this inferential step is viewed as closed; since we have seen that
this structure is normally open, formal invalidity seems relatively un¬
important.
A conclusion may not follow because it does not follow any more than
some other different propositions. This occurs primarily with explanatory
arguments, where the conclusion is an explanation of the premises, and there
is no reason to prefer it to some specifiable alternative explanation. There is
active evaluation here insofar as the alternative conclusion needs to be
426 CHAPTER 17

formulated and shown capable of accounting for the premises. We may call
this type inductive invalidity, though of course one would need to show that
other types of inductive arguments (such as analogy and generalizations) can
be reduced to explanatory arguments.
It may be that the conclusion does not follow because of the falsity of
some presupposed premise (i.e., some latent proposition required to make the
structure more explicit and less open). Since it is rare, or rather unrealistic,
for such falsity to be clear, obvious, or uncontroversial, it is normally the
conclusion of some argument constructible in the context. Thus the falsity of
such a presupposed premise corresponds to the correctness of some counter¬
argument adducible in the context, and the correctness of the latter is
nothing but its contextual unobjectionableness. The construction of such
counter-arguments means that this type of evaluation is active, in the sense
defined earlier.
It may be that the conclusion does not follow because some presupposed
premise is groundless, i.e., because there is no reason in the context to assert
that presupposed proposition, i.e., because in the context no argument
supporting this proposition is correct. This corresponds to complicating the
structure of the argument under consideration by making the premise under
consideration an intermediate proposition supported by some subargument(s)
in a bigger argument; the incorrectness of these subarguments is equivalent to
the groundlessness of the premise in question. This type of criticism is much
more significant when the critic himself takes up the burden of proof, in the
sense that, instead of merely asserting that no reason is explicitly provided
within the original argument for the presupposed premise, he shows that
some reason(s) could be given in the context, but that this reason is subject to
some insuperable objection. In turn, the construction and criticism of these
subarguments involves reasoning at the same level as the argument'being
evaluated; thus we have here a type of active evaluation.
The last two cases may be labeled structural invalidity because the
invalidity derives essentially from the openness and complexity of the pro-
positional structure of reasoning.
Another reason for invalidity may be that what does follow from the
premise(s) is some specifiable proposition different from or inconsistent with
the conclusion. Two special cases fall under this heading. One occurs when
the counter-conclusion can be seen to follow rather immediately and directly
from the premise(s) explicitly stated in the original argument. The more
important case occurs when the counter-conclusion can be arrived at on the
basis of these explicit premises together with other propositions which are
 CRITICISM, COMPLEXITY, AND INVALIDITIES 427

independently justifiable or contextually acceptable; here we have more active

involvement in the sense of more elaborate reasoning at the object level. We
may label these cases of constructive invalidity since the invalidity is the
result of the contextual construction of counter-arguments indicating which
counter-conclusion does follow.
How does invalidity relate to the other evaluation categories? The ground¬
lessness of an argument reduces to groundlessness of its final reasons, and
the latter corresponds to the invalidity of contextual arguments supporting
them. An incomplete argument is essentially one which is constructively
invalid. The fallacy of irrelevant conclusion is also a case of constructive
invalidity. The fallacy of composition is a special case of formal invalidity.
Uselessness refers to the fact that the given argument is in the context part of
a bigger one formed by using the final conclusion of the given argument as
an intermediate proposition to arrive at some other conclusion; the original
argument is useless when the subargument added to it has some type of
invalidity; that is, the uselessness of an argument corresponds to the invalidity
of the constructed additional subargument grafted onto it to define a fuller
contextual argument.
Equivocation is really a type of structural invalidity, but deserves a special
name because it is especially important. Its importance can be seen from the
frequency of this evaluation category in our data basis. Its structural basis
can be seen as follows. The simplest case of a complex argument will have
one final conclusion C, one intermediate proposition I, and one final reason
R:

I
I
R

Equivocation occurs when it cannot be both that I follows from R and C

follows from I, since if I follows it acquires a meaning such that C is seen
rather obviously not to follow, whereas if C follows from I then I must have
a meaning such that it rather obviously does not follow from R. We may say
also that equivocation derives from the semantical ambiguity of some term(s)
t in I, where the ambiguity is such that if t has one meaning then one of the
inferential connections holds but not the other, whereas if t has the second
meaning then the other inferential connection holds but not the first.
428 CHAPTER 17

Thus we might regard equivocation as semantical invalidity. This term has

the advantage of stressing the connection between equivocation and invalidity,
and the special significance of the structural problem generating the invalidity.
However, we must not forget that the semantical ambiguity of I does not
derive from the abstract possession of a double meaning by t, but rather from
the contextual double meaning of t, one that connects it to C and the other
with R.
The primacy of complex structure is beautifully illustrated by equivocation
that occurs with simple arguments. Suppose we have two premises PI and P2,
and one conclusion C:

Here equivocation can occur when some term(s) t of one of the premises,
say PI, has two meanings such that, if given one meaning, C does follow from
PI and P2 but PI is rather obviously not true, whereas given the second
meaning, PI is true but C rather obviously does not follow. The problem then
arises, how can anyone ever fall into such equivocation which involves giving
an argument which is obviously either valid but unsound, or sound but
invalid. I believe that if this were all that was happening in equivocation, it
would seldom occur. However, the fallacy becomes more comprehensible
when we realize that the above simple argument would normally be an
injudicious abstraction from a complex one, where PI would be an inter¬
mediate proposition supported by some reason, say Pll. Then it is easy to
see how in the context of the subargument ‘Pll,PI’ one will attach to PI
an appropriate meaning so that it does follow from Pll, whereas in the con¬
text of combining PI with P2, if there is a crucial term with a double
meaning, that the combination (which yields C) will be possible only if one
uses the second meaning of the term, different from the previous one.
The other evaluation categories in our list were not cases of invalidity.
An infinitely progressing argument (e.g., some versions of the liar’s paradox)
is one which has no nonarbitrary final conclusion; the explicit conclusion in
the actual argument does indeed follow, but from it there follows its denial
as well, and so on without end. One might say that we have a degenerate case
of invalidity where the conclusion does not follow because there is no final
conclusion.
A self-contradictory argument is one with inconsistent premises. Such
 CRITICISM, COMPLEXITY, AND INVALIDITIES 429

arguments automatically excape formal invalidity, though they may not

escape the other types. However, since not all premises can be true simulta¬
neously, it will never happen that they must (all) be accepted, and hence it
will never happen that the conclusion must be accepted because of the pre¬
mises. So we might say that the conclusion does not follow from the premises
because there are no (practically realizable) premises. This would yield
another degenerate case of invalidity, the source of the degeneracy (‘no
premises’) being opposite to that which generates infinite progression (‘no
conclusion’).
Question-begging is a case of circularity in the sense that a question¬
begging argument has some fmal premise(s) whose justification requires in
the context that one assume what is being proved (i.e., the fmal conclusion
of the question-begging argument), and this means that the bigger argument
consisting of the original one plus the justification of the crucial premise(s) is
such that the same proposition appears both as final conclusion and as a fmal
reason. The latter is precisely a circular argument.
Circularity, too, may be regarded as a degenerate case of invalidity, for
we may say that the conclusion does not follow from the premise because it
is one of them. This would yield a type of degeneracy intermediate between
the previous two.
We thus have six types of invalidity, each deriving from the reasons why
the conclusion may not follow from the premise(s): formal, inductive,
structural, constructive, semantical, and degenerative.
Let us now examine the elements of reasoning listed above as one of the
main topics extracted from our data basis. Deduction may be viewed as the
process of avoiding formal invalidity, and induction as the process of avoiding
inductive invalidity.
Explanation may be equated with induction in two senses. First many
inductive arguments may be viewed as explanatory ones; for example,
generalizations may be conceived as inferences to the best explanation,18
whereas arguments from analogy may be conceived as generalizations about
the properties shared by two similar entities, the range of the generalization
being restricted to a special class of relevant properties.19 Moreover, inductive
arguments are often, and perhaps normally, reports (from the context of
discovery, rather than justification) of how one arrived at the conclusion in
the first place.20 This means that they are (causal) explanations of one’s own
thoughts.
The linguistic expression of reasoning, by means of reasoning indicator
words, or by the more elaborate reconstructions and structure diagrams
430 CHAPTER 17

found in our data basis, probably corresponds to the avoidance of structural

invalidity. This is to be taken in the sense of reducing the amount of structural
invalidity and disciplining the attempts to avoid it, for instances of structural
invalidity are always potentially present, so that its elimination is an in¬
definitely long process andean be achieved only temporarily and contextually.
Persuasion corresponds to constructive invalidity in the sense that the
explosure of constructive invalidity is the attempt to persuade (oneself or
others) that even if we accept the premises in question, we need not accept
the conclusion. The articulation of constructive invalidity is a form of per¬
suasion in the sense that the criticism begins by accepting the same premises
and attempts to show that something different follows.
Persuasion is also partly the attempt to avoid degenerative invalidity since
the problem with the latter is that the reasoning being criticized can never
lead from a state of nonacceptance to a state of acceptance of the conclusion.
Regarding equivocation or semantical invalidity, it has been pointed out
by Hamblin21 that equivocation typically occurs when there is a failure of
persuasion in spite of any discernible invalidity. That is, the premises seem
acceptable, the inferential connections seem also acceptable, but the conclu¬
sion is not. The way to solve the problem is to detect an ambiguous term, and
a consequent equivocation.
Finally, evaluation corresponds to the avoidance of no one particular type
of invalidity, but to all of them collectively. For the process of avoiding these
invalidities is implicitly a form of evaluation, and so in that sense deduction,
induction, persuasion, and expression are forms of evaluation.
The evaluation methods listed earlier may be analyzed as follows. The
method of counter-example is obviously the method for showing formal
invalidity; the method of alternative conclusion is the one that shows
inductive invalidity. Active evaluation and ad hominem argument are essen¬
tially the same, and they are the method for proving constructive invalidity.
The principle of charity is not exactly a method, but it does relate to
structural invalidity, for the exposure of the latter is impossible without a
good deal of structural reconstruction work, which requires application of the
principle. And conversely, if the argument under criticism has first been
reconstructed in a charitable way, then the argued falsity or groundlessness of
one of its propositions acquires force.
The methods for showing semantical and degenerative invalidity (i.e.,
equivocation and circularity) do not appear in our list above because they are
equivalent to the method of analyzing reasoning into its propositional struc¬
ture. Nothing special is involved here other than the principles for under-
 CRITICISM, COMPLEXITY, AND INVALIDITIES 431

standing propositional structure, and an understanding of the concepts of

equivocation and degenerative invalidity, especially its most important case
of circularity.
In summary, our main result seems to be that reasoning is a form of
evaluation in which propositions (ideas, thoughts) are arranged in such a way
as to avoid invalidities of the kinds discussed above.

NOTES

1 Philosophy and Argument, pp. 57-92.

2 Fallacies, p. 229.
3 See Chapter 15.
4 In Chapters 11 to 13.
5 Philosophy and Argument, p. 82.
6 Cf. Hamblin’s Fallacies, p. 245.
7 H. W. Johnstone, Jr.,Philosophy and Argument, p. 58; G. Ryle, ‘Proofs in Philosophy’,
p. 321.
8 See above, chapter on ‘The Evaluation of Reasoning’.
9 The Story of Art, p. 5; italics in the original.
10 See his Reason and Emotion in Psychotherapy, A Guide to Rational Living, and
Growth Through Reason.
11 A. Ellis, Guide to Rational Living, pp. 69-70.
12 Cf. K. Popper, The Logic of Scientific Discovery, and Objective Knowledge.
13 K. Popper, The Open Society and Its Enemies, Vol. 2, pp. 224-58; W. W. Bartley,
III, The Retreat to Commitment, and ‘Rationality versus the Theory of Rationality’.
14 J. Agassi, Towards an Historiography of Science.
15 ‘Bartley’s Theory of Rationality’, p. 79.
16 See, for example, A. Taski, Introduction to Logic, and B. Mates, Elementary Logic.
17 See, for example, M. Black, Critical Thinking; D. B. Annis, Techniques of Critical
Reasoning', W. E. Moore, Creative and Critical Thinking', A. Michalos, Improving Your
Reasoning', M. C. Beardsley, Thinking Straight', S. M. Engel, With Good Reason: An
Introduction to Informal Fallacies', W. W. Fearnside and W. B. Holther, Fallacy, The
Counterfeit of Argument', and N. Capaldi, The Art of Deception.
18 R. H. Ennis, ‘Enumerative Induction and Best Explanation’.
19 See M. Scriven, Reasoning, pp. 210-15.
20 The essence of this suggestion may be found in Hamblin, Fallacies, pp.. 224-52.
21 Fallacies, pp. 283—303.
 CONCLUDING REMARKS

TOWARD A GALILEAN THEORY OF RATIONALITY

The preceding investigations have an internal structure and unifying topic.

That is, they constitute a concrete theory of rationality, concrete because
the case study of Galileo is considered; theoretical because this Galilean
material is studied not for its own sake, or with a tabula rasa, but with an
eye toward the theoretical lessons one can learn; and rationality-oriented
because these theoretical conclusions and suggestions deal with the nature of
human rationality. I speak here of rationality simpliciter, and not of scientific
rationality, because the latter has been treated not as something sui generis,
but merely as a special case of rationality in general. In this regard, Chapters
7—17 show in what sense and exactly how there is nothing above and beyond
ordinary reasoning in Galileo’s Dialogue.
From this point of view, the Introduction argues that Galileo’s Dialogue
has a special place for anyone interested in being rational and in reflecting
and learning about rationality. Chapters 8 and 9 may be regarded as a
formulation of the problem of rationality by reference to Galileo’s book, as
well as a discussion of the problem in two selected but especially important
fields, the philosophy and the history of science. Chapters 1—7 deal con¬
cretely with those features of rationality that might be labeled macroscopic,
and Chapters 10—17 with those features that might be labeled microscopic.
The macroscopic features are primarily three, namely rhetoric, method, and
reasoning. Rhetoric refers separately to communication with others (Chapter
1), to the structuring of one’s throughts (Chapter 2), and to the discursive
handling of emotions (Chapter 3). Method pertains collectively to effective
means for the establishment of aims and achievement of results (Chapter 4);
to self-reflection on one’s practice (Chapter 5); and to the judicious balancing
of potentially conflicting activities, aims, and requirements (Chapter 6).
Reasoning refers to a special activity of crucial importance (Chapter 7).
The microscopic features of rationality may be referred to as its micro¬
structure. Thus the structure of reasoning studied in Chapters 10—17 is also
the microstructure of rationality, since reasoning is one of the elements of
rationality. Insofar as reasoning is the central element to which all others can
be reduced (Chapter 7), then its structure is all there is to the microstructure
of rationality. Elements of the (macro) structure of reasoning are deduction,

432
 CONCLUDING REMARKS 433

explanation, persuasion, expression, and evaluation (Chapter 17). Of course,

it would be possible to continue and refine the analysis, and deal with the
finer features of reasoning, which would be simultaneously the microstructure
of reasoning and the micro-microstructure of rationality. This has not been
done in my investigations, except implicitly and incidentally in the course of
the analysis in Chapter 16. The reason was that I thought one needed to
understand the bigger features of reasoning before going on to the finer ones.
What is the general relationship between these two levels, macro and
micro, of the structure of rationality? I believe that the facts of the micro-
structure of rationality (i.e., the macrostructure of reasoning) are less
immune, if not completely so, to historical change and development and to
social conditions and situations, than the facts about rhetoric and method.
That is, rhetorical techniques and methods may change without any change in
the ratiocinative techniques underlying them. Relativism would thus be false
at the level of reasoning, but possible at the level of rhetoric and method. It
might be useful to make this definitionally true by defining ratiocinative
techniques to be those that remain and have remained constant in the midst
of changes in rhetoric and methodology.
Similarly with progress. My study provides no evidence that at this level
(reasoning) Galileo’s rationality was in any way inferior to present-day
rationality. However, the fact that he had to use the rhetorical techniques
that he did, and to emphasize the methods he did, is primarily a reflection of
his historical and social conditions. Our rhetorical and methodological proce¬
dures can be taken to have grown since Galileo’s time and perhaps to excel
his, at least in sophistication and complexity. But no similar growth is dis¬
cernible in reasoning.
Though theoretically reducible to reasoning, rhetoric and methodology
are not dispensible in practice, however. Hence a growth of rationality is
definable, deriving from the progress in methodology and rhetoric. I would
also be willing to admit, as suggested above, that there is a deeper level of
rationality than that of reasoning considered here. For example, deduction,
besides being one of the macroscopic elements of reasoning, might be the
basis of all others. Hence this distinction between the macro- and the micro¬
structure of rationality should not be taken as absolute, but rather as a matter
of depth of analysis, with the understanding that such depth is a matter of
degree. However, just as the rhetorical and methodological levels of rationality
are not dispensable for human beings, in favor of mere reasoning, neither is
the logical level here explored, i.e. the one relating to reasoning, dispensable
in favor of mere deduction. And this is perhaps one of the most important
434 CONCLUDING REMARKS

conclusions supported by these investigations, namely that there is, both in

theory and in practice, a logic which is intermediate between mere rhetoric
and mere methodology on the one hand, and pure deduction on the other.
This is the logic most operative both in science and in daily life, and hence
the truly central feature of rationality.
Such a double-tiered theory of rationality could also properly be labeled
“Galilean” in the double sense that Galileo’s Dialogue has provided us both
with the data, study materials, and subject matter of rationality, and with
the inspiration, model, and approach to follow. That is, our investigations
follow a demonstrably Galilean approach because we have applied to our
study of rationality the “dialectical” methodology that Galileo used in his
study of natural phenomena (Chapter 6). To this demonstration we now turn.
We began with an examination of various aspects of Galileo’s Dialogue
which are interesting and important both for their own sake and for their
relevance to the problem of rationality.
In Chapter 1 were analyzed the book’s rhetoric, in the sense of the term
‘rhetoric’ which refers to the appearances and pretensions found in verbal
discourse, and the various impressions conveyed by it, as opposed to its
substance. We examined the evidence giving the impression that Galileo is
engaged in religious apologetics relating to the Church’s anti-Copernican
decree of 1616, the evidence hinted at in the book’s title giving the impression
that it is a two-sided presentation of all the available arguments for and
against Copernicanism so that the issue could be decided by the proper
authorities, and the evidence that the book is an attempted demonstrative
proof of the earth’s motion. We concluded that it is in reality a justification
of Copernicanism in the sense of an attempt to produce adherence or to
increase assent to it. The justification proceeds on three levels, rhetorical,
logical-scientific, and philosophical-methodological.
In Chapter 2 we presented the evidence that in the Dialogue Galileo gives
arguments and evidence designed to support the theory that the earth moves;
this is the logical aspect of Galileo’s justification, in the sense that logic is the
theory and practice of reasoning; it is also its ‘scientific’ aspect in the sense of
the term that refers to the objective presentation and analysis of arguments
and evidence; consequently we have here the logical-scientific part of the
book’s ‘rhetoric’, in the sense of the latter term that refers to the general
theory of argumentation. This chapter also included a discussion of the intrin¬
sically interesting and important problem of the unity and structure of the
Dialogue, and its logical structure was presented as a solution to the problem.
In Chapter 3 we discussed the rhetorical aspect of Galileo’s justification of
 CONCLUDING REMARKS 435

Copemicanism, in the sense of the term ‘rhetoric’ that refers to verbal

techniques of persuasion which operate on human feelings and emotions.
These are alogical but not illogical techniques, from the point of view of a
strict sense of logic which would restrict it to the purely intellectual aspects
of reasoning. However, these techniques are logical in the broad sense in
which logic refers to the theory and practice of any type of argumentation,
designed to induce assent, as long as a distinction is made between what’s
proper and what’s improper. The distinction is made by reference to how the
arguments operate on human feelings and emotions. For this reason the
discussion here had aesthetic overtones and implications, to the extent that
aesthetics is the study of the proper linguistic expression of feelings and
emotions. The discussion was thus also a study of the literary aspects of
Galileo’s book since it proceeded by analyzing all passages significant from a
literary point of view. This chapter also relates to one aspect of the problem
of scientific rationality discussed in Chapter 8, since it suggests that rhetoric
(in the sense of Chapter 3) and literary art have a role to play in science,
specifically to make possible radical scientific changes involving transitions
from one very fundamental theory to another.
In Chapter 4 we analyzed the book’s scientific content, scientific in the
sense of pertaining to the concerns of present-day scientists. We argued that
the book has a richer scientific content than most historians of science are
inclined to believe, that their judgments contrast somewhat with scientists’,
that this contrast is partly the result of a tension that exists between scientific
relevance and historical accuracy, and that there are general reasons for
thinking that this problem is insuperable. We then gave a concrete discussion
of this tension by analyzing various interpretations of Galileo’s account of the
tides in the Fourth Day of the book, and we concluded that the acuteness of
the problem is thereby confirmed. In an attempt to combine scientific
relevance and historical accuracy, we gave a reconstruction of the First Day;
then we argued that, in spite of the novelty of such a reconstructed content,
its accuracy compares favorably with typical interpretations, using as a con¬
crete example the various attempts to view the beginning passages of the First
Day from the point of view of the law of inertia. We concluded that the most
important scientific content that the book possesses is its methodological
content, that is to say the lessons, ideas, and suggestions that scientists can
get from it about the nature and proper methods for the acquisition of
knowledge.
Chapter 5 provided a systematic and complete reconstruction of Galileo’s
book from the point of veiw of methodology, exhibiting the interplay between
436 CONCLUDING REMARKS

scientific practice and philosophical reflection. It is also a reconstruction of

Galileo’s clarification of the epistemological and methodological concepts
needed, above and beyond the objective presentation of the evidence, for a
full justification of Copernicanism.
After these analyses of the rhetorical, logical, literary, scientific, and
methodological aspects of Galileo’s Dialogue, an attempt was made at a
systematization of Galileo’s methodology, with an eye both toward gaining
some theoretical understanding and toward making some practical use of it.
In Chapter 6 we argued that the essential feature of Galileo’s methodology
is the judicious synthesis of such opposites as experience and reflection,
observation and speculation, quantitative analysis and qualitative consider¬
ations, causal explanation and phenomenological description, and anti¬
verbalism and logical analysis. This is the sense in which Galileo’s method¬
ology may be termed ‘dialectical’. Chapter 7 argued that at a deeper level
Galileo’s dialectical methodology is logical, in the sense that the just-
mentioned judicious synthesis of opposites ultimately reduces to reasoning.
Thus reasoning may be regarded as a microstructural level of rationality,
while at a macrostructural level rationality takes the form of various con¬
trasting procedures, such as quantitative and qualitative analysis, all of which
need to be kept in mind but none of which must be allowed to gain absolute
predominance.
These theoretical lessons in methodology were then put to use in
examining a number of leading Galilean interpreters and a series of concrete
methodological problems, and in suggesting critiques and solutions from the
point of view of the theory and practice of reasoning. Chapter 8 examined
the problem of the role of reason in science and Paul Feyerabend’s view that
Galileo’s work shows that ‘anything goes’ in scientific inquiry; we argued that
Feyerabend sees a conflict between rhetoric and reason which does not exist,
and that he fails to distinguish between proper and improper rhetoric; we also
argued that the sound part of his account shows that rhetoric has a non-
negligible role to play in science, while at the same time logic has an essential
role to play in rhetoric.
Chapter 9 suggested the application of the principle that “a science which
fails to forget its founders is lost” to the study of the history of science and
to Alexandre Koyre, whose pioneering Galileo Studies has exerted a formative
influence on the discipline. Without denying the past value of Koyr^’s
work, we argued that now is no longer the time for continued reverence, but
rather for a display of Galilean independent-mindedness. And just as Galileo
was able to argue that he was being more Aristotelian than the Peripatetics,
 CONCLUDING REMARKS 437

by refuting some of Aristotle’s substantive conclusions while following the

essence of his approach, so in this chapter we refuted Koyre’s claim that
Galileo’s work shows that good physics is made a priori, by following more
rigorously Koyre’s own method of logical analysis.
Chapter 10 showed that the logical neglect of Galileo’s work has deep
scholarly and historical roots by criticizing certain aspects of the erudition of
four leading Galilean scholars, Emil Strauss, Antonio Favaro, Stillman Drake,
and Maurice Clavelin. It was suggested that my critique points in the direction
of a way to overcome the dichotomy between logic and history.
These critiques (Chapters 8, 9, and 10) were conducted primarily from
the point of view of the most basic feature of Galileo’s methodology, namely
its logical character or emphasis on reasoning. In fact, these philosophers and
historians were criticized primarily for their inattention to matters of reason¬
ing. Such neglect, though perhaps understandable in the past, is no longer
justified. Next we undertook a critique of several groups of scholars who
certainly cannot be blamed for their neglect of reasoning; they are logicians
and other theorists of reasoning, who are vitally concerned with the topic.
However, it turned out that their work could be criticized from the point of
view of the other essential feature of Galileo’s methodology, namely its
dialectical component (discussed in Chapter 6).
From this point of view, that methodology suggested the following
approach to the scientific study of reasoning: identify a number of relevant
opposites and then attempt a judicious arbitration, and identify existing
excesses in current practice and then attempt to correct them. Since the
greatest contemporary excess in the scientific study of reasoning is the
abstractness of formal logic, we generally advocated a more realistic, con¬
crete, historical, and pragmatic approach (Chapter 13), analogous to what is
found in the psychology of reasoning (Chapter 11), in the ‘new rhetoric’
(Chapter 12), in the linguistics of reasoning (Chapter 13), and in works by
Stephen Toulmin and by Michael Scriven (Chapter 13). We argued, however,
that all of these particular approaches need certain emendations in order to
be acceptable. In fact, the relevant dichotomies in this field are such
opposites as logic and psychology, logic and rhetoric, understanding and
evaluation, and theory and practice, but they have not been properly handled.
Chapter 11 made a methodological criticism of recent work in the
psychology of reasoning by arguing that experimentation and formal logic
cannot be combined in the way attempted by psychologists of reasoning,
whose combination turns out to be both insufficiently formal and improperly
empirical; it was argued that, as a way out of their own difficulties, the
438 CONCLUDING REMARKS

correct synthesis of a logical and a psychological approach would be a

historical orientation. Chapter 12 made a methodological criticism of the
exponents of the “new rhetoric” by arguing that they fail to achieve a proper
synthesis of logical and of rhetorical analysis. Chapter 13 explored the
possibility that the science of reasoning may He in the practice of reasoning,
at the metalevel of reasoning about reasoning.
Finally, in Chapters 14—17 we attempted a constructive sketch of such a
concrete, Galilean theory of reasoning. Chapter 14 formulated a number of
basic principles for the logical analysis of reasoning by using as illustrative
examples various arguments and passages from Galileo’s Dialogue. The main
result was that for the understanding of reasoning, which is a pre-condition
for its evaluation, one needs to identify its propositional structure, which
refers to the inferential interrelationships among its propositions. The con¬
cept of proposition here developed was one derivative from that of reasoning,
so that the identity of individual propositions is dependent on the role they
play in the various steps of an argument.
Chapter 15 contains a criticism of textbook accounts of fallacies, since it
is perhaps the most widely-practiced approach to the criticism of arguments.
The inadequacy of the fallacy approach led us to suggest a more practical-
oriented one, in accordance with the program of Chapters 11 — 13. The
suggestion was to examine, in the light of the principles of Chapter 14, the
critical arguments contained in Galileo’s book.
In Chapter 16 we carried out this suggestion for 16 main arguments which
constitute about the first half of the book, thus providing a data basis or
reasoning materials for further theorizing later (Chapter 17). The principles
of selection of these arguments were stated and defended, and it was also
shown how the arguments in the rest of the book do not differ significantly
from these sixteen. The chapter consists of sixteen sections, each containing
a discussion of an argument together with relevant logical concepts and
theoretical ideas. Each discussion was systematically divided into three parts,
the first being a reconstruction of an object-argument criticized by Galileo,
the second being a reconstruction of Galileo’s own critical meta-argument,
and the third being a series of comments by myself designed to clarify con¬
cepts, formulate principles, and illustrate ideas, relevant for the theory of
reasoning.
Chapter 17 conducted an analysis of the data of the previous chapter and
led to such conclusions as that the most important distinction for the classifi¬
cation of arguments is the distinction between critical and constructive argu¬
ments, that critical arguments are prior, that reasoning normally consists of
 CONCLUDING REMARKS 439

complex arguments, that criticism (negative evaluation) has an essential and

constitutive role to play in reasoning, and that several related but distinct
types of invalidities are definable.
To summarize, we may say that Galileo’s Dialogue suggests new ways in
which science is logical and logic can be scientific. The evidence from Galileo
indicates that science is primarily a method rather that a set of abstract
truths; that method involves not the fixed adherence to some formal universal
rules, but rather the judicious balancing of such opposites as speculation and
observation, quantitative analysis and qualitative considerations, causal
explanation and phenomenological description, anti-authoritarianism and
traditionalism, and practical involvement and philosophical reflection; that
in particular such judgment does not automatically exclude rhetorical per¬
suasion or aesthetically expressed emotion; but that such exercise of judg¬
ment reduces ultimately to a matter of reasoning, namely the drawing of
conclusions from premises and the formulation of reasons for claims.
Moreover, Galileo’s book constitutes a unique source material and data
basis for a reform of logic, in the sense of the study of reasoning. I have
argued that in order to become more scientific, the study of reasoning needs a
greater empirical orientation toward reasoning as it actually occurs in the
world; that such empirical emphasis cannot be satisfied by the experimental
approach of psychology, but rather by a broadly conceived historical
approach; that such a historical approach must be able to use but ultimately
to transcend high standards of scholarly erudition; that it must recognize the
distinct importance of rhetorical analysis (i.e., the theory and practice of per¬
suasion), while retaining the supremacy of logical analysis; and that it must
not shy away from attempts at concrete theoretical systematizations,
however provisional or limited their value may be.
 APPENDIX

PAGE CONCORDANCE OF DIALOGUE EDITIONS

The purpose of the table below is to facilitate cross-references among the

following editions of Galileo’s Dialogue]the original edition of 1632, the one
published as volume VII of the ‘National Edition’ of his works edited by
Favaro, the Pagnini edition, Drake’s English translation, and Santillana’s
revision of Salusbury’s 17th century English translation. Though Favaro’s
edition may be regarded as definitive, the 1632 edition is still useful since it is
now easily available in a fascimilie reprint and since Favaro’s modernization
of the original punctuation (though usually an improvement) occasionally
obscures the logical structure of the original text,1 so that consultation of the
1632 edition is often valuable for a better understanding of the more intricate
passages. The Pagnini edition, though its text is taken from Favaro, can still
be read with profit because of its excellent notes. Drake’s translation is the
most easily available one in any language and not easily excelled; nevertheless
its imperfections2 can often be remedied for the English-speaking reader by
judiciously consulting Santillana’s edition, which also contains excellent
notes. In paginating the latter I discovered that it is somewhat incomplete and
that many passages (ranging from a few lines to a few dozen lines) are simply
not translated; occasionally Santillana notes the fact, but usually he does not.3
The table has been constructed as follows. The first column contains page
numbers from the Favaro edition, and the other columns the corresponding
page numbers for the other editions; thus anyone in possession of any of
these other editions can write on their pages the Favaro page numbers, which
may be taken as standard. What I have done is to indicate the line of the page
of those other editions that corresponds to the first line on the corresponding
Favaro page. In the case of the two Italian editions this line correspondence
is exact, down to the particular word or part of a word with which the page
begins; these beginning words are found in the column after the 1632 edition
page and line numbers, and though they were taken from that edition they of
course refer also to the Pagnini edition. In the case of the two English trans¬
lations the line numbers are often approximate to plus or minus one line, on
account of the difference in sentence structure; however, on lines where a
new interlocutor intervenes in the dialogue the correspondence is exact. The
line numbers are denoted as decimals, so that the number preceding the

440
 APPENDIX 441

decimal point is a page number, and the number following it is a line number;
the line number is arrived at by counting from the top of the page, except
for those cases in the Santillana column where a ‘b’ follows the decimal point,
in which case the counting begins with the last line of text. (This extra
notation was unnecessary for the other editions since, unlike Santillana’s,
their lines are evenly spaced and so can be counted ‘mechanically’ from the
top, e.g. by means of an appropriately graduated scale.) In the partial words
after the 1632 edition numbers, left hand hyphens indicate that the first part
of the word appears on the last line of the preceding Favaro page, and right
hand hyphens indicate that the word either continues on the following line
of the 1632 edition or is too long to be printed in this table in its entirety.
The word ‘missing’ in the Santillana column indicates that the first line of the
corresponding Favaro page is part of a passage left untranslated. Other
clarifications are made, as the need arises, by means of explanatory footnotes.

F avaro Edition of 1632 Pagnini Drake Salusbury-

Santillana

33 1.1 GIORNATA 93.1 9.1 12.1

34 2.8 di lunghezza 94.25 9.37 12.23
35 3.9 SALV. 96.13 11.1 14.17
36 4.9 prodotte 98.2 12.3 15.19
37 5.12 dal 99.19 13.5 missing
38 6.11 retti 100.31 14.3 missing
39 7.15 chiama 102.11 15.5 17.10
40 8.14 -rarmi 104.1 16.5 19.1
41 9.15 -che 105.17 17.8 20.20
42 10.14 -colare 106.30 18.8 21.b5
43 11.15 potessimo 108.6 19.5 22.28
44 12.15 de 109.25 20.6 24.16
45 13.16 -nevolmente 111.8 21.6 25.30
46 14.3 per 112.26 21.39 26.19
47 15.3 -desima 114.8 22.36 28.7
48 16.8 SALV. 116.3 23.36 30.4
49 17.14 SIMP. 117.17 24.35 31.6
50 18.15 difettosa 118.28 25.31 32.2
51 19.21 dimonstrate 120.9 26.35 33.1
52 20.27 tirando 121.23 27.35 34.4
53 21.28 qual 122.36 28.35 34.30
54 22.30 velocita 124.25 29.36 36.2
55 23.2 la 127.297 31.2111 36.13
56 23.6 mobile 127.35 31.24 36.16
57 24.8 globo 129.19 32.25 38.3
58 25.9 del 131.11 33.25 40.17
 APPENDIX

ha Edition of 1632 Pagnini Drake Salusbury-

Santillana

59 26.8 non 132.33 34.24 41.b2

60 27.8 -bricaxe 134.11 35.23 43.10
61 28.8 dire 135.24 36.22 44.5
62 29.8 per 137.16 37.22 45.11
63 30.9 natura 138.33 38.20 46.bl3
64 31.9 -trari 140.11 39.19 47.bl2
65 32.11 -guire 141.26 40.20 48.23
66 33.10 -cedervi 143.2 41.18 49,b5
67 34.7 li 144.16 42.18 50.b6
68 35.7 del 145.31 43.16 52.9
69 36.7 SIMP. 147.13 44.16 53.12
70 37.8 di- 148.26 45.15 54.8
71 38.10 della 150.7 46.14 55.2
72 39.10 b 151.20 47.13 56.13
73 40.9 cum 152.34 48.14 57.15
74 41.11 altro 154.11 49.15 58.14
75 42.11 il 155.23 50.14 59.14
76 43.11 -tagora 157.2 51.14 60.10
77 44.13 perch6 158.18 52.15 61.b3
78 45.13 macchia 160.26 53.16 62,bl
79 46.14 vi 162.3 54.13 63,b7
80 47.15 vuole 164.3 55.13 65.12
81 48.12 merck 166.15 56.11 66.15
82 49.14 silenzio 167.30 57.12 67.14
83 50.14 prossime 169.6 58.15 68.13
84 51.16 -che 170.22 59.15 69.8
85 52.16 non 172.4 60.16 70.b4
86 53.17 piante 173.20 61.14 71,b3
87 54.18 enco- 175.1 62.12 72.b3
88 55.17 e- 176.16 63.10 73.blO
89 56.18 Luna 177.33 64.9 74.21
90 57.19 della 179.8 65.10 75.16
91 58.20 -l’incidenza 180.25 66.11 76.17
92 59.21 -camento 183.11 67.11 77.b2
93 60.21 acquisto 185.5 68.10 78.bl
94 61.22 suo 186.20 69.9 79.b5
95 62.22 opache 187.37 70.8 80.b2
96 63.23 -tuosa 189.15 71.10 82.18
97 64.24 persuaso 190.29 72.9 83.14
98 65.25 specchio 192.7 73.10 84.20
99 66.26 -cessitata 193.21 74.10 85.14
67.25 la riflessione 194.35 75.7 86.6
68.25 in 196.8 76.4 87.3
69.26 a 197.22 77.3 87.b3
 APPENDIX 443

Edition of 1632 Pagnini Drake Salusbury-

Santillana

103 70.23 SIMP. 199.5 77.36 88.27

104 71.23 -streras- 200.21 78.36 89.24
105 72.25 neri 202.1 79.33 90.19
106 73.26 SAGR. 203.18 80.30 91.bl4
107 74.28 e 205.12 81.28 92.23
108 75.28 larga 206.26 82.25 93.24
109 76.30 SAGR. 208.3 83.26 94.12
110 77.28 -rata 209.31 84.23 95.9
111 78.28 pub 211.11 85.23 96.10
112 79.28 e meno 212.26 86.24 97.9
113 80.30 -minazione 214.5 87.24 99.2
114 81.31 ch’io 215.18 88.23 99.35
115 82.31 arriva 216.33 89.24 100.b5
116 83.33 SIMP. 218.12 90.24 101.bl
117 84.33 esser 219.20 91.21 102.bl
118 85.32 della 220.31 92.21 104.4
119 86.31 perchb 222.9 93.22 105.1
120 87.32 nostro 223.25 94.23 105.b3
121 88.33 miglia 225.2 95.22 107.1
122 89.33 facendo 226.16 96.22 107.b2
123 90.33 piu 227.27 97.22 108.b5
124 91.34 volerne 229.18 98.23 109.b8
125 92.23 le 230.23 99.15 110.b 18
126 93.25 -samento 232.6 100.16 111.19
127 94.26 la 233.31 101.17 112.15
128 95.25 questa 235.10 102.14 113.13
129 96.26 matematiche 236.24 103.14 114.b7
130 97.27 -tate 238.12 104.14 115.b5
131 98.28 una 240.7 105.16 117.4

132 99.1 GIORNATA 255.I8 106.1 118.1

133 99.314 a 256.9 107.2 119.b5
134 100.36 -mento 257.21 108.1 120.b4
135 101.36 e 259.11 109.3 121.b5
136 102.36 SALV. 261.3 110.3 123.4
137 103.36 a 262.17 111.2 124.7
138 104.38 perchb 263.30 112.4 125.5
139 105.38 -tre 265.13 113.5 126.7
140 106.40 -sario 266.29 114.5 127.5
141 108.1 SALV. 269.12 115.3 127.b2
142 109.2 ha 271.10 116.3 128.b2
143 110.3 si 272.26 117.2 129.bl
144 111.4 SALV. 274.8 118.4 130.b5
145 112.4 -terato 275.21 119.3 131.b2
 APPENDIX

Edition of 1632 Pagnini Drake Salusbury-

S antiliana

146 113.6 di 277.9 120.2 132.b4

147 114.7 -sano 278.26 121.3 134.5
148 115.6 -sime 280.6 122.4 135.4
149 116.7 tosto 281.21 123.4 136.bl6
150 117.8 -due 282.36 124.10 137.bl6
151 118.7 Con- 284.12 125.10 139.2
152 119.6 tempo 285.27 126.8 140.11
153 120.7 -glieria 287.5 127.7 141.18
154 121.6 -clusioni 288.20 128.4 142.4
155 122.7 Copernico 290.4 129.4 143.1
156 123.8 SIMP. 291.18 130.3 144.4
157 124.6 non 292.33 131.1 145 .b 11
158 125.2 -sentazione 294.4 131.36 146.b3
159 126.3 foglie 295.20 132.35 147.b9
160 127.3 modi 296.34 133.36 148.bl3
161 128.3 SIMP. 298.11 134.37 149.b4
162 129.4 bisogna 299.26 135.38 150.b8
163 130.2 eadem 300.37 136.35 151.b9
164 131.5 apparisco- 302.15 138.4 152.b5
165 132.7 Terra 303.30 139.6 153.b4
166 133.9 SALV. 305.6 140.5 154.b7
167 134.9 6 306.19 141.3 15 5 .b 13
168 135.10 CO SI 307.35 142.4 156.bl5
169 136.11 -sasse 309.11 143.5 157.23
170 137.11 sta 311.2 144.5 158.21
171 138.13 SALV. 312.18 145.7 159.bl0
172 139.14 -posizioni 313.33 146.7 160.bl3
173 140.13 piano 315.8 147.4 161.M9
174 141.14 SIMP. 316.22 148.7 162.11
175 142.15 di 318.1 149.3 163.8
176 143.15 oltre 319.14 150.6 164.bl
177 144.13 -der 320.33 151.8 165.bl
178 145.13 leggieri 322.12 152.4 166.bl 1
179 146.14 SAGR. 323.25 153.4 167.b6
180 147.14 nave 325.2 154.5 168.bl2
181 148.15 l’havra 326.15 155.5 169.bl2
182 149.16 nave 328.2 156.5 170.21
183 150.16 -giu- 329.13 157.3 171.17
184 151.16 de 330.26 158.3 172.bl3
185 152.14 l’inferiori 331.37 159.3 173.bl 1
186 153.13 SAGR. 333.10 160.3 174.bl4
187 154.13 SAGR. 334.21 161.3 175 .b 17
188 155.13 lasci 335.33 162.1 176.10
189 156.15 -vimento 337.24 163.2 177.6
 APPENDIX 445

Edition of 1632 Pagnini Drake Salusbury-

Santillana
157.14 altxo 338.36 164.2 178.3
191 158.15 Stante 340.17 165.3 179.12
192 159.20 -nita 341.30 166.4 180.16
193 160.22 b 343.8 167.4 181.b8
194 161.22 della 346.9 168.7 183.8
195 162.24 SIMP. 347.25 169.7 184.7
196 163.24 per 348.36 170.6 184.blO
197 164.23 -sofare 350.11 171.6 185.bl9
198 165.23 fusse 351.25 172.6 186.bl6
199 166.23 parte 352.37 173.4 187.12
200 167.25 SAGR. 354.14 174.3 188.1
201 168.26 SIMP. 355.35 175.3 188.b3
202 169.33 terra 357.15 176.2 190.4
203 170.33 b eguale; 358.26 177.8 190.b8
204 171.34 -ginato 360.2 178.7 191.bll
205 172.36 SAGR. 361.18 179.10 192.18
206 173.37 -lela 362.33 180.9 193.17
207 174.38 SALV. 364.10 181.6 194.14
208 175.40 trenta 365.25 182.2 195.8
209 177.1 perchk 366.34 182.41 195.b3
210 178.1 con- 368.17 183.38 196.b4
211 179.2 perch& 369.30 184.38 198.6
212 180.2 -tamente 371.6 185.39 199.2
213 181.3 -gusta 372.22 186.37 199.b9
214 182.4 di 373.34 187.35 200.8
215 183.4 -tire 375.10 188.34 201.2
216 184.5 per 376.25 189.31 201.20
217 185.5 -che 378.1 190.27 202.15
218 186.4 SALV. 379.14 191.27 203.b9
219 187.4 col 380.26 192.28 204.b8
220 188.1 nuova 382.1 193.27 205 ,b4
221 188.40 SALV. 383.13 194.27 207.2
222 189.40 -zione 385.3 195.28 208.1
223 191.3 SALV. 386.18 196.28 209.3
224 192.4 SALV. 387.27 197.28 210.1
225 193.14 -gente 389.8 198.26 211.8
226 194.18 -stato 390.22 199.28 212.7
227 195.19 fatta 392.3 200.31 213.b8
228 196.21 per 393.16 201.31 214.b8
229 197.22 pure 394.28 202.29 215.blO
230 198.23 e 396.4 203.27 216.b7
231 199.25 SALV. 397.20 204.27 217.b8
232 200.25 SALV. 398.36 205.30 219.3
233 201.29 SIMP. 400.12 206.30 220.11
446 APPENDIX

Favaro Edition of 1632 Pagnini Drake Salusbury-

Santillana

234 202.29 -gioni 401.27 207.29 222.13

235 203.28 sopra 403.2 208.28 223.8
236 204.28 b 404.17 209.23 224.11
237 205.26 for- 405.30 210.20 225.10
238 206.27 noi 407.7 211.21 226.bl9
239 207.27 -peto 408.21 212.18 227.12
240 208.29 SALV. 409.37 213.16 228.7
241 209.30 SAGR. 411.15 214.21 229.14
242 210.28 romano 412.26 215.22 230.17
243 211.33 SAGR. 414.17 216.27 232.4
244 212.33 ritenere 416.11 217.26 233.4
245 213.32 che 417.23 218.23 233.bl
246 214.32 k 419.4 219.20 235.5
247 215.32 per 420.16 220.22 236.4
248 216.31 si 421.28 221.22 237.2
249 217.33 SAGR. 423.13 222.23 238.3
250 218.31 -tete 424.26 223.19 239.8
251 219.31 SALV. 426.12 224.17 240.9
252 220.37 Luna 427.28 225.21 241.b23
253 221.39 cosa 429.7 226.21 242.15
254 222.40 -menti 430.23 227.20 243.17
255 224.2 anco 431.36 228.20 244.bl5
256 225.7 al 433.14 229.22 245.M2
257 226.6 -rebbe 434.30 230.23 246.b3
258 227.11 sia 436.14 231.29 247.b2
259 228.12 72 437.34 232.37 249.1
260 229.4 medesima 438.36 233.28 249.b3
261 230.5 eccettuatone 440.13 234.31 250.b5
262 231.5 SIMP. 441.28 235.35 252.8
263 232.5 sopra 443.5 236.37 253.b6
264 233.6 SIMP. 444.17 238.1 254.b2
265 234.8 viventi 445.30 239.10 256.7
266 235.8 hoc 447.6 240.15 257.6
267 236.10 al 448.19 241.23 258.9
268 237.13 ciok 449.32 242.39 259.13
269 238.14 non 451.8 244.4 260.11
270 239.16 -torno 452.21 245.4 261.10
271 240.17 nk 453.36 246.15 262.12
272 241.9 uno 455.4 247.8 263.12
273 242.10 da 456.23 248.14 264.14
274 243.12 -tissima 457.37 249.14 265.19
275 244.6 perchk 459.12 250.11 266.5
276 245.7 del 460.28 251.12 266.b2
277 246.8 SAGR. 462.5 252.12 missing
 APPENDIX 447

Edition of 1632 Pagnini Drake Salusbury-

Santillana
278 247.7 la 463.15 253.10 267.bl2
279 248.9 -conferenza 464.28 254.14 268.bl5
280 249.10 se 466.16 255.13 269.b21
281 250.11 tal 467.32 256.19 270.15
282 251.12 -che 469.11 257.18 272.3
283 252.13 a 471.2 258.21 273.7
284 253.15 SIMP. 472.21 259.22 274.12
285 254.15 SIMP. 474.1 260.24 275.13
286 255.16 -nente 475.16 261.24 276.7
287 256.14 somma 476.29 262.24 277.6
288 257.15 a 478.7 263.25 278.8
289 258.16 SIMP. 480.32 264.27 279.bl7
290 259.16 -vengo- 482.11 265.25 280.bl9
291 260.17 convengono 483.27 266.25 281.19
292 261.18 SIMP. 485.6 267.26 282.16
293 262.18 -stare 487.2 268.32 283.M6
294 263.19 -plero 488.16 269.36 284.21
295 264.21 -sario 489.35 271.1 285 .bl4
296 265.22 awenga 491.15 272.1 286.M5
297 266.23 contro 493.21 273.1 287.bl2
298 267.25 -nen- 495.3 274.4 288.bll

299 269.1 GIORNATA 5.1 276.1 290.1

300 269.31 loro 6.11 277.1 291.17
301 270.37 il tempo 7.23 278.2 292.14
302 271.36 traportate 9.7 278.39 293.b5
303 272.30 necessario 10.16 279.31 295.3
304 273.26 che 11.24 280.26 296.11
305 274.24 SIMP. 13.11 281.25 297.b4
306 275.19 nel 14.16 282.17 298.15
307 276.17 3 15.36 283.17 299.19
308 277.3 12 17.3 284.7 299.b2
309 278.1 SAGR. 18.13 285.6 300.b5
310 278.38 fame 19.21 286.6 301.b8
311 280.5 ben 20.31 287.5 302.b6
312 281.3 che 22.6 287.41 303.bl0
313 282.2 svanisca 23.20 288.38 304.bl5
314 283.1 nota 25.5 289.34 305.16
315 284.1 SALV. 26.17 290.33 306.b20
316 284.29 dal 27.14 291.20 307.7
317 285.29 egli 28.26 292.20 308.2
318 286.27 accadere 29.34 293.17 308.b3
319 287.27 Ticone 31.11 294.17 310.1
320 288.19 Agecio 32.11 295.1 310.bl8
448 APPENDIX

F avaro Edition of 1632 Pagnini Drake Salusbury-

Santillana

321 289.17 SIMP. 33.30 295.37 311.17

322 290.11 Dove 35.11 296.35 312.10
323 291.10 della 36.18 297.36 312.bl
324 292.1 Ma 37.15 298.20 313.20
325 293.1 Regola 38.22 299.10 314.12
326 294.1 Quest- 39.32 300.1 315.1
327 295.1 Regola 41.5 300.33 315.b9
328 296.1 SAGR. 42.15 301.30 316.23
94910 302.27 317.11
329 296.30 1 7200
43.25
330 298.19s bisogna 45.18 303.22 318.6
331 299.13 Nella 46.1 304.17 318.b5
332 300.23 Nell- 47.1 305.6 319.bl7
333 301.26 ciok 48.5 306.1 320.20
334 302.37 La 49.14 306.37 321.15
335 304.1 Di 50.19 307.28 322.1
336 304.38 Ticone 51.35 308.25 322.bl8
337 305.39 Landgr. 53.20 309.22 323.16
338 306.33 qual 54.30 310.14 323 .b 1
339 307.32 -siopea 56.5 311.14 324,bl2
340 308.30 l’eccesso 57.17 312.19 326.1
341 309.39 e 58.28 313.19 327.3
342 311.10 Angoli 60.4 314.19 327.bl6
343 312.13 sopra 61.12 315.17 328.M2
344 313.11 appunto 62.22 316.13 329.22
345 314.11 -golo 63.33 317.15 330.bl7
346 315.6 -stum a 65.2 318.9 331.16
347 316.2 sin 67.11 319.8 332.14
348 317.2 niente 70.8 320.8 333.b9
349 317.40 -cipio 71.24 321.7 334.M0
350 318.40 e 73.10 322.5 335.M2
351 320.1 SIMP. 75.3 323.6 336.M3
352 321.25 -ch6 76.23 324.22 338.6
353 322.25 -gior 78.12 325.19 339.1
354 323.25 maniera 79.29 326.20 339.b4
355 324.24 -star 81.8 327.22 340.b3
356 325.24 ragione 82.23 328.22 342.2
357 325.31 Copernicano 90.39 334.212 342.8
358 325.34 ci 90.8 334.5 342.11
359 325.37 -nissimo 90.12 334.7 342.13
360 325.40 congiunzio- 90.16 334.10 342.17
361 326.4 & 90.21 334.13 342.20
362 326.7 che 90.25 334.17 342.23
363 326.35 strut- 91.28 335.6 343.19
364 327.36 si 93.7 336.6 344.14
 APPENDIX 449

Edition of 1632 Pagnini Drake Salusbury-

Santillana
365 328.34 -tro 94.27 337.6 345.14
366 329.34 libero 96.10 338.2 346.9
367 330.34 oltre 97.28 339.4 347.7
368 331.35 solamente 99.7 340.2 348.10
369 332.37 SALV. 100.21 341.4 349.b6
370 333.37 -traiii 102.1 342.6 351.5
371 334.37 terra 103.27 343.30 352.bl5
372 336.26 la 104.33 344.24 353.b5
373 337.27 erano 106.32 345.26 355.3
374 338.27 alcun 108.8 346.26 356.5
375 339.29 -feriore 109.24 347.27 357.bll
376 340.30 e 111.2 348.26 358.M3
377 341.39 i 112.16 349.30 359.23
378 342.40 prima 113.28 350.30 missing
379 344.5 andar 115.4 351.35 361.11
380 345.5 abitatori 116.20 352.34 362.8
381 346.7 a 118.4 353.35 363.7
382 347.7 -renze 119.17 354.35 364.5
383 348.10 ci 120.31 355.38 365.8
384 349.12 -nico 122.13 356.39 366.11
385 350.13 -sime 123.27 358.2 367.bl7
386 351.13 esser 126.5 359.3 368.bl7
387 352.9 una 127.18 359.39 369.16
388 353.5 quali 128.32 360.32 370.b5
389 354.6 verso 130.33 361.34 371.b5
390 355.8 SAGR. 132.11 362.35 372.bl
391 356.9 carta 134.7 363.32 373.b3
392 357.10 SIMP. 135.21 364.32 375.3
393 358.10 appunto 137.11 365.31 376.6
394 359.10 -tribuirli 138.26 366.31 377.7
395 360.10 ci 140.10 367.34 378.7
396 361.13 per 141.26 368.34 379.bl4
397 362.15 Stella 143.8 369.35 380.bl5
398 363.16 che 144.25 370.36 381.bl7
399 364.16 adoperiamo 146.5 371.37 382.bl3
400 365.16 -ture 147.22 372.38 383.bl3
401 366.18 SALV. 148.36 374.1 385.3
402 367.20 SAGR. 150.14 375.6 385 .b 1
403 368.20 che 151.25 376.3 386.b6
404 369.18 la 153.3 377.1 387.b8
405 370.18 -ticolare 154.17 378.2 388.b8
406 371.19 si 156.15 379.2 390.2
407 372.20 obliquo 157.32 380.4 391.2
408 373.356 punto 159.12 381.4 392.4
450 APPENDIX

F avaro Edition of 1632 Pagnini Drake Salusbury-

S antiliana

409 374.38 SALV. 160.27 382.8 393.3

410 375.38 di 162.9 383.8 394.2
411 377.6 remoto 163.27 384.5 395.5
412 378.25 raggio 165.22 385.11 396.1
413 379.24 che 167.5 386.12 396.b5
414 380.22 -cate 168.19 387.9 397.bl0
415 381.23 necessario 170.1 388.11 398.b7
416 382.25 -zione 171.15 389.13 400.2
417 383.25 -tano 173.4 390.13 401.8
418 385.7 Hora 174.32 391.34 402.3
419 386.1 Cap. 175.37 392.29 403.10
420 387.1 archi 177.13 393.30 missing
421 388.2 niente 178.27 394.33 404.bl2
422 389.1 e 180.2 395.31 missing
423 390.2 le velocita 181.16 396.31 406.2
424 391.4 SIMP. 182.31 397.33 407.5
425 392.4 acquista 184.10 398.36 408.2
426 393.5 ogni 185.29 399.37 409.3
427 394.5 conceduto 187.18 400.38 410.7
428 395.3 -riori 189.6 401.38 411.13
429 396.2 esse re 190.16 402.32 412.10
430 397.3 della 192.2 403.31 413.10
431 398.5 -verta 193.21 404.32 414.7
432 399.5 ne 195.10 405.31 415.4
433 400.7 esercitar 196.26 406.34 416.9
434 401.8 -teria 198.6 407.35 417.6
435 402.9 del 199.21 408.37 418.6
436 403.9 -stenu- 201.11 409.38 419.1
437 404.10 -pernico 202.27 410.38 419.bl
438 405.10 -trina 204.3 411.37 420.b8
439 406.9 -trasse 205.22 412.39 421.b8
440 407.7 -tiquattr- 206.32 413.38 422.M3
441 408.7 vasi 208.18 414.37 423.bl0

442 409.1 GIORNATA 223.110 416.1 424.1

443 409.30 globo 224.13 417.1 425.18
444 410.36 -vamento 226.9 417.40 426.18
445 411.36 il 227.24 418.40 427.15
446 412.37 Luna 229.6 419.39 428.14
447 413.39 le 231.12 421.1 429.bl6
448 415.1 SALV. 232.29 421.40 430.b9
449 416.1 -vilii 234.6 422.38 431.b5
450 417.3 SALV. 235.19 423.40 432.b5
451 418.4 -riamoci 236.36 425.2 433.21
 APPENDIX 451

Edition of 1632 Pagnini Drake Salusbury-

Santillana
419.5 SIMP. 238.14 426.7 434.16
453 420.23 destra 239.33 427.7 435.b7
454 421.25 Hor 241.13 428.8 436.bl9
455 422.26 in 242.31 429.6 437.M1
456 423.27 e 244.31 430.3 437.b3
457 425.7 SAGR. 246.32 431.5 439.1
458 426.7 -menti 248.9 432.3 439.M1
459 427.8 osserva 249.23 433.5 440.18
460 428.9 Sicilia 251.3 434.5 441.19
461 429.9 altre 252.16 435.6 442.14
462 430.11 b 253.30 436.9 443.15
463 431.12 6 255.8 437.10 444.14
464 432.12 monti 256.24 438.14 445.10
465 433.11 d’ 258.2 439.13 446.4
466 434.12 effetti 260.7 440.14 446.b2
467 435.13 differenza 261.23 441.13 448.5
468 436.13 il 262.37 442.14 449.7
469 437.14 sia 264.17 443.15 450.bl8
470 438.14 di 265.31 444.17 451.21
471 439.16 co’l 267.9 445.21 452.bl0
472 440.18 ritener 268.27 446.23 453.30
473 441.19 de 270.3 447.21 454.b7
474 442.21 dall’ 271.26 448.27 455,b6
475 443.20 e 273.4 449.28 456.b2
476 444.22 si 274.28 450.30 457.bl
477 445.24 e 276.5 451.33 459.b2
478 446.25 senza 278.3 452.33 460.4
479 447.26 cagione 279.16 453.35 461.1
480 448.27 desidero 281.6 454.35 461.b2
481 449.29 al 282.22 455.36 463.2
482 450.30 -lissima 284.2 456.39 464.23
483 451.31 intenda 285.28 457.37 465.b5
484 453.7 incammi- 286.35 459.12 466.blO
485 454.9 Ma 288.10 460.14 467.M0
486 455.10 haveste 289.23 461.17 468.bl3
487 456.12 -dere 291.7 462.21 469.M3
488 457.13 -metter 292.27 463.24 470.blO
489 458.14 SALV. 294.12 463.23 471.b2
452 APPENDIX

NOTES

1 See above Chapters 10 (Favaro section) and 16 (Section 15).

2 See above Chapter 10. Moreover, in the course of compiling this appendix I found
two more, which, though perhaps not as logically consequential as the imperfections
previously noted, nevertheless are relatively clearcut: (1) ‘scienze matematiche pure’
(F128.37-129.1) does not mean ‘mathematical sciences alone’ (D103), but rather ‘pure
mathematical sciences’; and (2) ‘in un altro vaso di angusta bocca (F212.37—213.1)
does not mean ‘into a wide vessel’ (D186), but rather ‘into a narrow vessel’.
3 Santillana indicates (p. 15) his omission of the second half of the discussion of three-
dimensionality with which the Dialogue begins. (The passage corresponds to F36.10-
38.15.) The following omissions, however, are not noted (where ‘SS’ denotes page
references to the Salusbury-Santillana translation, and ‘b’ indicates that lines are counted
from the bottom):

F 156.32—157.3 cf. SS145.M1 F276.13-277.29 cf. SS267.12

F 164.7—165.5 cf. SS153.b4 F310.33—311.1 cf. SS302.b6
F172.35-173.8 cf. SS161.M9 F342.24—343.1 cf. SS328.M2
F174.3-174.4 cf. SS162.13-162.14 F377.32-378.3 cf. SS360.b20
F195.5-195.13 cf. SS184.11-184.12 F418.36-419.1 cf. SS403.10
F195.21-195.30 cf. SS 184.17 F419.34-420.2 cf. SS403.M
F199.13-199.24 cf. SS187.M5 F421.34-422.4 cf. SS405 .bl7
F202.29-202.35 cf. SS190.M 1 F422.10-422.24 cf. SS405.bl1
F204.6-204.7 cf. SS192.17-192.18 F422.28-422.30 cf. SS405.b7
F208.2-209.9 cf. SS195.9 F450.25-450.34 cf. SS433.18
F212.20-212.26 cf. SS199.20 F457.17-457.21 cf. SS439.18
F213.15-213.34 cf. SS200.4 F457.26-457.30 cf. SS439.22
F214.33-215.1 cf. SS201.1 F458.8-458.16 cf. SS439.b5
F215.18-215.33 cf. SS201.15-201.16 F475.35-475.37 cf. SS457.bl-457.b2

4 Here (and in other similar circumstances) I am counting in such a way that the title
is taken to occupy two lines, and the dialogue is taken to begin on the third line.
5 The computation on the top half of this page (above line 19) is also on p. 330 in
Favaro.
6 Here (and in other similar circumstances) I am counting in such a way that the
diagram is taken as occupying the same number of lines of text that would be printed in
its place if it were not there.
7 The text in the Pagnini edition from p. 125.9 to p. 126.23 appears in Favaro in a
footnote on p. 54, and the text in Pagnini from p. 126.23 to 127.29 appears in Favaro in
a footnote on p. 55.
8 Pp. 241-254 in Pagnini contain a note on Aristotelian physics.
9 Pp. 83.1-90.2 in Pagnini correspond to the footnote text in Favaro pp. 356 — 362,
which in turn correspond to Galileo’s addendum handwritten in his own copy of the
Dialogue. Pagnini’s p. 83.1-83.36 appears on p. 356 in Favaro; Pagnini’s pp. 83.37 — 85.1
on Favaro’s p. 357; Pagnini’s pp. 85.2—86.8 on Favaro’s p. 358; and so on for Pagnini’s
pp. 86.8-87.15, 87.15-88.21, 88.21-89.27, and 89.27-90.2.
 APPENDIX 453

10 Pp. 209-222 in Pagnini contain a note on Copernicus’ third motion.

11 Pp. 30.8-31.19 in Drake correspond to the footnote text on pp. 54-55 in Favaro.
Pp. 328.29-333.40 in Drake correspond to Galileo’s addendum handwritten in his
copy of the Dialogue, which is printed as a footnote in Favaro on pp. 356-362. The
correspondence is as follows:

Page in Favaro Pages in Drake

356 328.29-329.15
357 329.16-330.9
358 330.9-331.3
359 331.3-331.39
360 331.40-332.34
361 332.34-333.30
362 333.30-333.40
 SELECTED BIBLIOGRAPHY

Adams, Ernest W., The Logic of Conditionals. Dordrecht: Reidel, 1975.

Agassi, Joseph, ‘The Role of Corroboration in Popper’s Methodology’, Australasian
Journal of Philosophy 39 (1961), 82-91.
Agassi, Joseph, Towards an Historiography of Science. History and Theory, Beiheft 2.
The Hague: Mouton, 1963.
Agazzi, Evandro, ‘Fisica galileiana e fisica contemporanea’, in Nel quarto centenano
della nascita di Galileo Galilei, pp. 1-51. Milan: Societa Editrice Vita e Pensiero,
1966.
Aiton, E. J., ‘Galileo and the Theory of the Tides’, Isis 56 (1965), 56-61.
Aiton, E. J., ‘Galileo’s Theory of the Tides’, Annals of Science 10 (1954), 44-5 7.
Aiton] E. J. ‘On Galileo and the Earth-Moon System’,Isis 54 (1963), 265-6.
Anderson, Alan R., Nuel D. Belnap, Jr., et al., Entailment: The Logic of Relevance and
Necessity, Vol. 1. Princeton: Princeton University Press, 1975.
Angell, Richard B., Reasoning and Logic. New York: Appleton, 1964.
Annis, David B., Techniques of Critical Reasoning. Columbus, Ohio: Charles E. Merrill,
1974.
Antoni, Carlo, Commento a Croce. Venice: Neri Pozza, 1965.
Barenghi, Giovanni, Considerazioni sopra il Dialogo. Pisa, 1638.
Bar-Hillel, Yehoshua, ‘Argumentation in Natural Languages’, in Akten des XIV Inter-
nationalen Kongresses fur Philosophic, II, 3-6. Vienna: Verlag Herder, 1969, Rpt.
Aspects of Language, pp. 202-5.
Bar-Hillel, Yehoshua, ‘Argumentation in Pragmatic Languages’, in Aspects of Language,
pp. 206—21.
Bar-Hillel, Yehoshua, Aspects of Language. Essays and Lectures on Philosophy of
Language, Linguistic Philosophy, and Methodology of Linguistics. Jerusalem: The
Magnes Press, The Hebrew University; Amsterdam: North-Holland, 1970.
Bar-Hillel, Yehoshua, ‘Comments’, in ‘Formal Logic and Natural Language (A
Symposium)’, edited by J. F. Staal.
Bar-Hillel, Yehoshua, ‘A Neglected Recent Trend in Logic’, Logique et Analyse 39
(1967), 235-8.
Bar-Hillel, Yehoshua, (ed.), Pragmatics of Natural Languages. New York: Humanities
Press, 1971.
Bartley, William, W., Ill, ‘Rationality versus the Theory of Rationality’, in The Critical
Approach to Science and Philosophy, edited by M. Bunge, pp. 3-31. New York:
Free Press, 1964.
Bartley, William W., Ill, The Retreat to Commitment. New York: Knopf, 1962.
Beardsley, Monroe C., Thinking Straight. 3rd edition. Englewood Cliffs, New Jersey:
Prentice-Hall, 1966.

454
 SELECTED BIBLIOGRAPHY 455

Black, Max, Critical Thinking. 2nd edition. Englewood Cliffs, New Jersey: Prentice-Hall
1952.
Br&rier, E., The History of Philosophy: The Seventeenth Century. Translated by W.
Baskin. Chicago: University of Chicago Press, 1966.
Brown, Harold I., ‘Galileo, the Elements, and the Tides’, Studies in History and Philos¬
ophy of Science 7 (1976), 337-51.
Burstyn, Harold L., ‘Galileo and the Earth-Moon System: Reply to Dr Aiton’ Isis 54
(1963), 400-1.
Burstyn, Harold L., ‘Galileo and the Theory of the Tides’, Isis 56 (1965), 61-3.
Burstyn, Harold L., ‘Galileo’s Attempt to Prove that the Earth Moves’, Isis 53 (1962)
161-85.
Burtt, E. A., Metaphysical Foundations of Modern Physical Science. Garden City, New
York: Doubleday, 1954.
Butts, Robert E., and Joseph C. Pitt (eds.), New Perspectives on Galileo. Dordrecht-
Reidel, 1978.
Chiaramonti, Scipione, Difesa . .. al suo Antiticone e Libro delle tre nuove Stelle.
Florence: Landini, 1633.
Clavelin, Maurice, The Natural Philosophy of Galileo. Translated by A. J. Pomerans.
Cambridge, Massachusetts and London: M.I.T. Press, 1974.
Clavelin, Maurice, La philosophic naturelle de GaliUe. Paris: Colin, 1968.
Coffa,Jos^ A., ‘Galileo’s Concept of Inertia’, Physis 10 (1968), 261-81.
Cohen, I. Bernard, ‘Galileo, Newton, and the Divine Order of the Solar System’, in
Galileo, Man of Science, edited by E. McMullin, pp. 207-31.
Cohen, 1 Bernard, ‘Newton’s Attribution of the First Two Laws of Motion to Galileo’, in
Symposium Internazionale di Storia, Metodologia, Logica e Filosofia della Scienza
pp. XXV-XLIV.
Cohen, L. Jonathan, The Probable and the Provable. Oxford: Oxford University Press
1977.
Cohen, Morris R. and Ernest Nagel, An Introduction to Logic and Scientific Method.
New York: Harcourt, 1934.
Copi, Irving M., Introduction to Logic. 4th edition. New York: Macmillan, 1972.
Copleston, F., History of Philosophy, Vol. Ill, Late Medieval and Renaissance Philos¬
ophy. Westminster, Maryland: Newman Press, 1963.
Croce, Benedetto, Historical Materialism and the Economics of Karl Marx. Translated by
C. M. Meredith. New York: Russell and Russell, 1966. First English edition, 1914.
First Italian edition, 1900.
Croce, Benedetto, History: Its Theory and Practice [1920], Translated by D. Ainslie,
New York: Russell and Russell, 1960.
Croce, Benedetto, Philosophy of the Practical. Translated by D. Ainslie. 1913; rpt. New
York: Biblio and Tannen.
Croce, Benedetto, La poesia. Bari: Laterza, 1936.
Croce, Benedetto, Storia dell’eta barocca in Italia. Bari: Laterza, 1929.
Croce, Benedetto, Teoria e storia della storiografla. Bari: Laterza, 1917.
Croce, Benedetto, Theory and History of Historiography. Translated by D. Ainslie.
London: Harrap, 1921.
Crombie, A. C., ‘Galileo Galilei: A Philosophical Symbol’, in Actes du VIIIe Congrks
International d’Histoire des Sciences, 1956,111, 1089-95.
456 SELECTED BIBLIOGRAPHY

Drake, Stillman, ‘The Evolution of De Motu', Isis 67 (1976), 239-50.

Drake’ Stillman, ‘Galileo and the First Mechanical Computing Device,” Scientific
American 234 (1976), 104-13.
Drake, Stillman, ‘Galileo and the Law of Inertia’, American Journal of Physics 32
(1964), 601-8.
Drake, Stillman, Galileo at Work: His Scientific Biography. Chicago: University of
Chicago Press, 1978.
Drake, Stillman, ‘Galileo in English Literature of the Seventeenth Century’, in Galileo,
Man of Science, edited by E. McMullin, pp. 415-31.
Drake, Stillman, ‘Galileo’s Discovery of the Law of Free Fall’, Scientific American 228
(1973), 84-92.
Drake, Stillman, ‘Galileo’s Experimental Confirmation of Horizontal Inertia: Un¬
published Manuscripts’, Isis 64 (1973), 291-305.
Drake, Stillman, ‘Galileo’s New Science of Motion’, in Reason, Experiment, and
Mysticism in the Scientific Revolution, edited by M. L. Righini Bonelli and W. R.
Shea, pp. 131-56.
Drake, Stillman, ‘Galileo’s “Platonic” Cosmogony and Kepler’s Prodromus’, Journal for
the History of Astronomy 4 (1973), 174-91.
Drake, Stillman, Galileo Studies. Ann Arbor: University of Michigan Press, 1970.
Drake, Stillman, ‘Impetus Theory and Quanta of Speed Before and After Galileo’,
Physis 16 (1974), 47-65.
Drake, Stillman, ‘Review of Clavelin’s Philosophie naturelle de Galilee', Isis 61 (1970),
275-7.
Drake, Stillman, ‘Semicircular Fall in the Dialogue’, Physis 10 (1968), 89-100.
Drake, Stillman, ‘The Uniform Motion Equivalent to a Uniformly Accelerated Motion
from Rest’, Isis 63 (1972), 28-38.
Drake, Stillman, ‘Velocity and Eudoxian Proportion Theory’, Physis 15 (1973), 49-64.
Drake, Stillman (ed. and trans.), Discoveries and Opinions of Galileo. Garden City, New
York: Doubleday, 1957.
Drake, Stillman (ed. and trans.), Galileo Against the Philosophers. Los Angeles: Zeitlin
and Ver Brugge, 1976.
Drake, Stillman and C. D. O’Malley (eds. and trans.), The Controversy on the Comets of
1618. Philadelphia: University of Pennsylvania Press, 1960.
Drake, Stillman and I. E. Drabkin (eds. and trans.), Mechanics in Sixteenth Century
Italy. Madison: University ofWisconsin Press, 1969.
Drake, Stillman, and J. MacLachlan, ‘Galileo’s Discovery of the Parabolic Trajectory’,
Scientific American 232 (1975), 102-10.
Duhem, Pierre, The Aim and Structure of Physical Theory. Translated by P. P. Wiener.
Princeton, New Jersey: Princeton University Press, 1954.
Duhem, Pierre, Essai sur la thiorie physique de Platon a GaliUe. Paris: Hermann, 1908.
Duhem, Pierre, fctudes sur Leonard de Vinci. 3 Vols. Paris: Hermann, 1905 — 1913.
Duhem, Pierre, Les origines de la statique, 2 Vols. Paris: Hermann, 1905 — 1906.
Duhem, Pierre, To Save the Phenomena. Translated by E. Doland and C. Maschler.
Chicago: University of Chicago Press, 1969.
Duhem, Pierre, Le systtme du monde. 10 Vols. Paris: Hermann, 1913ff., and 1954ff.
Einstein, Albert, Foreword to Dialogue Concerning the Two Chief World Systems, by
G. Galilei. Edited by S. Drake. Berkeley: University of California Press, 1953.
 SELECTED BIBLIOGRAPHY 457

Ellis, Albert, Growth Through Reason: Verbatim Cases in Rational-Emotive Psycho¬

therapy. Palo Alto, California: Science and Behavior Books, 1971.
Ellis, Albert, Reason and Emotion in Psychotherapy. New York: Lyle Stuart, 1962.
Ellis, Albert, and R. A. Harper, A Guide to Rational Living. Hollywood^ California:
Wilshire Book Company, 1968.
Engel, S. Morris, With Good Reason: An Introduction to Informal Fallacies. New York:
St. Martin’s Press, 1976.
Ennis, Robert H., ‘Enumerative Induction and Best Explanation', Journal of Philosophy
65 (1968), 523-29.
Fearnside, W. Ward and W. B. Holther, Fallacy: The Counterfeit of Argument. Engle¬
wood Cliffs, New Jersey: Prentice-Hall, 1959.
Feyerabend, Paul K., ‘Against Method’, in Minnesota Studies in the Philosophy of
Science, Vol. 4, edited by M. Radner and S. Winokur, pp. 17-130. Minneapolis:
University of Minnesota Press, 1970.
Feyerabend, Paul K., Against Method. Atlantic Highlands, New Jersey: Humanities Press
Inc., 1975.
Feyerabend, Paul K., Ausgewdhlte Aufsatze. Brunswick: Vieweg, 1974.
Feyerabend, Paul K., ‘Consolations for the Specialist’, in Criticism and the Growth of
Knowledge, edited by I. Lakatos and A. Musgrave, pp. 197-230.
Feyerabend, Paul K., Contro il metodo. Milan: Lampugnani Nigri, 1973.
Feyerabend, Paul K., Einfuhrung in die Naturphilosophie. Brunswick: Vieweg, 1974.
Feyerabend, Paul K., ‘Explanation, Reduction, and Empiricism’, in Minnesota Studies in
the Philosophy of Science, Vol. 3, edited by H. Feigl and G. Maxwell, pp. 28-97.
Minneapolis, Minnesota: University of Minnesota Press, 1962.
Feyerabend, Paul K., ‘How to be a Good Empiricist’, in Philosophy of Science: The
Delaware Seminar, Vol. 2, edited by B. Baumrin, pp. 3-39. New York: Interscience
1963.
Feyerabend, Paul K., ‘Machamer on Galileo’, Studies in History and Philosophy of
Science 5 (1975), 297-304.
Feyerabend, Paul K.,1 problemi dell’empirismo. Milan: Lampugnani Nigri, 1971.
Feyerabend, Paul K., ‘Problems of Empiricism’, in Beyond the Edge of Certainty, edited
by R. Colodny, pp. 145-260. Englewood Cliffs, New Jersey: Prentice-Hall, 1965.
Feyerabend, Paul K., ‘Problems of Empiricism, Part II’, in The Nature and Function of
Scientific Theories, edited by Robert G. Colodny, pp. 275-353. Pittsburgh, Pennsyl¬
vania: University of Pittsburgh Press, 1970.
Feyerabend, Paul K., Wider den Methodenzwangtheorie. Frankfurt: Suhrkamp, 1976
Feynman, Richard P., R. B. Leighton, and M. Sands, The Feynman Lectures on Physics.
3 Vols. Reading, Massachusetts: Addison-Wesley, 1963.
Finocchiaro, Maurice A., ‘The Concept of Ad Hominem Argument in Galileo and
Locke’, The Philosophical Forum 5 (1974), 394-404.
Finocchiaro, Maurice A., ‘Dialectical Aspects of the Copernican Revolution: Conceptual
Elucidations and Historiographical Problems’, in The Copernican Achievement,
edited by Robert S. Westman, pp. 204-12. Berkeley: University of California Press,
1975.
Finocchiaro, Maurice A., ‘Essay-review of Criticism and the Growth of Knowledge', in
Studies in History and Philosophy of Science 3 (1973), 357-72.
Finocchiaro, Maurice A., ‘Galileo and the Philosophy of Science’, in PSA 1976
458 SELECTED BIBLIOGRAPHY

(Proceedings of the 1976 Biennial Meeting of the Philosophy of Science Association),

Vol. 1, edited by F. Suppe and P. D. Asquith, pp. 130-9. East Lansing, Michigan:
Philosophy of Science Association, 1976.
Finocchiaro, Maurice A., ‘Galileo as a Logician ,Physis 16 (1974), 129—48.
Finocchiaro, Maurice A., ‘Galileo’s Space-Proportionality Argument: A Role for Logic in
Historiography’,Physis 15 (1973), 65—72.
Finocchiaro, Maurice A., History of Science as Explanation. Detroit, Michigan: Wayne
State University Press, 1973.
Finocchiaro, Maurice A., ‘Logic and Rhetoric in Lavoisier’s Sealed Note: Toward a
Rhetoric of Science', Philosophy and Rhetoric 10 (1977), 111-22.
Finocchiaro, Maurice A., ‘Philosophizing About Galileo, British Journal for the
Philosophy of Science 26 (1975), 255-64.
Finocchiaro, Maurice A., ‘Review of Butts and Pitt’s New Perspectives on Galileo',
Philosophy of the Social Sciences, forthcoming.
Finocchiaro, Maurice A., ‘Review of Clavelin’s Natural Philosophy of Galileo', Review of
Metaphysics 29 (1976), 544.
Finocchiaro, Maurice A., ‘Review of G. Gentile’s La filosofia di Marx', The Thomist 39
(1975), 423-26.
Finocchiaro, Maurice A., 'Vires Acquirit Eundo: The Passage Where Galileo Renounces
Space-Acceleration and Casual Investigation’, Physis 14 (1972), 125-45.
Fogelin, Robert J., Understanding Arguments: An Introduction to Informal Logic. New
York: Harcourt Brace Javanovich, 1978.
Funkenstein, Amos, ‘The Dialectical Preparation for Scientific Revolutions’, in The
Copernican Achievement, edited by Robert S. Westman, pp. 165-203. Berkeley:
University of California Press, 1975.
Galilei, Galileo, Dialogo di Galileo Galilei Linceo .. . Florence, 1632; rpt. Brussels:
Editions Culture et Civilisation, 1966.
Galilei Galileo, Dialogo sopra i due massimi sistemi del mondo, tolemaico e
copernicano. Edited by Libero Sosio. Turin: Einaudi, 1971.
Galilei, Galileo, Dialog iiber die beiden hauptsachlichsten Weltsysteme. Translated by
Emil Strauss. Leipsig: Teubner, 1891.
Galilei, Galileo, Dialogue Concerning the Two Chief World Systems. Translated by S.
Drake. Berkeley: University of California Press, 1953, 1962, and 1967.
Galilei, Galileo, Dialogue on the Great World Systems. Salusbury’s translation revised by
G. de Santillana. Chicago: University of Chicago Press, 1953.
Galilei, Galileo, On Motion and on Mechanics. Translated by I. E. Drabkin and S. Drake.
Madison, Wisconsin: University of Wisconsin Press, 1960.
Galilei, Galileo, Opere. 5 Vols. Edited by P. Pagnini. Florence: Salani, 1964.
Galilei, Galileo, Opere. Volumes II and III, Dialogo sopra i due massimi sistemi. Edited
by P. Pagnini. Florence: Salani, 1964.
Galilei, Galileo, Le opere di Galileo Galilei. 20 Vols. Edizione Nazionale by A. Favaro
et al. Florence: Barbera, 1890-1909, 1929-1939, and 1968.
Galilei, Galileo, Le opere di Galileo Galilei. Vol. VII, Dialogo sopra i due massimi sistemi.
Edizione Nazionale by A. Favaro et al. Florence: Barbera, 1890—1909, 1929-1939,
and 1968.
Galilei, Galileo, Two New Sciences. Translated with Introduction and Notes by S. Drake.
Madison, Wisconsin: University of Wisconsin Press, 1974.
 SELECTED BIBLIOGRAPHY 459

Garin, Eugenio, ‘Chi legga di A. Koyrd . . . Giornale critico della filosofia italiana 11
(1957), 406-408.
Gellner, Ernest, ‘Beyond Truth and Falsehood’, British Journal for the Philosophy of
Science 26 (1975) 331-42.
Gentile, Giovanni, La filosofia diMarx. 5th edition by V. BeUezza. Florence: Sansoni,
1974. First edition, 1899.
Geymonat, L. Galileo Galilei: A Biography and Inquiry into his Philosophy of Science.
Translated by S. Drake. New York: McGraw-Hill, 1965. First Italian edition
1957.
Gillispie, Charles C., The Edge of Objectivity: An Essay in the History of Scientific
Ideas. Princeton, New Jersey: Princeton University Press, 1960.
Gombrich, Ernst H., The Story of Art. 11th edition. New York: Phaidon Publishers
1966.
Hall, A. R., ‘Essay Review of Clavelin’s La philosophic naturelle de Galilie', British
Journal for the History of Science 6 (1972), 80-4.
Hamblin, C. L., Fallacies. London: Methuen, 1970.
Hartner, Willy, ‘Galileo’s Contribution to Astronomy’, in Galileo, Man of Science,
edited by E. McMullin, pp. 178-94.
Heilbron, John L., ‘Review of Clavelin’s Philosophic naturelle de Galilie', Journal of
the History of Philosophy 8 (1970), 341-43.
Hempel, Carl G., Philosophy of Natural Science. Englewood Cliffs, New Jersey: Prentice-
Hall, 1966.
Hendel, C. W., Studies in the Philosophy of David Hume. Indianapolis, Indiana: Bobbs-
Merrill, 1963.
Hesse, Mary B., The Structure of Scientific Inference. Berkeley: University of California
Press, 1974.
Hintikka, Jaakko, The Semantics of Questions and the Questions of Semantics. Acta
Philosophica Fennica, Vol. 28, No. 4. Amsterdam: North-Holland, 1976.
Hume, David, Dialogues Concerning Natural Religion. Edited by N. Kemp-Smith.
Indianapolis, Indiana: Bobbs-Merrill, n.d. First Edition, 1779.
Husserl, Edmund, The Crisis of European Sciences and Transcendental Phenomenology.
Translated by D. Carr. Evanston, Illinois: Northwestern University Press, 1970.
Hurlbutt, R. H., III. Hume, Newton, and the Design Argument. Lincoln, Nebraska:
University of Nebraska Press, 1965.
Jaspers, Karl, The Great Philosophers. 2 volumes. Translated by R. Mannheim. New
York: Harcourt, Brace and World, Inc., 1962. First German edition, 1957.
Jones, W. T., A History of Western Philosophy. New York: Harcourt, Brace and World,
1952.
Johnstone, Henry W., Jr., Philosophy and Argument. University Park, Pennsylvania:
Pennsylvania State University Press, 1959.
Kahane, Howard, Logic and Contemporary Rhetoric: The Use of Reason in Everyday
Life. Belmont, California: Wadsworth, 1971.
Kalish, D. and R. Montague, Logic: Techniques of Formal Reasoning. New York:
Harcourt, Brace and World, 1964.
Kant, Immanuel, Critique of Pure Reason. Translated by N. Kemp Smith. New York: St.
Martin’s Press, 1965. First German edition, 1781.
Kedrov, B. M. and B. H. Kouznetsov, ‘La Logique de Galilee et la logique de la physique
460 SELECTED BIBLIOGRAPHY

actuelle’, in Symposium Internazionale di Storia, Metodologin, Logica e Filosofia

della Scienza, pp. IXC-XCVII.
Koertge, Noretta, ‘Bartley’s Theory of Rationality’, Philosophy of the Social Sciences 4
(1974), 75-81.
Koestler, Arthur, The Sleepwalkers. New York: Grosset and Dunlap, 1959.
Kouznetsov, Boris, ‘Galilee et Einstein. Prologue et Epilogue de la Science Classique , in
Actes du XIIe Congris International d’Histoire des Sciences (1968), Volume 5, pp.
59-63. Published 1971.
Kouznetsov, Boris, ‘L’id6e d’homog6n<iit£ de l’espace dans le Dialogo de Galilee’, in
Actes du Symposium International des sciences physiques et mathematiques dans la
premiere moitii du XVIIe siecle, Pisa-Vinci, 1958, pp. 133-41. Paris, 1960.
Kouznetsov, Boris, ‘Le soleil comme centre du monde, et l’homog6n6it6 de 1 espace chez
Galilee’, in Le Soleil a la Renaissance, pp. 73-88. Brussels: Presses Universitaires de
Bruxelles, 1965.
Kouznetsov, Boris, ‘Style et Contenu de la Science , Diogenes, No. 89, Spring 1975, pp.
55-75.
Koyrd, Alexandre, Etudes GaliUennes. 3 Vols. 1939; rpt. Paris: Hermann, 1966.
Koyrei, Alexandre, Galileo Studies. Translated by J. Mepham. Hassocks, England:
Harvester Press, 1978.
Kreyche, Robert J., Logic for Undergraduates. 3rd edition. New York: Holt, Reinhart
and Winston, 1970.
Kuhn, Thomas S., The Structure of Scientific Revolutions. 2nd edition. Chicago: Univer¬
sity of Chicago Press, 1970. 1st Edition, 1962.
Kyburg, Henry E„ Jr., Probability and Inductive Logic. New York: Macmillan,
1970.
Lakatos, Imre and Alan Musgrave (eds.), Criticism and the Growth of Knowledge.
Cambridge: Cambridge University Press, 1970.
Lyons, John, Structural Semantics: An Analysis of Part of the Vocabulary of Plato.
Oxford: Basil Blackwell, 1963.
Mach, Ernst, The Science of Mechanics. Translated by T. J. McCormack. La Salle,
Illinois: Open Court, 1960. 1st German edition, 1883.
Machamer, Peter K., ‘Feyerabend and Galileo: The Interaction of Theories and the
Reinterpretation of Experience’, Studies in History and Philosophy of Science 4
(1973), 1-46.
MacLachlan, James, ‘The Test of an “Imaginary” Experiment of Galileo’s’, Isis 64
(1973), 374-9.
Mahoney, M. S., ‘Galileo’s Thought’, Science 187 (1975), 944-5.
Mates, Benson. Elementary Logic. 2nd edition. New York: Oxford University Press,
1972.
McMullin, Ernan, ‘The Conception of Science in Galileo’s work’, in New Perspectives on
Galileo, edited by Robert E. Butts and J. C. Pitt, pp. 209-5 7.
McMullin, Ernan, ‘Introduction: Galileo, Man of Science’, in Galileo: Man of Science,
edited by E. McMullin, pp. 3-51.
McMullin, Ernan (ed.), Galileo: Man of Science. New York: Basic Books, 1968.
Merton, Robert K., On Theoretical Sociology. New York: Free Press, 1967.
Michalos, Alex C., Improving Your Reasoning. Englewood Cliffs, New Jersey: Prentice-
Hall, 1970.
 SELECTED BIBLIOGRAPHY 461

Montague, Richard, Formal Philosophy. New Haven, Connecticut: Yale University Press
1974.
Moore, W. Edgar., Creative and Critical Thinking. Boston: Houghton, 1967.
Nakayama, Shigeru, ‘Galileo and Newton’s Problem of World Formation’, Japanese
Studies in the History of Science 1 (1962), 76-82.
Natanson, Maurice and H. W. Johnstone, Jr. (eds.), Philosophy, Rhetoric, and Argu¬
mentation. University Park, Pennsylvania: Pennsylvania State University Press,
1965.
Naylor, Ronald H., ‘Galileo and the Problem of Free Fall’, British Journal for the
History of Science 7 (1974), 105-34.
Naylor, Ronald H., ‘Galileo: Real Experiment and Didactic Demonstration’, Isis 67
(1976), 398-419.
Naylor, Ronald H., ‘Galileo’s Simple Pendulum’, Physis 16 (1974), 23-46.
Newton, Isaac, Mathematical Principles of Natural Philosophy. Motte’s translation
revised by F. Cajori. Berkeley: University of California Press, 1934. 1st Latin edition,
1687.
Nobile, Vittorio, ‘Sull’argomento galileiano della quarta giornata dei Dialoghi e sue
attinenze col problema fondamentale della geodesia’, Atti dell Accademia Nazionale
dei Lincei, Gasse di Scienze Fisiche 16 (1954), 426-33.
Oregius, Augustinus, De Deo uno. Rome, 1629.
Ortega y Gasset, Jos^, En torno a Galileo. Madrid: Revista de Occidente, 1956.
Ortega y Gasset, Jos^, Man and Crisis. Translated by M. Adams. New York. Norton,
1958. Spanish title: En torno a Galileo.
Osherson, D„ ‘Models of Logical Thinking’, in Reasoning: Representation and Process in
Children and Adults, edited by Rachel J. Falmagne, pp. 81-91. Hillsdale, New
Jersey: Lawrence Erlbaum Associates, 1975.
Pap, A., An Introduction to the Philosophy of Science. New York: The Free Press of
Glencoe, 1962.
Perelman, Chaim, The Idea of Justice and the Problem of Argument. Translated by J.
Petrie. New York: Humanities Press, 1963.
Perelman, Chaim, The New Rhetoric’, in Pragmatics of Natural Languages, edited by
Y. Bar-Hillel, pp. 145-49.
Perelman, Chaim, ‘A Reply to Henry W. Johnstone, Jr.’, in Philosophy, Rhetoric, and
Argumentation, edited by Maurice Natanson and H. W. Johnstone, Jr., pp. 135 7.
University Park, Pennsylvania: Pennsylvania State University Press, 1965.
Perelman, Chaim and L. Olbrechts-Tyteca, The New Rhetoric: A Treatise on Argumenta¬
tion. Translated by J. Wilkinson and P. Weaver. Notre Dame: University of Notre
Dame Press, 1969.
Popper, Karl, The Logic of Scientific Discovery. New York: Harper and Row, 1959. 1st
German edition, 1935.
Popper, Karl, Logik der Forschung. Vienna: Julius Springer, 1935.
Popper, Karl, Objective Knowledge: An Evolutionary Approach. New York: Oxford
University Press, 1972.
Popper, Karl, The Open Society and Its Enemies. Vol. 2. Princeton, New Jersey:
Princeton University Press, 1971. First published 1945.
Quine, Willard V., Methods of Logic. 3rd edition. New York: Holt, Rinehart and
Winston, 1972.
462 SELECTED BIBLIOGRAPHY

Remusat, C. de, Bacon, sa Vie, son Temps, et son Influence jusqu’a nos Jours. 2nd
edition. Paris, 1858.
Righini Bonelli, M. L. and W. R. Shea (eds.), Reason, Experiment, and Mysticism in
the Scientific Revolution. New York: Science History Publications, 1975.
Ryle, Gilbert, ‘Proofs in Philosophy’, Revue internationale de philosophie 8 (1954),
152ff. Reprinted in G. Ryle, Collected Papers, Vol. 2. New York: Barnes and Noble,
1971.
Saarinen, Esa (ed.), Game-Theoretical Semantics. Dordrecht: Reidel, 1978.
Salmon, Wesley C., Logic. 2nd edition. Englewood Cliffs, New Jersey: Prentice-Hall,
1973.
Santillana, Giorgio de, The Crime of Galileo. Chicago: University of Chicago Press, 1955.
Schmitt, Charles B., ‘Review of Shapere’s Galileo’, Renaissance Quarterly 28 (1975),
362-3.
Scriven, Michael, ‘Definitions, Explanation, and Theories’, in Minnesota Studies in the
Philosophy of Science, Vol. 2, edited by H. Feigl, M. Scriven, and G. Maxwell, pp.
99-195. Minneapolis: Minnesota: University of Minnesota Press, 1958.
Scriven, Michael, ‘The Frontiers of Psychology: Psychoanalysis and Parapsychology’, in
Frontiers of Science and Philosophy, edited by Robert G. Colodny, pp. 79-129.
Pittsburgh, Pennsylvania: University of Pennsylvania Press, 1962.
Scriven, Michael, ‘A Possible Distinction Between Traditional Scientific Disciplines and
the Study of Human Behavior’, in Minnesota Studies in the Philosophy of Science,
Vol. 1, edited by H. Feigl and M. Scriven, pp. 330—9. Minneapolis, Minnesota:
University of Minnesota Press, 1956.
Scriven, Michael, ‘Psychology Without a Paradigm’, in Clinical Cognitive Psychology,
edited by Louis Breger.
Scriven, Michael, Reasoning. New York: McGraw-Hill, 1976.
Scriven, Michael, ‘Science: The Philosophy of Science’, International Encylopedia of the
Social Sciences, 1968, Vol. 14, pp. 83-92.
Scriven, Michael, ‘A Study of Radical Behaviorism’, in Minnesota Studies in the Philos¬
ophy of Science, Vol. 1, edited by H. Feigl and M. Scriven, pp. 88-130. Minneapolis,
Minnesota: University of Minnesota Press, 1956.
Scriven, Michael, ‘Verstehen Again’, Theory and Decision 1 (1971), 382-6.
Scriven, Michael, ‘Views of Human Nature’, in Behaviorism and Phenomenology, edited
by T. W. Warm, pp. 163—90. Chicago: University of Chicago Press, 1964.
Seeger, Raymond J., Men of Physics: Galileo Galilei, His Life and Works. New York:
Pergamon Press, 1966.
Seeger, Raymond J., ‘On the Role of Galileo in Physics’, Physis 5 (1963), 5-38.
Settle, Thomas B., ‘An Experiment in the History of Science’, Science 133 (1961)
19-23.
Shapere, Dudley, Galileo: A Philosophical Study. Chicago: University of Chicago Press
1974.
Shea, William R., Galileo’s Intellectual Revolution. New York: Science History Publica¬
tions, 1972.
Shea, William R., ‘Review of Clavelin’s Philosophie naturelle de GaliUe’, British Journal
for the Philosophy of Science 21 (1970), 124-5.
Skyrms, Brian, Choice and Chance: An Introduction to Inductive Logic. Belmont,
California: Dickenson Publishing Company, 1966.
 SELECTED BIBLIOGRAPHY 463

Sosio, Libero, ‘Galileo e la cosmologia’, in G. Galilei, Dialogo sopra i due massimi

sistemi, edited by L. Sosio.
Staal, J. F. (ed.), ‘Formal Logic and Natural Languages (A Symposium)’, Foundations
of Language 5 (1969), 256—84.
Stein, Howard, ‘Maurice Clavelin on Galileo’s Natural Philosophy’, British Journal for
the Philosophy of Science 25 (1974), 375—97.
Strawson, P. F., Introduction to Logical Theory. London: Methuen, 1952.
Str^mholm, P., ‘Galileo and the Scientific Revolution’, Inquiry 18 (1975), 345-53.
Symposium Internazionale di Storia, Metodologia, Logica e Filosofia della Scienza.
Atti. ‘Galileo nella storia e nella filosofia della scienza’, Manifestazioni celebrative del
IV centenario della nascita di Galileo (Florence-Pisa, 1964). Collection des Travaux
de l’Academie Internationale d’Histoire des Sciences, 16. Florence: Gruppo Italiano
di Storia delle Scienze, 1967.
Tarski, Alfred, Introduction to Logic and to the Methodology of Deductive Sciences.
New York: Oxford University Press, 1965.
Thomason, R. H. (ed.), Formal Philosophy: Selected Papers of Richard Montague. New
Haven, Connecticut: Yale University Press, 1974.
Toulmin, Stephen, Human Understanding. Vol. 1, The Collective Use and Evolution of
Concepts. Princeton, New Jersey : Princeton University Press, 1972.
Toulmin, Stephen, The Philosophy of Science: An Introduction. New York: Harper and
Row, 1960. 1st edition 1953.
Toulmin, Stephen, The Uses of Argument. Cambridge: Cambridge University Press,
1964. First published, 1958.
Ueberweg, F„ History of Philosophy. Vol. U, Modern Philosophy. Translated by G. S.
Morris from the 4th German edition. New York, 1876.
Uzdilek, S. M., ‘Galileo Galilei, The Founder of Experimental Philosophy’, in Symposium
Internazionale di Storia, Metodologia, Logica e Filosofia della Scienza.
Vlastos, Gregory (ed.), The Philosophy of Socrates. Garden City, New York: Doubleday,
1971.
Wallace, William A., ‘Galileo and Reasoning Ex Supposition: The Methodology of the
Two New Sciences', in PSA 1974, edited by R. S. Cohen etal., pp. 79-104. Boston
Studies in the Philosophy of Science, Vol. 32. Dordrecht, Reidel, 1976.
Wallace, William A., ‘Galileo Galilei and the Doctores Parisienses', in New Perspectives on
Galileo, edited by R. E. Butts and J. C. Pitt, pp. 87-138.
Wallace, William A., ‘Mechanics from Bradwardine to Galileo', Journal of the History of
Ideas 32 (1971), 16-28.
Wason, P. C., The Drafting of Rules’, New Law Journal 118 (1968), 548-9.
Wason', P. C. and P. N. Johnson-Laird, Psychology of Reasoning: Structure and Content.
Cambridge, Massachusetts: Harvard University Press, 1972.
Weinreich, U., ‘Explorations in Semantic Theory’, in Current Trends in Linguistics, Vol.
Ill, Theoretical Foundations,edited by T. A. Sebeok, pp. 395-477. The Hague, 1966.
Westman, Robert S. (ed.), The Copemican Achievement. Berkeley, California: Univer¬
sity of California Press, 1975.
Wisan, Winifred L„ ‘Galileo’s Scientific Method: A Reexamination’, in New Perspectives
on Galileo, edited by R. E. Butts and J. C. Pitt, pp. 1-57.
Wisan, Winifred L., The New Science of Motion: A study of Galileo’s De motu locali,
Archive for History of Exact Sciences 13 (1974), 103-306.
 INDEX

Abstraction: and concretion, 299

Abstract vs concrete reasoning, 258, 259, 269
Acceleration: falling bodies, 53-54; importance, 80; continuity, 80-81; contrasted to
speed, 81-82; acceleration motion, 351-52
Accuracy of reconstructions, 339-41
Active evaluation: 425; defined, 339; of reasoning, 350, 427; vs method of counter¬
example, 374-75; and invalidity, 430
Adams, Ernest W., 291n<?, 4 llnid
Ad hoc\ theories, 129—30; procedure, 190; arguments, 198, 284
Ad hominem argument: 131-32, 231-32, 368-70; as a so-called fallacy, 48-48, 334;
in 17th century meaning, 58,402-03;and invalidity, 430
Aesthetic content, 435
Aesthetics, 5
Aesthetics and method, 190
Affirming the consequent: 296, 336, 385-86; vs affirming the antecedent, 378-79
Agassi, Joseph, 73, 94, 202-03
Agnew, Spiro, 342n2
Air: unable to conserve acquired impulse image, 52—53
Aiton, E. J.: on Galileo’s tidal theory, 74—76
Alleged indeterminacy of argument in Dialogue, 13—15
Alternative conclusion method: vs method of counterexample, 375
Alternative explanation criticism, 388-89
Ambiguity: in reasoning experiments, 266-67; and rhetoric and logic, 288; and invalid¬
ity, 427
Analogy, 388
Analogy arguments, 199, 295, 296, 389-91
Analytical Reconstruction: vs analytical summary, 30
Anarchism: 203; and judgment, 165n25, 200; distinguished from counterinductivism,
165n25, 200n75; vs method, 188-89;and dialectical methodology, 200
Anderson, A. R., 291n<S, 411nJ0
Annual motion, 39-42, 357
Anthropocentrism: 109-110, 131, 328, 329; and reasoning, 175-76
Anthropological approach, 187
Anti-Copernican decree of 1616, 6-8, 16-17
Anti-Tycho, 373
Apodicticity, 173
A posteriori argument, 372-77
Apparent vertical fall, 242-43
Appeal to fear: how answered, 47
Appeal to force, 334, 337

464
 INDEX 465

Appeal to pity, 334, 337

Application of formal logic, 300
Apriorism: 205; vs reasoning, 176, 202; vs begging the question, 209; alleged necessity
of, 211 —13; as an alleged assumption of arguments, 213-16
Aquinas, St. Thomas, 224
Achimedes, 59
Argument against the man, 296, 297
Argument from authority, 295, 296, 297
Arguments: sound vs unsound, 126; vs rhetoric, 126; argument analysis, 302; as units of
reasoning, 311; and propositions, 312; propositional components, 312; complex
(defined), 313-14; vs nonarguments, 337
Ariosto, Ludovico, 64
Aristarchus, 128
Aristotle: Galileo as an Aristotelian, 22; classification of motions, 33-34; authority of,
35, 48, 113, 128, 174; concept of natural motion, 44; no cannons at his time, 51;
and rhetoric, 70; authority as a logician, 107-08, 226; his actual procedure, 108-09
Armatured loadstones, 136—37
Art history and criticism, 421
Astronomy: computational vs philosophical, 129, 178; instruments, 133; and reasoning,
178
Asymmetry between favorable and unfavorable evaluation, 340-41
Augustine, Saint, xiii
Authority: role in natural philosophy, 108—09; and reasoning, 174. See also Aristotle

Backward step in scientific inquiry, 190

Bacon, Francis, 152
Bar-Hillel, Yehoshua, 297—99, 301, 303
Beardsley, Monroe, 333
Begging the question: 193, 195, 228—29, 275—77, 286, 356, 393, 394—95;vs apriorism,
209; Antiphon’s speech on murder of Herodes, 275—76; and invalidity, 429
Belnap, N. D., 291n<9, 412n70
Bellarmine, Cardinal, 17
Bible, 14, 18,41, 130, 174
Birds argument, 119,243—44,403—06
Boccadiferro, Ludovico, 204
Bodies, vs parts of bodies, 358-59
Bonamici, Francesco, 204
Br^hier, E., 152
Brewster, Sir David, 152
British Association for the Advancement of Science: Presidential Address by Arthur
Schuster, 67
Buccafiga, Lodovico, 204
Burtt, E. A., 152
Burstyn, Harold: on Galileo’s tidal theory, 74-76

Card experiments, 258-59

Catholic Church, xiv, 6-12
466 INDEX

Cause: 135-36; causal investigation, 138; and effect, 138-39; and reasoning, 171-72;
causal arguments, 295, 296
Celestial inalterability, 320—25
Celestial mechanics, 302
Celestial-terrestrial dichotomy, 34-35, 359
Center of celestial revolutions, 128
Center of universe: argument, 108
Centrifugal force, 120—122 passim
Centrifugal force argument, 38, 407-11
Charity, Principle of, 291n<5, 240—41, 378-79, 425, 430
Chairamonti, Scipione: 211; ridiculed by being made a star, 56; his argument about the
novas ridiculed, 56; Galileo’s rhetorical ambivalence, 56-57, 59—60; his alleged
deception, 58-59; pitied by Galileo, 59-60
Christ, Jesus, 14,130
Churchill, Winston, 334
Circular inertia, 400-401
Circular motion, 87—92, 379—80, 380—82
Circular reasoning: 330, 382-84; vs aphorism, 209; and question-begging, 285-88; and
invalidity, 429
Classification of fallacies, 334-37
Clavelin, Maurice: 72, 225, 437; account of Dialogue as physics, 28; synthetic character
of his interpretation, 161-63; epistemological interpretation of Galileo’s logical
remarks, 246-53
Clavius, Christopher, 203
Classical physics: versus contemporary physics, 71-72
Classics, xiii
Coercion and propaganda, 185
Cohen, I. Bernard, 67
Cohen, L. Jonathan, 305n 1
Cohen, Morris R., 333
Combination of reason and observation, 162
Comets, 373
Complexity of reasoning: 426; complex arguments and critical ones, 419-20; primacy
of complex structure, 428
Composition of motion, 36-37
Comprehensibility: and reasoning, 171-72
Conceptualization: conceptual frameworks, 107; and aphorism, 176-77; and reasoning
176-77
Conclusions: 313; vs reasons vs causes, 346—49
Concomitant variations, method of, 140
Concreteness: 150; and Socratic sincerity, 155-56; in Feyerabend’s approach, 182;
vs abstract reasoning, 271n5; abstraction and concretion, 299
Conditionals: empirical inadequacy of material conditional, 257; standard vs nonstandard
truth-table of material conditionals, 267-68; material conditional, 289-90, 365;
material vs other types, 291n<5; material vs inferential, 383, 397, 411n70; and formal
implication, 411 ni 0
Conservation of motion, 36-37, 117, 216-17, 319, 389-90, 400-01
 INDEX 467

Content vs form in reasoning, 257—58

Context of discovery: vs context of justification, 94-95
Context of justification: versus context of discovery, 94-95
Contextual method: vs linear history of science, 76-79; in logical theory, 305
Continuity of acceleration: Galileo’s view, 80—81 ;Shapere on, 91-92
Contrapositive, 390, 391
Contrariety argument, 34, 108, 317 — 18, 357—72
Contrary motions, 357-58
Conversion: as a logical operation, 258, 268, 391
Copernicanism: decree of 1616, 3, 275 Justification, 22—24; as an element of Dialogue,
44; Dialogue viewed as chapter of, 68-70
Copernican revolution, xviinS, 4
Copernican system: Galileo’s musical metaphor, 60—61
Copernicus: 13, 14, 19, 20, 231; on heliocentrism being more fitting, 126; insight into
heliocentrism, 128-29; third motion, 134
Copi, Irving M., 334, 336
Copleston, Frederick, 152
Cotunio, Giovanni, 211
Counter-example method: 409 — 10; and invalidity, 430
Counter-examples: and negative evaluation, 423-24
Counterinduction: 158, 184, 190; distinguished from anarchism, 165n25, 200ni5
Cretan liar’s paradox, 226-27
Critical arguments: 421; vs constructive arguments, 418
Criticism: 132; and understanding, 179; and reasoning, 179; of reasoning, 303-05, 332
Croce, Benedetto: 338; synthesis of theory and practice, 104; conception of philosophy,
151-52, 154; on Galileo as a philosopher, 152; Crocean interpretation of Socrates
and Galileo, 163—64
Curiosity, 134—35

Darrow, Clarence, 334

Darwin, Charles, 303
Dead cat falling: image exploited by Galileo, 56
Deception of the senses, 38—39, 124-25, 169—70, 195—96, 215 — 16, 394—95
Deduction: deductive arguments, 335; fallacies, 336;invalidity, 394; vs reasoning, 433—
34
Definition by genus and difference, 244
Degenerate cases of invalidity, 428-29
Demonstration ex suppositione, 158
Demonstrative proof: science equated to, 69-70
Density and rarity, 358
Denying the consequent, 198
Descartes, Rend, 152, 204
Description vs prescription in logical theory, 305
Dewey,John, 338
Dialetical methodology: 190; and anarchism, 200; defined, 436
Dialogue', analytical summary, 30—31, 33—44; logical outline, 32—33, dramatic
structure, 46-47; methodological outline, 105-106; methodological analysis and
468 INDEX

reconstruction, 106—141; frequency of philosophical discussions, 145; variety of

individual philosophical discussions, 145-47; main philosophical themes, 145, 147—
48; distribution of philosophical discussions, 148-49; 1632 edition typographical
errors, 234-35; text corrections by Favaro, 234-35; 1st sentence mistranslated by
Drake, 239; index of main logical topics, 415-16; index of individual logical topics,
414-415
Diodati, Elia, 16
Discovery vs justification, 94—95, 204-05
Distinctions: conceptual and empirical, 290
Diurnal motion of earth: arguments in favor, 35-36, 113-14; classical objections, 36-
38; 17th century objections, 38-39
Diurnal motion of heavens: ridiculed, 51
Double-distance rule, 38, 123
Drake, Stillman: 72, 205, 225, 246, 412n13, 437; translation of Dialogue, 29, 32—33;
on Galileo’s inertial views, 88-89; on Galileo studies vs Galilean studies, 204-205;
improvements in his translation, 237-45; popular appeal of his translation, 254n21

Earth: natural motion of, 232—33, 239—40

Earth-heaven distinction, 34-35,47, 359
Earth’s motion: absolute vs hypothetical vs indeterminate discussion, 17-18; proof,
69-70; Aristotle’s argument from violent motion, 115; Aristotle’s argument from
vertical fall, 116; Aristotle’s argument from natural motion, 115-16; gunshots
objection, 118, 235-36; cause of rotation, 124; point-blank gunshots objection,
206; east-west gunshot objection, 208; natural motion objection, 208; ship experi¬
ment objection, 208; parallax objection, 227-28, 231-32, 235-36, 236-37;
impossibility of multiple motions, 229-30; birds objection, 233-34; diurnal vs
annual, 235-36; centrifugal force objection, 236, 237; falling bodies objection,
240—243; arguments and their logic, 301-02
East-west gunshot argument, 221, 391-94
Editing of reasoning text, 232-37
Einstein, Albert, 67, 97, 158
Elements of reasoning, 416-18, 429-30
Ellis, Albert, 421
Empiricism: empiricist interpretation of Galileo, 159; as the import of aphilosophical
historians, 166n30; critical, 216
Encoding and decoding of reasoning, 300
Engineering as a model science, 302
Entailment: vs material conditionals, 291n<9. See also Conditionals
Envelope experiments, 259
Epistemological modesty, 177
Epistemological realism, 21
Equivocation: 209, 354, 370-71, 376, 379-80, 410-11, 430; and invalidity, 427-28;
and complex structure, 428
Error analysis, 202-03, 222
Erudition: erudite commentary on reasoning, 225; and logic 253
Euclid, 203, 224
Evaluation: vs understanding of reasoning, 311, 338-41; of reasoning, 332-41; active,
 INDEX 469

339; logical, 349-53; and reasoning, 376, 424; categories, 416-18; methods, 416-
18; evaluative nature of reasoning, 425; evaluative nature of understanding, 425; and
invalidity, 430
Evil: and fallacies, 338
Existence of God: Hume’s arguments, xv
Experience: and reasoning, 168—71
Experimentation, 110, 251
Explaining erroneous reasoning: vs describing, 409—10
Explanation: 429; and reasoning, 171—72; as rational reconstruction of phenomena,
246-47; logic vs methodology, 249; vs inference, 398-99
Explanatory arguments: 374, 388-89, 394, 425-26; vs explanations, 398-99
Explicit and implicit philosophy, 104
Explicit propositional structure: defined, 327

Fallacies: 332-38, 424, 438; equivocation, 115; concept of, 332, 337; composition,
356, 427; irrelevant conclusion, 427
Fall from moon, 122—23
Falling bodies: objection to earth’s motion, 36; acceleration, 122-23; law of, 204
Falsifiability, 422-23
Favaro, Antonio, 29, 32-33, 72, 225, 232-37
Fearnside, W. Ward, 333
Feeling: and reasoning, 176
Feynman, Richard P., 69
Feyerabend, Paul: 4, 44, 182, 436; on incommensurability, 24n<5; on scientific role of
art and of rhetoric, 66; on propaganda for Copernicanism, 70-71; on Galileo’s
counterinductivism, 159-60; distinction between counterinductivism and anarchism,
165n25\ on scientific rationality, 182-87; on Galileo in general, 188-92; on the
tower argument, 194-96; and Koyrd, 203; on the ship experiment argument, 254n25
Ficino, Marsilio, 203
Field-dependence of validity and structure, 304
Fifth element: ridiculed, 50—51
Final conclusion: defined, 314
Final reason: defined, 314
Flight of birds, 37—38
Fluid properties of water, 43
Fogelin, Robert J., 342n2
Forced motion argument, 312—13
Formalism: and reasoning, 175
Formal logic: 293, 366, 437; critiques, 256—57; and actual reasoning, 269; and mathe¬
matics, 273; and natural language argumentation, 297-99; and everyday arguments,
300; formal validity, 364, 365; formal invalidity, 425
Frege, Gottlob, 150
Freud, Sigmund, 303

Galilean theory of rationality, 434

Garin, Eugenio, 203—204, 205
Geisteswissenschaften, 3
470 INDEX

General hypothesis: testing of, 258-69

General logic: possibility of, 304, 305
General relativity: similarity with argument by Galileo, 67
Genus and difference in definition, 244
Geocentric universe: compared to lazaretto in middle of town, 57—58
Gilbert, William, 61, 135-36, 137
Gillispie, Charles C., 203
God: omnipotence, 8-10, 141; divine vs human understanding, 112
Gombrich, E. H., 421
Gordian knot: compared to parallax objection, 61
Grand Duke of Tuscany, 6,12
Greatest scandal in Christendom, xiv, 4
Grosseteste, Robert, 158
Groundlessness: 389; and invalidity, 380,427
Growth of knowledge, 73
Gunshot arguments: 37; wittily described, 51

Hall, A. R., 246

Hamblin, C. L., 341ni,418
Heavens: unchangeability, 49-50, 108, 372-79;cause of heavenly motion, 124;perfec-
tion, 377-79
Hegel, Georg W. F., 151
Heliocentrism of planetary motions, 40
Hesse, Mary, 182, 305n7
Hintikka, Jaakko, 306nJ?0
History: Orgeta’s comparison with physics, xvi-xvii; historical accuracy vs scientific
relevance, 68-73; history of science versus history of philosophy of science, 70;
historical versus scientific attitude, 72-73; historiography of philosophy, 150-54;
history of philosophy vs history of science, 153; historical attitude, 222; vs logic,
224-25; uncontextual interpretations of historical method, 246
Hobbes, Thomas, 152
Holther, W. B., 333
Hopkins, Harry, 334
Horizontally latent structure: defined, 327
Human understanding: 112; and the divine understanding, 252
Hume, David: xv, 152, 154, 224; on induction, 194
Hunter’s argument, 398-99
Husserl, Edmund, 152
Hypotheses, 7—8
Hypothetico-deductive arguments, 296
Hypothetico-deductive explanation, 10-12

If-therefore, 361—63
Ignorance: 140; and reasoning, 172, 177
Ignotum per aeque ignotum, 284-85
Ignoratio elenchi, 409
Implication: vs support, 384, 385-86
 INDEX 471

Implicit vs explicit philosophy, 104

Incommensurability, 24n6, 186-87
Incomplete arguments: 403, 411; and invalidity, 427
Inconceivability: and reasoning, 172
Independent-mindedness: and reasoning, 174
Induction: 429; problem of induction, 10; inductive logic, 293-97; induction by
enumeration, 294, 295, 296, 297; inductive arguments, 294-95, 336; inductive
correctness, 295, 297; inductive form, 296; inductive fallacies, 335; inductive invalid¬
ity, 394,426
Inertia, 87-92, 207, 289
Infinitely progressing arguments: and invalidity, 428
Insight: 128-29; and reasoning, 170, 178
Insipid subtleties: dismissed by Galileo, 54
Intellectual cowardice, 134—35
Interdisciplinary reasoning, 304-05
Intermediate proposition: defined, 314
Invalidity: deductive vs inductive vs rhetorical, 394; types of, 403; general concept
defined, 425; structural, 426; constructive, 427; semantical, 428; degenerative, 429
Irrationalism, 185—86, 274
Isotropy of space: Galileo on, 84

Jaspers, Karl, xiii

Johnstone, Henry W., Jr., 24n<5, 418,419, 420
Johnson-Laird, P. N., 257-70, 303
Jones, W. T., 152
Joshua Biblical passage, 14, 130
Judgment: 150;vs method, 156; and anarchism, 165n25,200
Justification vs discovery, 94-95, 204-205

Kahane, Howard, 342n2

Kant, Immanuel, xv, xvi, 152
Kepler, Johann: 127, 303; third law, 99n30
Kidd, Thomas I., 334
Knowledge: unconscious, 118
Koertge, Noretta, 422
Koestler, Arthur: 70; on Galileo’s tidal theory, 77
Kouznetsov, Boris, 67—68
Koyr^, Alexandre: 72, 159, 162, 238, 246, 436-37; on aspects of Dialogue, 28; on
scientific content of Dialogue, 68; on Galileo’s inertial principle, 89; pioneer of
professional history of science, 202; on Galileo’s apriorism, 202-222; and Feyera-
bend, 203; as a step forward, 222
Kreyche, R. J., 333
Kuhn, Thomas, 24n<5, 94, 202

Lakatos, Imre, 189

Latent structure: defined, 327
Latent propositional structure: defined, 327
472 INDEX

Lavoisier, Antoine, 303

Law of Clavius, 363
Law of squares, 38
Legal arguments: vs mathematical reasoning, 274
Leibniz, Gottfried W., 152
Levels of validity, 423
“Letter to the Discerning Reader”, 6-8
“Letter to the Grand Duchess Christina”, 22
Liar’s paradox, 366-67
Linguistic expression of reasoning, 429—30
Linguistics of reasoning, 297-99
Literary art: as having a role in science, 66
Loadstones, 15,42, 61-62, 135-36
Locke, John,152
Logic: and science, 23-24; logical point of view, 29, 167;Galileo’s appreciation of, 44;
theory vs practice, 107—08, 224—25, 226, 299—302, 344—45; logical analysis, 108,
161, 203, 222, 275, 437; as the theory of reasoning, 167-68; vs methodology, 170;
logicism and methodologism, 175; logicism vs rationalism, 177; vs history, 224-25;
historiography of logic, 224-25; history of, 224-25; and mathematics, 252-53,
257; logical structure, 261-66, 269; and rhetoric, 273-77; logical probability, 293;
logical evaluation, 349-5 3; form of arguments, 410
Luminosity of heavenly bodies, 126
Lyncean Academician, 3, 12, 46
Lyons, John, 298

Mach, Ernst: on Galileo’s theory of tides, 74; criticism of Galileo misinterpreted by

Shea, 77-78
Machamer, Peter: on ship experiment argument, 254n25
Machiavelli, Niccolo, 224
Magnetism: figurative use of the term, 62. See also Loadstones
Mahoney, Michael, 246
Maieutic approach, 155
Man: as a rational animal, 230-31
Marx, Karl, xiii, 151, 224
Material conditional. See Conditionals
Mathematics: mathematical reasoning, 106; certainty, 136, 173-74; and reasoning,
173-74; Platonism, 203-204; mathematical intelligibility, 246; and physics, 250—
52; and logic, 252-53; mathematization in physics and in logic, 305
Mazzoni, Jacopo, 205
McMullin, Ernan, 20—22, 68-70
Measurement in astronomy, 127-28
Mechanical properties: primacy of, 84
Medicine, as a model science, 302
Medieval content of Dialogue, 71
Merton, Robert K., 203
Meta-arguments: 343; vs object arguments, 339, 341
Metaphorical use of language by Galileo, 238
 INDEX 473

Metaphysics: Kant’s analogy with physics, xvi

Methodology: methodological content, 95-96, 435-36; conceptions of, 103, 175, 188;
and systematization, 103-104; versus substantive claims, 108-09, 135, 141; dis¬
covery vs justification, 109; cognitive awareness, 111-12; error, 112; authority,
112-13; independent-mindedness, 112-13; epistemological modesty, 113;academic
opportunism, 113; probability, 113-14; simplicity, 113-14; logical analysis, 114-
16; rational-mindedness, 114-15; open-mindedness, 114-15; experience, 116-17;
reasoning, 116-17, 118, 175; Socratic method, 116-17; understanding, 117, 120;
digressions, 118; hypothetical vs circular reasoning, 118; knowledge, 118, 120;
critical experimentalism, 119; experiments, 119; criticizing, 120; recollection, 120;
mathematics and physical reality, 120-21, 122-23; causal investigations, 121-22;
comprehensibility, 123-24; truth, 123-24; inconceivability, 125; primacy of nature
over man, 125-26; logical distinctions, 126; superficiality of abstract answers, 126—
27; method-explanations, 160, 189; method of concomitant variations, 172; anarchy,
188—89; method of counterexample, 374—76; method of alternative conclusion,
430; method vs rhetoric vs reasoning, 432-34
Milton, John, xiv
Mirror experiments, 110
Mixed motion argument, 210-211, 213, 240-43, 387-89
Model-theory, 423-24
Modus ponens, 386
Modus tollens, 198, 215, 397
Moon: similarities with earth, 35; visibility of surface, 85-86; mountains, 86-87, 110;
large dark spots, 87; secondary light, 87, 112;and earth, 110; telescopic appearance,
111; reflection of light, 111-12; life on, 112
Montague, Richard, 299, 306n30, 340
Multiple natural motions, 39, 125

Nagel, Ernest, 333

Naive realism, 194
Natural interpretations, 195
Natural language argumentation, 297
Natural motion: 33, 359—60, 379-80; and circular motion, 44, 350-53; Aristotle’s
theory, 107;geostatic objection, 349-53;as spontaneous or nonviolent, 353
Negation: and reasoning, 257-58
Negative evaluation, 421
New rhetoric, 273—77, 293, 296
Newton, Isaac: 158, 246, 303; references to Galileo, 67; Newtonian interpretation of
Dialogue, 71; similarity with Galileo, 73; theory of tides, 75; methodological imita¬
tion of Galileo, 96—97
Nietzsche, Friedrich W., 155
Nobile, V.: on Galileo’s theory of tides, 74
North-south gunshot argument, 208
Novas: 373; nova of 1572, 39-40, 127, 128
Number mysticism, 106
Number three: alleged special status, 346-49
Numerology, 203—204
474 INDEX

Object arguments, 343

Objectivity: 130-31;and reasoning, 176
Observational-theoretical distinction, 132-33, 183-84
Occult qualities, 140, 172
Only, 371-72
Open-mindedness, 134-35, 177
Openness to criticism, 422
Openness of reasoning, 420—21,426
Oregius, Augustinus, 10
Orlando Furioso, 64
Ortega y Gasset, Jos6, xiv, xvi, 152

Pagnini, Pietro, 29, 32-33

Paper-reflection experiments, 110
Paradox: liar’s, 226—27
Paradoxical proof of earth’s motion: 201n79, 290, 396-97
Parallax: 1572 nova, 40; fixed stars, 42; geostatic argument compared to Gordian knot,
61;data, 127;annual, 132-33;geostatic argument, 245, 312, 386
Paralogisms, 207—211, 386
Parts vs wholes, 137—38, 239-40
Paul VI, Pope, xiv
Pendulum vibrations, 123
Perelman, Chaim, 274-76, 288, 293
Perfection: of heavens, 110; of universe, 346-49
Peripatetics: image ridiculed, 55; unprosaically called “animals”, 60, 230-31; explana¬
tions in terms of sympathy, 62
Perpetual motion, 379-80
Persuasion: and invalidity, 430
Petitio principii. See Begging the question
Philosophy: and methodology, 95-96; implicit vs explicit, 104, 148, 150; systematic
vs unsystematic, 163—64
Philosophy of science: Galileo’s unique position, 157—58; Galileo’s vs Galilean, 163;
anthropological approach, 182-83; phenomenological method, 182-83; historical
approach, 183; normative approach, 183
Physical arguments: and meta-arguments, 343-44
Physico-mathematical synthesis, 169, 73
Physics and astronomy, 63
Pico, Francesco, 204
Piety, 6-7
Planes, and spheres, 121
Plato: xiii, 224, 334; Platonist interpretation of Galileo, 77, 159; Platonic mathe-
maticism, 106; Platonism, 203—204
Point-blank gunshot argument, 208, 240, 399-403, 412n77
Pope Urban VIII: his favorite argument, 9—10
Popper, Karl: 158, 182, 422;his three worlds, 181; on induction, 294
Posterior Analytics, 20, 22
Post hoc ergo propter hoc, 336-37
 INDEX 475

Practical approach to logic, 341

Predicate calculus: empirical inadequacies, 258
Prediction: and explanation, 247
Premise, 313
Primum mobile, 380-82
Probability: probabilism, 21; and reasoning, 178
Procedure vs results, 128
Projectiles: objection to earth’s motion, 37; properties, 117; projectile motion, 390;
vs birds argument, 403—404
Propaganda: in the Dialogue, 4; in Galileo’s view of circular motion, 88—89; and
coercion, 185; and rhetoric, 200
Propositional calculus: empirical inadequacies, 257-58
Propositions: and propositional components, 312; propositional structure defined, 314;
propositional structure and invalidity, 430—31
Psychology: psychology of science, 181-82; psychology of reasoning, 270, 293, 437-
38; rational-emotive psychotheraphy, 421-22
Pseudo-irrationalism, 191, 203
Ptolemaic system: compared to lazaretto in midtown, 126
Punctuation: significance of, 233-34; logic of, 404-06
Pythagorean mathematicism, 106

Qualitative conclusion vs quantitative data, 127, 178

Quantitative invalidity, 407, 409
Question-begging arguments. See Begging the question

Ratiocinative nature of evaluation, 424

Ratiocinative nature of understanding, 425
Rationalism: and irrationalism, 7; rationalist interpretation of Galileo, 159; and logicism,
177; vs reasoning, 170, 176, 177, 202
Rationality: 422—23; role of rhetoric, 202; role of method, 202; human and scientific,
432; macroscopic vs microscopic features, 432; macrostructure vs microstructure,
432-33
Real: and rational, 338
Reasoning about reasoning, 301
Reasoning experiments: explanation by Wason and Johnson-Laird, 259-61; formal logic
explanation, 269; argument-tests vs answer-tests, 270
Reasoning ex suppositione, 139
Reasoning indicators, 311
Reasons: nonrational vs irrational, 185; as element of arguments, 313; independence vs
interdependence, 322-23; vs causes, 349
Reconstruction of arguments: defined, 318-19
Reductio ad absurdum, 56
Reform of logic, 439
Relativism, 433
Relativity of motion: 195-96, 217-19, 392—93;illustrated by painter on ship, 54;held
as an approximation by Galileo, 76; and principle of composition, 217; mechanical
vs optical, 219—20; supported not assumed by Galileo, 220—21
476 INDEX

Results vs procedure, 128

Retrograde planetary motion, 40
Rhetoric: and science, 4, 66; different concepts of, 23-24; form vs content, 46-47,
434-35; lack of rhetoric vs bad rhetoric, 56; rhetorical force defined, 65; logical and
emotional aspects, 65-66; vs arguments, 126; and method, 190; rhetorical considera¬
tions, 197-99; and logic, 200, 279, 282-83, 286, 370-71;rhetorical invalidity, 394;
vs method vs reasoning, 432—34
Rhetorical interpretation of Dialogue, 44-45, 70-71
Roland: character in Ariosto’s poem, 64
Ryle, Gilbert, 420

Sacrobosco, 138
Salmon, Wesley, 294—97, 333, 336
Salusbury, Thomas, 21
Santillana, Giorgio de: 21, 27, 70, 219, 225, 228; omissions in his edition of Dialogue,
452n3
Science: and rhetoric, 4; scientific rationality, 4-5, 180-87, 256; different concepts
of “scientific”, 23-24, 95-96; scientific relevance vs historical accuracy, 68-73;
scientific content, 93-94, 434-35; change vs progress vs rationality, 180-82, 199—
200; scientific change, 180-82; scientific progress, 180-82; scientism, 222; scientific
study of reasoning, 273, 437; science of logic, 301, 302; and logic, 439
Scriven, Michael, 300—302, 303, 340, 342n2, 437
Self-contradictory arguments: 363-66, 384-85; and invalidity, 428—29
Semantics, 298
Semicircular fall: 118; digression within digression, 53-54; viewed from post-classical
physics, 99n29
Sense experience, 107, 128—29
Shapere, Dudley: 161; on Galileo’s inertial views, 89-92
Shea, William: 70; evidence about conclusiveness of tidal argument, 16; account of
content of Dialogue, 28;on Galileo’s tidal theory, 76-78
Ship argument: 36, 116-17, 197-98, 199, 211, 213-14; 243, 254n25, 316, 319, 387,
389-91; Galileo’s boast, 51-52; Simplicio’s incredulity, 51-52; ship and earth
analogy, 243, 397; and vertical fall, 250-51
Simple motion: and natural motion, 349-53
Simple vs complex arguments, 418
Simplicity: 128-29, 133-34, 246; and reasoning, 177; different concept of, 178; and
the apparent motion of sunspots, 248—49; and probability, 249-50
Size, concept of, 130-31, 176
Socrates: xiii, 164, 334; Socratic demon, 53; synthesis of theory and practice, 104;
similarity with Galileo, 154-57; Socratic paradox, 154-55
Socratic method, 154-57: 171, 172-73
Solar system: Galileo’s view of origin, 82-84, 99a30
Sophistical arguments, 229-30
Sorites, 226-27
Sosio, Libero: on Galileo’s circular motion, 101n7<§
Spheres, and planes, 121
Spinoza, Benedictus de, 152
 INDEX 477

Spontaneous motion: and natural, 356

Squares: law of, 122—23
Stalin, Joseph, 334
Stars: stellar dimensions, 41, 245; apparent diameter, 130
Statistical syllogism, 295, 296, 297
Steelyard, 122
Stein, Howard, 159, 246
Straight motion: and circular motion, 34, 349—53; up and down motion, 107; natural
motion, 316—17, 325—26
Strauss, Emil, 72, 225-32, 258, 437
Strict demonstration, 18—22
Structure diagrams: rules, 314 — 15; standard labeling, 315
Structure of arguments, 304
Structure of Dialogue', problem of, 27-29
Sunspot motion, 40—41, 84—85, 129—30, 244—45, 373
Sunspots: wittily labeled “dirt”, 47-48; explanation and prediction of apparent motion,
247-49
Support: vs implication, 394
Syllogism: middle term of, 228—29
Symbolic logic: used to analyze reasoning experiments, 261-68
Syntactical vs logical structure, 212n8
Synthesis: of theory and practice, 103, 104, 150; vs eclecticism, 157, 161; synthetic
approach to Galileo, 161; of mathematics, physics, and astronomy by Galileo,
304
Systematization of Galileo’s methodology, 167

Tannery, Adam, 162

Teleology: 109-10, 131; teleological argument, 35,48-49, 320-25, 327-29, 377-79;
and reasoning, 175—76
Telescope: Galileo on, 84—85
Terrestrial-celestial dichotomy, 34-35, 359
Terrestrial rotation, 381-86
Theoretical explanation, 129-30
Theoretical-observational distinction, 132-33, 183-84
Theories vs observations, 132—33, 183-84
Theory of reasoning, 257, 438-39
Thomistic Aristotelianism, 158
Three-dimensionality argument, 106, 346-49
Thought-experiments: as hypothetical reasoning, 169
Tides: 138-40; Galileo’s argument not conclusive, 16-18; cause of, 42-44; some
explanations “fished out” of Aristotle, 62-63; lunar heat explanation ridiculed,
63; explanation of monthly and annual periods compared to situation in Ariosto’s
poem, 63-64; limitations of his view subtly admitted by Galileo, 64; historical and
scientific interpretations of Galileo’s theory, 74-79; and Newton’s gravitational
theory, 75; daily period, 78; reaction in 17th century France to Galileo, 78
Tiring of animals: anti-Copernican objection ridiculed, 58
Toulmin, Stephen, 181, 182, 303-05, 437
478 INDEX

Tower argument: 192—99; vs deception of the senses, 201n79; refutation vs proof of

earth’s motion, 201ni9. See also vertical fall argument
Translation of reasoning, 237-45
Trickery, 195-96
Truth: and comprehensibility, 171-72;and acceptance of propositions, 274
Tunneled earth experiment, 123
Two-motions argument, 380-86
Two New Sciences', historians’ preference, 68
Tycho, Brahe: 41, 131—32, 231 ;Tychonic system, 44

Ueberweg, F., 152, 153

Unconscious knowledge: and reasoning, 172-73
Understanding: and reasoning, 171-72, 424; vs evaluation of reasoning, 311, 338-41;
and evaluation, 425
Universe-center argument, 353-56
Universal circular motion: Galileo’s theory, 107
Unless: 371-72
Urban VIII: argument by, 11
Useless arguments, 360-61, 427

Validity: concept of, 304

Vertical gunshot argument, 208, 394-97
Vertically latent structure, 327
Vertical fall: actual vs apparent, 116,193-94, 198, 201n79
Vertical fall argument: 210, 220-21, 240-43; Strauss on, 228-29; and ship experiment,
250-51; and concept of question-begging, 277-90; detailed textual-logical analysis,
277-82;evaluation, 282-85;reconstruction, 285-88, 319;diagram and latent struc¬
ture, 329-30
Vico, Giambattista, 151
Violent motion argument, 209, 379-80
Vlastos, Gregory, 154

Wason, P. C., 257-70, 303, 305

Weinreich, U., 298
Whirling argument. See Centrifugal force argument
Whitehead, Alfred North, 203
Whole vs parts, 239-40
Winds, 43, 139
Wisan, Winifred: on Two New Sciences, 68
Witticism: about definition by genus and difference, 230-31

Yalta, 334
 BOSTON STUDIES IN THE PHILOSOPHY OF SCIENCE

Editors:
ROBERT S. COHEN and MARX W. WARTOFSKY
(Boston University)

1. Marx W. Wartofsky (ed.), Proceedings of the Boston Colloquium for the Philosophy
of Science 1961-1962. 1963.
2. Robert S. Cohen and Marx W. Wartofsky (eds.). In Honor of Philipp Frank. 1965.
3. Robert S. Cohen and Marx W. Wartofsky (eds.). Proceedings of the Boston Collo¬
quium for the Philosophy of Science 1964-1966. In Memory of Norwood Russell
Hanson. 1967.
4. Robert S. Cohen and Marx W. Wartofsky (eds.), Proceedings of the Boston Collo¬
quium for the Philosophy of Science 1966-1968.. 1969.
5. Robert S. Cohen and Marx W. Wartofsky (eds.). Proceedings of the Boston Collo¬
quium for the Philosophy of Science 1966-1968. 1969.
6. Robert S. Cohen and Raymond J. Seeger (eds.), Ernst Mach: Physicist and Philos¬
opher. 1970.
7. Milic Capek, Bergson and Modern Physics. 1971.
8. Roger C. Buck and Robert S. Cohen (eds.), PSA 1970. In Memory of Rudolf
Carnap. 1971.
9. A. A. Zinov’ev, Foundations of the Logical Theory of Scientific Knowledge
(Complex Logic). (Revised and enlarged English edition with an appendix by
G. A. Smirnov, E. A. Sidorenka, A. M. Fedina, and L. A. Bobrova.) 1973.
10. Ladislav Tondl, Scientific Procedures. 1973.
11. R. J. Seeger and Robert S. Cohen (eds.). Philosophical Foundations of Science. 1974.
12. Adolf Griinbaum, Philosophical Problems of Space and Time. (Second, enlarged
edition.) 1973.
13. Robert S. Cohen and Marx W. Wartofsky (eds.), Logical and Epistemological
Studies in Contemporary Physics. 1973.
14. Robert S. Cohen and Marx W. Wartofsky (eds.), Methodological and Historical
Essays in the Natural and Social Sciences. Proceedings of the Boston Colloquium
for the Philosophy of Science 1969-1972. 1974.
15. Robert S. Cohen, J. J. Stachel and Marx W. Wartofsky (eds.), For Dirk Struik.
Scientific, Historical and Political Essays in Honor of Dirk Struik. 1974.
16. Norman Geschwind, Selected Papers on Language and the Brain. 1974.
18. Peter Mittelstaedt, Philosophical Problems of Modern Physics. 1976.
19. Henry Mehlberg, Time, Causality, and the Quantum Theory (2 vols.). 1980.
20. Kenneth F. Schaffner and Robert S. Cohen (eds.). Proceedings of the 1972 Biennial
Meeting, Philosophy of Science Association. 1974.
21. R. S. Cohen and J. J. Stachel (eds.), Selected Papers of Leon Rosenfeld. 1978.
22. Milic Capek (ed.), The Concepts of Space and Time. Their Structure and Their
Development. 1976.
23. Marjorie Grene, The Understanding of Nature. Essays in the Philosophy of Biology.
1974.
24. Don Ihde, Technics and Praxis. A Philosophy of Technology. 1978.
25. Jaakko Hintikka and Unto Remes, The Method of Analysis. Its Geometrical Origin
and Its General Significance. 1974.
26. John Emery Murdoch and Edith Dudley Sy 11a, The Cultural Context of Medieval
Learning. 1975.
27. Marjorie Grene and Everett Mendelsohn (eds.), Topics in the Philosophy of Biology.
1976.
28. Joseph Agassi, Science in Flux. 1975.
29. Jerzy J. Wiatr (ed.), Polish Essays in the Methodology of the Social Sciences. 1979.
32. R. S. Cohen, C. A. Hooker, A. C. Michalos, and J. W. van Evra (eds.), PSA 1974:
Proceedings of the 1974 Biennial Meeting of the Philosophy of Science Association.
1976.
33. Gerald Holton and William Blanpied (eds.). Science and Its Public: The Changing
Relationship. 1976.
34. Mirko D. Grmek (ed.), On Scientific Discovery. 1980.
35. Stefan Amsterdamski, Between Experience and Metaphysics. Philosophical Problems
of the Evolution of Science. 1975.
36. Mihailo Markovic and Gajo Petrovic (eds.), Praxis. Yugoslav Essaysin the Philosophy
and Methodology of the Social Sciences. 1979.
37. Hermann von Helmholtz: Epistemological Writings. The Paul Hertz/Moritz Schlick
Centenary Edition of 1921 with Notes and Commentary by the Editors. (Newly
translated by Malcolm E'. Lowe. Edited, with an Introduction and Bibliography,
by Robert S. Cohen and Yehuda Elkana.) 1977.
38. R. M. Martin, Pragmatics, Truth, and Language. 1979.
39. R. S. Cohen, P. K. Eeyerabend, and M. W. Wartofsky (eds.), Essays in Memory
of Imre Lakatos. 1976.
42. Humberto R. Maturana and Francisco J. Varela, Autopoiesis and Cognition.
The Realization of the Living. 1980.
43. A. Kasher (ed.), Language in Focus: Foundations, Methods and Systems. Essays
Dedicated to Yehoshua Bar-Hillel. 1976.
46. Peter L. Kapitza, Experiment, Theory, Practice. 1980.
47. Maria L. Dalla Chiara (ed.), Italian Studies in the Philosophy of Science. 1980.
48. Marx W. Wartofsky, Models: Representation and the Scientific Understanding
1979.
50. Yehuda Fried and Joseph Agassi, Paranoia: A Study in Diagnosis. 1976.
51. Kurt H. Wolff, Surrender and Catch: Experience and Inquiry Today. 1976.
52. Karel Kosik, Dialectics of the Concrete. 1976.
53. Nelson Goodman, The Structure of Appearance. (Third edition.) 1977.
54. Herbert A. Simon, Models of Discovery and Other Topics in the Methods of Science
1977.
55. Morris Lazerowitz, The Language of Philosophy. Freud and Wittgenstein. 1977.
56. Thomas Nickles (ed.), Scientific Discovery, Logic, and Rationality. 1980.
57. Joseph Margolis, Persons and Minds. The Prospects of Nonreductive Materialism
1977.
58. Gerard Radnitzky and Gunnar Andersson (eds.), Progress and Rationality in Science
1978.
59. Gerard Radnitzky and Gunnar Andersson (cds.). The Structure and Development
of Science. 1979.
60. Thomas Nickles (ed.), Scientific Discovery: Case Studies. 1980.
61. Maurice A. Finocchiaro, Galileo and the Art of Reasoning. 1980.
'
 Date Due

•0ET=*r198*
 Q 175 . B73 v. 61 010101 000
Finocchiaro, Maurice A.,
Galileo and the art of reason

63 0149040
TRENT UNIVERSITY

Q175 .B73 v. 61
Finocchiaro, Maurice A., 1942- —
Galileo and the art of reasoning
: rhetorical foundations of logic
and scientific method / _

iccnmm O / / /->

33o HoQ