2005 Book MathematicsAndTheHistorianSCra PDF

Editors-in-Chief
Rédacteurs-en-chef
J. Borwein
K. Dilcher
Advisory Board
Comité consultatif
P. Borwein
R. Kane
S. Shen
1 HERMAN/KUČERA/ŠIMŠA Equations and Inequalities
2 ARNOLD Abelian Groups and Representations of Finite Partially
Ordered Sets
3 BORWEIN/LEWIS Convex Analysis and Nonlinear Optimization
4 LEVIN/LUBINSKY Orthogonal Polynomials for Exponential Weights
5 KANE Reflection Groups and Invariant Theory
6 PHILLIPS Two Millennia of Mathematics
7 DEUTSCH Best Approximations in Inner Product Spaces
8 FABIAN ET AL. Functional Analysis and Infinite-Dimensional Geometry
9 KŘÍŽEK/LUCA/SOMER 17 Lectures on Fermat Numbers
10 BORWEIN Computational Excursions in Analysis and Number Theory
11 REED/SALES (Editors) Recent Advances in Algorithms and Combinatorics
12 HERMAN/KUČERA/ŠIMŠA Counting and Configurations
13 NAZARETH Differentiable Optimization and Equation Solving
14 PHILLIPS Interpolation and Approximation by Polynomials
15 BEN-ISRAEL/GREVILLE Generalized Inverses, Second Edition
16 ZHAO Dynamical Systems in Population Biology
17 GÖPFERT ET AL. Variational Methods in Partially Ordered Spaces
18 AKIVIS/GOLDBERG Differential Geometry of Varieties with Degenerate
Gauss Maps
19 MIKHALEV/SHPILRAIN/YU Combinatorial Methods
20 BORWEIN/ZHU Techniques of Variational Analysis
21 VAN BRUMMELEN/KINYON Mathematics and the Historian’s Craft: The
Kenneth O. May Lectures
Glen Van Brummelen Michael Kinyon
Editors
Mathematics and the

Historian’s Craft
The Kenneth O. May Lectures
With 91 Figures
Glen Van Brummelen Michael Kinyon
Bennington College Department of Mathematical Sciences
Bennington, VT 05201 Indiana University South Bend
USA South Bend, IN 46634-7111
[email protected] USA
[email protected]
Editors-in-Chief
Rédacteurs-en-chef
Jonathan Borwein
Karl Dilcher
Department of Mathematics and Statistics
Dalhousie University
Halifax, Nova Scotia B3H 3J5
Canada
[email protected]
Mathematics Subject Classification (2000): 01-06, 01A85
Library of Congress Control Number: 2005923503
ISBN-10: 0-387-25284-3 Printed on acid-free paper.

ISBN-13: 978-0387-25284-1
© 2005 Springer Science+Business Media, Inc.

All rights reserved. This work may not be translated or copied in whole or in part without the
written permission of the publisher (Springer Science+Business Media, Inc., 233 Spring Street, New
York, NY 10013, USA), except for brief excerpts in connection with reviews or scholarly analysis.
Use in connection with any form of information storage and retrieval, electronic adaptation, com-
puter software, or by similar or dissimilar methodology now known or hereafter developed is for-
bidden.
The use in this publication of trade names, trademarks, service marks, and similar terms, even if
they are not identified as such, is not to be taken as an expression of opinion as to whether or not
they are subject to proprietary rights.
Printed in the United States of America. (EB)
9 8 7 6 5 4 3 2 1
springeronline.com
To Miriam May, in memory of Ken
Preface
1974 was a turning point for the history and philosophy of mathematics in
North America. After years of planning, the first issue of the new journal
Historia Mathematica was printed. While academic journals are born and die
all the time, it was soon clear that Historia Mathematica would be a major
factor in shaping an emerging discipline; shortly, it became a backbone for a
global network of professional historians of mathematics. In the same year, the
Canadian Society for History and Philosophy of Mathematics (CSHPM) was
founded, adopting Historia Mathematica as its official journal. (In the 1990s,
the CSHPM recognized its broader mission by naming Philosophia Mathemat-
ica as its official philosophical journal, rechristening Historia Mathematica as
its historical journal.) Initially consisting almost entirely of Canadian mem-
bers, the CSHPM has become in practice the North American society for the
scholarly pursuit of history and philosophy of mathematics. The joint estab-
lishment of society and journal codified and legitimized the field, commencing
what has become a renaissance of activity for the past 30 years.
These initiatives were begun by, and received much stimulus from, one
man: Kenneth O. May, of the Institute for History and Philosophy of Science
and Technology at the University of Toronto. May was a brilliant researcher,
but he recognized that the viability of the fledging discipline required ad-
ministrative leadership as well. In the introduction that follows, Amy Shell-
Gellasch, CSHPM archivist, describes May’s life and some of his achievements.
Central to May’s vision of the history of mathematics was the dichotomy be-
tween the role of the historian and the use that a mathematician might find
for history. Mathematical practitioners, for reasons of pedagogy or in order to
contextualize their own work, tend to focus on finding the antecedents for cur-
rent mathematical theories in a search for how particular sub-disciplines and
results came to be as they are today. On the other hand, historians of mathe-
matics eschew the current state of affairs, and are more interested in questions
that bear on the changing nature of the discipline itself. How, for instance,
have the standards of acceptable mathematical practice differed through time
and across cultures? What role do institutions and organizations play in the
VIII Preface
development of the subject? Does mathematics naturally align itself with the
sciences or the humanities, or is it its own creature, and do these distinc-
tions matter? The lead article in this volume, by Ivor Grattan-Guinness, is a
strong statement on what makes history of mathematics unique, and reflects
well May’s own vision for our field.
May passed away, too early, in 1977. However, his legacy lives on partly
through our thriving community; the continued prosperity of the CSHPM,
Historia Mathematica, and Philosophia Mathematica are a resounding testa-
ment to that. In 2002, on the 25th anniversary of his passing, the CSHPM
held a special meeting in May’s honour. One of our actions at this meeting
was to re-christen the keynote addresses at our annual general meetings as
the “Kenneth O. May Lectures”. Each of our annual meetings is a special oc-
casion: while also providing a forum for presentations on all aspects of history
and philosophy of mathematics, each meeting focuses on a specific theme,
with activity revolving around an invited keynote address by a scholar of in-
ternational repute. The diversity of these sessions over the years, witnessed
in the table below, is a clear testament to the breadth and significance of the
CSHPM’s activities.
Since 1988 the CSHPM has preserved a record of the scholarly activities
of the annual general meeting through the production of a volume of Pro-
ceedings, to which all speakers are invited to contribute. These Proceedings,
distributed internally to Society members, are by now a repository of a great
deal of valuable research. Some of these works have appeared elsewhere but
many which deserve wider exposure have not; this volume represents our first
attempt to correct this state of affairs. By printing the Kenneth May Lectures
since 1990, we hope not only to choose some of the finest work presented at
CSHPM meetings but also to present ourselves to the broader scholarly com-
munity. This volume represents by example who we are, how we approach
the disciplines of history and philosophy of mathematics, and what we find
important about our scholarly mission.
Many things happen over fifteen years. The editors attempted to reach all
May lecturers since 1988, but were not wholly successful. Also, some of their
lectures appeared later in formal scholarly journals (which the Proceedings is
not), and some of these later versions incorporated improvements. In these
cases we have chosen to reprint the polished final articles rather than the
original lectures. One implication of this is that the bibliographic standards
vary from article to article, reflecting the different sources in which the arti-
cles appeared. We are grateful to the following organizations that granted us
permission to reprint articles free of charge from the pages of their journals
and books: the Association for Symbolic Logic, the Canadian Mathematical
Society (CMS), the Mathematical Association of America, and Philosophia
Mathematica.
As editors of this volume, we have received a great deal of support from
many people. The CSHPM, both its executive and its members, has been
pivotal in working with us over the past year to produce the best possible
Preface IX
public imprint for the Society. The authors of the papers in this volume and
archivist Amy Shell-Gellasch have combined to produce a truly admirable
body of work. The editors of the CSHPM Proceedings over the years, listed
below, have moved mountains to produce these volumes. Jonathan and Peter
Borwein, editors of the CMS Books in Mathematics, provided highly valued
encouragement and advice. Ina Lindemann, Mark Spencer, and Anne Meagher
of Springer Verlag helped tremendously in bringing this volume to fruition.
Thanks also go to Dennis Richter for technical support. Our families have
sacrificed in their own ways, putting up with late dinners and with occasion-
ally absent parents; we thank them especially for their patience. Finally, our
greatest gratitude is due to the man to whom this volume is dedicated. Ken,
your vision lives and prospers in the 21st century. Without your insight and
formative efforts, the CSHPM might not be here today. Thank you.
Glen Van Brummelen and Michael Kinyon
A note on the title. Ken May considered the practice of the history of math-
ematics to be a unique melding of the crafts of mathematician and historian.
This entails sensitivity both to the mathematical content of the subject, and
to the various contexts in which it can be understood. Our daily work is con-
stantly informed by our attempts to achieve this delicate balance. In Ken’s
words:
“Clearly in historical work the danger in missing the mathematical point
is matched by the symmetric hazard of overlooking a historical dimension.
The mathematician is trained to think most about mathematical correctness
without a time dimension, i.e., to think ahistorically. Of course it is interesting
to know how a historical event appears when viewed by a twentieth century
mathematician. But it is bad history to confuse this with what was meant at
the time. The historian concentrates on significance in the historical context
and on the historical relations between events. And this is equally interesting
to the mathematician who wishes to understand how mathematics actually
developed.
“One could continue indefinitely, but the essential point is that the best his-
tory requires sensitivity to both mathematical and historical issues, a respect
for good practice of the crafts of both the historian and the mathematician. It
may even be that the best mathematical research is aided by an appreciation
of historical issues and results. I know of many instances and hope that the
work of historians may contribute to increasing their frequency.”1
1
Kenneth O. May, “What is good history and who should do it?”, Historia
Mathematica 2 (1975), 453.
X Preface
Annual Meeting Themes & Kenneth O. May Lecturers Since 1990
2003: Maritime Mathematics (Halifax, NS)

– Jim Bennett, Geometry, Instruments and Navigation:
Agendas for Research, 1500-1800
2002: In Memory of Kenneth May (Toronto, ON)

– Ivor Grattan-Guinness, History or Heritage? Historians and
Mathematicians on the History of Mathematics
2001: French Mathematics (Québec, PQ)

– Jean Dhombres, The Applied Mathematics Origins of Lebesgue
Integration Theory and Why it was Read as Pure Mathematics
During the First Years of the 20th Century
2000: History of Mathematics at the Dawn of a New Millennium

(Hamilton, ON)
– Rüdiger Thiele, Hilbert and his 24 Problems
1999: Joint meeting with the British Society for

History of Mathematics (Toronto, ON)
1998: Late 19th-Century Mathematics (Ottawa, ON)

– Volker Peckhaus, 19th-Century Logic:
Between Philosophy and Mathematics
1997: Science and Mathematics (St. John’s, NF)

– Rüdiger Thiele, The Mathematics and Science of Leonhard Euler
1996: Ancient Mathematics (St. Catharines, ON)

– Alexander Jones, Greek Applied Mathematics
1995: Mathematics Circa 1900 (Montreal, PQ)

– Joseph W. Dauben, Cantor and the Epistemology of Set Theory
1994: History of Mathematics in the United States and Canada

(Calgary, AB)
– Thomas Archibald (co-author Louis Charbonneau),
Mathematics in Eastern British North America in the
Nineteenth Century: Some Preliminary Remarks
– Karen Hunger Parshall,
The Emergence of the American Mathematical Research
Community 1876-1900
Preface XI
1993: Philosophy of Mathematics (Ottawa, ON)

– Stuart Shanker, Turing and the Origins of Artificial Intelligence
1992: Ethnomathematics (Charlottetown, PEI)

– Michael Closs, The Ancient Maya: Mathematics
and Mathematicians
1991: Women in Mathematics (Kingston, ON)

– Ann Hibner Koblitz, Women in Mathematics: Historical and
Cross-Cultural Perspectives
1990: History and Pedagogy (Victoria, BC)

– Judith Grabiner, Was Newton’s Calculus a Dead End?
A New Look at the Calculus of Colin Maclaurin
CSHPM/SCHPM Presidents
1974 – Charles V. Jones
1975, 1976 – Viktors Linis
1977, 1978 – J. L. Berggren
1979, 1980 – G. de B. Robinson
1981, 1982 – Wesley Stevens
1983, 1984, 1985 – Edward J. Barbeau
1986 – Marshall Walker
1987 – Louis Charbonneau
1988, 1989 – J. L. Berggren
1990, 1991 – Craig Fraser
1992, 1993, 1994, 1995 – Thomas Archibald
1996, 1997 – Robert Thomas
1998, 1999 – James J. Tattersall
2000, 2001 – Glen Van Brummelen
2002, 2003 – J. L. Berggren
2004, 2005 – Robert Bradley
CSHPM/SCHPM Proceedings Editors

1988, 1989 – Tasoula Berggren
1990 – Francine Abeles, Victor Katz, Robert Thomas
1991 – Hardy Grant, Israel Kleiner, Abe Shenitzer
1992-1999 – James J. Tattersall
2000, 2001 – Michael Kinyon
2002-present – Antonella Cupillari
XII Preface
Copyright Permissions
The following articles, based on May lectures, have appeared previously.
Our thanks go to the respective publishers (Canadian Mathematical Society,
Mathematical Association of America, Kluwer Academic Publishers, Associ-
ation for Symbolic Logic, and Philosophia Mathematica) for granting permis-
sion for us to reprint the papers in this volume. The original copyright holders
retain all rights.
Thomas Archibald and Louis Charbonneau. Mathematics in Canada before
1945: A preliminary survey, in Peter Fillmore, ed., Mathematics in Canada,
vol. I, Ottawa, ON: Canadian Mathematical Society, pp. 1-90. The article
appears in both English and French; only the English version (pp. 1-43) is
reprinted here.
Judith V. Grabiner. Was Newton’s calculus a dead end? The continental in-
fluence of Maclaurin’s treatise of fluxions, American Mathematical Monthly
104 (5) (1997), 393-410.
Ivor Grattan-Guinness. History or heritage? An important distinction in
mathematics and for mathematics education, American Mathematical Monthly
111 (1) (2004), 1-12.
Ann Hibner Koblitz. Mathematics and gender: Some cross-cultural observa-
tions, in Gila Hanna, ed., Towards Gender Equity in Mathematics Education,
Dordrecht: Kluwer, 1996, pp. 93-109.
Volker Peckhaus. 19th century logic between philosophy and mathematics,
Bulletin of Symbolic Logic 5 (4) (1999), 433-450. Copyright held by the Asso-
ciation for Symbolic Logic.
Stuart Shanker. Turing and the origins of AI, Philosophia Mathematica 3 (1)
(1995), 52-85.
Contents
Introduction: The Birth and Growth of a Community

Amy Shell-Gellasch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1 History or Heritage? An Important Distinction in
Mathematics and for Mathematics Education
Ivor Grattan-Guinness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2 Ptolemy’s Mathematical Models and their Meaning

Alexander Jones . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3 Mathematics, Instruments and Navigation, 1600-1800

Jim Bennett . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
4 Was Newton’s Calculus a Dead End? The Continental

Influence of Maclaurin’s Treatise of Fluxions
Judith V. Grabiner . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
5 The Mathematics and Science of Leonhard Euler
(1707–1783)
Rüdiger Thiele . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
6 Mathematics in Canada before 1945: A Preliminary Survey

Thomas Archibald, Louis Charbonneau . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
7 The Emergence of the American Mathematical Research

Community
Karen Hunger Parshall . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
8 19th Century Logic Between Philosophy and Mathematics
Volker Peckhaus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203
9 The Battle for Cantorian Set Theory
Joseph W. Dauben . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
XIV Contents
10 Hilbert and his Twenty-Four Problems

Rüdiger Thiele . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
11 Turing and the Origins of AI

Stuart Shanker . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297
12 Mathematics and Gender: Some Cross-Cultural
Observations
Ann Hibner Koblitz . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 347
List of Contributors
Thomas Archibald Judith V. Grabiner

Dept. of Mathematics and Statistics, Pitzer College,
Acadia University, Wolfville, N. S. 1050 North Mills Avenue,
B4P 2R6 Canada Claremont, CA 91711 USA
[email protected] [email protected]
Jim Bennett Ivor Grattan-Guinness

Museum of the History of Science, Middlesex University at Enfield,
University of Oxford, Middlesex EN3 4SF England
Broad Street, Oxford. OX1 3AZ [email protected]
England
[email protected]
Alexander Jones
Department of Classics,
Louis Charbonneau
University of Toronto,
Département de Mathématiques,
97 St. George Street,
Université de Québec a Montréal,
Toronto, ON M5S 2E8 Canada
Montréal, QC H2X 3Y7 Canada
[email protected]
[email protected]
Joseph W. Dauben Ann Hibner Koblitz

Department of History, Arizona State University
Herbert H. Lehman College, CUNY, P.O. Box 873404,
250 Bedford Park Blvd West, Tempe, AZ 85287-3404 USA
Bronx, NY 10469-1589 USA [email protected]
Ph.D. Program in History, Karen Hunger Parshall

The Graduate Center, Departments of History and
City University of New York, Mathematics, University of Virginia,
365 Fifth Ave, P. O. Box 400137, Charlottesville,
New York, NY 10016-4309 USA VA 22904-4137 USA
[email protected] [email protected]
XVI List of Contributors
Volker Peckhaus
Universität Paderborn, Fakultät für
Kulturwissenschaften – Philosophie,
Warburger Str. 100, D–33098
Paderborn, Germany
[email protected]
Stuart Shanker
Departments of Philosophy and
Psychology,
Atkinson College, York University,
North York, ON M3J 1P3 Canada
[email protected]
Rüdiger Thiele
Karl-Sudhoff-Institut
für Geschichte der Medizin und der
Naturwissenschaften,
University of Leipzig, D-04109
Leipzig, Germany
[email protected]
Kenneth O. May (1915-1977)
Originally appeared in Historia Mathematica 5 (1) (1978), 2.
Reprinted with permission of Elsevier Science.
Introduction: The Birth and Growth of a
Community
Amy Shell-Gellasch
CSHPM/SCHPM Archivist
The Canadian Society for History and Philosophy of Mathematics, or Société

d’Histoire et de Philosophie des Mathématiques, affectionately known as
CSHPM/SCHPM, is a society of those interested in the history or philos-
ophy of mathematics. Our constitution states that “the aim of the society is
to promote throughout Canada discussion, research, teaching and publishing
in the history and the philosophy of mathematics. Any person with interest
in the history or in the philosophy of mathematics is eligible to become a
member.” Those statements are clearly obvious and necessary; however, they
do not convey the depth, breadth or quality of the society and its members.
Currently the society has over two hundred members in nineteen countries,
including Brazil, Sweden, Bangladesh and Japan. Though most of our mem-
bers are academics, some do not work in academia but are simply consumers
of the subject, either personally, as educators, or through professional interest
from other disciplines. The diversity of the CSHPM is also its strength: our
different motives and perspectives combine to produce richer portraits of the
history of mathematics than we could achieve individually.
Our primary goal is to provide our members with the means to both
present and receive current research in the field. This is done primarily
through our Annual Meetings and the resulting Proceedings, as well as through
our official historical journal Historia Mathematica and philosophical journal
Philosophia Mathematica. Our semi-annual newsletters allow members to keep
abreast of events in the field as well as interact with one another. Occasion-
ally we hold joint meetings with our sister organization, the British Society
for the History of Mathematics, as well as with the Canadian Mathematical
Society. Our underlying goal, possibly the more important of the two, is to
establish a community of scholars, practitioners and consumers of the history
and philosophy of mathematics, with all the qualities and interactions that
the word “community” implies.
The groundwork for establishing that community was laid in 1972. In that
year Kenneth May sent letters to several colleagues inquiring into the desire
among practitioners to organize a society in the history and philosophy of
4 Amy Shell-Gellasch
mathematics. The responses that May received show an enthusiastic reception

to the idea. In May 1973 the first meeting of the new organization took place
at the Learned Societies Congress (often shortened to the “Learneds”) at
Queen’s University in Kingston, Ontario. The society was officially formed
the following summer when the society’s constitution was approved.
Kenneth Ownsworth May (1915-1977), at the time of the founding of the
society, was at the Institute for History and Philosophy of Science (IHPS) at
the University of Toronto. In addition he was the editor of the journal Historia
Mathematica, which he officially launched in 1974, having been in newsletter
form for the previous two years. May was an accomplished mathematician,
historian and educator. He studied mathematics and economics at the Uni-
versity of California at Berkeley, receiving his A. B. in 1936, and his M. A.
in 1937. En route to his doctorate under Griffith Evans, his life took many
turns. At the recommendation of Evans, May became a fellow of the Institute
of Current World Affairs in 1937, studying economic, social, and political con-
ditions in Europe. He traveled to England and Russia to conduct his research.
The next year May married and resigned his fellowship since his position and
its funding were unsure. He and his wife then studied at the Sorbonne in
Paris and became active in the workers movement. By 1939, he returned to
Berkeley to resume work on his doctorate in mathematics with applications
to social theory. However, in 1940 he was dismissed from his teaching duties
at the University due to his involvement in the Communist Party. In 1942
May ran unsuccessfully for State Treasurer of California on the Communist
ticket, nevertheless gaining 44% of the vote.
Just before finishing his thesis, May’s life changed again. With Russia allied
with the U.S. for the war, May sought to join the service; however, married
men were not accepted in the service at that time. When his wife filed for
divorce in mid-1942, May was able to enlist. May joined the 87th Mountain
Infantry (10th Mountain Division), and served in the Aleutians (1943) and in
Italy (1945). In 1944 he remarried, and after the war he and his second wife
stayed in Italy, where May taught mathematics at the Army University Study
Center. He returned to California in 1946 and defended his thesis, “On the
Mathematical Theory of Employment”, under Evans. May then accepted an
assistant professorship at Carleton College in Northfield, Minnesota. During
the late 1940s he published and presented his research in mathematics and
industrial theory with titles such as his 1947 “The Aggregate Effect of Tech-
nological Changes in a Two-Industry Model”. Throughout the 1950s, May’s
research focused on election theory, in which he published extensively. In the
1960s May’s interests were directed towards the history of mathematics. In
1966 he moved to the University of Toronto’s Institute for the History and Phi-
losophy of Science, where he promoted the history of mathematics and science
through his involvement in the IHPS, his founding of Historia Mathematica,
and the founding of the CSHPM in 1973.
At the last moment, May did not attend the 1973 Queen’s University
meeting at which the society was founded. Charles V. Jones of York University
Introduction: The Birth and Growth of a Community 5
chaired the meeting. At that meeting, the name of the society was agreed
upon and Historia Mathematica was selected as its official journal. Jones was
elected President, along with Thomas Settle as Vice President and J. Lennart
Berggren as Secretary-Treasurer. At this meeting a modest set of papers in
the history and philosophy of mathematics was presented. During 1973 and
early 1974, Jones and Settle drafted the original bylaws by which the society
still operates (with only slight modifications).
The following year the first official meeting of the society was held at the
Learneds, with sixty charter members. A more extensive program of papers
was presented at this meeting, including invited papers from all three of the
executive members. These were the first annual guest presentations, of which
this volume contains a sampling. After May’s death in 1977 the Kenneth O.
May fund was established, which helps to bring noted historians to the annual
meetings as guest speakers.
Over the years, the society has traditionally held its annual meeting at the
Learneds Congress (now the Congress of the Humanities and Social Sciences)
every spring. From time to time we sponsor joint sessions with the Canadian
Society for the History and Philosophy of Science, the first in 1974. In 1996
reciprocal memberships between the two organizations were introduced. Our
most recent meeting with the Canadian Mathematical Society occurred in
2000 in Hamilton; another is planned for the year 2005. On a grander scale,
at about the same time CSHPM and the British Society for History of Mathe-
matics (BSHM) became sister organizations, with joint meetings held in 1997
in Oxford, 1999 in Toronto and most recently, 2004 in Cambridge.
As our membership continues to grow and diversify, so does interest in
history and philosophy of mathematics from the mathematical community at
large. In the past few years, interest in using the history of mathematics in
teaching to motivate learning at both the school and collegiate levels, and
an interest in the subject in its own right, has increased dramatically. To
facilitate this new interest from those outside of the specialty, the History
of Mathematics Special Interest Group of the Mathematical Association of
America (HOM SIGMAA) was formed in 2002. Initial discussions leading to
the formation of this new organization occurred during the annual CSHPM
meeting in Quebec, 2001. Two members of the society drafted the constitution
of this new group. Though HOM SIGMAA and CSHPM are not officially
affiliated, they maintain a close informal working relationship. The CSHPM
focuses on scholarly activity in the history and philosophy of mathematics, and
HOM SIGMAA focuses primarily on the pedagogical aspects of the history of
mathematics. The goal is for a symbiotic relationship that will promote not
competition but complementary pursuits. Currently, all the HOM SIGMAA
executive members are also CSHPM members.
This volume represents the next major project undertaken by the CSHPM.
Since 1973, a wide variety of original work in the history and philosophy
of mathematics has been presented at our annual meetings. That work has
been recorded since 1988 in our internally produced annual Proceedings. After
6 Amy Shell-Gellasch
sixteen years, it is time to present some of that material to a wider audience.

Though all the papers presented in the Proceedings deserve wider attention,
this volume will showcase the papers presented by the keynote speakers at the
Annual Meetings. These papers are of a consistently high quality by known
experts in the field.
On behalf of the Society, I would like to thank Michael Kinyon and Glen
Van Brummelen for their time and energy in seeing this project to completion.
I would also like to thank all the speakers over the years who have shared their
research and interest in the history of mathematics with us at our Annual
Meetings. Of course these meetings, and the Proceedings that result, could
not happen without those who devote many hours to making sure that the
meetings run smoothly and that the Proceedings are published. We hope to
share our love of the history of mathematics with you through this sampling
of CSHPM activities.
1
History or Heritage? An Important Distinction
in Mathematics and for Mathematics
Education∗
Ivor Grattan-Guinness
Middlesex University at Enfield
To the fond memory of John Fauvel (1947–2001)
1.1 Interest and Disagreements
During recent decades there has been a remarkable increase in work in the his-
tory of mathematics, including its relevance to mathematics education. But at
times considerable differences of opinion arise, not only about its significance
but even concerning legitimacy–that is, whether or not an historical interpre-
tation counts as history at all. In this paper I consider the latter issue, and
also note some consequences for education.
The disagreements are general, in that they may arise for any branch of
mathematics in any period or culture; so they need a general resolution. I
offer one in the form of a distinction in the ways of interpreting a piece of
mathematics of the past. Take such a mathematical notion N; it could be
anything from one notation through a definition, proof, proof–method or al-
gorithm to a theorem, a wide-ranging theory, a whole branch of mathematics,
and ways of teaching it. By its ‘history’, which becomes a technical term, one
considers the development of N during a particular period: its launch and
early forms, its impact, and applications in and/or outside mathematics, and
so on. It addresses the question ‘What happened in the past?’ by offering de-
scriptions. Maybe some kinds of explanation will also be attempted to answer
the companion question ‘Why did it happen?’.
History should also regard as important two companion questions, namely
‘What did not happen in the past?’ and ‘Why not?’. The reasons may involve
the other side of this distinction, which I call ‘heritage’. There one is largely
concerned with the effect of N upon later work, during any relevant period
including that of its launch. Some modernised versions of N are likely to be
∗
First published in the American Mathematical Monthly 111 (1) (2004) 1–12.
8 Ivor Grattan-Guinness
taken, for heritage is largely concerned with the question ‘How did we get
here?’, that is, to some current version of the context in question.
The distinction between history and heritage is often sensed by people who
study some mathematics of the past, and feel that there are fundamentally
different ways of doing so. Hence the disagreements can arise; one man’s read-
ing is another man’s anachronism, and his reading is the first one’s irrelevance.
The discords often exhibit the differences between the approaches to history
usually adopted by historians and those often taken by mathematicians.
The claim put forward here is that both history and heritage are legitimate
ways of handling the mathematics of the past; but muddling the two together
or asserting that one is subordinate to the other, is not. Many consequences
flow from this stance, which will be treated in sections 3 and 4; first let us
take a simple and well-known example, from the distant past.
1.2 Pythagoras’s Theorem, Euclid Style

One of the best-known theorems in Euclid’s Elements (fourth century B.C.E.)
concerns the sides of a right-angled triangle ABC in Figure 1.
C
B
Fig. 1.1.
We recognise it as saying of the sides AB, AC, and BC that
AB 2 + AC 2 = BC 2 ; (1.1)
but Euclid actually says something quite different [11, Book 1, Proposition
47J]: in right-angled triangles the square on the side subtending the right
angle is equal to the squares on the sides containing the right angle. There is
an attached diagram of which Figure 1 is part, and the differences between it
and (1) are basic. Not only is (1) algebraic whereas the figure is geometric: the
diagram shows the squares outside the triangle, which (1) does not convey.
1 History or Heritage? 9
Were any of the squares to lie over the triangle, then both (1) and the theorem
would still be true: but the complicated proof, not shown in the figure, could
not be effected. The algebraic character of (1) emerges further when, as was
and is commonly done, the letters ‘a’, ‘b’, and ‘c’ are used for the sides: for
algebra is the branch of mathematics in which special words and especially
symbols are used to a significant extent to represent constants, unknowns,
variables, and operations.
Another important difference concerns the word ‘on’. Euclid never used
the phrase ‘side squared’, for in his geometrical Books he never multiplied
geometric magnitudes together, either in the statement of theorems or (more
importantly) in any proof. For example, he did not draw upon side-squaring
when proving Pythagoras’s theorem, either in the complicated proof just men-
tioned, which relies upon congruence, or in a more elegant one for the more
general theorem about rectangles with the same ratio of sides set upon the
sides of the triangle, where the proof deploys similar triangles and ratio the-
ory [11, Book 6, Proposition 31]. Thus ‘BC 2 ’ is already a transgression from
his geometry (and the frequent use in diagrams of small letters such as ‘a2 ’
even more so). Instead Euclid constructed a square on a given line–indeed,
in the proposition immediately preceding Pythagoras’s theorem [11. Book 1,
Proposition 46].
The issue is more profound than it may seem. Both here and everywhere
else in the Elements Euclid works with lines rather than lengths, the latter
being lines upon which some arithmetical measure has been imposed. Euclid
presented geometry without arithmetic in the sense just explained; numbers
are also present, but for other purposes, such as saying that this line is twice
that line, or that the ratio of two lines is the same ratio as 5 : 7. In the same
way he worked with planar regions but not (measured) areas, with solids but
not volumes, with angles but not in degrees. By contrast, which is sometimes
overlooked, in the arithmetical Books 7–9 multiplication of integers themselves
occurs as usual [15].
These remarks concern the history of Euclid. When one moves to its her-
itage, then a quite different situation arises, in which (1) and many other
such equations are prominent. For the Elements played a major role in the
development of common algebra among some of its Arabic initiators, and a
still greater one when Europe at last woke up during the twelfth century and
began to elaborate that algebra with symbols introduced both for unknown
quantities and for operations. Both (1) and Pythagoras’s theorem as shown
by the figure are legitimate readings of Euclid, but are quite different from
each other.
The Elements is a particularly interesting historical example, because com-
mon algebra as in (1) became the dominating reading of Euclid (including in
mathematics education) to such an extent that during the nineteenth century
it also became the normal historical interpretation; apparently Euclid had
been a ‘geometric algebraist’, talking geometry but really practising common
algebra. A supporter of this reading was T. L. Heath, whose English edition
and translation, first published in 1908, is still the most widely used, usually
now in the second edition [11]. Greek specialists tell me that his translation
is very reliable both to the language and to the mathematics; in particular,
for Pythagoras’s theorem and all other contexts he says there ‘square on the
side’, not ‘square of the side’ as many earlier translations had rendered (the
word ‘apo’ can admit both ‘on’ and ‘of’ as translations) but which can easily
lead to the algebraic ‘side squared’. Nevertheless, Heath added to his trans-
lation many algebraic versions of the propositions without seeming to notice
the differences entailed.
While some historians of that time did not follow the algebraic interpre-
tation of Euclid—for example, the Dutchman E. J. Dijksterhuis [26, chap.
5]—the standard view came under severe challenge only from the 1960s on-
wards. In particular, in the mid 1970s the historian Sabatei Unguru attacked it
strongly, to the opposition of some mathematicians interested in history. Un-
guru’s charges of anachronism and ahistory are largely vindicated: his math-
ematician opponents were inheritors [20].
We shall take another Euclid example in section 8. First, though, let us
explore some general consequences of the distinction.
1.3 Some Principal Differences between History and

Heritage
The distinction between the history of a notion N and its heritage obviously
involves its respective pre- and post-histories; but much more is at hand, for
history has to use post-history also. To see this, let us consider the advice,
which is quite often put forward for history of all kinds, about a way of
being ‘history-minded’ about N (say, Pythagoras’s theorem in Euclid); namely,
forget everything that has happened since N was formed, and read Euclid with
the eyes with which he wrote it. But this advice begs the question at hand.
For in order to forget everything E that has happened since N, then one has
to know E already; however, to do that one needs to be able to distinguish E
from the history and pre-history of N; but this is the task to be attempted.
Thus the distinction between history and heritage rests in part upon the
ways in which notions later than N are to be used. When they are determined
to be later notions, the view urged here is this: by all means bring them to
bear, and deploy them to understand the heritage from N, but avoid feeding
them back to appraise its history (such-and-such did not happen). Further,
when considering periods intermediate between that of N and some later ones
such as now, apply the distinction carefully. Thus, in our example the equation
(1) is not only part of the history of René Descartes and the heritage of Euclid
but also belongs to the heritage from, among others, the algebraist François
Viète in the sixteenth century, whose work also belongs to the history of
Descartes. Note also that history is usually a history of heritages; it is a tale
of mathematicians taking and modifying notions from the past (often pretty
recent) without enquiring about the history of those notions.
Various other matters can be explored; a more detailed discussion, largely
focussed upon history, is given in a companion paper [16]. The following table
summarises the main features of handling past notions N in the two different
ways suggested. An apparent contradiction between the third and fourth rows
needs to be addressed. When the historian reconstructs past muddles, he
will conflate notions that we now know to be different, a feature that the
inheritor will stress. But the difference that the reconstruction exposes is that
between past ignorance of the distinction, which is different from our (and the
inheritor’s) present knowledge of it.
Feature History Heritage
Motivation(s) to N Important issue; maybe Probably only of

hard to find (for example, minor interest
for Euclid’s Elements)
Types of influence Can be negative as well Likely to draw only

as positive; both should upon the positive cases
be noted
Relationships of N Major issue; differences Important issue:

to earlier and to stressed as much as similarities stressed
later notions similarities, maybe more more than differences
Handling unclarities Reconstruct them, and Recognise them, but

evident in N as clearly as possible clean them up
Successful Very important: but also Likely to be the

developments study failures, delays, main concern
missed opportunities,
and late arrivals
Role of chronology Usually important; can Beyond broad details,

be hard to establish not likely to matter
so much
Historical May try to reconstruct May try to construct
consequences the foresight (hopes, and hindsight and historical
so on) for N held by the perspective of the
historical figures developments after N
Determinism? Preferably not claimed: May carry a determinist

the actual developments flavour; we had to get
were so-and-so, but not here (but see the history
necessarily so-and-so column!)
Foundations Dig down to them, and Lay them down and

of a theory build upon a swamp build up from them,
like on solid ground
Level of importance Can vary over time, Not normally considered;

or popularity of N independently of content; current importance
should be noted (and assigned
maybe explained)
1.4 Changing Habits

A further type of issue, which is not susceptible to tabular expression, concerns
the use of notions that have become standard and therefore are now used
habitually. Such habits may well help in determining the heritage from N: but
historical anachronism can easily arise, which needs to be controlled. I note
three important examples.
First, after an interesting history of its own from the 1870s [9], Georg
Cantor’s set theory has been part of our mathematical furniture for just over
a century; so for the mathematics of this period its use may well be faithful.
Now collections of things have been handled in mathematics since at least
Greek antiquity; but the earlier theory of so doing was part-whole theory,
where (say) British women form part of the class of women, membership is not
distinguished from inclusion, and an object is not distinguished from its unit
class. The differences between part-whole and set theories are considerable,
both technically and philosophically, and the historian needs to mark them
carefully. By contrast, the inheritor can deploy set theory with little chance
of deception.
Second, while the influence of Euclid was great in Western mathematics,
his stress on axioms and common notions was rarely imitated (though to some
extent Newton’s Principia is an example). The axiomatisation of mathemat-
ical theories became more prominent only during the late nineteenth cen-
tury, especially in connection with the axioms of Euclidian and non-Euclidian
geometries, and the emergence of abstract algebras [8]. Both developments
attracted the attention of David Hilbert, and led him to launch the wide-
ranging use of axiomatisation during the first half of the twentieth century,
an attitude that has now become pretty standard: a clear path of heritage
can be traced up to present-day practises. But the historian should be careful
when looking at the structure of earlier mathematical theories, for axiomati-
sation may well not be prominent beyond specifying basic principles or laws.
Cantor’s set theory is a good example: while it too was developed during the
late nineteenth century, he showed little interest in the axioms that it may
require.
Third, vector and matrix theory have become standard fare in mathemat-
ics, though (especially in the second case) only from the l930s onwards and
after rather scrappy historical developments in various contexts during the
nineteenth century. Once again, care should be exercised in applying them to
earlier work. For example, much of the mechanics developed by figures such
as L. Euler, J. L. Lagrange, and P. S. Laplace can be rewritten in vectorial
and matricial forms, but historical understanding will not profit. For none of
these figures knew that their theories could be developed in terms of strings
or arrays of scalar elements; they worked instead in terms of collections of
simultaneous linear or differential equations, or quadratic and bilinear forms
[14, chaps. 5–6]. The introduction of vectors or matrices is not merely a mat-
ter of changing notation; new theories are involved. It is of course nice to save
such space, for one thing; but if the historian does deploy these theories, then
a chronological health warning should be appended.
By contrast to all these cautions to the historians, the inheritors can ex-
ecute all these reformulations of theory quite legitimately; indeed, much nice
heritage mathematics may emerge. Further, some history of mathematics pro-
duced after the initial period under study might be created; for, as was men-
tioned in section 1, mathematicians normally read the past in a heritage spirit.
As an example, take Lagrange and others in mechanics. A major prob-
lem, which he formulated in the 1770s, was to prove mathematically that the
planetary system was stable. (Previous figures such as Newton and Euler had
relied on God to watch out for danger; that is, a religion influenced mathe-
matics.) In terms of matrix theory, Lagrange’s brilliant theory sought proof
of the reality of all the eigenvalues and eigenvectors (to use modern terms)
of a certain matrix. But he had no such theory, and worked with the corre-
sponding quadratic forms; so did Laplace, who adapted his results to some
extent; neither man found a watertight proof. The next major contribution
came in 1829 from (surprisingly) A. L. Cauchy, and in 1829 he did formulate
‘tableaux’ of scalar entries in his own work on this problem [17]. Thus matrix
theory may—indeed, should—be used to describe Cauchy’s contribution, and
thus to help us to grasp an important part of his heritage from his predeces-
sors. And we also have a nice example of the ‘What did not happen?’ question;
for Cauchy never realised the significance of his achievement and rarely used
it later, so that unfortunately he was not an influential founder of the spectral
theory of matrices.
1.5 Some Philosophical Background

It is obvious that this talk of earlier and later notions, the development of the-
ories, and so on, is not confined to mathematics: such features occur also in the
histories of other sciences (including technology, engineering, and medicine),

and indeed elsewhere (for example and a nice one, practices to be adopted
and avoided in the so-called authentic performance of older music). The main
general principles that underlie the foregoing discussion are as follows.
First, history is unavoidable, whether one likes it or not. A mathematician
who presents his theory without concern with history is not thereby immune
from it. For example, an enthusiast for axiomatics mentioned in section 3
will lay out his theory in a very formal way without reference to predecessors
or precedents; but they will be there, including previous formal theories laid
out by preceding axiomatists without reference to their own predecessors or
precedents. Thus the question of whether or not one can use history in math-
ematics is miscast: it is rather the question of whether it is done consciously
or not. Indeed, independent of the content of this paper, it is useful to have
some general historical idea of a topic of interest, whatever it may be.
Second, knowledge and ignorance go together. This symbiosis has not re-
ceived the general philosophical attention that it deserves. In particular and
of special significance for mathematics, there is knowledge of ignorance, espe-
cially when one formulates a problem. When, for example, J. P. G. Dirichlet
studied the convergence problem of Fourier series in the late 1820s, he knew
that he did not know sufficient conditions on a function to establish con-
vergence to it: finding some was precisely his problem. Having done so, he
knew that he did not know whether or not they could be weakened, thereby
setting the next problem in this chain (to which the first answer was the Lip-
schitz condition, by the way). One can also have ignorance of ignorance, or
unawareness, where people do not know that they do not know something be-
cause the required connections between notions have not yet been laid down.
Thus Dirichlet did not know that he did not know how his proof bore upon
the specification of function spaces, because that notion did not emerge until
the late nineteenth century [21].
Third, and following from the preceding line of thought, knowledge of all
kinds is stratified into theory, metatheory, . . . . For mathematics this means
not only metamathematics of the technical kind that Hilbert launched, but
also informal kinds. In particular, the history of notion N is one kind, its her-
itage is another, manners of its possible teaching a third, heuristic strategies
to explain its significance a fourth, and there may well be others. The rela-
tionship between knowledge and ignorance just outlined lie in the metatheory
of the notions involved. Similarly, metatheory requires metametatheory as its
own forum for discussion, and so on upwards as far as is needed. An example
of metametatheory is the history of the history of mathematics, an interesting
story recently recorded in detail in [10]; the comments on Heath in section 2
form an example of it; and this paper itself is a self-referring example, with
its heritage (if any) awaited!
The recognition of history and heritage as metatheoretic also releases both
historians and inheritors from the need to like what they find in the past that
they study. Why should they? After all, they were not there (as a rule). The
point seems obvious enough; after all, one can be a good historian (or inher-
itor) of, say, military history without being a militarist. Yet not infrequently
historians and inheritors become overly attached to their objects and figures
of study, in any kind of history, and feel that they have to defend what they
find. While of course such attachment can be felt if it arises naturally, no
compunction to it should even be encouraged.
The generality of stratification is an insight forged in connection with
symbolic logic in the early 1930s, thanks principally to Kurt Gödel and Al-
fred Tarski. In logic the distinction of (object-level) logic itself from metalogic
is especially tricky but thereby all the more important; as was known al-
ready in Greek times, failure to make a distinction of some kind admits nasty
paradoxes. Gradually stratification spread into other disciplines, especially
mathematics and some types of philosophy. One follower, inspired by Tarski
in the mid 1930s, was Karl Popper. Several parts of his philosophy of fallibil-
ism are metaphilosophical; for example, his preference for indeterminism over
determinism [19]. Of particular relevance to this paper is his essay ‘On the
Sources of Knowledge and Ignorance’ [18, introduction], for it contains an in-
sight largely missing from other kinds of philosophy; that ignorance is nice, for
it is the site (in metatheory) of our problems when construed as knowledge
of ignorance. In most other philosophies ignorance is a disease to be cured
by the acquisition of knowledge however that acquisition is claimed to occur
(see [25, chaps. 1–6] for the various forms of this view maintained within the
sceptical tradition of philosophy). So far explicit use of stratification has not
been widely canvassed among prevalent philosophies of history (which are well
surveyed in [23]); but it seems worthy of further elaboration.
1.6 General Remarks about History in Mathematics

Education
In recent decades a considerable and international increase has developed in
the use of history in mathematics education, in order to temper and challenge
the normal picture of mathematics as a human-free zone, all answers but no
questions, all solutions but no problems. Several edited or authored books and
special issues of journals have appeared containing material of various kinds:
textbooks significantly informed by the relevant history; summary histories of
particular developments; surveys of the lives and works of important historical
figures; international and/or multicultural comparisons of the development of
(more or less) the same theories; translations of original texts with commen-
tary; and suggested strategies for using history in teaching practice, both in
specific contexts and in general. The emphasis often falls upon motivation
and context, on showing that mathematics is after all human activity despite
appearances, and moreover that much of it is not Western in origin. The range
of concerns is well captured in a recent volume [12].
Most attention seems to have fallen on teaching at school and college level,
but the university level has also been addressed. Much more work has been
done on pure mathematics than on applied or applicable mathematics, or on
probability and statistics; a redress of balance would be most welcome. I do
not attempt to review this literature here, but I consider the place and utility
of the distinction between history and heritage in mathematics education in
general.
As with researchers in history mentioned in section 1, there is an evident
sense of the distinction in this kind of educational literature, or at least an
intuition that the mathematics of the past can be used in different ways.
Where is mathematics education to be found between history and heritage?
My answer is that that is exactly where it should be found, so that it can
profit from both sides. In particular, if notion N is to be taught, then both
its history and its heritage can be used. Euclid’s Elements is a good example,
where the inherited use of algebra has been well used quite frequently. In
addition, the historical Euclid deserves attention, with its geometry presented
without arithmetic with lines rather than lengths, and the beautiful theory of
ratios used in both his geometry and his arithmetic.
1.7 History-Satire and the Calculus

In the paper [13] I introduced long ago the term ‘history-satire’ to charac-
terise a way in which history and also heritage can be used in mathematics
education. Under it the broad features of the historical record are respected
and used; but usually many detours and complications occur that, while they
attract the historian, will impede teaching and so should be set aside or at
most treated only in passing. The ‘genetic method’ of Otto Toeplitz, which
he introduced initially in the late 1920s in connection with teaching the dif-
ferential and integral calculus, is similar in sentiment [24]. More recently the
Mathematical Association of America published a novel and important text-
book in real-variable mathematical analysis by David Bressoud, in which he
gives prominent places to the main developments, especially of the nineteenth
century, such as Fourier series [5].
As Bressoud duly notes, a major innovation of the century was the found-
ing of analysis in the 1820s by Cauchy. His approach was based upon a newly
sophisticated theory of limits, not with limit left as an intuitive notion. Un-
doubtedly it was much superior to the preceding versions in the organisation
of the subject and statements and proofs of the theorems; however the loss in
heuristics was heavy, and both his colleagues and students objected forcefully
to it [14, chaps. 10–11 passim, and 20.8].
For an explicit example, here is a use of history-satire that I found helpful
in my own teaching. In a remarkable analogy, Cauchy adapted his real-variable
analysis to complex variables and their functions and thereby introduced a
major new subject into mathematics. But it seems a strange subject when
first learnt: it uses the corresponding expressions as in real-variable analysis,

but there are no curves, tangents to them, or areas underneath them to think
about or look at. Among the many theorems that Cauchy proved, a main
one is now named after him: namely, that the integral of a single-valued and
differentiable function with a continuous derivative around and inside a closed
contour C is zero. To students, including me long ago, it seemed to be a pecu-
liar result: and a quick and doubtless valid proof using the Cauchy-Riemann
equations and Green’s theorem did not assuage the perplexity.
Cauchy developed his theory fitfully from the 1810s to the 1840s [22], and
this version of his theorem is the last one, with the complex plane available
as the site for C. I found that an earlier stage of his theory helped in under-
standing the theorem. In his treatment of the real-variable integral of f (x)
over the range x0 ≤ x ≤ X (I use his symbols) he formed the area sum S for
a partition of values of x over the range, took successive subpartitions and
formed the corresponding sums, and defined the integral as the limiting value
of the sequence if it existed at all. This manner of defining the integral has
long been standard, and his version is still worth reading and teaching [6,
lecture 21].
Soon afterwards Cauchy deployed his analogy. He defined the integral
√ of
a finite-valued and continuous complex-variable function ‘f (x +√y −1)’ by
forming the expression corresponding to S for f (x)√but with x + y −1 taking
√
a sequence of values between the limits A = ‘x0 +y0 −1’ and B = ‘X+Y −1’
for which both x and y were continuous functions of a parametric variable t.
Then, drawing upon integration by parts and the calculus of variations, he
proved that the value of this integral between A and B ‘is independent of the
nature of the functions’ involved [7, sec. 3]. The closed contour theorem then
follows by taking the integral along one sequence of values between A and B
and back along another sequence under which the required conditions obtain;
the two integrals for the two sequences cancel out, so that the value of the
integral over C is zero.
Working through the theorem this way certainly took more than a few
lines; but the understanding increased substantially, especially as the defi-
nition of the real-variable integral had already been taught elsewhere. My
account follows Cauchy historically to the extent of deliberately avoiding di-
agrams, for both types of integral: at that stage in his career he regarded
geometric notions as unrigorous and so wished to avoid them. The status of
geometry makes a nice point to debate in the classroom, and in fact I increased
the measure of satire by using diagrams myself. I also ignored several special
cases and other details of the theory as Cauchy was then developing it. But
I raised questions such as whether or not Cauchy assumed the derivative of
f to be continuous (yes, but implicitly); and I also taught his 1825 version of
the residue
√ theorem, noting that, contrary to most later practice,
√ he allowed
x + y −1 to go through, and not just round, a pole of f (x + y −1) [6, sec.
8].
The considerations of this section have used the calculus and mathemat-
ical analysis because these case studies happen to come from it. But genetic
approaches and history-satire can be applied to any mathematical notion or
level of teaching.
1.8 The Proposals of Bashmakova

The relationships between knowledge and ignorance outlined in section 5 de-
serve serious consideration, including the niceness of ignorance as the source
of problems (big or small) to tackle. One important area of education where
these relationships are prominent is the design of a course syllabus and the
manner and order of teaching the topics proposed, when in effect the designer
is considering the stage at which the pupils or students should cease to be
ignorant of some specific notions.
Let us take an example, examining the historiography proposed in recent
years by the Russian historian I. G. Bashmakova, for two of her books have
recently been translated into American and published by the Mathematical
Association of America for their utility in mathematics education. While deal-
ing with the history of common algebra, her position is put forward in a general
way, most explicitly in a joint paper with I. M. Vandaloukis [4]. For them,
there are two main stages in handling an historical text. ‘First the text should
be “translated” into the [sic] contemporary mathematical language, i.e. an
adequate model for it should be constructed. This is absolutely necessary in
order to understand the text, to reveal its mathematical content’ (p. 251). In
the next stage ‘it is necessary to embed the considered work in the context of
science of its day’ (p. 252).
The authors state that the second stage is ‘more difficult’ than the first:
in my view it might well be impossible, since the first stage will have put
so much heritage in place that the historical context could be masked. They
state the aim of heritage very clearly: ‘the mathematicians of every new age
reconsider the previous material and restate it in new terms, thus making it
readily available and applicable for the contemporary scientist’ (p. 250).
The examples given in Bashmakova’s writings seem to exhibit the con-
flation of history and heritage, without the stress on the distinction between
them that was argued in section 7. For example, she takes Proposition 4 of
Book 2 of Euclid’s Elements to express the quadratic identity
(a + b)2 = a2 + 2ab + b2 (1.2)
as a legitimate prime reading [1, p. 88], [3, p. 165].

In a more recent short history of algebra, coauthored with G. Smirnova,
(2) is held to be ‘equivalent’ to the diagram [3, chap. 2]; and throughout this
book the modern notations dominate, although the older terms and symbols
are also presented in some detail [3, chaps. 4–5]. The dominance of heritage
Fig. 1.2.
is clear in a general historiographical appendix, where in specifying the term

‘geometric algebra’ the authors characterise algebra as an historical category
drawn from ‘the class of problems associated with algebra today’ (p. 164). For
them, therefore, Euclid’s Book 2 is concerned with algebraic identities such as
(2) (see especially Bashmakova in [2]): indeed, her most recent stance is to im-
pose algebraic readings onto ancient arithmetic and geometry for all cultures
(see [3, pp. 163–172], where Bashmakova and Smirnova vote for the math-
ematicians and against Unguru in the disagreement noted earlier in section
2).
The preference for modern notations in the book fits its primarily educa-
tional purpose well, exposing an important chain of heritage influences. But
the quoted general statements of historical interpretation seem to involve her-
itage mistaken as history. For me, in Book 2 Euclid presents theorems relating
subregions of planar rectilinear constructions involving rectangles, squares,
and triangles (as in the cited example, where the relative locations of the sub-
squares and subrectangles are lost in the ubiquitous sign ‘+’); the algebraic
content is empty, as also in all his other geometry Books. By contrast, algebra
looms very large in the post-Grecian heritage from Euclid’s geometry. Both
readings are valuable to mathematics education, though better presented as
distinct sources. Indeed, like Euclid himself the history of the theory of poly-
nomial equations is especially suitable for historical satire.
1.9 Concluding Remark

In this paper, and in more detail in its companion [16], I assert that the his-
tory of mathematics differs fundamentally from heritage studies in the use
of the mathematics of the past, and that both are beneficial in mathematics
education when informed by the mathematics of the past. The majority of the
examples presented come from fairly modern periods. This is no accident, for
they constitute my specialist areas; thus the examples as such have no partic-
ular significance. Indeed, since the distinction between history and heritage is
held to be general, then indefinitely many more examples could be presented;

the reader is invited to construct some of his or her own. A rich resource
comes from considering the many ways in which notions are changed, espe-
cially when they are (major) theorems or theories. These include the alteration
of known results by extension, generalisation, and/or abstraction; reaction to
counterexample; the exposure as axioms or as procedures of assumptions pre-
viously taken for granted; the adaptation of algorithms; the introduction, or
maybe removal, of connections between branches (such as geometry with or
without arithmetic); classifications into kinds of objects in a theory; switches
between axiom, theorem, and definition; and new applications, both within
mathematics and to other disciplines.
More attention has been paid in this paper to issues concerning history and
historiography than to heritage and heritage studies; but no value judgement is
involved, for, as stated in section 1, neither activity is subordinate to the other
one. A companion paper concentrating on good and bad practices in heritage
work could be written. The two activities are distinct but they interact in
fruitful ways, each posing questions for the other to address.
Acknowledgment. This paper is based upon a plenary lecture delivered to
the joint annual meeting of the Mathematical Association of America and the
American Mathematical Society that was held in Baltimore in January 2003.
Thanks are offered to the former organisation for the invitation.
References
1. I. G. Bashmakova, Diophantine equations and the evolution of algebra, Transl.
Amer. Math. Soc. 147 (2) (1990) 85-100.
2. , Diophantus and Diophantine Equations (trans. A. Shenitzer), Math-
ematical Association of America, Washington, D.C., 1997.
3. I. G. Bashmakova and G. Smirnova, The Beginning and Evolution of Algebra
(trans. A. Shenitzer), Mathematical Association of America, Washington, D.C.,
2000.
4. I. G. Bashmakova and I.M. Vandaloukis, On the justification of the method of
historiographical interpretation, in Trends in the Historiography of Science, K.
Gavroglu et al., eds., Kluwer. Dordrecht, Boston and London, 1994, pp. 249–264.
5. D. M. Bressoud, A Radical Approach to Real Analysis, Mathematical Association
of America. Washington, D.C., 1994.
6. A. L. Cauchy, Résumé des Lecons Données a l’Ecole Polytechnique sur le Calcul
Infinitesimal, vol. 1 [and only], de Bure, Paris, 1823; also in Oeuvres Complètes,
ser. 2. vol. 4. Gauthier-Villars, Paris, 1898, pp. 5–261.
7. , Memoire sur les intégrales définies, prise entre des limites imagi-
naires, de Bure, Paris, 1825; also in Oeuvres Complètés, ser. 2, vol. 15, Gauthier-
Villars, Paris, 1974, pp. 41–189.
8. J. Cavaillès, Methode axiomatique et formalisme, 3 vols., Hermann, Paris, 1938.
9. J. W. Dauben, Georg Cantor, Harvard University Press, Cambridge, Mass.,
1979; reprinted by Princeton University Press, Princeton, 1990.
10. J. W. Dauben and C. J. Scriba. eds., Writing the History of Mathematics: Its
Historical Development, Birkhäuser, Basel, 2002.
11. Euclid, Elements: edition used: The Thirteen Books of Euclid’s Elements, 2nd
ed.. 3 vol.,. ed. and trans. T. L. Heath). Cambridge University Press, Cambridge,
1926: reprinted by Dover, New York, 1956: 1st ed. 1908.
12. J. Fauvel and J. van Maanen, eds., History in Mathematics Education. The
ICME Study, Kluwer, Dordrecht, 2000.
13. I. Grattan-Guinness, Not from nowhere. History and philosophy behind math-
ematical education, Int. J. Math. Edu. in Science and Technology 4 (1973)
421–453.
14. , Convolutions in French Mathematics, 1800–1840, 3 vols., Birkhäuser,
Basel, and Deutscher Verlag der Wissenschaften, Berlin, 1990.
15. , Numbers, magnitudes, ratios and proportions in Euclid’s Elements:
How did he handle them? Historia Mathematica 23 (1996) 355–375; printing
correction in 24 (1997) 213.
16. , The mathematics of the past. Distinguishing its history from our
heritage, Historia Mathematica 31 (2004) 163–185.
17. T. W. Hawkins, Cauchy and the spectral theory of matrices, Historia Mathe-
matica 2 (1975) 1–29.
18. K. R. Popper, Conjectures and Refutations, Routledge and Kegan Paul, London,
1963.
19. , The Open Universe. An Argument for Indeterminism, Hutchinson,
London, 1982.
20. D. Rowe, New trends and old images in the history of mathematics, in Vita
Mathematica. Historical Research and Integration with Teaching, R. Calinger,
ed., Mathematical Association of America, Washington, D.C., 1996, pp. 3–16.
21. R. Siegmund-Schulze, Die Anfänge der Functionalanalysis, Archive for History
of Exact Sciences 26 (1982) 13–71.
22. F. Smithies, Cauchy and the Creation of Complex Function Theory, Cambridge
University Press, Cambridge, 1997.
23. M. Stanford, An Introduction to the Philosophy of History, Blackwells, Oxford,
1997.
24. O. Toeplitz, The Calculus. A Genetic Approach, University of Chicago Press,
Chicago, 1963.
25. P. Unger, Ignorance. A Case for Scepticism, Clarendon Press, Oxford, 1975.
26. K. van Berkel, Dijksterhuis. Een biografie, Bert Bakker, Amsterdam, 1996.
2
Ptolemy’s Mathematical Models and their
Meaning
Alexander Jones
Department of Classics, University of Toronto
In the middle decades of the second century of our era, a Greek-speaking

Egyptian living in the town of Canopus close to Alexandria carried out a
massive scientific program centring on the writing of about a dozen books
on astronomy, astrology, optics, harmonics, and cartography.1 Unlike his near
contemporary Galen, Ptolemy evidently did not lead the sort of career, and
certainly did not have the self-trumpeting personality, that would procure no-
toriety among one’s contemporaries, and so we know scarcely anything about
his life. But his works were well enough appreciated, in spite of their severe
style and uncompromising technicality, so that the great part of them were
preserved, almost the sole remnants of their kind of scientific writing from an-
tiquity. Though ranging widely in subject matter, these books revolve around
two great themes: mathematical modelling of phenomena, and methods of vi-
sual representation of physical reality. In the following, I wish to consider what
Ptolemy thought the relationship was between his models and the physical
nature that he was describing.
To begin, let us look briefly at what his predecessors made of this ques-
tion. The explanations of phenomena offered by Greek physical science varied
greatly, but they were often framed in terms of two broad principles: first,
that change in matter can be reduced to the operations of a small number
of fundamental qualities, typically hot, cold, wet, and dry; and secondly, that
the phenomena can be modelled by mathematical objects. These principles
tended to be regarded as mutually exclusive, so that certain phenomena were
referred to qualitative and others to quantitative explanation.
What decided which kind of rationale was appropriate in any particular
situation? The historical reality was surely that people stuck to whichever
1
For the biographical data on Ptolemy see Toomer 1987. (Given the informal
nature of the present paper, I have thought it appropriate to furnish the text with
references only to translations of the pertinent works and to a few particularly helpful
works of modern scholarship. The translations of passages quoted in this paper are
my own.)
24 Alexander Jones
kind seemed to work best for the subject matter. Certain areas of experience
lent themselves more obviously to mathematical modelling than others; in
particular, the motions of the heavenly bodies, the esthetics of musical inter-
vals, and the visual perception of shape, distance, and motion were sensed to
have a quantitative regularity not shared by physical change in materials such
as heating, melting, or burning, which on the contrary seemed to have a fairly
direct connection with transference of the qualities hot, cold, wet, and dry.
Aristotle’s cosmology, with its inner globe of more or less stratified earth,
water, air, and fire enclosed in an outer spherical shell of ether, was in part
motivated by this polarity, and in return gave it an a priori rationalization.
The matter of the heavens–the part of the cosmos where the stars, planets, sun,
and moon dwell–is of a different kind from the four mundane elements, subject
to a different natural motion (circular revolution as opposed to motion towards
the cosmic centre), and not subject to any other kind of change. Aristotle’s
ether has the power to force change in other things, but considered by itself,
its only property is eternally regular circular motion. Hence an Aristotelian
astronomy has everything to do with mathematics, and nothing to do with
elementary qualities. Earth, air, fire, and water, on the other hand, can be
forced by an external agent to move in any direction or to change properties,
and, moreover, these processes vary unpredictably in degree and duration.
This is why, even if the continual changes among the four elements–including
life itself–can be traced back through a chain of cause and effect to the physical
action of the heavenly bodies (most importantly the sun’s annual revolution
alternating between north and south), terrestrial phenomena are not as regular
and periodic as the celestial revolutions:
We see that when the sun comes closer, coming-into-being takes place,
and when it recedes, ceasing-to-be takes place, and each happens in
equal time. . . . But it often happens that things cease to be in a shorter
time because of the mixture of things with one another; for since their
matter is not uniform and not the same everywhere, necessarily their
comings into being too are not uniform, and some are faster and some
slower. . . . (Aristotle, De Gen. et Corr. 336b16)
Aristotle’s cosmology thus explains why we can have a mathematical as-

tronomy. It does not, however, account for the possibility of mathematical sci-
ences dealing with special aspects of the world of the four elements, although
Aristotle recognized that possibility, since he classified optics and harmon-
ics, along with astronomy, as sciences embedding mathematics, or indeed as
branches of mathematics (Physics 194a6). Here and there in the Aristotelian
corpus we encounter obiter dicta confirming that Aristotle recognized that
mundane phenomena could be subject to mathematical constraints, for ex-
ample in the following passage where he speculates on a possible analogy
between harmonic theory and colour theory:
2 Ptolemy’s Mathematical Models and their Meaning 25
We have to discuss the other colours [besides white and black], dis-
tinguishing the number of ways that they can arise. Now white and
black can be placed side by side in such a way that each one cannot
be seen because of its tiny size, but the product of the two becomes
visible in this way. This cannot appear either as white or as black.
But since it must have some colour, and it cannot be either of these,
it must be a mixture and some different form of colour. In this way
one can suppose that there are more colours besides white and black,
and that they are numerous in accordance with ratio. For they can lie
next to each other in the ratio three to two, and three to four, and
in ratios of other numbers; and others can be wholly in no ratio, but
incommensurable by some excess and defect. And it is possible that
these things subsist in the same manner as (musical) concords; for the
colours that are in numbers that form good ratios, just like concords
in the other context, would seem to be the most pleasant of colours,
for example sea-purple, red, and a few others like these, for the same
reason that there are just a few concords, while those that are not in
such numbers are the other colours. (De Sensu 439b19)
But an analogy is not an explanation, and we are left in the dark as to why
simple ratios of whole numbers should have a special status in a world of
geometrically continuous matter and change. Similarly one is left wondering
why vision follows straight lines if it is in fact a process of continuous change
in nonuniform matter.
For a working scientist of the Hellenistic or Roman periods in search of a
broad rationalizing framework in which to set his own theorizing, Aristotle’s
cosmology and conception of matter were not the only ones on offer. In the
first place, Epicurus revivified atomism into an elaborate, strictly materialistic
physics in which all matter and change are reduced to the chance motions of
eternal atoms, endowed with a minimum of properties (shape and size), in an
infinite void. Epicurus has sometimes been portrayed as a prophet of science;
in reality he was no friend to the sciences of his time. He endeavoured to show
how the phenomena for which the astronomers sought unique explanations
could result from numerous different physical situations, any of which might
be temporarily valid at some time and place within his boundless universe;
his theory that vision occurs by means of films of atoms that continually peel
off bodies and fly off in all directions would not have stood up long to the
scrutiny of a practitioner of geometrical optics; and in general he contemned
any inquiry into nature that was not subordinated to his ethical goals, freeing
humanity from avoidable pain and fear.
The physics of the Stoics was closer to Aristotle’s. We find again an in-
sistence that matter is geometrically continuous and reducible to variable
mixtures of a restricted number of fundamental stuffs, which at one level of
analysis prove to be the familiar earth, air, fire, and water. The Stoic cos-
mos is finite and spherical, but there is no outer shell of special unchanging
26 Alexander Jones
matter for the heavenly bodies, and the cosmos in its present differentiated
state has a finite span of life. On the other hand, although Stoic physics is
strictly materialist, its cosmos is orderly and rational. The organization of the
cosmos and its parts is effected by pneuma, a vital mixture of fire and air that
extends in varying degrees throughout the cosmos and that has the power to
“tense” the bodies with which it is intermixed. In place of a reductionist expla-
nation of the mathematical behaviour of phenomena, we encounter a deistic
appeal to the will of the cosmic mind. (It should, however, be kept in mind
that our sources for Stoic physics are less satisfactory and more controversial
than those for Epicurean physics, and in any case Stoicism was considerably
more open to innovation than Epicureanism with its ipse dixit deference to
its founder’s pronouncements.)
Alongside these more or less coherent systems there existed a looser tra-
dition of physical speculation, which we call “Peripatetic” because its most
prominent known advocates, in particular Theophrastus and Strato, were close
associates and followers of Aristotle. This was an eclectic approach, grounded
in observation and analogy, and again materialistic. Properties of matter and
processes of change are explained in fairly mechanical terms, for example by
supposing that materials can be composed of particles that can be packed
loosely or tightly, but the particles lack the permanence of true atoms and
are less denuded of innate characteristics. Aristotle’s fifth element seems to
have won no following; the heavens were instead supposed to be composed
mostly or entirely of fire. This fire might be endowed with special proper-
ties, perhaps, but the divide between the celestial and mundane spheres was
inevitably blurred.2
Such were the main lines of physical thought evolving during the century
following Aristotle’s death. It was also at this time that the earliest surviving
works that treat physical problems using mathematical models were written.
These include works on astronomy by Autolycus, Euclid, and Aristarchus,
works on statics by Archimedes, and works on optics and harmonics by (or
at least ascribed to) Euclid. What is striking about these works is not only
the attempt to deduce phenomena through explicit axiom and theorem struc-
tures, but also the fact that these works seem deliberately to evade physical
interpretation of the axioms.
One would dearly like to know what developments the subsequent three
and a half centuries brought. The state of evidence is far from encouraging.
Thus, of the numerous books written by undoubtedly the most important
mathematical scientist of this period, Hipparchus, we possess only one, and
with scarce gratitude and less justice we tend to dismiss that work as atypical
and uninteresting. Among the philosophers, Posidonius stands out as a writer
who undoubtedly exerted a considerable influence on physical thought. One
recognizes in some of the reports of his lost writings the tincture of Peripatetic
2
The rejection of Aristotle’s fifth element is ably discussed by Falcon 2001, 121-
183.
physics in his Stoicism for which he was later criticized; it is harder to discern
a serious engagement with the mathematical sciences.
For that, we must turn to Theon of Smyrna, a Platonist philosopher of
far lesser distinction than Posidonius, but one with the accidental merit that
a large part of one of his books, The Mathematics Useful for Reading Plato,
has come down to us.3 Theon’s mathematics embraces harmonics and astron-
omy, and the long astronomical section is of particular interest here. Theon
exposes, with geometrical demonstrations, the epicyclic and eccentric models
as assemblages of circular paths in the plane; but he insists that these circles
are not mere abstract conceptions but stand for spheres of ether such that,
for example, an epicycle is a rotating sphere nested in the gap between two
concentric spherical shells which revolve together, bearing the epicycle with
them. Theon was a mere generation older than Ptolemy, but this is enough to
establish that the revival of Aristotle’s etherial spheres and their adaptation
to non-homocentric models was not due to Ptolemy, though it may have been
fairly new science in his time.
It makes sense in several ways to begin considering Ptolemy’s attitude to
mathematical models in the context of his astronomy. This was the science
closest to his heart, the only one on which he is known to have written a
multiplicity of books. His central treatise on astronomical modelling, known
to us as the Almagest, preceded most of the others, yet it followed upon a
quarter-century of personal observation and analysis.4 It is also a monumental
piece of reasoning, much more complex and at the same time more structurally
unified than his other large works.
The models with which the Almagest is concerned are kinematic geometri-
cal constructions built up from circular motions representing the paths trav-
elled by the heavenly bodies (the sun, moon, planets, and fixed stars). Most
of the bulk of the Almagest, and most of its mathematics (in the usual sense
of the word), is devoted to determining the radii, rotational velocities, and
orientations of the components of each model. These parts, taken in isolation,
leave open the question whether the circles in the diagrams stand for some
sort of physical bodies in motion, or whether they are just abstract analytical
components of a complex motion which the heavenly bodies perform due to
undetermined physical causes.5 We can at least dismiss a third option, that
they are mere computational devices with no necessary relation to what the
heavenly bodies really do, but by which one can reproduce the phenomena
seen by a terrestrial observer; Ptolemy’s treatment of parallax and eclipses
depends on the assumption that his lunar and solar models correctly describe
the distances of the sun and moon from the earth as well as their directions
from the observer.
3
The most reliable translation is Dupuis 1892.
4
Toomer 1984.
5
On the question of Ptolemy’s realism in the Almagest and Planetary Hypotheses
see Lloyd 1978.
28 Alexander Jones
However, the broader deductive structure of the Almagest decisively com-

mits Ptolemy to believing that his circles stand somehow for material bodies,
even if it is not made explicit precisely how they do so. This may be seen by
examining how Ptolemy arrives at each model before turning to the deduction
of its numerical parameters. I take as an illustration Ptolemy’s model for the
moon, since Ptolemy’s presentation of this model in Books 4 and 5 of the Al-
magest is particularly explicit about the stages by which the model is worked
out.
Ptolemy starts out in Almagest 4.5 with a working hypothesis, which he
warns us will later be disproved, that the moon has a “single and invariant”
anomaly, that is, that its apparent progress along the ecliptic has a periodic
variation that always repeats exactly. He asserts that two models identically
produce this phenomenon. In one model (Fig. 1), the centre of an epicycle E
travels eastward along a circular deferent concentric with the earth T with
uniform angular velocity (relative to an arbitrary stationary radius from the
earth’s centre), while the moon M travels along the epicycle uniformly in
the opposite direction (relative to the radius from the earth’s centre to the
epicycle’s centre). The angular velocity of the moon on its epicycle is slightly
less than that of the epicycle’s centre on the deferent. In the other model (Fig.
2), the centre of an eccentric circular orbit C revolves with a slow uniform
westward motion along a circle concentric with the earth T , while the moon M
travels along the orbit with a uniform eastward motion (relative to the radius
from the earth’s centre to the eccentre’s centre). The two models, as Ptolemy
proves, are kinematically interchangeable; that is, any set of positions in space
of the moon for specific dates generated by the one model can be generated
identically by the other. Moreover, Ptolemy knows already that they are both
incorrect, because the lunar anomaly is not simply periodic. Nevertheless,
Ptolemy selects the epicyclic model as the basis for a preliminary lunar theory
in which the numerical parameters are determined by analysis of observations
of lunar eclipses. It is noteworthy that Ptolemy makes a point of showing that
the same parameters result from several different sets of observation reports,
thus establishing that the preliminary model is computationally valid for all
eclipses (and by extension, all oppositions).
The motivation for Ptolemy’s selection of the epicyclic model only becomes
fully evident when he shows in Almagest 5.2 how it is defective. He finds that
the “equation,” or difference, between the moon’s observed position and its
mean position (that is, the direction to the epicycle’s centre according to the
model) is in general greater than the model predicts, with the discrepancy
vanishing when the moon is at 0◦ or 180◦ elongation from the sun and max-
imum when it is at ±90◦ elongation. In an epicyclic model the equation is
explained by the planet’s motion on the epicycle, so that the new phenom-
enon (essentially equivalent to “evection” in later lunar theory) would amount
M
E
Fig. 2.1. Ptolemy’s simple epicyclic model for the moon.
C
T
Fig. 2.2. Eccentre model equivalent to the model of Fig. 1.

30 Alexander Jones
to an apparent enlargement of the epicycle.6 Ptolemy accounts for this easily

(Fig. 3) by replacing the concentric deferent in the preliminary model with an
eccentric deferent, the centre C of which revolves around the earth T at a rate
such that the epicycle’s centre E (which is still revolving uniformly as seen
from T ) is furthest from the earth whenever the mean moon and the mean
sun are aligned or diametrically opposite. Now it is true that if Ptolemy had
employed an eccentric orbit to effect the anomaly in Book 4, he could have
corrected the model in Book 5 by adding an epicycle (or even a second, inde-
pendent eccentricity), but the relation of the components to the phenomena
would have been much less intuitive. And in any case Ptolemy fine-tunes the
model, for closer agreement with observations, by stipulating that the moon’s
motion on its epicycle is uniform as measured relative to a revolving radius,
not drawn from the centre of the deferent C or from T , but from a distinct
point D such that T is always the midpoint of C and D. This could not, I
believe, be translated in any straightforward way into a model in which the
primary anomaly is produced by an eccentre.
E M
D T C
Fig. 2.3. Ptolemy’s eccentre-and-epicycle model for the moon.
Thus all the stages from the selection of a basic model type to the final
model are motivated in Ptolemy’s exposition by the requirements of agreement
with observations, simplicity, and a clear one-to-one correspondence of the
6
For the relationship between Ptolemy’s so-called second anomaly of the moon
and the component called “evection” in modern lunar theory, see Neugebauer 1975
v. 3, 1108–1109.
elements of the model to the basic facts about the moon’s motion. A similar
account could be given for Ptolemy’s deduction of the models for the sun
and the five planets. Ptolemy makes no appeal in these parts of the Almagest
to physical constraints arising from the corporeal nature of the model. But
this is because those constraints have already been taken into account at a
still earlier stage, the decision to build all the models out of uniform circular
motions, which is made once and for all in Almagest 3.3, just before the first
discussion of the sun’s model. Here Ptolemy writes:
The next task being to exhibit also the apparent anomaly of the sun,
the assumption must first be made that the shiftings of the planets
[including the sun and moon] in the trailing direction of the heav-
ens [i.e., westward] are uniform, just like the movement of the whole
[heavens] in the leading direction [i.e., the daily eastward rotation of
the heavens], and they are circular by nature, that is, the straight
lines that are imagined as leading the heavenly bodies or their cir-
cles in their revolutions sweep out in all cases equal angles in equal
times with respect to the centres of each one’s revolutions, while the
apparent anomalies pertaining to them are produced by the positions
and arrangements of the circles on their spheres, by means of which
they make their motions, and nothing in nature really occurs that is
foreign to their eternity in connection with the imagined irregularity
of the phenomena.
This is one of only a handful of references in the Almagest to the circles in the
models as being on the surfaces of spheres; when he does this, it is always in
a matter-of-fact way, implying that the reader will already be familiar with
the conception. In this particular passage Ptolemy uses language connecting
the idea of uniform circular motion with physical nature and eternity, so that
ether, though not explicitly named, is inevitably called to mind.
And this brings us back to Ptolemy’s very first chapter, Almagest 1.1.
Here he defines the science of which his subject matter is a part, which he
calls “mathematics” (the Almagest’s real title is Mathematical Composition),
as the study of shapes and spatial movements in all kinds of bodies, whether
eternal and etherial or perpetually changing and composed of the four ele-
ments. Mathematics offers “sure and unshakeable knowledge,” and when con-
cerned with the etherial heavens, this knowledge is as eternal as its objects.
In other words, the conviction that the heavens are composed of etherial bod-
ies, which are by their composition both eternal and subject to no kind of
change except circular revolution, guarantees the legitimacy and truth of the
kind of reasoning that the Almagest embodies. It is noteworthy that, while
practically every other theoretical hypothesis in the Almagest is justified by
some empirical argument, the hypothesis of the etherial nature of the heavens
is given axiomatically at the beginning.
His claim to be arriving at “sure and unshakeable knowledge” in the Al-
magest turns out in practice to have certain limitations. Numerical parame-
32 Alexander Jones
ters are, by his confession, knowable only approximately, and in particular

the rates of rotation become more precisely known as we accumulate a longer
temporal span of observation reports. Ptolemy does not say outright whether
he believes that his specific model structures (exclusive of their numerical
parameters) are certainly valid. What the Almagest does affirm through its
broad plan is that Ptolemy’s models suffice to explain all the known phenom-
ena of the heavenly bodies, including eclipses, planetary retrogradations, and
visibility conditions. But the impression of finality is moderated, not only by
the way that Ptolemy recounts his discovery of the moon’s evection (might
there be other phenomena waiting to be noticed?), but also by his passing ref-
erence to alterations he had only lately made in his models for Mercury and
Saturn. In a famous passage towards the end of the work (Almagest 13.2), he
justifies the complexity of his models for the latitudinal motion of the planets
by affirming that the principle of simplicity in models should not be allowed to
override the necessity to account for the phenomena, since what seems com-
plex to us with our experience of the imperfections of mechanisms built from
the four mundane elements may be simple to essences that are eternal and
free from hindrance. Implicit in this is his confidence that his models really
are the simplest that can be brought into agreement with observation.
Ptolemy’s reticence regarding any but the most fundamental properties of
ether and regarding the way in which the geometrical objects that constitute
the Almagest models are supposed to be instantiated in etherial “spheres” in
the actual heavens may be due partly to a reluctance to digress from the core
subject matter of the book, but another reason may be that he had not yet
given these topics much thought (just as he tells us in Almagest 2.13 that he
has not yet worked out the list of geographical locations that he eventually
delivered in the Geography). In a much later work, the Planetary Hypotheses,
Ptolemy has considerably more to say about the spheres.7
The Planetary Hypotheses is ostensibly an exposition of the Almagest mod-
els, with some revisions, described in a manner that will be helpful for people
who wish to make demonstration models or planetaria, with the parts either
manually adjustable to their positions at any date or driven by a mechanism.
After a first book in which Ptolemy sets out the parameters of all the models
individually and proposes a scheme for nesting the models one inside the next,
from the moon’s model outwards to those of Saturn and finally the fixed stars,
he turns in Book 2 to a consideration of the models as three-dimensional cor-
poreal objects, that is, the “spheres” alluded to in the Almagest. Here Ptolemy
engages in an extended discussion of his notion of how etherial matter works.
7
The original Greek text of the Planetary Hypotheses is extant only for the first
part of Book 1, for which see Heiberg 1907, 70–106; there is no reliable modern
translation from the Greek. The whole of Book 1 in the medieval Arabic translation
is edited and translated into French in Morelon 1993. For the Arabic text of Book 2,
one currently depends on the German translation of L. Nix in Heiberg 1907, 111–145
and the facsimile of a manuscript in Goldstein 1967. Murschel 1995 is an excellent
synopsis of the work.
Etherial bodies, he says, are subject to no external influence or alteration.

To each independent motion in the kinematic models there corresponds a
rotating etherial body incited into motion by the power of the visible heavenly
body that it bears. These visible bodies (i.e., the sun, moon, planets, and
stars) are the same in composition as the matter that surrounds them. They
differ, however, in that they issue rays that have a power to penetrate other
bodies, analogous to our intellects and vision. Moreover, their ability to set
their spheres in motion is analogous to the power of our minds to cause our
bodies to move; but in the celestial case the movement is utterly effortless.
Aristotle’s cosmology had been strongly influenced by Eudoxus’ astronom-
ical models, in which the motions of the heavenly bodies were produced by
combinations of circular motions all concentric with the centre of the cos-
mos; hence he could ascribe to ether a “natural motion” always perpendicular
to any radius from the cosmic centre (in contrast to the natural motion of
the four elements, which is always rectilinearly centripetal). For Ptolemy this
cannot do, but he proposes a novel principle, that what Aristotle had charac-
terized as natural rectilinear motion is in fact only natural to a body that has
been removed from its “natural place.” The etherial bodies, being already in
their natural place, are subject to no tendency to migrate up or down, but are
free to stand still or rotate effortlessly. “Mathematics” (i.e., deductive math-
ematical astronomy in the style of the Almagest) allows for two possibilities
for the shapes of the etherial bodies. On the one hand, they can be spherical
shells and solid spheres; but if so, they do not have to be imagined as being
driven in their rotations by their axes, as in a mundane machine. Indeed, the
entire polar regions of the spheres seem to Ptolemy to be superfluous to their
motions, so that he is prepared to restrict the mobile bodies to equatorial
slices of spheres and spherical shells, so-called “tambourines” and “rings,”
which are presumably sandwiched between regions of stationary ether. Since
distinct etherial bodies can slide freely against each other, there is no need to
imagine “unwinding” spheres that cancel out the revolutions of outer spheres,
such as Aristotle had imposed on his mechanistic interpretation of Eudoxus
in Metaphysics Λ.
When we come to the detailed description of each heavenly body’s physical
model, we find that the basic conception is similar to Theon of Smyrna’s, but
extended to include eccentric as well as epicyclic motions. Fig. 4 (a cross-
section through the plane of the moon’s orbit) shows how Ptolemy conceives
of the arrangement of etherial bodies that bring about the moon’s motion,
on the assumption that the bodies are complete spheres or spherical shells.
The entire apparatus must be thought of as being spun about the poles of the
celestial equator with the daily rotation of the heavens. The outermost shell
A rotates around the poles of the ecliptic with the slow motion of the moon’s
nodes. Within this is a shell B rotating around the poles of the inclined plane
of the moon’s orbit at the rate that, in the model of Almagest Book 5, the
centre of the moon’s eccentre revolves around the earth relative to the nodes.
Cut out of shell B (and actually dividing it into two noncontiguous parts) is an
34 Alexander Jones
eccentric shell C that has embedded within it the solid epicyclic sphere D. C
and D together revolve uniformly as seen from the centre of the cosmos T with
the rate that the centre of the epicycle revolves around the earth in Almagest
5. Finally, the epicyclic sphere rotates, carrying embedded close to its surface
the moon M itself, producing the primary anomaly. This physical model is
wholly consistent with the Almagest model, except that Ptolemy abandons
the special radius with respect to which the moon’s regular revolution on the
epicycle is reckoned, instead stipulating that the moon’s revolution is uniform
relative to the radius from the centre of the cosmos. At the beginning of the
Planetary Hypotheses, Ptolemy writes that the models as set out in this work
incorporate some revisions to the Almagest models based on newer analysis
of observations, but also that he is making some minor simplifications purely
for the sake of an easier construction of demonstration models; one is left
uncertain which kind of change is being made here in the lunar model.
D M
C
A B
T
Fig. 2.4. Cross-section of Ptolemy’s etherial-spheres model for the moon.

To the extent that the Planetary Hypotheses is intended as a description

of the reality of the heavens (as opposed to its professed purpose of giving
designs for didactic three-dimensional illustrations that we can construct), the
pronouncements in the book are more equivocal than those of the Almagest.
Ptolemy is quite sure of the etherial composition of the heavens, and also
quite sure of the fundamental geometrical structures of the celestial motions;
but the specific way that these geometrical structures are embodied in the
etherial matter is open to alternatives (complete spheres or equatorial slices,
spinning driven by cosmic souls or by planetary rays or by the axes). Ptolemy
generally tells us which way he is inclined to choose, but these topics are not
the province of the “unshakeable knowledge” of mathematics.
With the Tetrabiblos (a work written after the Almagest but probably well
before the Planetary Hypotheses), Ptolemy turns from pure contemplation of
the celestial realm of ether to an investigation of the action of the heavens upon
the world of the four elements.8 The fundamental assumption, comparable in
its role in Ptolemy’s astrology to the hypothesis of the uniform circular motion
of ether in his astronomy, is propounded in Tetrabiblos 1.2:
The fact would appear utterly obvious to everyone through even a few
considerations that some power is given forth and reaches from the
etherial and eternal nature to all the region around the earth, which
is in all respects subject to change, with the first elements below the
moon, i.e., fire and air, surrounded and directed by the movements in
the ether and surrounding and directing all the rest, i.e., earth and
water and the plants and animals within them.
There are, however, important differences between these fundamental hy-

potheses. In both, the etherial matter is simply a given. But the property of
uniform circular motion in the Almagest is justified on a priori grounds (cir-
cular rotation being the only kind of eternally unchanging motion that can
be conceived), whereas the property of exerting a power of change on the four
elements is argued directly from empirical facts. Ptolemy backs it up with a
series of examples of situations where laymen know perfectly well, and act
on the knowledge, that the motions of the heavenly bodies affect (or at least
predict) mundane phenomena such as seasons and weather, floods and tides,
and the generation of plants and animals. Secondly, and more significantly for
our topic, uniform circular motion is a mathematical behaviour, which leads
immediately to the modelling of the Almagest, while power to change the ele-
ments is by its nature not mathematical, since it operates with qualities such
as hot and cold, wet and dry. And this creates a problem: how can cause-
and-effect relations operating at the qualitative level, and largely within the
“irregular” sublunary part of the cosmos, be well described by the predictive
mathematical models of astrology?
8
Robbins 1940.
36 Alexander Jones
It must be confessed that Ptolemy evades this problem. Essentially Ptolemy

relies on the orderliness of the heavens to justify the mathematical structure
of the predictive schemes of his astrology, but appeals to the disorderliness
and complexity of the mundane environment to explain why astrological pre-
dictions, even when made according to the most correct principles, are not
certain to be borne out. Moreover, the schemes that Ptolemy sets out to ratio-
nalise are in great part the rather chaotic traditional practices of the astrology
of his time. Though he allows himself to reform or suppress some of this tra-
dition in accordance with his physics, he can only go so far in that direction
since his claim that astrology is a valid science depends heavily on the as-
sumption that the traditional practices really work. One can sense his delight
in finding here and there some apparent pattern in the jumble, for exam-
ple when he finds harmonically significant ratios embedded in the “aspects”
(astrologically significant linkages of zodiacal signs forming sides of triangles,
squares, or hexagons), or when–indulging in a topos beloved of authors–he
recovers from an old, neglected, and nearly illegible manuscript a gloriously
complicated rationale for the seemingly nonsensical but empirically verified
“Egyptian” system of terms (divisions of zodiacal signs associated with indi-
vidual planets). Elsewhere Ptolemy almost seems to give up trying to explain,
and lapses into catalogues of astrological associations scarcely distinguishable
from the manuals of astrologers who were less sophisticated from a scientific
point of view.
Optics, which in antiquity meant the study of visual perception, was a
more fruitful subject for the interplay between mathematical and physical
modelling. As in astronomy, there existed a range of well-established phenom-
ena that lent themselves to explanation in terms of a geometrical model, in
this case the “visual ray,” diagrammed as a straight line extending from the
viewer’s eye to a point on an object. The hypothesis is that when a visual
ray exists between the eye and a point on a body, that point is seen. The eye
(or mind) always perceives the seen point as being in the direction in which
the ray sets out from the eye, even if the ray is reflected or refracted at the
interface between two bodies. This directional information provides the eye
with indications of the shape, position, and movement of bodies; on the other
hand, the ray conveys to the eye either limited knowledge or no knowledge at
all about the distances to the point it perceives.
But what are these lines really? The classic exposition of Greek geomet-
rical optics, repeatedly cited or paraphrased by later authors, was Euclid’s
Optics. This treatise does not explain the physical nature of the visual rays
but does specify that they are discrete, with spaces between the individual
rays that grow wider as the rays fan out towards more distant objects; more-
over, some of the explanations of visual phenomena appear to assume that
the rays are somehow attached to the eye (so that as the eye moves, the
rays move accordingly). The gaps between the rays provide an explanation of
the fact that objects are seen less clearly, or not seen at all, as they become
more distant. But the gaps also lent themselves to a physical interpretation of
the rays that is found in Peripatetic texts approximately contemporary with

Euclid. According to this interpretation the eye emits, through pores in its
surface, exquisitely thin and straight rods of matter (typically composed of
fire) that extend with unimaginable swiftness until they encounter a body, at
which point their progress may be impeded (in which case a visual perception
of the body occurs) or reflected (if the surface is smooth enough) or refracted
(if the body is porous).9
In his Geography Ptolemy treats the problem of planar map projections
as essentially one of optics: how can one devise an appropriate framework of
lines representing parallels and meridians to give the illusion of a part of a
spherical surface?10 Ptolemy takes for granted many elementary perspective
consequences of the hypothesis of rectilinear visual rays. In one passage con-
cerning the relation between the appropriate size of a map and the expected
distance of the spectator from it, Ptolemy invokes the Euclidean gaps, which
suggests that at this stage in his career (not long after the Almagest and Tetra-
biblos), if he had any opinion at all about the physical nature of vision, it was
not far removed from the Peripatetic notion of discrete material emanations.
When he came to write his Optics (a work that I suspect was among his
last writings), Ptolemy had changed his mind.11 He now speaks of the eye
as emitting an entity conventionally translated as the “visual flux,” a cone
comprising a geometrical continuum of rectilinear rays that are stronger or
weaker in perceptive power both to the extent that they have to extend a
shorter or longer distance from eye to object, and to the extent that they are
nearer to or further from the central axis of the cone.12 It is to this weakening
of the rays, rather than any supposed gaps between them, that fuzzy vision
of distant or peripheral objects is due. Thus Ptolemy’s geometrical treatment
of visual phenomena thus preserves the parts of the Euclidean scheme that
depend on the rectilinearity of the visual rays (namely, perspective phenom-
ena, reflections, and refractions) but replaces the somewhat clumsy Euclidean
handling of visual resolution with a more flexible and powerful hypothesis.
Unfortunately the entire first book of the Optics is lost, and with it
Ptolemy’s discussion of the physical makeup of the visual flux. Obviously
he cannot have thought of it as a body, at least not the kind that displaces
other bodies that formerly occupied its space, which is the only kind of body
9
The Peripatetic texts and their possible relation to the Euclidean model are
discussed in Jones 1994.
10
Berggren and Jones 2000.
11
The Optics survives, lacking its beginning and end, only in a medieval Latin
translation of an Arabic translation, a circumstance that causes great difficulties of
interpretation. The French translation in Lejeune 1989 and the English one in Smith
1996 are both useful, though under the circumstances neither can claim to represent
Ptolemy’s meaning exactly throughout.
12
The Latin term rendered as “visual flux” is uisus, which almost certainly repre-
sents the same Greek word opsis that, in Euclidean optics, refers to the single visual
rays; but Ptolemy used a different word when he meant an individual line of sight.
38 Alexander Jones
that is envisioned in Aristotle’s or in Peripatetic physics. It seems likely that

Ptolemy resorted to ideas from Stoic physics, which allowed for having distinct
elements occupy the same space as if in layers. In this manner the Stoics could
hypothesise that the entire cosmos was pervaded and regulated by pneuma.
Ptolemy may have suggested that the eye issues the visual flux as an over-
lapping layer of matter in the space between eye and object; or perhaps more
likely, he could have attributed to the eye a faculty of radiating a tensing
power, creating the flux by means of the pneuma already present in the inter-
vening space. We recall that in the Planetary Hypotheses he asserted a kinship
between the motive power of the heavenly bodies and the analogous power in
living things. As it happens, one of the very few references to Ptolemy’s Optics
in other authors that appear to pertain to its lost first book is a sentence in
a work on physical topics by the eleventh-century Byzantine writer Simeon
Seth: “Ptolemy says in his Optics that the visual pneuma is etherial and com-
posed of the fifth element.” This “visual pneuma” is probably the substance of
the cone of the visual flux, and so we have a remarkable fusion of Aristotelian
and Stoic element theory. The sixth-century philosopher Simplicius gives us
a further clue when he writes:
It should be noted that Ptolemy in his book On the Elements and in

his Optics, and the great Plotinus, and Xenarchus in his Difficulties
Addressing the Fifth Element, assert that motion in a straight line
belongs to the elements when they are still in a place that is not
natural to them, but (such motion) no longer belongs to them when
they have assumed their natural place.... Manifestly they do not move
when they are completely in their natural state, but, as the aforesaid
men, i.e., Ptolemy, Xenarchus, and Plotinus, say, when they are in
their natural state and in their proper places the elements either stand
still or move in a circle.
This is precisely the notion that we have seen Ptolemy putting forward
in the Planetary Hypotheses—which Simplicius does not cite here. It is ob-
vious why Ptolemy would have repeated it in a (no longer extant) work on
the elements; but what relevance can it have had in the Optics? I suspect
that Ptolemy was invoking it here for a purpose converse to his purpose in
the Planetary Hypotheses. There, the point was that etherial bodies in the
heavens can spin freely and effortlessly even if their revolution is not concen-
tric with the centre of the cosmos; in the Optics, perhaps Ptolemy claimed
that the etherial matter of the visual flux, connected with our sight and thus
displaced from its natural place in the heavens, travels in straight lines. A final
observation worth making is that the Xenarchus cited by Simplicius as sharing
this idea was active about the late first century B.C., so that here we may be
able to identify the source of a principle that, in Ptolemy’s hands, simultane-
ously accounts for the geometrical properties of the models of astronomy and
of optics.13
I have saved for last what may have been Ptolemy’s first major effort at
mathematical modelling, the Harmonics.14 The subject of this work calls for
some explanation. Ancient Greek music was essentially melodic unison melody,
occasionally employing singing or playing at the octave or the sounding of
simultaneous distinct notes as an effect, but free of harmony in the modern
sense. There existed numerous systems of relative pitches (i.e., scales) in which
melodies could be composed, none of which involved a sequence of intervals
quite like the diatonic scales on which most modern Western music is based.
The science of harmonics, as Ptolemy presents it, investigates models that
explain why certain intervals and combinations of intervals are esthetically
pleasing and hence exist as constituents of the music actually produced in
Ptolemy’s time.
Unlike the kinematic models of the Almagest and the visual rays of the
Optics, the models of the Harmonics are not geometrical but arithmetical. The
model for any interval between musical pitches is a ratio of whole numbers,
the question at issue being what rules determine the whole-number ratios
that correspond to the intervals of existing musical scales. Ptolemy credits
the ratio model to the Pythagoreans, though he disagrees with what he sees
as their tendency to develop a priori modelling principles that are not referred
to empirical evidence in an appropriate manner. In the course of criticizing
the Pythagoreans (and the more fundamentally wrong-headed Aristoxeneans)
and evolving his own models, Ptolemy makes more explicit pronouncements
about the interplay between a priori and empirical reasoning in science than
in any of his other works.
Ptolemy’s harmonic models are built up from three kinds of esthetically
satisfying intervals: (a) homophones, i.e., intervals between notes that sound
nearly alike, being identical in pitch or separated by one or more octaves,
modelled by ratios always of the type m : 1, e.g., 1 : 1 or 2 : 1 or 4 : 1; (b)
concords, i.e., intervals between notes that sound different but akin, and that
form the more stable larger intervals in scales, modelled by ratios of the type
m : n such that m is often but not always equal to n + 1, e.g., 3 : 2 or 4 : 3
or 8 : 3; and (c) the smaller melodic intervals between consecutive notes of a
scale, which are almost always modelled by ratios of the type (n + 1) : n, e.g.,
9 : 8.
The ratios are observable through the devices or instruments that make
the notes. This is clearest in cases where the difference between notes follows
from a difference between lengths in an instrument. For example, in wind
13
The “fragments” of the lost part of Ptolemy’s Optics (there are only four
known) are collected in Lejeune 1989, 271. On Xenarchus, see Falcon 2001, 272,
s.v. “Senarco.”
14
Barker 1989, 270-391 provides the best of the existing translations. West 1992
is a splendidly lucid introduction to all aspects of Greek music.
40 Alexander Jones
instruments one can measure the length of the pipe, say from the reed of an
aulos (conventionally rendered by tin-eared classicists as “flute,” but actually
a double reed like an oboe or shawm) to one of the finger-holes. For his
harmonic demonstrations, Ptolemy prescribes instruments involving tensed
strings, since these allow the maximum control and precision in the tunings
and measurements. Thus it is by dividing a tensed string with a bridge into
two parts in the ratio 4 : 3 that Ptolemy establishes the association of this
ratio with the tetrachord, the principal fixed interval in the Greek scales (in
modern terminology, a “fourth”).
But Ptolemy knows that length is not the only factor contributing to pitch.
Thickness and density, among other characteristics of the bodies that produce
the notes, are other variables that determine pitch; for this reason, before
allowing us to try out ratios on a tensed string, Ptolemy instructs us to conduct
a careful check of each part of the string to ensure that equal short lengths
sound equal notes. Hence it is not at all easy to give a physical interpretation
to the numbers in the modelling ratios that fully explains the musical intervals.
Somehow a multiplicity of quantitative properties of a sounding body, some
of them more straightforwardly measurable than others, give rise to a single
abstract magnitude in the air in which the sound subsists.
In the chapters where he discusses the nature of sound and musical tone
(Harmonics 1.3-4), Ptolemy does not try to explain the nature of sound more
deeply than his initial definition that it is “a modification (pathos) of air when
it is struck” (Harmonics 1.1), except for the conclusion that differences in
pitch (“sharpness” and “heaviness”) are a form of quantity. He does, however,
restrict the scope of harmonic science to the study of sequences of discrete
sounds, each of which has a constant pitch, so that one may speak of stable
relations or “ratios” between the notes. The special status of whole-number ra-
tios enters the discussion circuitously, by way of the review of the Pythagorean
model, and although Ptolemy uses divisions of a tensed string to provide em-
pirical justification that the homophones and concords are modelled by ratios
of small whole numbers, he provides no a priori justification of this fact.
But patience is rewarded. When Ptolemy has completed his set task of
deducing a more or less complete set of models to describe the systems of
tuning current in his time (Harmonics 3.2), he embarks on a new project of
describing how harmonic theory illuminates our understanding of aspects of
the cosmos that have no direct connection with sound, namely the behaviour
of human beings and of the heavens. It turns out that harmonics is not really
a science concerning sound at all. It is a science that discovers far deeper and
more general truths about our world, exploiting one specific part of it that
happens to be exceptionally well adapted to the interplay between sensory
observation and rational deduction that, for Ptolemy, constitutes scientific
method. The true subject of harmonics is harmonia, “the form of rational
causation (i.e., causation arising from reason and intellect) that concerns good
ratios of motions,” and this is necessarily present in all things that can move
themselves, and above all in the most rational self-movers, namely, people and
celestial spheres.
What this means is that the special status of whole-number ratios is a man-
ifestation of the Good (in the Platonic sense) that the intellect apprehends and
puts into action. One way that our intellects do this is by constructing musical
instruments to produce sounds that fit the ideal ratios (since, after all, the
sounds spontaneously produced by natural objects would not be recognized
as music). Because of the close correspondence between measurable quantities
in the instruments and the notes that we hear (which we can compare but not
measure), we can discover the laws governing the order that our souls impose
on this external matter. But these same laws are also recognizable, Ptolemy
maintains, in the arrangement, motions, and powers of the heavenly bodies,
which we discover through astronomy and astrology, and they must exist in
our own characters, virtues, and emotions, where the quantitative relations
are not apparent to our senses.15
These closing chapters of the Harmonics have received faint praise from
modern readers, and it is undoubtedly true that the identification of detailed
correspondences between the elements of his theory of musical tunings and an
assortment of ethical, astronomical and astrological concepts is not Ptolemy’s
forte. But there can be no doubt that the principle motivating this péché de
jeunesse was close to Ptolemy’s heart, the conviction that the mathematical
behaviour that we find here and there in the cosmos is structure imposed for
the sake of the Good by minds upon a world that would otherwise be governed
by disorder.
Bibliography
Barker, A. 1989. Greek Musical Writings. Vol. 2. Harmonic and Acoustic The-
ory. Cambridge.
Berggren, J. L., and A. Jones. 2000. Ptolemy’s Geography: An Annotated
Translation of the Theoretical Chapters. Princeton.
Dupuis, J. 1892. Théon de Smyrne philosophe platonicien: Exposition des con-
naissances mathématiques utiles pour la lecture de Platon. Paris.
Falcon, A. 2001. Corpi e movimenti. Il De caelo di Aristotele e la sua fortuna
nel mondo antico. Napoli.
Goldstein, B. R. 1967. The Arabic Version of Ptolemy’s Planetary Hypotheses.
Transactions of the American Philosophical Society 57.4. Philadelphia.
Heiberg, J. L. 1907. Claudii Ptolemaei Opera quae exstant omnia. Vol. 2.
Opera astronomica minora. Leipzig.
15
See Swerdlow 2004 for a discussion of this part of the Harmonics, with emphasis
on Ptolemy’s celestial harmonics.
42 Alexander Jones
Jones, A. 1994. “Peripatetic and Euclidean Theories of the Visual Ray.” Physis
31, 47-76.
Lejeune, A. 1989. L’Optique de Claude Ptolémée dans la version latine d’après
l’arabe de l’émir Eugène de Sicile. 2nd edition. Collection de travaux de
l’Académie internationale d’histoire des sciences 31. Leiden.
Lloyd, G. E. R. 1978. “Saving the Appearances.” Classical Quarterly 28, 202–
222. Reprinted with preface in G. E. R. Lloyd, Methods and Problems in Greek
Science, Cambridge, 1991, 248-277.
Morelon, R. “La version arabe du Livre des hypothèses de Ptolémée.” Mélanges
de l’Institut dominicain d’études orientales du Caire 21 (1993) 7–85.
Murschel, A. 1995. “The Structure and Function of Ptolemy’s Physical Hy-
potheses of Planetary Motion.” Journal for the History of Astronomy 26, 33–
61.
Neugebauer, O. 1975. A History of Ancient Mathematical Astronomy. 3 vols.
Berlin.
Robbins, F. E. 1940. Ptolemy: Tetrabiblos. Loeb Classical Library 435. Cam-
bridge, Mass.
Smith, A. M. 1996. Ptolemy’s Theory of Visual Perception: An English Trans-
lation of the Optics with Introduction and Commentary. Transactions of the
American Philosophical Society 86.2. Philadelphia.
Swerdow, N. M. 2004. “Ptolemy’s Harmonics and the ‘Tones of the Universe’
in the Canobic Inscription.” Studies in the History of the Exact Sciences in
Honour of David Pingree, ed. C. Burnett, J. P. Hogendijk, K. Plofker, and
M. Yano. Islamic Philosophy Theology and Science: Texts and Studies 54.
Leiden. 137–180.
Toomer, G. J. 1984. Ptolemy’s Almagest. London.
. 1987. “Ptolemy.” Dictionary of Scientific Biography 15, 207–224.
West, M. L. 1992. Ancient Greek Music. Oxford.
3
Mathematics, Instruments and Navigation,
1600-1800
Jim Bennett
Museum of the History of Science, University of Oxford
This article will consider four episodes in the history of navigation that are
not part of the customary story. They will be set within a broad overview of
developments in the period 1600 to 1800. Will this topic qualify as history of
mathematics, as it must if it is to fall within the rubric of the Kenneth May
Lecture? On our contemporary understanding of the mathematical discipline
that may seem doubtful, and we may have to expand how we view mathe-
matics and alter our assumptions about the identity of the mathematician,
if we are to admit navigation in the seventeenth and eighteenth centuries.
The expanded view we will need sits more comfortably in the sixteenth cen-
tury, but the attitudes it contains are still sufficiently strong to illuminate the
seventeenth and to have at least some relevance to the eighteenth. However,
increasingly there were other visions of mathematics in play; there were chal-
lenges to the mathematical legitimacy of professional groups and a general
shift away from the identity characteristic of the earlier period.
That earlier identity saw mathematics — mainly geometry — as engaged
in a world of action and as central to success in that world. It was regarded as
a vital tool for practising in a range of professional arenas. Navigation is our
focus on this occasion, but the same was true of astronomy, surveying, archi-
tecture, warfare, engineering, and so on. The mathematician was one whose
mode of life engaged with these worlds and whose skills and originality were
valued there. While this was the arena of mathematical practice, mathemati-
cians were not engaged with natural philosophy, as they would become later.
They were not concerned with causal explanations of how the world operates
or with what might be its material constitution. That mathematical culture
— mathematics as action — was dominant in the sixteenth century, remained
strongly present in the seventeenth, and was subordinate in the eighteenth.
Taken as a whole, the four episodes of this article construct a narrative of
decline.
What characterises the period around 1600 in the generally received his-
tory of navigation? Emphasis is placed on the effective and commonly adopted
technique for finding latitude, based on altitude measurements of either the
44 Jim Bennett
stars (notably the Pole Star) or the sun. This was sufficiently successful to
alter navigational practice, at least in the open ocean, and to displace the
bearing and distance technique (based on the magnetic compass, the log for
measuring speed, the sand-glass for time, the traverse board for recording and
the chart for plotting position). ‘Bearing and distance’ was replaced by lati-
tude sailing, where a course would be set significantly to the east or west of
the target destination and followed until the relevant latitude was achieved.
This latitude would then be maintained while sailing west or east to landfall.1
As with other assumptions we will encounter, this one is only partly true,
because without a corresponding technique for finding longitude and in the
absence of permanently clear skies, the seaman still required dead reckoning
based on a bearing and distance technique. Latitude sailing and dead reckon-
ing were complementary and were used together, and this would be the case
for a couple of centuries to come.
It may be that the apparent solution for latitude has thrown our attention
too strongly on the complementary problem of longitude. We might be better
advised to be guided by a contemporary opinion of the inadequacies in navi-
gational practice, and to take Edward Wright’s Certain Errors in Navigation,
published in 1599, as a starting point.2 The errors in question were:
• problems with the magnetic compass, particularly in relation to magnetic
variation;
• problems with the design and use of the cross-staff, one of the instruments
used for measuring the altitudes required in latitude sailing;
• the inadequacies of the plane chart; and
• the inadequacies of astronomical tables used for navigation; in particular,
tables of declinations of stars and of the sun.
Wright does not nominate the longitude. At this stage there was no direct
method for finding longitude; the hope of finding a predictable relationship
between magnetic variation and longitude, based on a geometric ‘theoric’ (as
it was called) after the manner of the astronomers, had failed. Wright was
at the centre of the recognition of that failure through his promotion of Si-
mon Stevin’s Haven-Finding Art and of William Gilbert’s De Magnete.3 Both
1
For general accounts, see E.G.R. Taylor, The Haven-Finding Art: a History of
Navigation from Odysseus to Captain Cook (London, 1971); D.W. Waters, The Art
of Navigation in England in Elizabethan and Early Stuart Times (London, 1978);
J.B. Hewson, A History of the Practice of Navigation (Glasgow, 1983).
2
E. Wright, Certaine Errors in Navigation, Arising Either of the Ordinarie Er-
roneous Making or Vsing of the Sea Chart, Compasse, Crosse Staffe, and Tables
of Declination of the Sunne, and Fixed Starres Detected and Corrected by E. W.
(London, 1599)
3
S. Stevin, trans. E. Wright, The Hayen-Finding Art, or, the Way to Find any
Hauen or Place at Sea, by the Latitude and Variation (London 1599); W. Gilbert,
De magnete, magneticisque corporibus, et de magno magnete tellure (London, 1600).
3 Mathematics, Instruments and Navigation, 1600-1800 45
works assume that the magnetic variation is totally dependent on local con-
tingencies and is not subject to any global pattern or theoric. Wright does not
cover longitude because he is dealing with the actual practice of navigation in
the absence of a longitude method, and with the inadequacies of that practice.
In fact Wright’s book reflects much of the agenda for the development of
navigation and navigational theory in the seventeenth century. This is not to
say that he was influential in setting that agenda, but rather that the set of
problems he identified were in fact addressed in the seventeenth century, which
reflects well on his ability to identify the central errors in the practice of his
time. Each of my four episodes begins with one of the errors that concerned
Edward Wright in 1599.
Magnetic Variation and Planetary Motion
Wright was at the centre of the English interest in a set of questions concerning
both the possible application of variation to position finding at sea, and the
need to manage variation so as to use the steering compass as accurately as
possible. One outcome of this interest was the discovery of secular changes in
variation, announced by Henry Gellibrand in 1635.4
Rather than ending any hope of position finding by variation, as might
have been expected now that it was known that variation in a given loca-
tion changes continuously, the discovery of this secular dimension to variation
stimulated interest in finding some predictive account of the changes in varia-
tion — some more complex theoric that included a time variable. This may be
less surprising than meets the eye. A wholly contingent variation, depending
simply on the irregularities of the local terrain and to be discovered only by
measurement and mapping, is a daunting (if not boring) prospect. However,
if this pattern changes with time, there may be an underlying pattern that
might be discovered and linked to some physical hypothesis. The prospect at
least holds some interest.
The almanac publisher and teacher of navigation Henry Bond began his
work on variation following Gellibrand’s announcement, and from 1636 he
published predictions that variation, then some degrees to the east, would
reduce to zero in 1657 and then increase to the west.5 He stated that his
prediction was based on an account of the earth’s magnetism that would yield
a longitude method. The prediction turned out to be true, and Christopher
Wren was sufficiently impressed that he stated in his inaugural address as
Gresham Professor of Astronomy in 1657 that the study of variation may well
4
H. Gellibrand, A Discourse Mathematical on the Variation of the Magneticall
Needle (London, 1635).
5
H. Bond, ed., The Seaman’s Kalender (London, 1636).
46 Jim Bennett
yield a method for longitude ‘than which, former Industry hath hardly left
any Thing more glorious to be aim’d at in Art.’6
This much is familiar, but I want to take as my first noteworthy episode
Wren’s other interest in variation pursued at the same time. Using a very long
magnetic needle, he was making a detailed and systematic study of changes in
variation — not just the long-term movements, such as the gradual decrease
to zero in 1657, but also the cycles of change that lay within these. Since he
was aiming to detect an annual cycle within the overall pattern of change
(‘I hope to discover the Annual Motion of Variation & Anomalies in it’7 ), he
must have thought he was dealing with a link between physical characteristics
of the earth, or local phenomena that were amenable to experimental philos-
ophy, and astronomy. This, of course, sounds very Newtonian, since it has the
character of what has been called the ‘Newtonian synthesis.’ A fundamen-
tal causal agency, located in the earth and in celestial bodies, has both local
consequences we can investigate on earth and celestial consequences we can
measure in the heavens. These sets of consequences are essentially the same,
but they manifest themselves at different distances.
Wren pointed out that Kepler himself acknowledged his debt to William
Gilbert for the physical explanation Kepler gives of planetary motion based on
an interaction between magnetic bodies. Wren said that this elliptical astron-
omy required ‘Perfection,’ i.e., further elaboration of both its geometry and its
causal explanatory account, and he believed that this refinement would come
from a study of the magnetic cycles of the earth, which would be reflected
in its orbital motion and thus in astronomical measurements. Wren says that
the study of the earth’s magnetism is ‘a Kind of Terrestrial Astronomy, an
art that tells us the Motions of our own Star we dwell on.’8
Wren’s programme of magnetic dynamics gave way to an account based on
a different influence at a distance, but still involving an attractive force — now
a single, central force, governed by a distance law and combined with a prin-
ciple of rectilinear inertia. It is worth remembering that Wren was involved,
along with Hooke and Halley, in formulating the principles of this programme
for planetary dynamics.9 What seems to have been overlooked in the standard
history is his earlier projected synthesis. While it had a very similar scope and
ambition, it depended on a different set of physical observations and specula-
tions; it emerged from a sustained tradition of work on magnetism, which had
been driven by its importance for the practice and development of navigation.
6
C. Wren, Jnr, Parentalia: or, Memoirs of the Family of the Wrens (London,
1750), p. 206.
7
J.A. Bennett, ‘A Study of Parentalia, with Two Unpublished Letters of Sir
Christopher Wren’, Annals of Science, 30 (1973), 129-47, see p. 147.
8
Wren, op. cit., p. 206.
9
J.A. Bennett, ‘Hooke and Wren and the System of the World’, British Journal
for the History of Science, 8 (1975), 32-61.
From Plane to Mercator Sailing

Returning to Wright’s Errors for the starting point of my second episode, we
find him exercised by the continued use of the plane chart, i.e., one where
there was no systematic accommodation of the curvature of the earth by
means of a geometrical projection. In 1569 Gerard Mercator had published a
chart based on a projection that was admirably suited for use at sea, especially
for plotting compass bearings since they were projected as straight lines, but
Mercator had not revealed the geometry on which this was based and, until
that was available, there was no way that his invention could be generally
used.
Wright explained the mathematics behind the Mercator chart in Certain
Errors, and the following generation of English mathematicians, based at Gre-
sham College in London, translated this mathematics into a practical tech-
nique for seamen. They did so by providing two things:
• protocols for each of the different types of calculations a navigator might
face in plotting a position, and
• instruments that made these protocols achievable through routines of ma-
nipulation adapted to each problem.
The mathematicians involved in this work were Henry Briggs, one of
Wright’s associates, the first Professor of Geometry at Gresham College, and
Edmund Gunter, the third Professor of Astronomy. Briggs contributed to the
development of John Napier’s logarithms, while Gunter was the foremost de-
signer of mathematical instruments in the England of his day. The problem
with sailing by the Mercator chart was that it inevitably involved some en-
tanglement with trigonometry. A consequence of keeping the rhumb lines or
compass bearings straight for the convenience of the navigator was that the
scale of the chart was a function of the secant of the latitude.
This was beyond the regular seaman, but Gunter devised instruments that
reduced operating with trigonometrical functions to a series of manipulations
with instruments, and he explained both the instruments themselves and the
protocols for their use in his book De sectore et radio of 1623.10 The ‘sector’
of the title was a calculating instrument that had been proposed or modified
by a number of mathematicians, including Galileo; Gunter’s contribution was
both to include trigonometrical functions on the lines of his sector, and to
translate the calculations required for ‘Mercator sailing’ into routines that
could be carried out with this instrument.
The basic principles of the sector relied on the proportional characteris-
tics of similar triangles, but underlying the ‘radius’ — the second calculating
instrument in Gunter’s book of 1623 — was logarithms. The logarithms of
trigonometrical functions were provided as lines on Gunter’s radius, or ‘rule’
as it became commonly known, and were added or subtracted (that is, the
10
E. Gunter, De sectore et radio (London, 1623).
48 Jim Bennett
trigonometrical functions were multiplied or divided) using a pair of dividers.

It was not long before, instead of a single rule and dividers, some instruments
had pairs of logarithmic scales sliding along each other. The ‘sliding Gunter’
of the seventeenth century is the origin of the ubiquitous logarithmic slide
rule.
This critical development in the history of calculation arises from the con-
temporary problems of navigation and the established mathematical practice
of rendering techniques accessible to navigators through the designing of ap-
propriate instruments. No other set of circumstances in the period would have
promoted such an outcome: only in navigation was there a pressing need for a
complex and totally unfamiliar technique involving the management of large
numbers to be rendered accessible to a large population of relatively unso-
phisticated practitioners. This episode is an important moment in the history
of practical mathematics, but it is also significant for the history of naviga-
tion. The general development of routines of observation and calculation on
board ship, the growing discipline in the management of these routines, and
the gradual adoption and mastery of the Mercator chart are processes as im-
portant to progress in navigation as the more glamorous business of latitude
finding and searching for longitude.
The ‘radius’ of Gunter’s title De sectore et radio was so called because
Gunter originally advised that the logarithmic scales should be marked on
the radius of a cross-staff. This makes the link between the genesis of the log-
arithmic rule and contemporary problems in navigation even more emphatic.
Altitude Measurement by Backstaff and Octant

The cross-staff brings us back to Wright’s Certain Errors and the beginning
of my third episode. Wright was exercised by several problems with the use
of the cross-staff: the user had to look in two directions simultaneously; if he
was sighting the sun, he was looking at the sun; and — the point that Wright
emphasises — the centre of the angle required is at the centre of the observer’s
eyeball, and this inaccessibility results in further inaccuracy.11
So far as solar sights were concerned, for finding latitude from the meridian
altitude of the sun, an effective and popular solution came in the form of
an alternative instrument, the backstaff (see figure 1). In the backstaff the
quadrant is divided into two portions: usually a 65-degree arc, and one of 25
degrees drawn to a much larger radius. A near-sight, held to the eye, can be
moved along the 25-degree arc, while the user views the horizon through this
sight and a far-sight (called the horizon vane) at the common centre of the
two arcs. A shadow vane can be set to any position on the 65-degree arc. In
practice this is done so that the reading on the scale is some 10 or 15 degrees
11
W.F.J. Mörzer Bruyns, The Cross-Staff: History and Development of a Navi-
gational Instrument (Amsterdam, 1994).
Fig. 3.1. Backstaff by J. Gilbert, London, 1745. Museum

c of the History of Science,
Oxford.
less than the expected altitude. The two scales diverge from a common zero,
so that adding the two readings gives the angle subtended at the horizon
vane by the shadow vane and the near-sight. In making his measurement, the
navigator has his back to the sun and, keeping the horizon in view and the
shadow of the shadow vane on the horizon vane, moves the near-sight down the
25-degree arc until the maximum angle is reached, and to continue he would
have to begin to move the sight back. The sum of the two scale readings is
the meridian altitude of the sun. To find the altitude of the equator, account
must be taken of the solar declination, i.e. the angle of the sun above or below
the equator, which varies by the time of year. For this the observer must
have a table linking date and solar declination; such a table is sometimes
found engraved on the instrument. The latitude is then the complement of
the altitude of the equator. Wright’s concerns have been met: the user looks
in a single direction (that of the horizon) and looks away from the sun, while
the angle being measured is external to his eye.
Backstaves are relatively common, which shows that they were much used,
for generally speaking the survival rate for wooden instruments such as these
is very low. They were robustly made, intended to cope with the rigours of
seaborne life. Nonetheless, despite this unpretentious working context, if we
look closely at individual examples we find that they have quite sophisticated
features, that they are ingenious in their design, and that they have surprising,
even impossible ambitions for accuracy.
The general design maximises the potential for precision by magnifying the
measuring scale differentially without enlarging the whole instrument. The
backstaff would be impossible to manage if an entire quadrant were made
50 Jim Bennett
to the radius of the 25-degree arc. Since the 65-degree arc is used only for
setting the shadow vane to a particular reading, it need only be divided to
single degrees, but enlarging the arc where the measurement is taken allows it
to carry a much more closely divided scale. In fact the scale used is imported
from astronomical instruments — used, for example by Tycho Brahe and
Johannes Hevelius — namely, the diagonal or transversal scale. The scale is
commonly divided to degrees, then to 30 minutes, 10 minutes and 5 minutes.
But diagonal lines between the 10-minute divisions and crossed by 10 equally-
spaced arcs concentric with the main scale, carry the sub-division down to
one minute — an unrealistic ambition, given the nature of the observation.
Two features demonstrate further refinements of the backstaff design.
While the 65-degree scale is marked on the face of the arc, it is usual for
the 5-degree divisions to be repeated on the rim. The rim divisions never
quite coincide with the scale on the face, being consistently offset slightly. At
first this may seem to be carelessness, but it is found on every instrument, so
must be deliberate. It reflects the concern that the observer needs to register
the position of the centre of the sun, but it is the limb of the sun that de-
termines the extent of the shadow. The apparent discrepancy in the positions
of the rim divisions is meant to give the user the option to correct for the
semi-diameter of the sun. Of course there are other uncertainties at play, with
penumbra effects and so on, so that this feature only indicates a concern over
the problem; it is not a full solution.
The problem is completely solved by a second refinement: a convex lens
can be fitted to the 65-degree arc as an alternative to the shadow vane. This
focuses the light of the sun onto the horizon vane. It is often said that this was
for use in hazy conditions, when sunlight was insufficiently strong to cast a
clear shadow, but given the concern about recording the position of the sun’s
centre, the lens might have been preferable under a range of conditions. The
vanes so rarely survive that it is difficult to know to what extent lenses were
used.
Despite its unpretentious origins, emerging from the world of mathemat-
ical practitioners, instrument makers and navigators rather than university
mathematicians, the backstaff has considerable sophistication and surprising
ambition in its concern for precision. It was replaced eventually in the eigh-
teenth century by the octant, also known as the ‘Hadley quadrant’ after one
of its designers, the mathematician, optical experimenter and Fellow of the
Royal Society John Hadley. His designs were published in the Philosophical
Transactions after he had presented them to a meeting of the Royal Society
in 1731, a very different context for the introduction of a navigational instru-
ment. Hadley began, not by describing the instrument, but with what he calls
a ‘Principle in Catoptrics,’ on which he develops a geometrical construction
related to two successive reflections and elaborates five corollaries.12 This was
12
J. Hadley, ‘The Description of a New Instrument for Taking Angles,’ Philosoph-
ical Transactions, 37 (1731), 147-57.
a presentation tuned to his audience and to Hadley’s standing in the Royal

Society.
When Hadley moved to the practical realisation of his ideas, he offered
two designs. One was much like a design later attributed to Newton, on the
basis of a drawing found in Edmond Halley’s papers. Hadley spends less time
on his second instrument, but it was much closer to the arrangement that was
commonly adopted. Sea trials were requested at the Royal Society, by James
Bradley and Halley, and accordingly Hadley had Jonathan Sisson make a brass
octant, which was tried at sea and the results reported to the Society.
This development took place in the context of the Royal Society. Beyond
this institutional context, the instrument makers already had a successful and
familiar design in the backstaff, and they did not receive the new instrument
in a passive way, but assimilated to it some of the features of the backstaff.
For example, they dispensed with Hadley’s telescopic sight, substituting the
pinhole sight with which they and their customers were familiar. They added
the diagonal scale from the backstaff. In fact, the scales on early octants
are very similar to the backstaff scale, which is detrimental to the putative
accuracy of the observation. A scale on an octant that is physically similar
to one on a backstaff must be numbered at twice the rate, because the index
arm which carries the mirror moves through only half the angle moved by
the reflected ray. Therefore on an octant, the standard backstaff scale will
measure to only two minutes of arc.
In fact this is only an apparent loss, since the whole measurement with
the octant is a more accurate procedure than with the backstaff. Nonethe-
less, these adaptations show that the design was not simply adopted by the
makers — it was modified, through the incorporation of features of the ex-
isting instrumentation for measuring angles. In fact, the next development in
the octant, namely, the adoption of the vernier instead of the diagonal scale,
again taken from astronomical instruments, seems to have been an initiative
of the makers; it is not present in Hadley’s designs.
During this period — from, say, 1730 to 1750 — it was by no means clear
that the octant would come to dominate altitude measurements at sea. There
were a number of proposed designs, both for reflecting instruments and for
developments in the backstaff. One feature common to the latter proposals
was the focusing lens (known as a ‘Flamsteed glass’); in the contest between
the two classes of instrument, the possibility of a lens for focusing sunlight
on a hazy day was thought to be an important competitive feature of the
backstaff and of neo-backstaves. Another feature was that some of the neo-
backstaves incorporated an artificial horizon, so that they could be used when
the real horizon was obscured either by land or cloud. John Elton’s quadrant,
for example, was a developed form of backstaff, also made by Sisson, published
in the same volume of the Philosophical Transactions as the one containing
52 Jim Bennett
Hadley’s quadrant.13 The inclusion of an index arm, which is horizontal when

the measurement is taken, allowed Elton to add a longitudinal bubble level
as an artificial horizon, and to attach a vernier, in order to dispense with the
diagonal scale. Other neo-backstaves, such as the similar designs promoted by
the makers George Adams and Benjamin Cole in 1748, included the focusing
lens and the vernier. In 1733 Hadley himself described to the Royal Society
what he called a ‘Quadrant for taking a Meridional Altitude at Sea, when the
Horizon is not visible.’14 It is completely different from his octant, focusing
the sun’s image onto a target and having a bubble-level artificial horizon. So,
even Hadley was part of the neo-backstaff discussion.
My third episode, then, shows again that the familiar account, in this
case the evident superiority of Hadley’s quadrant and its rapid and untrou-
bled assimilation into the instrumentation of navigation, is inadequate. The
reality was more complex. The makers and practitioners already had a sta-
ble, sophisticated and successful instrument in the backstaff, and aspects of
this technology were incorporated into a fluid and active exchange between
mathematicians and makers in this period of evolution in the instrumental
techniques for finding latitude.
Astronomy and Longitude

The message from my first three episodes, that the past is more complex and
interesting than we might have imagined and that this complexity and interest
come to light when you look more closely, applies especially to my final topic,
the longitude. In this case it is astonishing that one can imagine that the past
can have as simple a story as has recently been told.15 While it is true that a
popular book must be less detailed and technical than a scholarly monograph,
here it is the morality of the tale, rather than its technical content, that has
been simplified to an implausible extent.
One of the more prominent and enduring seventeenth-century responses
to the inadequacy of astronomical tables - the fourth source of error identi-
fied by Wright - was the foundation of the Royal Observatory at Greenwich,
whose immediate aim was an astronomical solution to the problem of finding
longitude at sea. The context that brought Wright’s final Error to the notice
of the political establishment in England takes us back to Henry Bond. In
1676 Bond had published a book with the striking and apparently authorita-
tive title The Longitude Found. Examined by Six Commissioners Appointed
13
J. Elton, ‘A Descrition of a New Quadrant for Taking Altitudes Without an
Horizon, either at Sea or Land,’ Philosophical Transactions, 37 (1731), 273-9.
14
J. Hadley, ‘A Spirit Level to be Fixed to a Quadrant for Taking a Meridional
Altitude at Sea, when the Horizon is not Visible,’ Philosophical Transactions, 38
(1733), 167-72.
15
D. Sobel, Longitude: the True Story of a Lone Genius who Solved the Greatest
Scientific Problem of his Time (London, 1996).
by the King’s Majesty.16 The title was accurate in one respect — the method
had been examined and recommended by a Royal Commission — but had
a technique for finding longitude at sea really been discovered, as Bond had
been asserting on his own behalf for many years?
Soon afterwards a book by Peter Blackborrow appeared with the title The
Longitude Not Found: or, an Answer to a Treatise by Henry Bond.17 Despite
the ruling of the distinguished commissioners, who included the President of
the Royal Society, a former Professor of Astronomy in Oxford now Bishop
of Salisbury, and the current Professor of Geometry in Gresham College, at
least one commentator — the unknown Blackborrow, who scarcely seemed
qualified to judge — remained unconvinced. Hindsight seems to hand the ver-
dict to him. Bond’s scheme, as revealed to the Royal Commissioners, was in
the general class of the longitude-by-variation solutions, though he brought
measurements of magnetic inclination, or dip, into the equation. His theory
involved magnetic poles rotating in the atmosphere, lagging behind the mo-
tion of the earth and, for this reason, moving in a circle displaced from the
geographical poles.
What was in Bond’s favour at the commission? Its members had some
impressive evidence that the theory worked, namely, the prediction of zero
variation in London in 1657. Hazardous prediction and subsequent confir-
mation are said to be the touchstone of the scientific method, but modern
scientists might have difficulty subscribing to Bond’s theory of rotating poles
in the atmosphere, despite its empirical success. We know, in fact, that the
Royal Commission itself harboured serious doubts. Robert Hooke confided to
his diary that he, probably in collusion with other commissioners, ‘Found it
ignorant and groundless and fals but resolved to speak favourably of it.’18
He, or they, may have adopted this less than candid policy as part of a more
complex stratagem within the fractured contemporary arguments over the
longitude, and it is here that Wright’s concern about astronomical tables is
relevant. While the Royal Commission was active, a proposal was made at
court by a French associate of the King’s mistress for a lunar method for
finding longitude, and Charles referred the matter to substantially the same
commissioners as were considering Bond’s solution. They may have wanted to
dispatch this foreign intrusion by supporting Bond’s ineffectual theory. It was
in response to this interest in the lunar method, and the absence of the neces-
sary astronomical data, that Charles was moved to establish the observatory
at Greenwich.
Was the otherwise obscure Blackborrow right? Did he see through an im-
possible theory? Had he detected a conspiracy of vested interest that had
16
H. Bond, The Longitude Found: or, a Treatise Shewing an Easie . . . Way . . . to
Find the Longitude (London, 1676).
17
P. Blackborrow, The Longitude Not Found: or an Answer to a Treatise by
. . . Henry Bond (London, 1678).
18
H.W. Robinson and W. Adams, The Diary of Robert Hooke (London, 1935), p.
97.
54 Jim Bennett
resulted in a favourable account from a commission whose members did not

believe their own report? No. In fact, Blackborrow’s main concern was to at-
tack the Copernican theory, and Bond had assumed a moving earth. Once we
begin to look beneath the initial appearance, a story of simple virtue is rarely
sufficient.
In the eighteenth century we encounter an even thicker longitude plot,
one that has attracted interest and spawned controversy and division from its
inception. The spin that has been popular recently offers a single, linear tale
with a consistent moral: the just and heroic struggle of a virtuous individual,
John Harrison, against the calumnies of powerful vested interests. A humble
but determined man, with right on his side, took on the prejudice of a self-
serving establishment and, after a prolonged campaign based on integrity and
justice, won through to recognition and reward.
In fact this story is not at all new: it was told at the time, as part of
the campaign waged for the reward. In its final stages it gained the eloquent
championship of Edmund Burke, when in Parliament in 1773, he berated the
Government’s response to Harrison’s claims:
Where, Sir, is the dignity, where is the sense, where even the justice
of the representative of a great, powerful, enlightened, and maritime
nation, when a petition of a man is laid before them, claiming not a
favour, but justice; claiming that reward which law would give him,
and to see it refused — upon what principle? Why, a man of 83 is
to make new watches; and he is not only to make them, but to make
new voyages to the Indies to try them. Good God, Sir, can this be a
British House of Commons?19
What could the Prime Minister, Lord North, do but make excuses about delays
and arrange for a direct subvention from Parliament, outside the provision of
the Longitude Act?
Here again, just as in the case of Henry Bond a century earlier, opinion was
strongly divided, even among those not directly involved. Toward the end of
the century Harrison was famous as the winner of the fabulous reward and as
an outstanding watchmaker, but it would have been difficult to have found any
informed commentator who thought that, in any effectual sense, he had solved
the longitude problem. At the height of the debate surrounding Harrison’s
watch, one writer in 1765 considered the very idea that this longitude solution
had been proved by a voyage to the West Indies ‘such an insult upon common
sense as cannot be read without indignation.’20 After this stage in Harrison’s
campaign had resulted in the award of £10,000, another writer concluded in
1770 that the longitude was ‘still a secret, and likely to continue so, for tho
many thousands of pounds have been paid for the pretended discovery thereof,
19
The Parliamentary History of England, vol. xvii, 1773 (London, 1813), columns
841-3.
20
The Gentleman’s Magazine, 35 (1765), p. 34.
we remain just as wise as we were before the discovery, except the ill success
of it happens to teach us so much wit as to take better care of our money for
the future.’21 Clearly very different judgements were possible.
We might wonder why it is that in our own cynical and sophisticated age,
a story of virtue versus jealousy and greed, of ingenuity versus ignorance and
prejudice, and of humility versus arrogance and disdain has struck a chord.
Despite all the worldly wisdom we use to assess stories today, the analysis
we expect from expert reporters, the weight we give to different explanations,
and the account we take of individual motives, we are prepared to accept an
idealistic story located safely in the past. Do we really regard our predecessors
in the eighteenth century so differently from ourselves that their story is one
of virtue and villainy? Can we really forget that historians are inclined to
consider Georgian England as an ‘age of jobbery’ ? We might then believe that
the fabulous longitude prize of £20,000 could be won by getting the answer
right, and by getting that right answer, so to speak, properly ‘marked’. What
we have seen here is that a closer look often reveals a more complex and
qualified story, but one that is richer and more interesting.
21
W. Emerson, The Mathematical Principles of Geography, Navigation and Di-
alling (London, 1770), p. 172. See J. Betts, ‘Arnold and Earnshaw: the Practical
Solution,’ in W.J.H. Andrewes, ed., The Quest of Longitude (Cambridge, Massa-
chusetts, 1996), pp. 312-28. For a general account, see J. Bennett, ‘The Travels and
Trials of Mr Harrison’s Timekeeper’, in M.-N. Bourguet, C. Licoppe and H.O. Sibum,
eds, Instruments, Travel and Science: Itineraries of Precision from the Seventeenth
to the Twentieth Century (London, 2002), pp. 75-95.
4
Was Newton’s Calculus a Dead End? The
Continental Influence of Maclaurin’s Treatise of
Fluxions∗
Judith V. Grabiner
Pitzer College
4.1 Introduction
Eighteenth-century Scotland was an internationally-recognized center of know-

ledge, “a modern Athens in the eyes of an enlightened world.” [74, p. 40] [81]
The importance of science, of the city of Edinburgh, and of the universities in
the Scottish Enlightenment has often been recounted. Yet a key figure, Colin
Maclaurin (1698–1746), has not been highly rated. It has become a common-
place not only that Maclaurin did little to advance the calculus, but that he
did much to retard mathematics in Britain—although he had (fortunately)
no influence on the Continent. Standard histories have viewed Maclaurin’s
major mathematical work, the two-volume Treatise of Fluxions of 1742, as
an unread monument to ancient geometry and as a roadblock to progress in
analysis. Nowadays, few people read the Treatise of Fluxions. Much of the
literature on the history of the calculus in the eighteenth and nineteenth cen-
turies implies that few people read it in 1742 either, and that it marked the
end—the dead end—of the Newtonian tradition in calculus. [9, p. 235], [49,
p. 429], [10, p. 187], [11, pp. 228–9], [43, pp. 246–7], [42, p. 78], [64, p. 144]
But can this all be true? Could nobody on the Continent have cared to
read the major work of the leading mathematician in eighteenth-century Scot-
land? Or, if the work was read, could it truly have been “of little use for the
researcher” [42, p. 78] and have had “no influence on the development of
mathematics”? [64, p. 144]
We will show that Maclaurin’s Treatise of Fluxions did develop important
ideas and techniques and that it did influence the mainstream of mathematics.
The Newtonian tradition in calculus did not come to an end in Maclaurin’s
Britain. Instead, Maclaurin’s Treatise served to transmit Newtonian ideas in
calculus, improved and expanded, to the Continent. We will look at what
these ideas were, what Maclaurin did with them, and what happened to this
work afterwards. Then, we will ask what by then should be an interesting
∗
First published in the American Mathematical Monthly 104 (5) (1997), 393–410.
58 Judith V. Grabiner
question: why has Maclaurin’s role been so consistently underrated? These

questions will involve general matters of history and historical writing as well
as the development of mathematics, and will illustrate the inseparability of
the external and internal approaches in understanding the history of science.
4.2 The Standard Picture

Let us begin by reviewing the standard story about Maclaurin and his Treatise
of Fluxions. The calculus was invented independently by Newton and Leib-
niz in the late seventeenth century. Newton and Leibniz developed general
concepts—differential and integral for Leibniz, fluxion and fluent for Newton—
and devised notation that made it easy to use these concepts. Also, they found
and proved what we now call the Fundamental Theorem of Calculus, which
related the two main concepts. Last but not least, they successfully applied
their ideas and techniques to a wide range of important problems. [9, p. 299]
It was not until the nineteenth century, however, that the basic concepts were
given a rigorous foundation.
In 1734 George Berkeley, later Bishop of Cloyne, attacked the logical valid-
ity of the calculus as part of his general assault on Newtonianism. [12, p. 213]
Berkeley’s criticisms of the rigor of the calculus were witty, unkind, and—with
respect to the mathematical practices he was criticizing—essentially correct.
[6, v. 4, pp. 65–102] [38, pp. 33–34] [82, pp. 332–338] Maclaurin’s Treatise was
supposedly intended to refute Berkeley by showing that Newton’s calculus
was rigorous because it could be reduced to the methods of Greek geometry.
[10, pp. 181–2, 187] [9, pp. 233, 235] Maclaurin himself said in this preface
that he began the book to answer Berkeley’s attack, [63, p. i] and also to rebut
Berkeley’s accusation that mathematicians were hostile to religion. [78, p. 50]
The majority of Maclaurin’s treatise is contained in its first Book, which
is called “The Elements of the Method of Fluxions, Demonstrated after the
Manner of the Ancient Geometricians.” That title certainly sounds as though
it looks backward to the Greeks, not forward to modern analysis. And the
text is full of words—lots of words. So much time is spent on preliminar-
ies that it is not until page 162 that he can show that the fluxion of ay is
a times the fluxion of y. Florian Cajori, whose writings have helped spread
the standard story, compared Maclaurin to the German poet Klopstock who,
Cajori said, was praised by all, read by none. [10, p. 1881] While British
mathematicians, bogged down with geometric baggage, studied and revered
the work and notation of Newton and argued with Berkeley over foundations,
Continental mathematicians went onward and upward analytically with the
calculus of Leibniz. The powerful analytic results and techniques in eighteenth-
century Continental mathematics were all that mathematicians like Cauchy,
Riemann, and Weierstrass needed for their nineteenth-century analysis with
its even greater power, together with its improved rigor and generality. [9,
ch. 7] [49, p. 948] This story became so well known that it was cited by the
4 Was Newton’s Calculus a Dead End? 59
literary critic Matthew Arnold, who wrote, “The man of genius [Newton] was
continued by. . . completely powerless and obscure followers. . . . The man of
intelligence [Leibniz] was continued by successors like Bernoulli, Euler, La-
grange, and Laplace —the greatest names in modern mathematics.” [1, p. 54;
cited by [61, p. 151]
Now since I myself have contributed to the standard story, especially in de-
lineating the links among Euler, Lagrange, and Cauchy. [38, chs. 3–61] I have
a good deal of sympathy for it, but I now think that it must be modified.
Maclaurin’s Treatise of Fluxions is an important link between the calculus of
Newton and Continental analysis, and Maclaurin contributed to key develop-
ments in the mathematics of his contemporaries. Let us examine the evidence
for this statement.
4.3 The Nature of Maclaurin’s Treatise of Fluxions

Why—the standard story notwithstanding—might Maclaurin’s Treatise of
Fluxions have been able to transmit Newtonian calculus, improved and ex-
panded, to the Continent? First, because the Treatise of Fluxions is not just
one “Book,” but two. While Book I is largely, though not entirely, geomet-
ric, Book II has a different agenda. Its title is “On the Computations in the
Method of Fluxions.” [my italics] Maclaurin began Book II by championing
the power of symbolic notation in mathematics. [63, pp. 575–576] He ex-
plained, as Leibniz before him and Lagrange after him would agree, that
the usefulness of symbolic notation arises from its generality. So, Maclaurin
continued, it is important to demonstrate the rules of fluxions once again,
this time from a more algebraic point of view. Maclaurin’s appreciation of
the algorithmic power of algebraic and calculus notation expresses a common
eighteenth-century theme, one developed further by Euler and Lagrange in
their pursuit of pure analysis detached from any kind of geometric intuition.
To be sure, Maclaurin, unlike Euler and Lagrange, did not wish to detach the
calculus from geometry. Nonetheless, Maclaurin’s second Book in fact, as well
as in rhetoric, has an algorithmic character, and most of its results may be
read independently of their geometric underpinnings, even if Maclaurin did
not so intend. (In his Preface to Book I, he even urged readers to look at Book
II before the harder parts of Book I.) [63, p iii] The Treatise of Fluxions, then,
was not foreign to the Continental point of view, and may have been written
in part with a Continental audience in mind.
Nor was this algebraic character a secret open only to the reader of Eng-
lish. There was a French translation in 1749 by the Jesuit R. P. Pézénas,
including an extensive table of contents. [62] Lagrange, among others, seems
to have used this French edition (since he cited it by the French title [58, p. 17]
though he cited other English works in English [58, p. 18]). Pézénas’ transla-
tion, moreover, was neither isolated nor idiosyncratic, but part of the activity
of a network of Jesuits interested in mathematics and mathematical physics,
especially work in English, with Maclaurin one of the authors of interest to

them. [84, pp. 33, 221, 278, 517, 655] For instance, Pézénas himself translated
other English works, including those by Desaguliers, Gardiner’s logarithmic
tables, and Seth Ward’s Young Mathematician’s Guide [83, pp. 571–2] Thus
there was a well-worn path connecting English-language work with interested
Continental readers. Furthermore, the two-fold character of the Treaties of
Fluxions was noted, with special praise for Book II’s treatment of series, by
Silvestre-François Lacroix in the historical introduction to the second edition
of his highly influential three-volume calculus textbook. [52, p. xxvii] Unfor-
tunately, though, recognition of the two-fold character has been absent from
the literature almost completely from Lacroix’s time until the recent work by
Sageng and Guicciardini. [42] [78] We shall address the reasons for this neglect
in due course.
4.4 The Social Context: The Scottish Enlightenment

Another reason for doubting the standard picture comes from the social con-
text of Maclaurin’s career. Eighteenth-century Scotland, Maclaurin’s home,
was anything but an intellectual backwater. It was full of first-rate thinkers
who energetically pursued science and philosophy and whose work was known
and respected throughout Europe. One would expect Scotland’s leading math-
ematician to share these connections and this international renown, and he
did.
Although Scotland had been deprived of its independent national gov-
ernment by the Act of Union of 1707, it still retained, besides its indepen-
dent legal system and its prevailing religion, its own educational system. The
strength and energy of Scottish higher education in Maclaurin’s time is owed
in large part to the Scottish ruling classes, landowners and merchants alike,
who saw science, mathematics, and philosophy as keys to what they called
the “improvement” of their yet underdeveloped nation. [65, p. 254] [80, pp.
7–8, 10–11] [17, pp. 127, 132–3] Eighteenth-century Scotland, with one-tenth
the population of England, had four major universities to England’s two. [80,
p. 116] Maclaurin, when he wrote the Treatise of Fluxions, was Professor of
Mathematics at the University of Edinburgh. Edinburgh was about to become
the heart of the Scottish Enlightenment, and Maclaurin until his death in 1746
was a leading figure in that city’s cultural life.
Mathematics played a major role in the Scottish university curriculum.
This was in part for engineers; Scottish military engineers were highly in
demand even on the Continent. [17, p. 125] Maclaurin himself was actively
interested in the applications of mathematics, and just before his untimely
death had planned to write a book on the subject. [36] [68, p. xix] In addition,
mathematics and Newtonian physics were part of the course of study for
prospective clergyman. [80, p. 20] The influential “Moderate” party in the
Church of Scotland appreciated the Newtonian reconciliation of science and

religion. [16, pp. 53, 57]
Maclaurin’s position in Edinburgh’s cultural life was not just that of a tech-
nically competent mathematician. For instance, he was part of the Rankenian
society, which met at Ranken’s Tavern in Edinburgh to discuss such things as
the philosophy of Bishop Berkeley; the society introduced Berkeley’s philoso-
phy to the Scottish university curriculum. [24, p. 222] [17, p. 133] [65, p. 197]
Maclaurin and his physician friend Alexander Monro were the founders and
moving spirits of the Edinburgh Philosophical Society. [65, p. 198] With New-
ton’s encouragement, Maclaurin had become the chief spokesman in Scotland
for the new Newtonian physics. His posthumously published book, An Ac-
count of Sir Isaac Newton’s Philosophical Discoveries, was based on material
Maclaurin used in his classes at Edinburgh, and the book was of great interest
to philosophers. [24, p. 137] That book became well known on the Continent.
It was translated into French almost as soon as it appeared, by Louis-Anne
Lavirotte in 1749, and the first part appeared in Italian in Venice in 1762.
Another branch of Scottish science, namely medicine, also had many links
with the Continent and was highly regarded there. Medical students went
back and forth between Scotland, Holland, and France. [17, p. 135] [80, p. 7]
The best-known figures of eighteenth-century Scotland had major interac-
tions with, and influence upon, Continental science and philosophy. [39] [81]
Let it suffice to mention the names of four: the philosopher David Hume, who
was a student at Edinburgh in Maclaurin’s time; the geologist James Hutton,
who attended and admired Maclaurin’s lectures; [34, pp. 577–8] and, a bit
after Maclaurin’s time but still subject to his influence on Scottish higher ed-
ucation, the chemist Joseph Black and the economic and political philosopher
Adam Smith. Maclaurin himself had twice won prizes from the Académie des
Sciences in Paris, once in 1724 for a memoir on percussion, and then in 1740
(dividing the prize with Daniel Bernoulli, P. Antoine Cavalleri, and Leonhard
Euler) for a memoir on the tides. [79, p. 611] [39, pp. 400–401]
Scotland in the eighteenth century nurtured first-rate intellectual work on
mathematics, philosophy, science, medicine, and engineering, and did it all
as part of a general European culture. [39, p. 412] [81, passim] The Treatise
of Fluxions was the major mathematical work of a Scottish mathematician
of considerable reputation on the Continent, a major work philosophically
attuned to the enormously influential Newtonian physics and the Continen-
tally popular algebraic symbolism. Such a work would certainly be of interest
to Continental thinkers. Social considerations may not suffice to determine
mathematical ideas, but they certainly affect the mathematician’s ability to
make a living, to get research support, and to promote contact and commu-
nication with other mathematicians and scientists at home and abroad. And
so it was with Maclaurin.
4.5 Maclaurin’s Continental Reputation

An even better reason for not accepting the traditional view of Maclaurin
is that his work demonstrably was read in the eighteenth century, and was
read by the big names of Continental mathematics. He had a Continental
acquaintance through travel and correspondence. Even before the Treatise of
Fluxions, his reputation had been enhanced by his Académie prizes and by
his books on geometry. He was thus a respected member of an international
network of mathematicians with interests in a wide range of subjects, and
the publication of the Treatise of Fluxions was eagerly anticipated on the
Continent.
The Treatise of Fluxions of 1742 was Maclaurin’s major work on analysis,
incorporating and somewhat dwarfing what he had done earlier. It contains an
exposition of the calculus, with old results explained and many new results in-
troduced and proved. Maclaurin seems to have included almost everything he
had done in analysis and its applications to Newtonian physics. In particular,
the findings of his Paris prize paper on the tides were included and expanded.
His other papers, the posthumous and relatively elementary Algebra, and his
works on geometry as such—though highly regarded—do not concern us here,
but his Continental reputation was enhanced by these as well.
Let us turn now to some specific evidence for the Continental reputation
of Maclaurin’s major work. In 1741, Euler wrote to Clairaut that, though he
had not yet seen the Paris prize papers on the tides, “from Mr. Maclaurin I
expect only excellent ideas.” [47, p. 87] Euler added that he had heard from
England (presumably from his correspondent James Stirling) that Maclaurin
was bringing out a book on “differential calculus,” and asked Clairaut to keep
him posted about this. In turn, Clairaut asked Maclaurin later in 1741 about
his plans for the book, [66, p. 348] which Clairaut wanted to see before pub-
lishing his own work on the shape of the earth. [47, p. 110] Euler did get the
Treatise of Fluxions, and read enough of it quickly to praise it in a letter to
Goldbach in 1743. [48, p. 179] Jean d’Alembert, in his Traité de dynamique
of 1743, [22, sec. 37, n.] praised the rigor brought to calculus by the Treatise
of Fluxions. D’Alembert’s most recent biographer, Thomas Hankins, argues
that Maclaurin’s Treatise, appearing at this time, helped persuade d’Alembert
that gravity could best be described as a continuous acceleration rather than
a series of infinitesimal leaps. [44, p. 167] D’Alembert’s general approach to
the foundations of the calculus in terms of limits clearly was influenced by
Newton’s and Maclaurin’s championing of limits over infinitesimals, in par-
ticular by Maclaurin’s clear description of limits in one of the parts of his
Treatise of Fluxions that explicitly responds to Berkeley’s objections (and
which incidentally may be the first explicit description of the tangent as the
limit of secant lines; see Section 7). [44, p. 23] [63, pp. 422–3] Lagrange in
his Analytical Mechanics [55, p. 243] said that Maclaurin, in the Treatise of
Fluxions, was the first to treat Newton’s laws of motion in the language of
the calculus in a coordinate system fixed in space. Though C. Truesdell [80,
pp. 250–3] has shown that Lagrange was wrong because Johann Bernoulli and
Euler were ahead of Maclaurin on this, the fact that Lagrange believed this
is one more piece of evidence for the Continental reputation of Maclaurin as
mathematician and physicist.
4.6 Maclaurin’s Mathematics and Its Importance

The previous points show that Maclaurin could have been influential, but
not that he was. Five examples will reveal both the nature of Maclaurin’s
techniques and the scope of his influence: a special case of the Fundamen-
tal Theorem of Calculus; Maclaurin’s treatment of maxima and minima for
functions of one variable; the attraction of spheroids; what is now called the
Euler-Maclaurin summation formula; and elliptic integrals.
a. Key Methods in the Calculus. Two methods were central to the study
of real-variable calculus in the eighteenth and nineteenth centuries. One of
these is studying real-valued functions by means of power-series representa-
tions. This tradition is normally thought first to flower with Euler; it is then
most closely associated with Lagrange, and, later for complex variables, with
Weierstrass. The second such method is that of basing the foundations of the
calculus on the algebra of inequalities—what we now call delta-epsilon proof
techniques—and using algebraic inequalities to prove the major results of the
calculus; this tradition is most closely associated with the work of Cauchy in
the 1820’s. I have traced these traditions back to Lagrange and Euler in my
work on the origins of Cauchy’s calculus. [38, chs. 3–6] It is surprising, at least
if one accepts the standard picture of the history of the calculus, that both
of these methods—studying functions by power series, basing foundations on
inequalities—were materially advanced by Maclaurin in the Treatise of Flux-
ions. It is especially striking that the importance of Maclaurin’s work on
series-work based, it is well to remember, on Newton’s use of infinite series—
was recognized and praised in 1810 by Lacroix, who also linked it with the
series—based calculus of Lagrange. [52, p. xxxiii]
Maclaurin skillfully used algebraic inequalities in his proof of a special case
of the Fundamental Theorem of Calculus. He showed, for a particular func-
tion, that if one takes the fluxion of the area under the curve whose equation
is y = f (x), one gets the function f (x). In his proof, Maclaurin adapted the
intuition underlying Newton’s argument for this fact in De Analysi [69]—that
the rate of change of the area under a curve is measured by the height of the
curve—but Maclaurin’s proof is more rigorous. Although Maclaurin’s argu-
ment proceeds algebraically, the concepts involved resemble those of the Greek
“method of exhaustion” (more precisely termed by Dijksterhuis “indirect pas-
sage to the limit”). [26, p. 130] A key step in this Greek work is first to assume
that two equal areas or expressions for areas are unequal, and then to argue
to a contradiction by using inequalities that hold among various rectilinear
areas. Newton in the Principia had based proofs of new results about areas
and curves on methods akin to those of the Greeks. Maclaurin carried this
much further. It was Maclaurin’s “conservative” allegiance to Archimedean
geometric methods that led him to buttress the kinematic intuition of New-
ton’s calculus with algebraic inequality proofs.
Fig. 4.1.
What Maclaurin proved in the example under discussion is that, if the

area under a curve up to x is given by xn , the ordinate of the curve must be
y = nxn−1 , which is known to be the fluxion of f . [63, pp. 752–754] Maclaurin’s
diagram for this is much like the one Newton gave in the De Analysi. [69, pp.
3–4] Maclaurin began by saying that, since x and y increase together, the
following inequality holds between the areas shown:
xn − (x − h)n < yh < (x + h)n − xn . (4.1)
(Maclaurin gave this inequality verbally; I have supplied the “<” signs; also, I
use “h” for the increment where Maclaurin used ”o”) Now Maclaurin recalled
an algebraic identity he had proved earlier: [63, p. 583; inequality notation
added]
If E < F, then nF n−1 (E − F ) < E n − F n < nE n−1 (E − F ). (4.2)
(It may strike the modern reader that, since nxn−1 is the derivative of xn , this
second inequality is a special case of the mean-value theorem for derivatives.
I shall return to this point later.)
Now, letting x − h play the role of F and x play the role of E, E − F is h
and the first inequality in (2) yields
n(x − h)n−1 h < xn − (x − h)n .
Similarly, if F = x and E = x + h, then E − F = h and the second inequality

in (2) becomes
(x + h)n − xn < n(x + h)n h.
Combining these with inequality (1) about the areas, Maclaurin obtained
n(x − h)n−1 h < yh < n(x + h)n−1 h.
Dividing by h produces
n(x − h)n < y < n(x + h)n−1 . (4.3)
Recall that, given that the area was xn , Maclaurin was seeking an expression
for y, the fluxion of that area. A modern reader, having reached the inequality
(3), might stop, perhaps saying “let h go to zero, so that y becomes nxn−1 ,”
or perhaps justifying the conclusion by appealing to the delta-epsilon charac-
terization of limit. What Maclaurin did instead was what Archimedes might
have done, a double reductio ad absurdum. But what Archimedes might have
done geometrically and verbally, Maclaurin did algebraically. He assumed first
that y is not equal to nxn−1 . Then, he said, it must be equal to nxn−1 + r for
some r. First, he considered the case when this r was positive. This will lead
to a contradiction if h is chosen so that y = n(x + h)n−1 , since, he observed,
inequality (3) will be violated when h = (xn−1 +r/n)1/(n−1) . Similarly, he cal-
culated the h that produces a contradiction when r is assumed to be negative.
Thus there can be no such r, and y = nxn−1 . [63, p. 753]
Maclaurin introduced this proof by saying something surprising for a Trea-
tise of Fluxions: that the use of the inequalities makes the demonstration of
the value of y “independent of the notion of a fluxion.” [63, p. 752] (Of course
one would need the notion of fluxion to interpret y as the fluxion of the area
function xn , but the proof itself is algebraic.) This proof was presumably part
of his agenda in writing the more algebraic Book II of the Treatise for an
audience on the Continent, where fluxions were suspect as involving the idea
of motion. Later Lagrange, in seeking his purely algebraic foundation for the
calculus, explicitly said he wanted to free the calculus from fluxions and what
he called the “foreign idea” of motion. It is thus striking that Lagrange’s
Théorie des fonctions analytiques (1797) gives a more general version of the
kind of argument Maclaurin had given, applying to any increasing function
that satisfies the geometric inequality expressed in (1). In place of the alge-
braic inequality (2), Lagrange used the mean-value theorem. [58, pp. 238–9]
[38, pp. 156–158] The similarity of the two arguments does not prove influ-
ence, of course, but it certainly demonstrates that Maclaurin’s work, which
we know Lagrange read (e.g., [58, p. 17]), uses the algebra of inequalities in a
way consistent with that used by Lagrange and his successors.
Maclaurin’s argument exemplifies the way his Treatise reconciles the old
and the new. The double reductio ad absurdum reflects his Archimedean
agenda. Treating the area as generated by a moving vertical line, and then
searching for the relationship between the area and its fluxion, are Newtonian.
Maclaurin did not have a general proof of the Fundamental Theorem in this
argument, but relied on an inequality based on the specific properties of a
specific function. Nonetheless, he had the precise bounding inequalities for
the area function used later by Lagrange, and he used an algebraic inequality
proof in a manner that would not disgrace a nineteenth-century analyst.
Inequality-based arguments in the calculus as used by Lagrange and
Cauchy owe a lot to the eighteenth-century study of algebraic approxima-
tions, and it once seemed to me that this was their origin. But the algebra of
inequalities as used in Continental analysis, especially in d’Alembert’s pioneer-
ing treatment of the tangent as the limit of secants in the article “Différentiel”
in the Encyclopédie, [19] must owe something also to Maclaurin’s translation
of Archimedean geometry into algebraic dress to justify results in calculus.
Throughout the eighteenth century, practitioners of the limit tradition on the
Continent use inequalities; a clear line of influence connects Maclaurin’s ad-
mirer d’Alembert, Simon L’Huilier (who was a foreign member of the Royal
Society), the textbook treatment of limits by Lacroix, and, finally, Cauchy.
[38, pp. 80–87]
Now let us turn to some of Maclaurin’s work on series. There is, of course,
the Maclaurin series, that is, the Taylor series expanded around zero. This
result Maclaurin himself credited to Taylor, and it was known earlier to New-
ton and Gregory. It was called the Maclaurin series by John F. W. Herschel,
Charles Babbage, and George Peacock in 1816 [51, pp. 620–21] and by Cauchy
in 1823. [14, p. 257] Since it was obvious that Maclaurin had not invented it,
the attribution shows appreciation by these later mathematicians for the way
Maclaurin used the series to study functions. A key application is Maclaurin’s
characterization of maxima, minima, and points of inflection of an infinitely
differentiable function by means of its successive derivatives. When the first
derivative at a point is zero, there is a maximum if the second derivative is
negative there, a minimum if it is positive. If the second derivative is also
zero, one looks at higher derivatives to tell whether the point is a maximum,
minimum, or point of inflection. These results can be proved by looking at
the Taylor series of the function near the point in question, and arguing on
the basis of the inequalities expressed in the definition of maximum and min-
imum. For instance (in modern [Lagrangian] notation), if f (x) is a maximum,
then 2
f (x) > f (x + h) = f (x) + hf (x) + h2! f (x) + · · · , and
2 (4.4)
f (x) > f (x − h) = f (x) − hf (x) + h2! f (x) − · · ·
if h is small. If the derivatives are bounded, and if h is taken sufficiently
small so that the term in h dominates the rest, the inequalities (4) can both
hold only if f (x) = 0. If f (x) = 0, then the h2 term dominates, and the
inequalities (4) hold only if f (x) is negative. And so on.
I have traced Cauchy’s use of this technique back to Lagrange, and from
Lagrange back to Euler. [38, pp. 117–118] [37, pp. 157–159] [58, pp. 235–61]
[29, Secs. 253–254] But this technique is explicitly worked out in Maclaurin’s
Treatise of Fluxions. Indeed, it appears twice: once in geometric dress in Book
I, Chapter IX, and then more algebraically in Book II. [63, pp. 694–696] Euler,
in the version he gave in his 1755 textbook, [20] does not refer to Maclaurin on
this point, but then he makes few references in that book at all. Still we might
suspect, especially knowing that Stirling told Euler in a letter of 16 April 1738
[91] that Maclaurin had some interesting results on series, that Euler would
have been particularly interested in looking at Maclaurin’s applications of the
Taylor series. Certainly Lacroix’s praise for Maclaurin’s work on series must
have taken this set of results into account. [52, p. xxvii] Even more important,
Lagrange, in unpublished lectures on the calculus from Turin in the 1750’s,
after giving a very elementary treatment of maxima and minima, referred to
volume II of Maclaurin’s Treatise of Fluxions as the chief source for more
information on the subject. [7, p. 154] Since Lagrange did not mention Euler
in this connection at all, Lagrange could well have not even have seen the
Institutiones calculi differentialis of 1755 when he made this reference. This
Taylor-series approach to maxima and minima (with the Lagrange remainder
supplied for the Taylor series) plays a major role in the work of Lagrange, and
later in the work of Cauchy. It is because Maclaurin thought of maxima and
minima, and of convexity and concavity, in Archimedean geometrical terms
that he was led to look at the relevant inequalities, just as the geometry of
Archimedes helped Maclaurin formulate some of the inequalities he used to
prove his special case of the Fundamental Theorem of Calculus.
b. Ellipsoids. We now turn to work in applied mathematics that constitutes
one of Maclaurin’s great claims to fame: the gravitational attraction of el-
lipsoids and the related problem of the shape of the earth. Maclaurin is still
often regarded as the creator of the subject of attraction of ellipsoids. [85,
pp. 175, 374] In the eighteenth century, the topic attracted serious work from
d’Alembert, A.-C. Clairaut, Euler, Laplace, Lagrange, Legendre, Poisson, and
Gauss. In the twentieth century, Subramanyan Chandrasekhar (later Nobel
laureate in physics) devoted an entire chapter of his classic Ellipsoidal Fig-
ures of Equilibrium to the study of Maclaurin spheroids (figures that arise
when homogeneous bodies rotate with uniform angular velocity), the condi-
tions of stability of these spheroids and their harmonic modes of oscillation,
and their status as limiting cases of more general figures of equilibrium. Such
spheroids are part of the modern study of classical dynamics in the work of sci-
entists like Chandrasekhar, Laurence Rossner, Carl Rosenkilde, and Norman
Lebovitz. [15, pp. 77–100] Already in 1740 Maclaurin had given a “rigorously
exact, geometrical theory” of homogeneous ellipsoids subject to inverse-square
gravitational forces, and had shown that an oblate spheroid is a possible fig-
ure of equilibrium under Newtonian mutual gravitation, a result with obvious
relevance for the shape of the earth. [39, p. 172] [86, p. xix] [85, p. 374]
Of particular importance was Maclaurin’s decisive influence on Clairaut.

Maclaurin and Clairaut corresponded extensively, and Clairaut’s seminal 1743
book La Figure de la Terre [18] frequently, explicitly, and substantively cites
his debts to Maclaurin’s work. [39, pp. 590–597] A key result, that the attrac-
tions of two confocal ellipsoids at a point external to both are proportional
to their masses and are in the same direction, was attributed to Maclaurin
by d’Alembert, an attribution repeated by Laplace, Lagrange, and Legendre,
then by Gauss, who went back to Maclaurin’s original paper, and finally by
Lord Kelvin, who called it “Maclaurin’s splendid theorem.” [15, p. 38] [85,
pp. 145, 409] Lagrange began his own memoir on the attraction of ellipsoids
by praising Maclaurin’s treatment in the prize paper of 1740 as a master-
work of geometry, comparing the beauty and ingenuity of Maclaurin’s work
with that of Archimedes, [57, p. 619] though Lagrange, typically, then treated
the problem analytically. Maclaurin’s eighteenth- and nineteenth-century suc-
cessors also credit him with some of the key methods used in studying the
equilibrium of fluids, such as the method of balancing columns. [39, p. 597]
Maclaurin’s work on the attraction of ellipsoids shows how his geometric in-
sights fruitfully influenced a subject that later became an analytic one.
c. The Euler-Maclaurin Formula. The Euler-Maclaurin formula expresses
the value of definite integrals by means of infinite series whose coefficients
involve what are now called the Bernoulli numbers. The formula shows how
to use integrals to find the partial sums of series. Maclaurin’s version, in
modern notation, is:
∞ a 1 1
h=0 F (a + h) = 0 F (x)dx + 2 F (a) + 2 F (a)
1 1
− 720 F (a) + (
30240 F v)(a) − ···
[35, pp.84 − −86]
James Stirling in 1738, congratulating Euler on his publication of that formula,

told Euler that Maclaurin had already made it public in the first part of the
Treatise of Fluxions, which was printed and circulating in Great Britain in
1737. [47, p. 88n] [91, p. 178] (On this early publication, see also [63, pp. iii,
691n]). P. L. Griffiths has argued that this simultaneous discovery rests on
De Moivre’s work on summing reciprocals, which also involves the so-called
Bernoulli numbers. [40] [41, pp. 16–17] [25, p. 19] In any case, Euler and
Maclaurin derived the Euler-Maclaurin formula in essentially the same way,
from a similar geometric diagram and then by integrating various Taylor series
and performing appropriate substitutions to find the coefficients. [31] [32] [33]
Maclaurin’s approach is no more Archimedean or geometric than Euler’s; they
are similar and independent. [63, pp. 289–293, 672–675] [35, pp. 84–93] [67] In
subsequent work, Euler went on to extend and apply the formula further to
many other series, especially in his Introductio in analysin infinitorum of 1748
and Institutiones calculi differentialis of 1755. [35, p. 127] But Maclaurin, like
Euler, had applied the formula to solve many problems. [63, pp. 676–693] For
instance, Maclaurin used it to sum powers of arithmetic progressions and to
derive Stirling’s formula for factorials. He also derived what is now called the
Newton-Cotes numerical integration formula, and obtained what is now called
Simpson’s rule as a special case. It is possible that his work helped stimulate
Euler’s later, fuller investigations of these important ideas.
In 1772, Lagrange generalized the Euler-Maclaurin formula, which he ob-
tained as a consequence of his new calculus of operators. [53] [35, pp. 169,
261] In 1834, Jacobi provided the formula with its remainder term, [46, pp.
263, 265] in the same paper in which he first introduced what are now called
the Bernoulli polynomials. Jacobi, who called the result simply the Maclaurin
summation formula, cited it directly from the Treatise of Fluxions. [46, p. 263]
Later, Karl Pearson used the formula as an important tool in his statistical
work, especially in analyzing frequency curves. [72, pp. 217, 262]
The Euler-Maclaurin formula, then, is an important result in the main-
stream of mathematics, with many applications, for which Maclaurin, both in
the eighteenth century and later on, has rightly shared the credit.
d. Elliptic Integrals. Some integrals (Maclaurin used the Newtonian term
“fluents”) are algebraic functions, Maclaurin observed. Others are not, but
some of these can be reduced to finding circular arcs, others to finding log-
arithms. By analogy, Maclaurin suggested, perhaps a large class of integrals
could be studied by being reduced to finding the length of an elliptical or
hyperbolic arc. [63, p. 652] By means of clever geometric transformations,
Maclaurin was able to reduce the integral that represented the length of a
hyperbolic arc to a “nice” form. Then, by algebraic manipulation, he could
reduce some previously intractable integrals to that same form. His work was
translated into analysis by d’Alembert and then generalized by Euler. [13, p.
846] [23] [27, p. 526] [28, p. 258] In 1764, Euler found a much more elegant,
general, and analytic version of this approach, and worked out many more
examples, but cited the work of Maclaurin and d’Alembert as the source of
his investigation. A.-M. Legendre, the key figure in the eighteenth-century
history of elliptic integrals, credited Euler with seeing that, by the aid of
a good notation, arcs of ellipses and other transcendental curves could be
as generally used in integration as circular and logarithmic arcs. [45, p. 139]
Legendre was, of course, right that “elliptic integrals” encompass a wide range
of examples; this was exactly Maclaurin’s point. Thus, although his successors
accomplished more, Maclaurin helped initiate a very important investigation
and was the first to appreciate its generality. Maclaurin’s geometric insight,
applied to a problem in analysis, again brought him to a discovery.
4.7 Other Examples of Maclaurin’s Mathematical

Influence
The foregoing examples provide evidence of direct influence of the Treatise
of Fluxions on Continental mathematics. There is much more. For instance,
Lacroix, in his treatment of integrals by the method of partial fractions, called
it “the method of Maclaurin, followed by Euler.” [52, Vol. II, p. 10] [63, pp.
634–644] Of interest too is Maclaurin’s clear understanding of the use of lim-
its in founding the calculus, especially in the light of his likely influence on
d’Alembert’s treatment of the foundations of the calculus by means of limits
in the Encyclopédie, which in turn influenced the subsequent use of limits by
L’Huilier, Lacroix, and Cauchy, [38, chapter 3] (and on Lagrange’s acceptance
of the limit approach in his early work in the 1750’s). [7] Although the largest
part of Maclaurin’s reply to Berkeley was the extensive proof of results in
calculus using Greek methods, he was willing to explain important concepts
using limits also. In particular, Maclaurin wrote, “As the tangent of an arch
[arc] is the right line that limits the position of all the secants that can pass
through the point of contact.. . though strictly speaking it be no secant; so
a ratio may limit the variable ratios of the increments, though it cannot be
said to be the ratio of any real increments.” [63, p. 423] Maclaurin’s statement
answers Berkeley’s chief objection—that the increment in a function’s value
is first treated as non-zero, then as zero, when one calculates the limit of the
ratio of increments or finds the tangent to a curve. Maclaurin’s statement is in
the tradition of Newton’s Principia (Book I, Scholium to Lemma XI), but is
in a form much closer to the later work of d’Alembert on secants and tangents.
[20] Maclaurin pointed out that most of the propositions of the calculus that
he could prove by means of geometry “may be briefly demonstrated by this
method [of limits].” [63, p. 87, my italics]
In addition, Maclaurin had considerable influence in Britain, on mathe-
maticians like John Landen (whose work on series was praised by Lagrange),
Robert Woodhouse (who sparked the new British interest in Continental work
about 1800), and on Edward Waring and Thomas Simpson, whose names are
attached to results well known today. [42] Going beyond the calculus, Maclau-
rin’s purely geometric treatises were read and used by French geometers of
the stature of Chasles and Poncelet. [90, p. 145] Thus, though Maclaurin may
not have been the towering figure Euler was, he was clearly a significant and
respected mathematician, and the Treatise of Fluxions was far more than an
unread tome whose weight served solely to crush Bishop Berkeley.
4.8 Why a Treatise of Fluxions?

The Treatise of Fluxions was not really intended as a reply to Berkeley.
Maclaurin could have refuted Berkeley with a pamphlet. It was not a stu-
dent handbook either; this work is far from elementary. Nor was it merely
written to glory in Greek geometry. Maclaurin wrote several works on geom-

etry per se. But he was no antiquarian. Instead, the Treatise of Fluxions was
the major outlet for Maclaurin’s solution of significant research problems in
the field we now call analysis. Geometry, as the examples I gave illustrate,
was for Maclaurin a source of motivation, of insight, and of problem-solving
power, as well as being his model of rigor.
For Maclaurin, rigor was not an end in itself, or a goal pursued for purely
philosophical reasons. It was motivated by his research goals in analysis. For
instance, Maclaurin developed his theory of maxima, minima, points of in-
flection, convexity and concavity, orders of contact, etc., because he wanted
to study curves of all types, including those that cross over themselves, loop
around and are tangent to themselves, and so on. He needed a sophisticated
theory to characterize the special points of such curves. Again, in problems
as different as studying the attraction of ellipsoids and evaluating integrals
approximately, he needed to use infinite series and know how close he was to
their sum. Thus, rigor, to Maclaurin, was not merely a tool to defend New-
ton’s calculus against Berkeley—though it was that—nor just a response to
the needs of a professor to present his students a finished subject—though it
may have been that as well. In many examples, Maclaurin’s rigor serves the
needs of his research.
Moreover, the Treatise of Fluxions contains a wealth of applications of
fluxions, from standard physical problems such as curves of quickest descent
to mathematical problems like the summation of power series—in the context
of which, incidentally, Maclaurin gave what may be the earliest clear definition
of the sum of an infinite series: “There are progressions of fractions which may
be continued at pleasure, and yet the sum of the terms be always less than
a certain finite number. If the difference betwixt their sum and this number
decrease in such a manner, that by continuing the progression it may become
less than any fraction how small soever that can be assigned, this number
is the limit of the sum of the progression, and is what is understood by the
value of the progression when it is supposed to be continued indefinitely.” [63,
p. 289] Thus, though eighteenth-century Continental mathematicians did not
care passionately about foundations, [38, pp. 18–24] they could still appreciate
the Treatise of Fluxions because they could mine it for results and techniques.
4.9 Why the Traditional View?

If the reader is convinced by now that the traditional view is wrong, that
Maclaurin’s Treatise did not mark the end of the Newtonian tradition, and
that not all of modern analysis stems solely from the work of Leibniz and his
school, the question arises, how did that traditional view come to be, and why
it has been so persistent?
Perhaps the traditional view could be explained as follows. Consider the
approach to mathematics associated with Descartes: symbolic power, not de-
bates over foundations; problem-solving power, not axioms or long proofs.

The Cartesian approach to mathematics is clearly reflected in the work and
in the rhetoric of Leibniz, Johann Bernoulli, Euler, Lagrange—especially in
the historical prefaces to his influential works—and even Cauchy. These men,
the giants of their time, are linked in a continuous chain of teachers, close
colleagues, and students. Some topics, like partial differential equations and
the calculus of variations, were developed mostly on the Continent. More-
over, the Newton-Leibniz controversy helped drive English and Continental
mathematicians apart. Thus the Continental tradition can be viewed as self-
contained, and the outsider sees no need for eighteenth-century Continental
mathematicians to struggle through 750 pages of a Treatise of Fluxions, which
is at best in the Newtonian notation and at worst in the language of Greek
geometry. Lagrange’s well-known boast that his Analytical Mechanics [55]
had (and needed) no diagrams, thus opposing analysis to geometry at the
latter’s expense, reinforced these tendencies and enshrined them in historical
discourse. But the explanation we have just given does not suffice to explain
the strength, and persistence into the twentieth century, of the standard in-
terpretation. The traditional view of Maclaurin’s lack of importance has been
reinforced by some other historiographical tendencies that deserve our critical
attention.
The traditional picture of Maclaurin’s Treatise of Fluxions radically sep-
arates his work on foundations, which it regards as geometric, sterile, and
antiquarian, from his important individual results, which often are mentioned
in histories of mathematics but are treated in isolation from the purpose of
the Treatise, in isolation from one another, and in isolation from Maclaurin’s
overall approach to mathematics. Strangely, both externalist and internalist
historians, each for different reasons, have reinforced this picture.
For instance, in the English-speaking world, viewing the Treatise as only
about Maclaurin’s foundation for the calculus, and thus as a dead end, has
been perpetuated by the “decline of science in England” school of the history
of eighteenth-century science, stemming from such early nineteenth-century
figures as John Playfair, and, especially, Charles Babbage. [77] [2] [4] Babbage
felt strongly about this because he was a founder of the Cambridge Analyt-
ical Society, which fought to introduce Continental analysis into Cambridge
in the early nineteenth century. This group had an incentive to exaggerate
the superiority of Continental mathematics and downgrade the British, as is
exemplified by their oft-quoted remark that the principles of “pure d-ism”
should replace what they called the “dot-age” of the University. [5, ch. 7] [10,
p. 274] The pun, playing on the Leibnizian and Newtonian notation in cal-
culus, may be found in [2, p. 26]. These views continued to be used in the
attempt by Babbage and others to reform the Royal Society and to increase
public support for British science.
It is both amusing and symptomatic of the misunderstanding of Maclau-
rin’s influence that Lacroix’s one-volume treatise on the calculus of 1802, [50]
translated into English by the Cambridge Analytical Society with added notes
on the method of series of Lagrange, [51] was treated by them, and has been
considered since, as a purely “Continental” work. But Lacroix’s short trea-
tise was based on the concept of limit, which was Newtonian, elaborated by
Maclaurin, adapted by d’Alembert and L’Huilier, and finally systematized by
Lacroix. [38, pp. 81–86] Moreover, the translators’ notes by Babbage, Her-
schel, and Peacock supplement the text by studying functions by their Taylor
series, thus using the approach that Lacroix himself, in his multi-volume trea-
tise of 1810, had attributed to Maclaurin. This is, of course, not to deny the
overwhelming importance of the contributions of Euler and Lagrange, both
to the mathematics taught by the Analytical Society and to that included
by Lacroix in his 1802 book, nor to deny the Analytical Society’s emphasis
on a more abstract and formal concept of function. But all the same, Bab-
bage, Herschel, and Peacock were teaching some of Maclaurin’s ideas without
realizing this.
In any case, the views expressed by Babbage and others have strongly
influenced Cambridge-oriented writers like W. W. Rouse Ball, who said that
the history of eighteenth-century English mathematics “leads nowhere.” [5,
p. 98] H. W. Turnbull, though he wrote sympathetically about Maclaurin’s
mathematics on one occasion, [88] blamed Maclaurin on another occasion for
the decline: “When Maclaurin produced a great geometrical work on fluxions,
the scale was so heavily loaded that it diverted England from Continental
habits of thought. During the remainder of the century, British mathematics
were relatively undistinguished.” [89, p. 115]
Historians of Scottish thought, working from their central concerns, have
also unintentionally contributed to the standard picture. George Elder Davie,
arguing from social context to a judgment of Maclaurin’s mathematics, held
that the Scots, unlike the English, had an anti-specialist intellectual tradition,
based in philosophy, and emphasizing “cultural and liberal values.” Wishing
to place Maclaurin in this context, Davie stressed what he called Maclau-
rin’s “mathematical Hellenism,” [24, p. 112] and was thus led to circumscribe
the achievement of the Treatise of Fluxions as having based the calculus “on
the Euclidean foundations provided by [Robert] Simson,” [24, p. 111] who
had made the study of the writings of the classical Greek geometers the “na-
tional norm” in Scotland. The “Maclaurin is a geometer” interpretation among
Scottish historians has been further reinforced by a debate in 1838 over who
would fill the Edinburgh chair in mathematics. Phillip Kelland, a candidate
from Cambridge, was seen as the champion of Continental analysis, while the
partisans of Duncan Gregory argued for a more geometrical approach. Wish-
ing to enlist the entire Scottish geometric tradition on the side of Gregory,
Sir William Hamilton wrote, “The great Scottish mathematicians, . . . even
Maclaurin, were decidedly averse from the application of the mechanical pro-
cedures of algebra.” [24, p. 155] Though Kelland eventually won the chair, the
dispute helped spread the view that Maclaurin had been hostile to analysis.
More recently, Richard Olson has characterized Scottish mathematics after
Maclaurin as having been conditioned by Scottish common-sense philosophy
to be geometric in the extreme. [70, pp. 4, 15] [71, p. 29] But in emphasizing
Maclaurin’s influence on this development, Olson, like Davie, has overstated
the degree to which Maclaurin’s approach was geometric.
By contrast, consider internalist historians. The treatment of Maclaurin’s
results as isolated reflects what Herbert Butterfield called the Whig approach
to history, viewing the development of eighteenth-century mathematics as a
linear progression toward what we value today, the collection of results and
techniques which make up classical analysis. Thus, mathematicians writing
about the history of this period, from Moritz Cantor in the nineteenth cen-
tury to Hermann Goldstine and Morris Kline in the twentieth, tell us what
Maclaurin did with specific results, some named after him, for which they
have mined the Treatise of Fluxions. [13, pp. 655–63] [35, pp. 126ff, 167–8]
[49, pp. 522–3, 452, 442] They either neglect the apparently fruitless work on
foundations, or, viewing it as geometric, see it as a step backward. It is of
course true that many Continental mathematicians used Maclaurin’s results
without accepting the geometrical and Newtonian insights that Maclaurin
used to produce them. But without those points of view, Maclaurin would
not have produced those results.
Both externalist and internalist historians, then, have treated Maclaurin’s
work in the same way: as a throwback to the Greeks, with a few good results
that happen to be in there somewhat like currants in a scone. Further, the fact
that Maclaurin’s book, especially its first hundred pages, is very hard to read,
especially for readers schooled in modern analysis, has encouraged historians
who focus on foundations to read only the introductory parts. The fact that
there is so much material has encouraged those interested in results to look
only at the sections of interest to them. And the fact that the first volume is
so overwhelmingly geometric serves to reinforce the traditional picture once
again whenever anybody opens the Treatise. The recent Ph.D. dissertation by
Erik Sageng [78] is the first example of a modern scholarly study of Maclau-
rin’s Treatise in any depth. The standard picture has not yet been seriously
challenged in print.
4.10 Some Final Reflections
Maclaurin’s work had Continental influence, but with an important excep-

tion—his geometric foundation for the calculus. Mastering this is a major
effort, and I know of no evidence that any eighteenth-century Continental
mathematician actually did so. Lagrange perhaps came the closest. In the
introduction to his Théorie des fonctions analytiques, Lagrange could say only,
Maclaurin did a good job basing calculus on Greek geometry, so it can be done,
but it is very hard. [58, p. 17] In an unpublished draft of this introduction,
Lagrange said more pointedly: “I appeal to the evidence of all those with the
courage to read the learned treatise of Maclaurin and with enough knowledge
to understand it: have they, finally, had their doubts cleared up and their
spirit satisfied?” [73, p. 30]
Something else may have blunted people’s views of the mathematical qual-
ity of Maclaurin’s Treatise. The way the book is constructed partly reflects
the Scottish intellectual milieu. The Enlightenment in Britain, compared with
that on the Continent, was marked less by violent contrast and breaks with
the past than by a spirit of bridging and evolution. [75, pp. 7–8, 15] Similarly,
Scottish reformers operated less by revolution than by the refurbishment of
existing institutions. [16, p. 8] These trends are consistent with the two-fold
character of the Treatise of Fluxions: a synthesis of the old and the new, of
geometry and algebra, of foundations and of new results, a refurbishment of
Newtonian fluxions to deal with more modern problems. This contrasts with
the explicitly revolutionary philosophy of mathematics of Descartes and Leib-
niz, and thus with the spirit of the mathématicien of the eighteenth century
on the Continent.
Of course Scotland was not unmarked by the conflicts of the century. Dur-
ing the Jacobite rebellion in 1745, Maclaurin took a major role in fortifying
Edinburgh against the forces of Bonnie Prince Charlie. When the city was sur-
rendered to the rebels, Maclaurin fled to York. Before his return, he became
ill, and apparently never really recovered. He briefly resumed teaching, but
died in 1746 at the relatively young age of forty-eight. Nonetheless, the New-
tonian tradition in the calculus was not a dead end. Maclaurin in his lifetime,
and his Treatise of Fluxions throughout the century, transmitted an expanded
and improved Newtonian calculus to Continental analysts. And Maclaurin’s
geometric insight helped him advance analytic subjects.
We conclude with the words of an eighteenth-century Continental math-
ematician whose achievements owe much to Maclaurin’s work. [39, pp. 172,
412–425, 590–597] The quotation [66, p. 350] illustrates Maclaurin’s role in
transmitting the Newtonian tradition to the Continent, the respect in which
he was held, and the eighteenth-century social context essential to under-
standing the fate of his work. In 1741, Alexis-Claude Clairaut wrote to Colin
Maclaurin, “If Edinburgh is, as you say, one of the farthest corners of the
world, you are bringing it closer by the number of beautiful discoveries you
have made.”
Acknowledgement. I thank the Department of History and Philosophy of
Science of the University of Leeds, England, for its hospitality while I was
doing much of this research, and the Mathematics Department of the Uni-
versity of Edinburgh, where I finished it. I also thank Professor G. N. Cantor
for material as well as intellectual assistance, and Professors J. R. R. Christie
and M. J. S. Hodge for stimulating and valuable conversations.
References
1. Arnold, Matthew, The Literary Importance of Academies, in Matthew Arnold,
Essays in Criticism, Macmillan, London, 1865, 42–79. Cited in [61].
2. Babbage, Charles, Passages from the Life of a Philosopher, Longman, London,
1864.
3. [Babbage, Charles], “Preface” to Memoirs of the Analytical Society Cambridge,
J. Smith, 1813. Attributed to Babbage by Anthony Hyman, Charles Babbage:
Pioneer of the Computer, Princeton University Press, Princeton, 1982.
4. Babbage, Charles, Reflections on the Decline of Science in England, and Some
of its Causes, B. Fellowes, London, 1830.
5. Ball, W. W. Rouse, A History of the Study of Mathematics at Cambridge, Cam-
bridge University Press, Cambridge, 1889.
6. Berkeley, George, The Analyst, or a Discourse Addressed to an Infidel Mathe-
matician, in A. A. Luce and T. R. Jessop, eds., The Works of George Berkeley,
vol. 4, T. Nelson, London, 1951, 65–102.
7. Borgato, Maria Teresa, and Luigi Pepe, Lagrange a Torino (1750–1759) e le sue
lezioni inedite nelle R. Scuole di Artiglieria, Bollettino di Storia delle Scienze
Matematiche, 1987, 7: 3–180.
8. Bourbaki, Nicolas, Elements d’histoire des mathématiques, Paris Hermann,
Paris, 1960.
9. Boyer, Carl, The History of the Calculus and Its Conceptual Development,
Dover, New York, 1959.
10. Cajori, Florian, A History of the Conceptions of Limits and Fluxions in Great
Britain from Newton to Woodhouse, Open Court, Chicago and London, 1919.
11. Cajori, Florian, A History of Mathematics, 2d. ed., Macmillan, New York, 1922.
12. Cantor, G. N., Anti-Newton, in J. Fauvel et al., eds., Let Newton Be!, Oxford
University Press, Oxford, 1988, pp. 203–222.
13. Cantor, Moritz, Vorlesungen über Geschichte der Mathematik, vol. 3, Teubner,
Leipzig, 1898.
14. Cauchy, A..L., Résumé des leçons données a l’école royale polytechnique sur le
calcul infinitésimal, in Oeuvres completes, Ser. 2, vol. 4, Gauthier-Villars, Paris,
1899.
15. Chandrasekhar, S., Ellipsoidal Figures of Equilibrium, Yale, New Haven, 1969.
16. Chitnis, Anand, The Scottish Enlightenment: A Social History, Croom Helm,
London, 1976.
17. Christie, John R. R., The Origins and Development of the Scottish Scientific
Community, 1680–1760, History of Science 12 (1974), 122–141.
18. Clairaut, A.-C., Théorie de la figure de la terre, Duraud, Paris, 1743.
19. d’Alembert, Jean, “Différentiel,” in [21].
20. d’Alembert, Jean, and de la Chapelle, “Limite,” in [21].
21. d’Alembert, Jean, et al., eds., Dictionnaire encyclopédique des mathématiques,
Hotel de Thou, Paris, 1789, which collects the mathematical articles from the
Diderot-d’Alembert Encyclopédie.
22. d’Alembert, Jean, Traité de dynamique, David l’Ainé, Paris, 1743.
23. d’Alembert, Jean, Récherches sur le calcul integral, Histoire de l’Académie de
Berlin (1746), 182–224.
24. Davie, George Elder, The Democratic Intellect: Scotland and her Universities
in the Nineteenth Century, The University Press, Edinburgh, 1966.
25. De Moivre, Abraham, Miscellanea analytica, Tonson and Watts, London, 1730.
26. Dijksterhuis, E. J., Archimedes, 2d. ed., Tr. C. Dikshoorn, Princeton University
Press, Princeton, 1987.
27. Enneper, Alfred, Elliptische Functionen: Theorie und Geschichte, 2d. ed.,
Nebert, Halle, 1896.
28. Euler, Leonhard, De reductione formularum integralium ad rectificationem el-
lipsis ac hyperbolae, Nov. Comm. Petrop. 10 (1764), 30–50, in L. Euler, Opera
Omnia, Teubner, Leipzig, Berlin, Zurich, 1911–, Ser. I, vol. 20, 256–301.
29. Euler, Leonhard, Institutiones calculi differentialis, 1755, sections 253–255; in
Opera, Ser. I, Vol. XI.
30. Euler, Leonhard, Introductio in analysin infinitorum, Lausanne, 1748; in Opera,
Ser. I, vols. 8–9.
31. Euler, Leonhard, Inventio summae cuiusque seriei ex data termino generali,
Comm. Petrop. 8 (1741). 9–22; in Opera, Ser. I, vol. 14, 108–123.
32. Euler, Leonhard, Methodus generalis summandi progressiones, Comentarii
Acad. Imper. Petrop. 6 (1738), 68–97; in Opera, Ser. II, vol. 22.
33. Euler, Leonhard, Methodus universalis serierum convergentium summas quam
proxime inveniendi, Comm. Petrop. 8 (1741), 3–9; in Opera, Ser. I, vol. 14,
101–107.
34. Eyles, V. A., Hutton, Dictionary of Scientific Biography, Scribner’s, New York,
1972, vol. 6, 577–589.
35. Goldstine, Herman, A History of Numerical Analysis from the 16th through the
19th Century, Springer-Verlag, New York, Heidelberg, Berlin, 1977.
36. Grabiner, Judith V., A Mathematician Among the Molasses Barrels: MacLau-
rin’s Unpublished Memoir on Volumes, Proceedings of the Edinburgh Mathemat-
ical Society 39 (1996), 193–240.
37. Grabiner, Judith V., The Calculus as Algebra: J.-L. Lagrange, 1736–1813, Gar-
land Publishing, Boston, 1990.
38. Grabiner, Judith V., The Origins of Cauchy’s Rigorous Calculus, M.I.T. Press,
Cambridge, Mass., 1981.
39. Greenberg, John L., The Problem of the Earth’s Shape from Newton to Clairaut,
Cambridge University Press, Cambridge, 1995.
40. Griffiths, P. L., Private communication.
41. Griffiths, P. L., The British Influence on Euler’s Early Mathematical Discoveries,
preprint.
42. Guicciardini, Niccolo, The Development of Newtonian Calculus in Britain,
1700–1800, Cambridge University Press, Cambridge, 1989.
43. Hall, A. Rupert, Philosophers at War: The Quarrel between Newton and Leibniz,
44. Hankins, Thomas, Jean d’Alembert: Science and the Enlightenment, Clarendon
Press, Oxford, 1970.
45. Itard, Jean, Legendre, Dictionary of Scientific Biography, vol. 8, 135–143.
46. Jacobi, C. G. J., De usu legitimo formulae summatoriae Maclaurinianae, Journ.
f. reine u. angew. Math. 18 (1834), 263–272. Also in Gesammelte Werke, vol. 6,
1891, pp. 64–75.
47. Juskevic, A. P., and R. Taton, eds., Leonhard Euleri Commercium Epistolicum,
Birkhauser, Basel, 1980; in Leonhard Euler, Opera, Ser. 4, vol. 5.
48. Juskevič, A. P., and Winter, E., eds. Leonhard Euler und Christian Goldbach:
Briefwechsel, 1729–1764, Akademie-Verlag, Berlin, 1965.
49. Kline, Morris, Mathematical Thought from Ancient to Modern Times, Oxford,
New York, 1972.
50. Lacroix, S. F., Traité élémentaire de calcul différentiel et de calcul intégral,
Duprat, Paris, 1802. Translated as [51].
51. Lacroix, S.-F., An Elementary Treatise on the Differential and Integral Calculus,
translated, with an Appendix and Notes, by C. Babbage, J. F. W. Herschel, and
O. Peacock, J. Deighton and Sons, Cambridge, 1816.
52. Lacroix, S.-F., Traité du calcul différentiel et du calcul integral, 3 vols., 2d. ed.,
Courcier, Paris, 1810–1819. Vol. 1, 1810.
53. Lagrange, J.-L., Sur une nouvelle espece de calcul rélatif a la differentiation
et a l’intégration des quantités variables, Nouvelles Memoires de l’académie de
Berlin, 1772, 185–221; in Oeuvres, vol. 3, 439–476.
54. Lagrange, J.-L., Leçons sur le calcul des fonctions, new ed., Courcier, Paris,
1806; in Oeuvres, vol. 10.
55. Lagrange, J.-L., Mechanique analytique, 2d. ed., 2 vols., Courcier, Paris, 1811–
1815; in Oeuvres, vols. 11–12.
56. Lagrange, J.-L., Note sur la métaphysique du calcul infinitesimal, Miscellanea
Taurinensia 2 (1760–61), 17–18; in Oeuvres, vol. 7, 597–599.
57. Lagrange, J.-L., Sur l’attraction des sphéroides elliptiques, Memoires de l’aca-
démie de Berlin, 1773, 121–148. Reprinted in Oeuvres de Lagrange, vol. III,
619ff.
58. Lagrange, Joseph-Louis, Théorie des fonctions analytiques, Imprimérie de
Ia République, Paris, An V [1797]; compare the second edition, Courcier,
Paris, 1813, reprinted in Oeuvres de Lagrange, pub. M. J.-A. Serret, 14 vol-
umes, Gauthier-Villars, Paris, 1867–1892, reprinted again, Georg Oms Verlag,
Hildesheim and New York, 1973, vol. 9.
59. Legendre, Adrien-Marie, Mémoires sur les integrations par arcs d’ellipse et sur
la comparaison de ces arcs, Mémoires de l’Academie des sciences, 1786, 616,
644–673.
60. Legendre, Adrien-Marie, Traite des fonctions elliptiques et des intégrales eulér-
iennes, avec des tables pour en faciliter le calcul numerique, 3 vols., Paris, 1825–
1828.
61. Loria, Gino, The Achievements of Great Britain in the Realm of Mathematics,
Mathematical Gazette 8 (1915), 12–19.
62. Maclaurin, Colin, Traité de fluxions, Traduit de l’anglois par le R. P. Pézénas,
2 vols., Jombert, Paris, 1749.
63. Maclaurin, Colin, A Treatise of Fluxions in Two Books, Ruddimans, Edinburgh,
1742.
64. Mahoney, Michael, Review of [42], Science 250 (1990), 144.
65. McElroy, Davis, Scotland’s Age of Improvement, Washington State University
Press, Pullman, 1969.
66. Mills, Stella, The Collected Letters of Colin Maclaurin, Shiva Publishing,
Nantwich, 1982.
67. Mills, Stella, The Independent Derivations by Leonhard Euler and Colin Maclau-
rin of the Euler-MacLaurin Summation Formula, Archive for History of Exact
Science 33 (1985), 1–13.
68. Murdoch, Patrick, An Account of the Life and Writings of the Author, in Colin
Maclaurin, Account of Sir Isaac Newton’s Philosophical Discoveries, For the
Author’s Children, London, 1748, i-xx; reprinted, Johnson Reprint Corp., New
York, 1968.
69. Newton, Isaac, Of Analysis by Equations of an Infinite Number of Terms, J.

Stewart, London, 1745, in D. T. Whiteside, ed., Mathematical Works of Isaac
Newton, vol. 1, Johnson Reprint, New York and London, 1964, 3–25.
70. Olson, Richard, Scottish Philosophy and British Physics, 1750–1880, Princeton
University Press, Princeton, 1975.
71. Olson, Richard, Scottish Philosophy and Mathematics, 1750–1830, Journal of
the History of Ideas 32 (1971), 29–44.
72. Pearson, Karl, The History of Statistics in the Seventeenth and Eighteenth Cen-
turies [written 1921–1933]. Ed. E. S. Pearson, Charles Griffin & Co., London
and High Wycombe, 1976.
73. Pepe, Luigi, Tre ‘prime edizioni’ ed un’ introduzione inedita della Fonctions
analytiqucs di Lagrange, Boll. Stor. Sci. Mat. 6 (1986), 17–44.
74. Phillipson, Nicholas, The Scottish Enlightenment, in [76], pp. 19–40.
75. Porter, Roy, The Enlightenment in England, in [76], pp. 1–18.
76. Porter, Roy, and Mikulas Teich, eds., The Enlightenment in National Context,
77. Playfair, John, Traité de Mechanique Celeste, Edinburgh Review 22 (1808), 249–
84.
78. Sageng, Erik Lars, Colin Maclaurin and the Foundations of the Method of Flux-
ions, unpublished Ph.D. Dissertation, Princeton University, 1989.
79. Scott, J. F., Maclaurin, Dictionary of Scientific Biography, vol. 8, 609–612.
80. Shapin, Stephen, and Arnold Thackray, Prosopography as a Research Tool in
History of Science: The British Scientific Community, 1700–1900, History of
Science 12 (1974), 95–121.
81. Stewart, M. A., ed., Studies in the Philosophy of the Scottish Enlightenment,
Clarendon Press, Oxford, 1990.
82. Struik, D. J., A Source Book in Mathematics, 1200–1800, Harvard University
Press, Cambridge, Mass., 1969.
83. Taton, Juliette, Pezenas, Dictionary of Scientific Biography, vol. 10, Scribner’s,
New York, 1974, 571–2.
84. Taton, René, ed., Enseignement et diffusion des sciences en France au XV III e
siècle, Hermann, Paris, 1964.
85. Todhunter, Isaac, A History of the Mathematical Theories of Attraction and the
Figure of the Earth, from the Time of Newton to That of Laplace, Macmillan,
London, 1873.
86. Truesdell, C., Rational Fluid Mechanics, 1687–1765, introduction to Euler
Opera, Ser. 2, vol. 12.
87. Truesdell, C., The Rational Mechanics of Flexible or Elastic Bodies, 1638–1788,
in Euler, Opera, Ser. 2, vol. 11.
88. Turnbull, H. W., Bicentenary of the Death of Colin Maclaurin (1698–1746), The
University Press, Aberdeen, 1951.
89. Turnbull, H. W., The Great Mathematicians, Methuen, London, 1929.
90. Tweedie, Charles, A Study of the Life and Writings of Colin Maclaurin, Math-
ematical Gazette 8 (1915), 132–151.
91. Tweedie, Charles, James Stirling, Clarendon Press, Oxford, 1922.
5
The Mathematics and Science of Leonhard
Euler (1707–1783)
Rüdiger Thiele
Karl-Sudhoff-Institut für Geschichte der Medizin und der Naturwissenschaften,

University of Leipzig
Laplace, qui pourtant n’avait pas pris, dans ses écrits, pour modèle
le célèbre géomètre de Bâle, ne cessait de répéter aux jeunes mathé-
maticiens ces paroles mémorables que nous avons entendues de sa
propre bouche: Lisez Euler, lisez Euler, c’est notre maı̂tre à tous.1
Guglielmo Libri-Carucci dalla Sommaja (1803-1869)
On 23 October 1783 a memorial session of the Imperial Academy of Sci-

ences in St. Petersburg, Russia took place. Professor Nicolas (Nikolaus) Fuss
(1755–1826), one of Euler’s assistants during Euler’s long period of blindness,
delivered his famous Éloge de Monsieur Léonard Euler. He began with the
following words:
Représenter le cours de la vie d’un grand homme qui a illustré son

Siècle en éclairant le monde, c’est faire l’éloge de l’esprit humain.2
At this time French was in common use not only in Paris, but as the
lingua franca in Berlin and in St. Petersburg as well. One may summarize
Fuss’ words in this way: the task of giving a survey of Euler’s life and work
means giving a survey of the human intellect.
In 1783, in the year of Euler’s death and 17 years before the turn of cen-
tury, one of the most influential figures of the German Enlightenment, Georg
Christoph Lichtenberg (1742–1799), devoted a paper to such a survey of the
human intellect but to a smaller one. He proudly imagined how coming cen-
turies might view his century. He set up an inventory of the scientists Isaac
Newton (1642–1727), Gottfried Wilhelm Leibniz (1646–1716) and Leonhard
1
Journal des Savants, 1846, p. 51: “Although Laplace in his writings did not
regard that famous Basel mathematician as a model, he never ceased repeating
these memorable words to young mathematicians which we [G. Libri-Carucci] heard
from Laplace himself: Read Euler, read Euler, he is the master of us all.”
2
Fuss 1783, p. 1; German translation in: EO I, 1, p. XLIII.
82 Rüdiger Thiele
Leonhard Euler
April 15, 1707 – September 9, 1783
Engraving after the plaster cast by J. Rachette (1781)
Courtesy Universitätsbibliothek Basel.
Euler (1707–1783) as well as the monarchs Peter the Great (1672–1725), Fred-
erick the Great (1712–1786) and Catherine the Great (1729–1796). Concern-
ing the year 1783 Lichtenberg mentioned a “huge new state” (referring to
England’s acceptance of America’s Declaration of Independence in 1776) and
5 The Mathematics and Science of Leonhard Euler (1707–1783) 83
Fig. 5.1. Nicolas Fuss. Silhouette by F. Anting (1784). Due to a recommendation

by Daniel Bernoulli, the Swiss Fuss became Euler’s assistant in 1772, later professor
of mathematics, in February 1783 member of the Petersburg Academy, and in 1784
Euler’s grandson-in-law. Courtesy Sudhoff-Institut, Universität Leipzig.
Fig. 5.2. N. Fuss’s eulogy on Euler delivered at a meeting of the Petersburg Acad-
emy in October 23, 1783. Courtesy Universitätsbibliothek Basel.
84 Rüdiger Thiele
an air balloon (Montgolfière).3 In the three memoirs of Catherine,4 however,

Euler, like any other scientist in Russia, is not mentioned, symbolic of the two
cultures in the sense of Charles Percy Snow (1905–1980).5 Snow argued that
in the major branches of Western culture (above all science and the human-
ities) little or nothing is known about the other one, and that between the
two camps communication is difficult if not impossible. Even the first edition
of the Encyclopaedia Britannica took no notice of Euler in 1771.6
A century after Euler, the German historian of mathematics Hermann
Hankel (1839–1873) properly remarked that Euler represented the scientific
consciousness of the 18th century in the best way.7 Recently, Clifford A. Trues-
dell (1919–2000) estimated that Euler wrote about one third of the mathe-
matical works in the 18th century (including mathematical physics). But in
general culture, the age of reason is regarded more as the age of John Locke
(1632–1704) or Voltaire (1694–1678), rather than as the age of the natural
sciences. A visit to any bookshop confirms this view. The author of a best-
selling book on the Enlightenment8 published in the promising Cambridge
series “New approaches to European history” managed not to mention Euler
in the entire book. However, it was the progress of the natural sciences that in-
fluenced more and more the rise of industry and, as a consequence, determined
the conditions of human living. In this culture Euler has been appreciated; in
his lifetime (in 1765) Euler was awarded with a prize (£300) by the British
Parliament for his elaboration of two lunar theories, among them one with
important consequences for navigation (1753, E 187); and in modern times
(in 1935) the International Astronomical Union honoured him with a lunar
crater, the diameter of which is 27 km wide, named after him (latitude 23.3
N and longitude 53.4 W).
3
Lichtenberg, 1972, 3, 62–63. “Und was ich [the 18th century] gesehen habe?
O genug. Ich habe Peter den Ersten gesehen und Katharina und Friedrich ... und
Leibniz und Newton und Euler ... Bist du [reader] damit zufrieden? Gut. Aber sieh
hier noch ein paar Kleinigkeiten: Hier habe ich einen neuen ungeheuren Staat [USA],
... und siehe endlich habe ich in meinem 83sten Jahr ein Luftschiff [Montgolfière]
gemacht.”
4
Katharina II, 1986.
5
Snow 1959. The Two Cultures and the Scientific Revolution (1959) and its sequel
The Two Cultures and the Scientific Revolution: A Second Look (1964). The first
heading is the title of C. P. Snow’s Rede Lecture at Cambridge on the gap between
science, and literature and religion.
6
Edinburgh, vol. 2. 1771.
7
Hankel 1982. “Euler, der das wissenschaftliche Bewußtsein in der Mitte des
vorigen Jahrhunderts am vollständigsten vertritt, definiert [als Funktion . . . ] ”, p.
64. Reprint of a speech delivered in Tübingen in 1870.
8
Outram 1996.
Youth
Precisely three towns determine the life of Leonhard Euler: Basel, a town
of Roman origin; Berlin, founded in the 13th century; and St. Petersburg,
founded in 1703.9 He was born in Basel on 15 April 1707, some days before
the Act of Union between England and Scotland constituted one Parliament
in the island. Basel was an independent city of the empire and a centre of
learning.
The Euler family is first mentioned in 1287 near Lindau, a town in southern
Germany on Lake Constance. The German name Euler [oyler] sounds like Eule
[oyle], the German word for owl, but the name refers to a wet meadow called
an Au (diminutive: Äule [oyle]) in German, strictly speaking to a possessor of
an Äule called an Äuler [oyler]. This Au may be found in the names of many
smaller German towns, as in Nassau and Dessau.
Euler’s father, Paul Euler (1670–1745), was a Protestant minister married
to another minister’s daughter, Margarete Brucker (1677–1761). The Euler
family was rather poor. Most of Paul Euler’s ancestors were comb-makers. In
1708 Paul Euler became pastor at the village of Riehen, two or three miles
away from Basel, and his family soon moved there.
At Riehen10 Euler grew up with his parents and later with two sisters in
two rooms of the parsonage. Nevertheless, he was surrounded by educated
people. Leonhard received his first instruction from his father. Paul Euler
taught mathematics to his son using the widespread Coss (1525) by Christoff
Rudolff (1500?–1549?) in Michael Stifel’s (1487?–1567) edition of 1553 — a
well-known book printed in many editions but difficult to read. Paul Euler
himself was also a mathematician, having been a pupil of the famous James
(Jakob) Bernoulli (1654–1705), and James’s successor was his younger brother
John (Johann) Bernoulli (1667–1748).
About 1713 — the time when Newfoundland was ceded to England —
Leonhard Euler moved back to Basel to attend a grammar school, in which
mathematics was not taught. Therefore Euler had private lessons from a
Calvinist priest Johannes Burckhardt (1691–1743), who is known as a sup-
porter of John Bernoulli in Bernoulli’s clash with Brook Taylor (1685–1731).
Burkhardt seems to have played an important role in Euler’s mathematical
education. Daniel Bernoulli (1700–1782), son of John Bernoulli, referred to
Burckhardt as “magni Euleri praeceptor in mathematicis” (the teacher of
the great Euler in mathematics).11 Incidentally, the treatment of children in
the 18th century is not generally regarded as having been lenient and kindly.
9
In 1710 St. Petersburg had 8,000 inhabitants; in 1725 already 70,000; and in
1796 a quarter of a million.
10
In the local Swiss dialect Rieche (pronounced ri:ch, ch as in the Scottish pro-
nunciation of Loch).
11
Fuss 1843, 2. Letter from Daniel Bernoulli to Euler (September 4, 1743), 529–
537. “Vor etlichen Tagen ist der grosse Burcard [Burckhardt], Magni Euleri prae-
ceptor in mathematicis, gestorben”, p. 535.
86 Rüdiger Thiele
Fig. 5.3. View of the city of Basel. Etching by W. Herrliberger (1761) after a
drawing by E. Büchel (1759). From the right: the end of the quarter Schifflände,
the Church of St. Martin, the Old University (on the bank of the river Rhine), the
Gothic Münster. Courtesy Universitätsbibliothek Basel.
Fig. 5.4. View of the Church of St. Martin in modern times, taken from the left
side of the bridge shown in fig. 3. Photo R. Thiele.
Fig. 5.5. Parish register of St. Martin in Basel for the year MDCCVII (1707). Entry
of Leonhard Euler’s christening ceremony on page 376, no. 1. Columns from left:
date and infant to be baptized, parents, godfathers (including Leonhard Respinger).
Courtesy Staatsarchiv Basel.
Pupils were frequently flogged by teachers, and occasionally an enraged father

would appear in the classroom and respond in kind.
Study
In October 1720, at the not unusual age of 13, Leonhard entered the University
of Basel. As a citizen of Basel he was allowed to do so whereas country folk
could not enter the University before 1798. The city’s university was founded
in 1460 by Pope Pius II (1405-1464, pope since 1458) and was the first in
Switzerland. The famous Dutch scholar Erasmus (1469?-1536) taught at this
university in the 16th century, making the city a centre of humanism. In 1723
Euler and John II (Johann II), son of John Bernoulli, took their Master’s
degrees. Paul Euler wanted his son to follow him into church service; hence,
Leonhard started to study theology.
The Calvinist clergy — even its most pious members — strongly advocated
“law and order,” mainly for political reasons. This created an ambiguous reli-
gious situation. Euler grew up in a spirit of submission to religious discipline.
He upheld this spirit of true piety and religious discipline, and later on in
Berlin and St. Petersburg he advocated parish affairs in this spirit to the very
last. He remained a devout Calvinist all his life, and he conducted family
prayers for his whole household, usually finishing off with a sermon.12
The University of Basel was very small in those days (19 professors and
about 100 students); however, it was a mathematical centre of Europe under
John I Bernoulli. At this small university it was inevitable that Euler and
Bernoulli would meet. Indeed, soon Euler’s mathematical abilities earned him
the esteem of John I Bernoulli, who advised him to study mathematics. Finally
Leonhard’s father relented, and Leonhard left the Faculty of Theology and
12
Cf. R. Thiele 2005a; M. Raith 1983, 459-470; F.G. Hartweg 1979.
88 Rüdiger Thiele
Fig. 5.6. View at Riehen, a village near Basel where Euler spent his childhood. The
village church is St. Martin where Euler’s father was a Protestant minister. Courtesy
Universitätsbibliothek Basel.
Fig. 5.7. St. Martin, indoor photograph. The sight is rather similar to that in
Euler’s day (extension of the church in 1694) despite the renovation in 1943. The
oldest parts of St. Martin date to 950. Courtesy M. Raith, Riehen.
Fig. 5.8. De rationibus et proportionibus (On ratios and proportions) by Paulus

Euler. Under the supervision of James Bernoulli, Paulus Euler (Baccalaureus in
1687) defended 50 propositions on October 8, 1688. After that examination he
started the study of theology, which was finished in 1693. It is remarkable that
the disputatio was printed; however it is not sure or not whether P. Euler is its
author (M. Raith). Courtesy Universitätsbibliothek Basel.
90 Rüdiger Thiele
studied mathematics together with John II. He also became acquainted with
Bernoulli’s other sons Daniel (1700–1782) and Nicolas (1695–1726). The closer
acquaintance with the three sons brought Euler into the house of Bernoulli.
In his words:
I soon found an opportunity to be introduced to a famous profes-
sor named John Bernoulli . . . True, he was very busy and so refused
outright to give me private lessons.13
However, Bernoulli supervised Euler’s study by posing problems to Euler and
by recommending mathematical reading. This was done on Saturday and
Euler spent the rest of the week solving the problems and trying to trou-
ble his teacher with as few questions as possible. Later on, in 1767, Leonhard
Euler remembered:
I was given permission to visit him freely every Saturday afternoon and
he kindly explained to me everything I could not understand . . . and
this, undoubtedly, is the best method to succeed in mathematical sub-
jects.14
In 1726 we first hear of Euler’s own mathematical researches. In his pa-
per “Constructio linearum isochronum . . . ” (Construction of isochrones in
a resistant medium; E 1; EO II, 6),15 Euler took up a special version of the
brachistochrone problem posed by John Bernoulli in 1696.16 Euler’s paper was
published in the journal Acta eruditorum in Leipzig in 1726. In the same year
he wrote his paper “Meditationes super problemate nautico” (Essay on navi-
gation; E 4; EO II, 20) dealing with the masting of ships — a prize problem of
the Paris Academy of Sciences. Although Euler’s paper failed to win the prize
he received an honourable mention, an accessit. Later he won academy prizes
twelve times, not to mention the eight prizes of his sons (substantially due to
the father). Among Euler’s prizes were five on navigation (an important topic
at this time), and he earned 30,000 livres, a huge amount of money.
That Euler failed is not surprising when you recall the jokes about the
nonexistent Swiss navy. Euler might have seen some boats on the river Rhine,
but he had never seen a ship. But it comes as a surprise to see how Euler
worked. He has been criticized for letting mathematics run away with his sense
13
Euler 1767/1995, 11-13. “ . . . wo ich bald Gelegenheit fand mit dem berühmten
Professori Johanni Bernoulli bekannt zu werden . . . Privat Lectionen schlug er mir
zwar wegen seiner Geschäfte gänzlich ab . . . ”.
14
Euler 1767/1995, 11-13. “ . . . gab er mir alle Sonnabend Nachmittag einen
freyen Zutritt bey sich, und hatte die Güte mir die gesammelte Schwierigkeiten
zu erläutern, . . . welches gewiß die beste Methode ist, um in den mathematischen
Wissenschaften glückliche Progressen zu machen.”
15
E refers to the number of the Eneström register (Eneström 1910–1913); EO
heads the series and the volume of the Opera omnia Euleri (Euler 1911– ) in Roman
and Hindu-Arabic figures respectively.
16
Bernoulli 1696; see also Thiele 2002, and 2004.
Fig. 5.9. Title page of the Coss (Strasbourg 1525) by Ch. Rudolff in M. Stifel’s
edition (in a reprint of 1615, first edition in 1553). Courtesy Universitätsbibliothek
Leipzig.
92 Rüdiger Thiele
Fig. 5.10. John Bernoulli (left) and his son Daniel (right). Paintings by R. Huber
(cut-out). Courtesy Universitätsbibliothek Basel.
of reality. The physical world was an occasion to apply mathematics, and if

it failed to fit his analysis, it was the physical world, not the mathematics,
which was in error. A well-known example is his investigation of a falling body
given in his paper on sound (“Dissertatio physica de sono . . . ”; E 2; EO III,
1) in 1727. Euler considered a body falling through a tunnel from one pole
of the earth to the other. His surprising and incorrect solution was that the
body falls only as far as the centre of the earth and then returns. Leonhard
Euler’s 1726 prize paper on navigation ends with the following words:
I did not consider it necessary to confirm my theory by any experi-
ments. For this theory is derived from the most certain principles of
mechanics. Hence there can be no doubt, whether it be true or have
a place in praxi.17
17
Final draft of the “Meditationes”, Archive of the Russian Academy of Sciences
in St. Petersburg, 136.1.213; quoted by G.K. Michailov in Sammelband 1959, p.
259. “Haud opus esse existimavi istam meam theoriam experiment confirmare, cum
However, in a draft of this paper found some years ago in the Archives in
St. Petersburg we have another wording:
I set to work to confirm my theory by experiments although this the-

ory is deduced correctly from most certain and unquestionable princi-
ples, as far as I could see. Hence, if anybody wants to investigate the
certainty of the principles or the correctness of the deduction of my
theory, he will ipso facto [i.e., by his very investigation] see the truth
of this theory confirmed. Consequently, no room will be left for doubt
or hesitation.18
Concerning his theory of music Euler wrote to Daniel Bernoulli in 1740 that
“experience will decide which theory is in better accordance with the truth.”19
In his late Petersburg years he still appreciated models of bridges made by
the gifted Russian inventor Ivan Petrovich Kulibin (1735–1818) and went on
to develop a mechanics of models (similarity mechanics).20
Euler, aware he was a born mathematician, applied for a professorship in
Basel — and he failed. There were few opportunities for mathematicians in
Switzerland. Therefore, in 1727 Daniel and Nicolas Bernoulli left Basel and
became members of the Russian Academy. They promised Euler to find a
position for him there. Euler received his call to Sankt Peterburg (Russian)21
or Saint Petersburg (English) in 1727, officially as an associate of the medical
section of the Academy. On April 5, 1727, a few days after Newton’s death,
he left Basel and he never returned to his home town, although he retained
citizenship all his life. It is worth noting that Euler enrolled in the department
of medicine at the University of Basel on April 2, exactly three days before
his departure for Russia.
St. Petersburg
Euler started his way to St. Petersburg on the river Rhein until Mainz, con-
tinued by land until the seaport Lübeck, and finished by sea. On his journey
integra et ex certissimis et irrepugnabilibus principiis mechnicis deducta, atque adeo
de illa dubitari, an vera sit ac an in praxi locum habere queat, minime possit.”
18
First draft of the “Meditationes”, describing experiments; cf. footnote 17. “Sed
istam meam theoriam, quamvis sit ex certissimis et de quibus neutiquam dubitar
potest principiis recte quantum perspicere potui deducta, tamen eam experimentia
quoque confirmare aggressus sum, quo, si quis de certitudine principiorum aut le-
gitima deductione meae theoriae ambigere velit, veritatem ipso facto videre possit
confirmatum iri, et ita, ulli dubitationi aut haesitationi nullus locus relinquatur.”
19
Eneström 1906. Letter to Daniel Bernoulli from September 15, 1740, 152.
“Durch die Experientz kann man also leicht determiniren, welche Theorie mit der
Wahrheit übereinkommt.”
20
Idey Eylera 1988, see N.M. Raskin, Eyler i Kulibin (in Russian), pp. 304-320.
21
Later Petrograd (1914–1917), from 1917 Leningrad, now St. Petersburg.
94 Rüdiger Thiele
Fig. 5.11. Europe in 1748. Based on a map of Sudhoff-Institut, Universität Leipzig.
Euler stopped in Marburg to meet the philosopher Christian Wolff (1679–

1754). Wolff, who was a pupil of Gottfried Wilhelm Leibniz, is best known
as the German spokesman of the Enlightenment. Wolff was banished from
Halle in 1723 by the Prussian king Frederick William I (Friedrich Wilhelm I)
(1688–1740) under penalty of hanging as a result of religious and philosophical
disputes, and he fled to Marburg. In Russia, however, due to Leibniz’s efforts
he was made a science advisor (in absentia) to tsar Peter the Great (1672–
1725, tsar from 1682, alone from 1696). He helped to establish the Russian
Academy of Sciences which was founded after Peter’s unexpected death in St.
Petersburg in 1725 by Catherine (1684–1727, tsarina from 1725), the widow
of Peter, in the same year. Wolff wrote to Euler that he would “travel into
the paradise of the scholars.”22 Euler himself confirmed this statement in a
letter written in Berlin in 1749:
I and all others who had the good fortune to be some time with
the Russian Imperial Academy cannot but acknowledge that we owe
everything we are and possess to the favourable conditions we had
there.23
At first, however, Euler’s expectations were quickly dashed. The day he

reached Russia Catherine had died and was succeeded by the boy tsar Peter
II (1715–1730, tsar from 1727), a grandson of Peter the Great. Despite the
regular succession, the question of succession was reopened, and Peter’s re-
forms — Europeanization — were called in question. Peter’s modernization
was seen as a serious break with the past (which it was), and as a betrayal
of Russia’s culture. The actual power passed into the hands of a brutal or-
thodox faction, the Bironovshchina,24 which looked upon the Academy as a
dispensable luxury and tried to send all its foreign members home.
The Academy somehow managed to survive. In the confusion, Euler
slipped into the mathematical section instead of the medical one. Political
affairs slowly improved, and Euler was able to get down to solid work. He
lectured to students, if there were any, and for pupils at the Gymnasium, and
he gave about 10 lectures a year in the Academy. But scientific relations re-
mained tense among the 17 members of the Academy (13 Germans, 3 Swiss,
and one Frenchman).25 The philosophical conflicts, that is to say, the contro-
versies between Newtonians, Cartesians, and Leibnizo-Wolffians, reached St.
Petersburg and came to a critical point there. Daniel Bernoulli, living with
Euler, complained:
I deplore my misfortune as much as ever I can.
22
Letter from 20. 4. 1727. “Sie reisen jetzt in das Paradiess der Gelehrten.”
Archive of Russian Academy of Sciences in St. Petersburg. Quoted in Sammelband
1959, 276.
23
Akademien im Briefwechsel 1961, II, 182. Letter to J. D. Schumacher from
7. (18.) 11. 1749. Johann Daniel (also Ivan Danilovich) Schumacher (1690–1761),
librarian and councillor of the St. Petersburg Academy. “. . . auch ich und alle übrige,
welche das Glück gehabt, einige Zeit bey der russischen Kaiserlichen Academie zu
stehen, müssen gestehen, daß wir alles, was wir sind, den vortheilhaften Umständen,
worin wir uns daselbst befunden, schuldig sind.”
24
Named after Ernst Johann Reichsgraf von [Imperial count of] Biron, from 1737
duke of Courland. German adventurer who became the influential chief adviser of
the Russian empress Anna. After her death in 1740 he was exiled to Siberia; in 1762
he was finally granted an amnesty.
25
Among them Jakob Hermann (1678–1733), Georg Bilfinger (also spelled
Bülfinger) (1693–1750), Christian Goldbach (1690–1764), and Gerhard Müller
(1705–1783).
96 Rüdiger Thiele
Fig. 5.12. Oldest known portrait of Leonhard Euler and the only one that shows
Euler with two healthy eyes. Mezzotint print by B. Sokolov (1737) after a lost
painting of J. Brucker. Courtesy Sudhoff-Institut, Universität Leipzig.
Consequently, when his contract ended Daniel Bernoulli returned to Basel

in 1733, as did other academicians. This was the first major change in Euler’s
scientific career. He succeeded Bernoulli and became professor of mathematics.
Euler, at 26, stepped into the leading position in the Academy. He himself was
conscious of his mathematical abilities. We read in a letter to the President
of the Academy Laurentius Blumentrost (1692–1755) that
in Europe there are only a few so advanced in mathematics and physics
as I.26
Euler decided to settle down in Russia. Meanwhile his younger brother

Johann Heinrich (1719–1750) worked as a painter in Petersburg. At the end
of 1733 (old style) Euler married Katharina Gsell (1707–1773), the daughter
of a Swiss painter, and he bought a wooden house. Euler was very fond of
children, and the couple were to have 13, all but five of whom died very
young. Johann Albrecht, Euler’s first son, was born in 1734, followed by three
daughters who died in infancy, and Karl in 1740.27 Euler was able to work
while children played around him.
Under Anna’s reign (1730–1740), that is, under her German favourite
Biron (1690–1772) (in German Bühren), Duke of Courland, Russia suffered
one of its bloodiest periods. However, Euler led a quiet life interrupted only by
one disaster. Since childhood he had suffered from the disease scrofula (kings’
evil), unknown today, which led to terrible consequences during a fever in
1738, destroying the sight in his right eye and weakening the other eye. The
common legends that this loss was caused by his geographical work or by an
exhausting astronomical calculation are not true.
In the 1730s the dominant Wolffian philosophical position received a seri-
ous setback in the Academy because of the departure of some Wolffian mem-
bers. However, Wolff continued to enjoy a position of high esteem in Peters-
burg and he maintained an active correspondence with the Academy. Euler
strongly opposed Leibnizo-Wolffian rationalism, and there is no question that
Wolff was aware of Euler’s opposition. This caused conflicts later in Berlin.
After Anna’s death, a boy tsar again succeeded the Empress in November
1740, and the political situation worsened again.
Things looked rather dubious,28
as Euler himself said. At this time the new Prussian king Frederic II (Friedrich
II) the Great (1712–1786, King from 1740), as young as Euler (28 and 33 years
26
Letter from 7. (18.). 9. 1730. Archive of the Russian Academy of Sciences in St.
Petersburg, AAN, f. 121, op. 2, no. 164.
27
Johann Albrecht (1734–1800), Karl Johann (1740–1790), Katharina Helene [m.
von Bell] (1741–1781), Christoph (1743–1808), Charlotte [m. van Delen] (1744–
1780).
28
Euler 1767/1995, 11-13. “ . . . es bey der darauffolgenden Regentschafft ziemlich
misslich auszusehen anfieng.”
98 Rüdiger Thiele
respectively), decided to reorganise the Berlin Academy, and Euler was invited
to work in this Academy. At that time Berlin was no place of scholarship.
Nevertheless, Euler accepted and finally departed with his family on 19 June
1741. During the years in Berlin, Euler’s scientific work was closely connected
with the Petersburg Academy to which he was appointed an honorary member
(he received a pension). In Petersburg Euler wrote between 80 and 90 papers
on number theory, analysis, and mechanics, of which 55 were published during
his first Petersburg period, among them the following:
1736 Mechanica sive motus scientia analytice exposita, 2 vols.
(Analytical Mechanics; E 15-16; EO II, 1-2),
1738 Einleitung zur Rechen-Kunst zum Gebrauch des Gymnasii bey der
Kaiserlichen Akademie in St. Petersburg (Introduction to the Art of
Arithmetic for Use in the Grammar School Affiliated to the Imperial
Academy in St. Petersburg, 2 vols. 1738-1740; E 17, 35; EO III, 2)
1739 Tentamen novae theoriae musicae (Theory of music; E 33; EO III, 1).
1749 Scientia navalis, 2 vols. (Naval Science [fluid mechanics]
prepared in 1738; E 110-111; EO II, 18-19)
Euler’s international reputation rapidly increased as he regularly won Paris
Academy prizes; in the period from 1737 to 1746 he won the Prix Paris for the
problems posed in the years 1737, 1738, 1739, 1742, 1743, 1746 and 1747;29
in total he received the prize a dozen times, mostly on navigation and naval
science, and another one under the name of his son Johann Albrecht.
It was John Bernoulli who started Euler on his researches in several
branches of mathematics. However, since the time of Euler’s marriage new
fields such as number theory and geography developed rapidly. Number the-
ory had attracted many mathematicians before Euler, above all Pierre de Fer-
mat (1601–1665). In a nutshell, while Pierre de Fermat had formulated and
conjectured, Leonhard Euler proved and refuted. Euler laid the foundation of
number theory as a science.
In 1725 Christian Goldbach (1690–1764), interested in number theory, be-
came the Secretary of Conferences at the Petersburg Academy, but already in
1728 he left Petersburg in order to educate the young tsar Peter (1715–1730,
tsar from 1727) at the court in Moscow. In 1729 Euler opened a very inter-
esting correspondence30 with Goldbach. In his first response Goldbach dealt
with Fermat’s conjecture that all numbers of the form
n
Fn = 22 + 1, n = 1, 2, 3, . . .
are prime. In 1732 Euler showed that F5 is composite:
F5 = 232 = 4, 294, 967, 297 = 641 × 6, 700, 417
29
E 34, E 78, E 57, E 108, E 109, E 150, and E 120.
30
Euler and Goldbach 1965.
Fig. 5.13. Euler’s Tentamen novae theoriae musicae (Essay on a New Theory of
Music; 1739, E 33). The title page shows the Petersburg Academy. Courtesy Uni-
versitätsbibliothek Basel.
100 Rüdiger Thiele
(published in “Observationes de theoremate,” 1738; E 26; EO I, 2). Soon he

extended the question to numbers of the form
a2 + 1,
and, moreover, he looked for prime factors of
a2 + b2 , a2 − b2 where (a, b) = 1 (a and b are relatively prime)
or even
ma2 ± nb2 where m, n are integers.
This investigation led to the basis of a general arithmetic theory of binary
quadratic forms, developed by Joseph Louis Lagrange (1736–1813) and above
all by Carl Friedrich Gauss (1777–1855). Euler’s most remarkable discovery
was the law of quadratic reciprocity, conjectured first in 1744 and completely
stated in a paper “Observationes” in 1772 (published 1783; E 552; EO I, 3).
But he was unable to prove it, so his conjecture was forgotten (cf. E 598 from
1775; EO I, 4).31
It is very impressive to see how Euler’s genius developed from the minor
impulse given by Goldbach to a general theory. And again it is impressive to
see how a given development in mathematics is inevitable and repeats itself.
We recognize the same procedure in Gauss who passed through the same
development up to the law of reciprocity (his theorema aureum, the golden
theorem), but more quickly and successfully, at 24 in 1801.
Incidentally, dealing with the mentioned problems Euler considered the
sums
02 + 41,
12 + 40,
22 + 39,
.. ..
. .
x2 + (41 − x)
(x + 1)2 + (41 − x − 1)
.. ..
. .
392 + 2,
402 + 1,
412 + 0;
as a by-product he gave this nice example:
P (x) = x2 + (41 − x) = x2 − x + 41
supplies primes for x = 0, 1, 2, . . . , 40, but P (41) is composite (since obviously

41 is a divisor of P (41)).
31
Weil 1984, §10.
Furthermore, we note the famous Königsberg32 bridge problem originally

concerning the branched river Pregel and its seven bridges. Euler generalized:
Whatever the arrangement and division of a river into branches may
be, and however many bridges there be, can one find out whether or
not it is possible to cross each bridge on one tour exactly once?33
Euler gave his solution in 1736, founding the new branch 100 years later named
topology (rather than the former analysis situs) by Johann Benedikt Listing
(1808–1882), and popularized by his book Vorstudien zur Topologie (1847).
Berlin
Frederick the Great (Friedrich der Große), or Frederick II, transformed Prus-
sia into an efficient and prosperous state. When his father, king Frederick
William I (Friedrich Wilhelm I), died in 1740, he left an army of about 83,000
soldiers from a population of about 2,000,000. Frederick II was a brilliant
military campaigner; he made Prussia the foremost military power on the
Continent, enlarging Prussia’s territory and undermining Austria’s reign. In-
deed, Frederick the Great was a great military leader but guided by a strategy
of power without regard for losses. However, Henry (Heinrich) Prince of Prus-
sia (1726–1802), a younger brother of Frederick II, preferred a more modern
warfare using tactical military manoeuvres which minimized losses. These dif-
ferences led to a sharp discord with the king.34 In scientific affairs the king
was later on strained terms with Euler.
On the other hand, Frederick II is often regarded as a philosopher on the
throne who introduced new traits of absolutist reign. Indeed, the francophile
Frederick was also interested in the sciences, although more in art and philos-
ophy. Among his first activities was an ambitious scientific aim: to transform
the unimportant Society of Science founded by his father in Berlin in 1700
into a modern academy to rival that in Paris. So, he invited top scholars to
become members in his academy. When the invited Leonhard Euler finally
arrived in Berlin on 25 July 1741, the Prussian king was at war and con-
quered Silesia (Silesian Wars 1740–1742, 1744–1745, the Seven Year’s War,
32
Former eastern outpost of Prussia, now Russia. Seaport. Birthplace of I. Kant
and D. Hilbert.
33
Euler 1736, p 129. “Quaecunque sit fluuii figura et distributio in ramos, atque
quicunque fuerit numerus pontium, invenire, utrum per singulos pontes semel tan-
tum transiri queat, an vero secus?”
34
Frederick William von Steuben (1730–1794), who served in the Prussian army
under Prince Henry, knew these quarrels. When Steuben served the cause of U.S.
independence he supported endeavours to install a constitution like the English one
rather than a republican one and, moreover, he tried to bring Henry to a leading
position as a King or a Governor of the U.S.A. (letter to Henry from November 2,
1786) — but in vain (Henry’s response from April 1787). Cf. von Krockow 1996.
Fig. 5.14. Map of Königsberg, East Prussia, about 1760. Owing to the walks of the
citizens of Königsberg over the river Pregel, the well-known Bridge Problem arose.
Courtesy by Landes- und Universitätsbibliothek Halle.
Fig. 5.15. Euler’s schematic figure in his paper “Solutio problematis ad geometriam
situs pertinentes” (Solution of a problem concerning geometrical position; 1736,
published in 1741, E 53). Courtesy by the library of the Academy Leopoldina, Halle.
1756–1763). Because of the war the Prussian Academy was not reorganized
until 1744 and formally opened in 1745.
In the 18th century universities were not centres of mathematical research,
the leadership taken instead by a few royal academies. In Euler’s case the
Academies of Berlin and St. Petersburg (founded in 1700 and 1725 respec-
tively), which both owed their existence to the restless ambition of Gottfried
Wilhelm Leibniz, gave Euler the chance to be the mathematician we know.
Euler’s friend Daniel Bernoulli congratulated him from the bottom of his
heart for the wonderful appointment in Berlin.35 Incidentally, in addition to
the academies in Paris and London it was above all these academies which
made possible a full century of mathematical progress.
Euler became director of the mathematical class of the Academie Royale
des Sciences et Belles Lettres in Berlin and even substituted for the president
Pierre-Louis Moreau de Maupertuis (1698–1759) when the latter was absent.
At the beginning of his Berlin period Euler was content. From the field the
king addressed a letter to Euler: “A mon professeur Euler” (To my professor
Euler). Euler felt flattered and wrote enthusiastically: “I can do just what I
wish. [. . . ] The king calls me his professor, and I think I am the happiest man in
the world.”36 Euler bought a house in the Bärenstraße 21 [now Behrenstraße]
which still exists. His mother arrived in Berlin in 1750 and stayed there until
her death in 1761.
Euler’s energy was inexhaustible. He supervised
• the library,
• the observatory,
• the botanical garden,
• the publication of scientific papers, and
• various financial matters including the publication of various calendars
and geographical maps (the sale of which was a source of income to the
Academy).
Submitting his papers to the Prussian and to the Russian Academies, Euler
worked in both academies as a mathematician. Nevertheless, his endless
stream of manuscripts overtaxed the publication capabilities of both acad-
emies. His busy pen left many manuscripts after his death, which the Peters-
burg Academy published during the next half century.
All in all, Euler led a peaceful life in Berlin for many years. However,
Euler took part in several sharp philosophico-theological debates, the most
famous of which was the controversial dispute on Maupertuis’s principle of
least action. This principle was published by Maupertuis in a paper entitled
“Accord de différentes Loix de la Nature qui avoient jusq’ici paru incompat-
ibles” (Harmony between different laws of nature which have, up to now,
35
Fuss 1843, 2. Letter to Euler, 28. 1. 1741, 466: “Zu der herrlichen Berliner
Vocation gratulire ich von Herzen.”
36
Euler and Goldbach 1965, cf. the letter to Goldbach, 2./13. 3. 1742.
Fig. 5.16. French Church in Berlin, Gendarmenmarkt. For the French refugiés
(Huguenots) built by L. Cayard from 1701 to 1705, modeled after a Huguenot church
in Chareton; in 1905 reconstruction changed the interior. Photo R. Thiele.
Fig. 5.17. From 1780 to 1885 the marvellous dome tower was added to the church
by K. von Gontard. Owing to the French name of the tower, dôme, which sounds like
the German word Dom (= Cathedral), in German the tower is named Französischer
Dom (French Cathedral) even though the tower is no church. The symmetrically
designed Gendarmenmarkt is one of the most beautiful squares in Germany; the
counterpart of the French Church is the German Church, and both complexes have
a tower and are related to the Theater (Schauspielhaus) as center. Courtesy Archive
of Berlin-Brandenburgische Akademie, Berlin.
Fig. 5.18. The Prussian Academy of Sciences in Berlin (1752). Courtesy Archive
of Berlin-Brandeburgischen Akademie, Berlin.
appeared incompatible) in 1744.37 Two years later he went on to state the

“principe general” in a paper “Les Loix du Mouvement et du Repos déduites
d’un Principe Metaphysique” (On the Laws of Motion and of Rest Deduced
by a Metaphysical Principle) in this way: “to produce some changes in na-
ture the necessary quantity of action is the smallest that is possible.”38 By
it Maupertuis hoped to unify the laws of physics and he even regarded the
principle as a proof of God’s existence (the supreme being). From a mathemat-
ical viewpoint, soon some scholars criticized the general principle (Gottsched
circle in Leipzig 1748, Patrick Comte d’Arcy (1725–1779) 1749, 1752), and
in 1751 Samuel Koenig (1712–1757) even accused Maupertuis of plagiarizing
Leibniz’s work.39
In the 1740s and 1750s this debate changed into a conflict and grew critical;
Voltaire and even Frederick II were involved. Euler supported Maupertuis,
but Maupertuis was finally ruined. Euler’s attitude was ambiguous: because
he interpreted the principle as a theological one, he was compelled to defend
religion against the hated ideology of free thought. On the other hand, he
37
Maupertuis 1744; cf. also Thiele 1996, 373-390, and 1999, 437-504.
38
Maupertuis 1746. “Lorsqu’il arrive quelque changement dans la Nature, la
Quantité d’Action, nécessaire pour ce changement, est la plus petite qu’il soit pos-
sible.”
39
Anonymous review of Maupertuis’s 1744 paper, in: Neuer Büchersaal, vol. 7,
Leipzig 1748, pp. 99-117; d’Arcy 1749 (published in 1753) and 1752; Koenig 1751.
Fig. 5.19. Frederick II (the Great), king of Prussia (1740-1786) (left); P.L.M.
de Maupertuis, President of the Berlin Academy (1746-1759). Courtesy Sudhoff-
Institut, Universität Leipzig.
correctly formulated the principle for some cases, and he believed that nature
operates in this way. From his Methodus inveniendi (1744):
For since the fabric of the Universe is most perfect and the work of a
most wise Creator, nothing at all takes place in the Universe in which
some rule of maximum or minimum does not appear.40
Euler was unpopular at Frederick’s court. “Nous avons ici un gros cyc-
lope de géomètre”, Frederick II tastelessly wrote to Voltaire.41 Euler’s and
the king’s personalities were too different; these two important figures of the
Enlightenment never became close friends. Frederick viewed science as the
servant of the state — his state. King Frederick II exclusively judged science
in view of its utilitarian aspects (not unlike the present arguments for cut-
ting budgets). Inasmuch as Frederick appreciated Euler’s scientific talents he
engaged him in a variety of practical problems. Examples are the projects to
correct the level of the Finow Canal and to build up a hydraulic system of
40
Euler 1744, appendix De curvis elasticis (also in: EO I, 24, p. 231). “Cum enim
Mundi universi fabrica sit perfectissima atque a Creatore sapientissimo absoluta, ni-
hil omnio in mundo contingit, in quo non maximi minimive ratio quaepiam eluceat.”
41
Frederick II, 1849, 11. Letter to Voltaire from November 19, 1748, p. 128. “Here
we have a great cyclops of mathematics [in Greek mythology cyclops are giants with
one eye; Euler was nearly blind in one eye].”
pumps and pipes at the royal summer residence. Unfortunately, the fountains
never worked satisfactorily.42 The King interpreted this technical failure as
a failure of science itself and commented maliciously: “Vanités des vanités!
vanité de la Géometrie!”43 Frederick preferred French culture and its repre-
sentatives, such as Voltaire or Maupertuis.
When Maupertuis died in 1759 (the year of the battle of Québec), Euler
continued to run the Academy, but Frederick never made him President. Euler
and the King differed more and more sharply, especially in financial affairs
and personnel matters concerning the Academy. Calendars were a source of
revenue for the Berlin Academy. On Euler’s mathematically unconvincing
accounts with the king dealing with the sale of calendars Frederick responded
arrogantly: “I who do not know how to calculate curves do know that sixteen
thousand écus [French coin] of receipts are preferable to thirteen thousand.”44
For this and other reasons Euler began to think of leaving Berlin in 1762,45
and during the Seven Years War he contacted the enemy Russians where
Catherine the Great (tsarina from 1762) had come to power. The war ended
in 1763, the same year Canada was ceded to the British by the Treaty of Paris.
In 1766, February 2, Euler pleaded for royal permission to leave Prussia, but
the King, now becoming aware of the immense loss, declined the request. Euler
insisted. Finally, on May 2, the King agreed with the following humiliating
words, showing no gratitude for Euler’s incomparable work:
42
Recently M. Eckert (Eckert 2002) pointed out that Euler is not to blame for
the failure of Frederick’s ambitious fountain project inspired by the French model
at Versailles. Eckert quoted a letter of Euler to the king dated October 17, 1749
(EO VIA, pp. 320-330, citation p. 322, VIII): “Car sur le pied qu’elles se trouvent
actuellement, il le bien certain, qu’on n’éleveroit jamais une goutte d’eau jusqu’au
reservoir, et tout la force ne seroit employée qu’à la destruction de la machine et
des tuyaux” (For in the situation in which they [the pumps] are at present, it is
quite certain that one would never raise one drop of water as far up as the reservoir,
and the entire force would be employed only for the destruction of the machine
and the tubes; Eckert’s translation, p. 457). But in 1778 Frederick II angrily wrote
to Voltaire on the presumed mathematical execution (“exécuté géométriquement”):
“Il [moulin] n’pu élever une goutte d’eau à cinquante pas du bassin” (The pumps
did not elevate one drop of water fifty steps over the basin). January 25, 1778. In:
Frederick II, 1853, 23, 421.
43
Frederick II, 1853, 23. Letter to Voltaire from January 25, 1778, p. 421.
44
Letter to Euler from June 16, 1765. In: EO VI A6, p. 390, and Frederick II,
1852, 20, 209. “Moi, qui ne sais point calculer des courbes, je sais pourtant que
seize mille écus de recette en valent mieux que treize mille.” As early as in 1743
Frederick II wrote to Euler: “Vous avez péché contre les règles ordinaires du calcul”
(You contravened elementary rules of calculation), letter from January 21, 1743, p.
303 and 200 resp.
45
Cf. K.-R. Biermann 1985, 91-99.
Fig. 5.20. Map showing the Northern hemisphere of the Geographischer Atlas beste-
hend in 44 Land=Charten, worauf alle Theile des Erd=Creyses vorgestellt werden
(Geographical Atlas Consisting of 44 Maps Showing All Parts of the World). The
map was drafted by the Count of Redern in 1754. The whole atlas was published by
the Berlin Academy by order of L. Euler in 1762. Already in St. Petersburg Euler
was involved in the edition of a map of the whole Russian Empire which appeared
in 1745. Courtesy the library of the Academy Leopoldina, Halle.
Je vous permets, sur votre lettre du 30 d’avril dernier, de quitter pour

aller en Russie.46
Euler, at 59, left Berlin on 9 June 1766. Joseph-Louis de Lagrange succeeded

him in the same year and remained in Berlin until the death of Frederick II
in 1786. At the same time James Cook (1728–1779) explored the seaways and
coasts of Canada (1763–1767) and began to prepare his first circumnavigation
of the world in 1768–1771.
During the Berlin period Euler prepared about 380 works of which 275
were published, including these books:
46
Letter from Frederick II to Euler. In: EO IV A6 Basel: Birkhäuser 1986, p. 393.
“Referring to your letter dated 30 of April I permit you to quit to depart for Russia.”
Fig. 5.21. Euler’s home in Berlin in the Behrenstraße (opposite the Komische Oper
(Comic Opera), present view). Photo R. Thiele.
1744 Methodus inveniendi

(calculus of variations; E 65; EO I, 24)
1744 Theoria motuum planetarum et cometarum
(calculation of orbits; E 66; EO II, 28)
1745 Neue Grundsätze der Artillerie
(translated and enlarged edition of New Principles of Gunnery
by B. Robins; E 77; EO II, 14)
(not before) 1745 Anleitung zur Naturlehre
(Introduction to Natural Sciences, published postum in
Opera posthuma II, 1862; E 842; EO III, 1)
1746 Gedancken von den Elementen der Körper
(objections to monadology which were based on physical and
theological nature; E 81; EO III, 2)
1747 Rettung der göttlichen Offenbarung gegen die Einwürfe
der Freygeister
[anonymous] (The Rescue of Divine Revelation from the Objections
of the Freethinkers; E 92; EO III, 12)
1748 Introductio in analysin infinitorum
(Elements of Analysis; E 101–102; EO I, 8-9)
1748 Réflexions sur l’Espace et le Tems [temps]
(Reflections on space and time; E 149; EO III, 2)
1749 Scientia navalis
(on shipbuilding and navigation, prepared in St. Petersburg;
E 110–111; EO II, 18–19)
1753 Theoria motus lunae
(first lunar theory; E 187; EO II, 23)
1755 Institutiones calculi differentialis
(Introduction to the Differential Calculus, prepared about 1748;
E 212; EO I, 10; his integral calculus written in 1763 appeared in
1768–1770; E 342, 366, 385, 660; EO I, 11–13)
1765 Theoria motus corporum
(second mechanics, mechanics of solids; E 289; EO II, 3–4)
[1765] Théorie général de la dioptrique
(General Theory of Lenses, prepared in Berlin, published in 1862;
E 844, EO III, 9)
[1768] Lettres à une Princesse d’Allemagne
(philosophical letters, popular science, and philosophy, prepared in
Berlin, but published in St. Petersburg; unusual success: at least
12 French, 9 English, 7 German, 4 Russian editions;
E 343, 344, 417; EO III, 11–12)
In Berlin Euler’s research was at the summit of his creative powers, mainly
devoted to analysis and mathematical physics with a practical orientation. In
his Introductio (E 101) Euler gave the so-called Moivre formula
eiϕ = cos ϕ + i sin ϕ,

Fig. 5.22. Title page of the Methodus inveniendi (calculus of variations; 1744, E 65).
“One of the most beautiful mathematical works ever written” (C. Carathéodory).
The poster on the tree shows a cycloid. Courtesy the library of Mathematische
Institut of Universität Leipzig.
Fig. 5.23. Title page of the Neue Grundsätze der Artillerie (New Principles of Gun-
nery; 1745, E 77). Essential extended translation of Benjamin Robins book with the
same title (1742), in a French translation (1751) read by Napoléon I Bonaparte. In
the publisher’s signet the slogan “Sapere aude (dare to be wise)” of the Enlighten-
ment appears. Courtesy Universitätsbibliothek Leipzig.
which in a slightly different form had been given by Roger Cotes (1682–1716)
in 1714.47 A special case appears in the paper “De la controverse entre Mrs.
Leibnitz et Bernoulli sur les logarithmes des nombres négatifs et imaginaires”
(On the controversy between Messers. Leibniz and John Bernoulli on the
logarithms of negative and imaginary numbers) (1749, 1751 printed; E 168;
EO I, 17, pp. 195–232): log(−1) = i (discovered already in 1727), or in another
form eiπ + 1 = 0, linking five important magnitudes in one equation.
St. Petersburg
Euler and his family (18 persons including 4 servants) arrived in St. Peters-
burg on 28 July 1766. He did not find any friends from his old St. Petersburg
days, but he was received overwhelmingly by Catherine the Great, born a
German princess. Again he lived a quiet life, and above all he lived for math-
ematical sciences. Clifford A. Truesdell remarked: “There is no evidence that
governmental ‘despotism’ had any influence on Euler, who, like most scien-
tists, accepted and rejected positions on the basis of the concrete conditions
of work and pay they involved.”48
However, he suffered some setbacks. Soon after his return he became al-
most completely blind. An operation only temporarily restored his sight in
1771. Owing to his phenomenal memory49 and his enormous powers of men-
tal calculation he was equal to the challenge and continued to work, dictat-
ing the results to his assistants, his sons Johann Albrecht (1734–1800) and
Christoph (1743–1808) as well as Anders Johan Lexell (1740–1784), Wolf-
gang Ludwig Krafft (1743–1814), Mikhail Evseyevich Golovin (1756–1790)
and Nicolas (Nicolaus) Fuss. Fuss prepared about 250 of Euler’s papers and
Golovin 70 which were written during Euler’s second time in St. Petersburg.
In 1771 Euler’s house on the embankment of the Great Neva was destroyed
in a great fire, but with the help of a Swiss servant blind Euler escaped and the
most of his manuscripts were saved. The house was rebuilt and is preserved,
but with some changes. Nevertheless Euler, then in his sixties and almost
blind, “was the principal light of Catherine II’s Academy of Science” (C. A.
Truesdell).50 His papers were generally short and devoted to a particular topic;
however, he completed voluminous books such as the Theoria motum lunae
(1772) with 775 pages and the three volume Dioptrica (1769–1771). Almost
half of his papers were written in St. Petersburg, among them:
47
Cf. Cotes 1714.
48
Truesdell 1984, 213.
49
Euler was able to recite Virgil’s Aeneid from beginning to end (almost 10,000
lines) in Latin by heart.
50
Cf. Truesdell 1984, 338.
Fig. 5.24. Euler’s investigations of elastic lines (of elastic materials as strips, beams,
etc.) by means of the calculus of variations in “De curvis elasticis” (On elastic
curves) led to nine species of which four are shown in table IV (Additamentum I
in the Methodus inveniendi; 1744, E 65). Euler regarded an ideal elastic line to be
one whose positions of stable equilibrium are characterized by the minima of the
potential energy (i.e., by John Bernoulli’s principal of virtual work). Cf. fig. 30.
Courtesy the library of the Mathematisches Institut of Universität Leipzig.
1768 Lettres à une Princesse d’Allemagne

(Letters to a German Princess [non-existent title], 1768–1772, 3 vols.;
E 343, 344, 417; EO III, 11-12; prepared in Berlin)
1768 Institutiones calculi integralis
(Introduction to the Integral Calculus, 3 vols., 1 vol. postum,
1768–1770, 1794; E 342, 366, 385, 660; EO I, 11–12)
1768 Dioptrica
(Optics, 1768–1770, 3 vols.; E 367, 386, 404; EO III, 3–4)
1770 Vollständige Anleitung zur Algebra
(Complete Introduction to Algebra, 2 vols.; E 387–388; EO I, 1)
1772 Theoria motuum lunaea
(second lunar theory; E 418; EO II, 22)
1773 Théorie Complette de la Construction et de la Manoeuvre des Vaissaux
(second naval theory; E 426; EO II, 21)
In 1773 Euler’s wife, who had managed the whole household, died. Euler,
at 69 and blind, intended to marry again. The children opposed their father’s
marriage. At first Euler gave in, but eventually he insisted. Euler announced
his wedding to his children on July 20, 1776. On July 28, 1776 the wedding
ceremony took place in his house and he married Salome Abigail Gsell (1723-
1793), a half-sister of his late wife.51
Leonhard Euler died on 18 September 1783 at the age of 76 years, 5 months
and 3 days, active to the end. On that very day he gave instruction to his
grandson in the morning, did some mathematical calculations on the motion
of balloons (recall Montgolfière), dined and discussed the orbit of Uranus
(discovered in 1781 by Frederick William [Friedrich Wilhelm] Herschel, 1738–
1822) with his assistants Nicolas Fuss and Andreas Johann Lexell (1740–1784).
Then Euler outlined his calculations, played with his grandson, whom he had
been teaching that morning, and smoked his pipe. About five o’clock Euler’s
pipe dropped from his hand and fell to the floor. Euler stooped, but straight-
ened up again without the pipe and said: “Ich sterbe” (I am dying). He suffered
a stroke and died the same night.
Euler ceased to calculate and to live,52
said the Marquis Marie-Jean-Antoine-Nicolas de Condorcet (1743–1794), a

French philosopher of the Enlightenment and Perpetual Secretary of the Paris
Academy in his Éloge de M. Euler at the Paris Académie des Sciences. Twenty-
two years earlier Euler had written in his Lettres à une Princesse d’Allemagne
on death:
Since death is a dissolution of the bond which links body and soul
during lifetime, [. . . ] we can derive an idea of the status of the soul
51
Cf. the contribution “Der Zwist um die zweite Ehe Eulers (The quarrel con-
cerning Euler’s second marriage)” by G.K. Michajlow in Fellmann 1995, 112-116.
52
Condorcet 1783. “Il cessa de calculer et de vivre.”
Fig. 5.25. Memorial meeting of the Petersburg Academy in honour of Euler in 1783.
Silhouette by F. Anting (1784). Euler’s son Johann Albrecht puts a bust of Leonhard
Euler on a base. Academicians of the Mathematico-Physical Department, from left
to right: A.J. Lexell (1740–1784), J.A. Euler (1734–1800) (with the bust), N. Fuss
(with the amphora), I.I. Lepechin (1740–1802), P.S. Pallas (1741–1811), W.L. Krafft
(1743–1814); the middle oval shadow figure on top shows the tszarina Catherine II
(1729–1796, empress from 1762). Courtesy Sudhoff-Instituts of Universität Leipzig.
after death. While the soul during life received all its information
from the senses, it will no longer receive any information. [. . . ] Sleep
provides an obvious picture and at the same time an experience of this
state. [. . . ] Hence after death we shall be in a state of most perfect
dreaming [. . . ] And, I think, this is almost all that we can say [on
death] in a distinct way.53
53
Euler 1768-1772, Part II, Letter no. 93. Also in: EO III, 11. “La mort n’est
donc autre chose que la destruction de cette liaison: . . . On peut se former quelque
idée de l’etat de l’ame après la mort. Comme l’ame pendant la vie tire toutes ses
connoissances par le moyen des sens, étant dépoulliée par la mort de ce rapport
des sens, elle n’apprend plus rien de ce qui se passe dans le monde matériel. . . . Le
Sommeil nous fournit aussi un bel échantillon de cet état. . . . Ainsi après la mort
nous nous trouverons dans un état des songes les plus parfaites, que rien ne sera plus
capable de troubler. . . . Et c’est à mon avis à peu près tout ce que nous saurions en
dire de positif.”
Fig. 5.26. St. Petersburg, about 1760. The view along the river Neva is downstream.
The Imperial Russian Academy on the right, the Imperial Winter Palais on the
left. Not far from the right side of the engraving on the embankment of the river
Neva (Vasiliostrov, Vassili Island) is Euler’s house in which he lived from 1766 to
1783; its present view differs essentially from the original. Copper engraving by
Makhayev. Courtesy Niedersächsische Staats- und Universitätsbibliothek Göttingen,
Handschriftenabteilung, gr. 2◦ H. Russ. 434.
Fig. 5.27. Conference Room in the Imperial Academy in St. Petersburg (recon-
struction). Courtesy the Archive of Berlin-Brandenburgische Akademie, Berlin.
The Concept of Function

In any lecture on Euler, who is known as analysis incarnate (l’analyse in-
carnée),54 it is indispensable to say something on his analysis. Incidentally,
the status of analysis at that time is described by Voltaire (1694–1778) in this
way:
The art of numbering and measuring exactly a thing whose existence
cannot be conceived,
and Hermann Weyl (1885–1955) spoke of “Euler’s era of happy-go-lucky analy-
sis”55 in order to point out the carelessness which was rooted in good and
reliable intuition.
It was Euler who established the concept of function at the heart of the
new branch of mathematics called analysis. Furthermore, analysis was not
only an application of algebra to geometry. It became a subject in its own
right, above all by making a formal theory of functions that had no need
of geometrical conceptions. Therefore no figures are to be found in Euler’s
textbooks on analysis as for example had been the case in Isaac Barrow’s
Lectiones geometricae (1670).
But, what is a function?
“Nobody can explain that,” wrote H. Weyl in Philosophy of Mathematics
and Natural Sciences, first published in 1928 and again in 1949.56 I will re-
spond to this question as a historian. To understand better why Euler created
functions I, therefore, provide a very brief survey of functional concepts.57
Since the time of the Greeks magnitude has played the central role in geome-
try, characterised by its ability to increase and to decrease. By constructions,
in general by ruler and compass, we get new magnitudes. This procedure might
be called “geometrical concept of a function” used in the time of Gottfried
Wilhelm Leibniz (1646–1716) and Isaac Newton.
However, new ideas arose. In Cartesian geometry (1637) in general ordi-
nate depends on abscissa; such a relation between magnitudes is graphically
represented by means of curves as well as arithmetically by means of pro-
portions. Instead of ordinate René Descartes (1596–1650) said “la ligne est
appliquée par ordre” (“the line is applied/drawn in given order”; in modern
terms this means by a “given function”).58 Also the mathematical description
of the real world needed concepts of relationship between observed quantities.
54
“Euler, qu’on aurait pu appeler presque sans métaphore et certainement sans
hyperbole, l’analyse incarnée.” (Euler, whom almost without metaphor and surely
without exaggeration, we can call analysis incarnate.) Arago 1854, p. 443.
55
Weyl 1968, 124.
56
Weyl 1928. “Niemand kann erklären, was eine Funktion ist,” p. 8. In the Ameri-
can edition (1949) he even added: “...but this is what really matters in mathematics.”
However, in the index there is no entry “function.”
57
Cf. Thiele 2000, 128–181.
58
Descartes 1637, 66.
The roots of these concepts are to be found in Galileo Galilei (1564–1642) or

Isaac Newton. The latter considered all magnitudes as time-dependent. Like-
wise the emergence of an analytical concept arose in the calculus of variations,
the typical problem of which is the determination of magnitudes, especially
curves (geometrical view) or functions (analytical view). Craig Fraser (born
1951) gave an evaluation in greater detail:
Although the theme of analysis was well established at that time
[about 1730] there was in his [Euler’s] work something new, the be-
ginning of an explicit awareness of the distinction between analytical
and geometrical methods and an emphasis on the desirability of the
former in proving theorems of the calculus.
Although Euler in 1744 clearly recognised the essential analytical char-
acter of the variational calculus his insight was not fully developed in
his treatise. . . . There was an increasing emphasis on analysis.59
In 1697 using a geometrical concept of function James Bernoulli posed the

famous isoperimetrical problems60 in response to his brother’s brachistochrone
problem. (Incidentally, the year 1997 was its tercentennial.) Dealing with his
brother’s new challenge John extended — or let me say transformed — this
geometrical concept into an analytical one, now representing the geometrical
curves analytically. In a paper published in 1706 but communicated to Leibniz
already in 1698 in a letter he said:
. . . to find the curve the ordinates of which are of given power [of a
line], or generally, which is the function of its ordinates expressed by
other ordinates.61
He repeated this definition in a paper in 1718:
Here, I call a function of one variable a quantity which is composed

in some way of this variable quantity and of constants.62
The geometrical relationship, the “geometrical concept of function,” disap-

peared. From this time forward it was Euler who developed the arithmetical
concept.
There is an unpublished Latin manuscript “Calculus differentialis” by
Euler written about 1727. In 30 pages he had briefly outlined the ideas which
59
Fraser 1997, 63.
60
Bernoulli, Jakob 1697, 211–217. Also in: Bernoulli 1991, 271–282.
61
Bernoulli, Johann 1706, 235: “ . . . trouver la Courbe telle que ses appliquées
élevées à une puissance donné, ou généralement telle que les fonctions quelconques
de ces appliquées, exprimées par d’autres appliquées”; already 1698 in a letter to
Leibniz. See also Bernoulli 1991, 515–526, quotation p. 515.
62
Bernoulli, Johann 1718. “On appelle ici fonction d’une grandeur variable, une
quantité composée de quelque manière que ce soit de cette grandeur variable et de
constantes.” See also Bernoulli 1991, 527–568, quotation p. 534.
Fig. 5.28. John Bernoulli’s concept of a function of one variable that arose by
solving the isoperimetric problems, contained in the paper “Remarques” (Op. CIII)
which was published in the Mémoires de l’Academie Royal des Sciences, Paris 1718,
pp. 100–138. “We call here a quantity composed somehow by this variable quantity
and by constants a function of one variable magnitude.” This is the starting point of
Bernoulli’s disciple Euler. The first line of Euler’s unpublished manuscript “Calculus
Differentialis” (about 1730) reads: “Quantitas quomodocunque ex una vel pluribus
quantitatibus composita appelatur ejus unius vel plurium functio” (A quantity which
is composed in some way from one or more quantities is called its or their function
respectively). The nature of the composition was incessantly extended by Euler.
Courtesy the library of Academy Leopoldina, Halle.
he gave in the Introductio in analysin infinitorum (1748, E 101) about 20 years

later. Euler began the manuscript by introducing the concept of function:
One quantity composed somehow from one or a greater number of
quantities is called its or their function.63
Then he pointed out that functions may be composed by means of algebraic

operations like addition, multiplication, etc., by extraction of roots, by in-
version, by taking logarithms, or by combinations of these operations. In his
Introductio he stressed the idea that analysis is a science of functions and
presented this statement:
§4. A function of one variable quantity is the analytic expression (an-

alytic formula) composed in any way from this variable quantity and
numbers or constant quantities.64
A function is an analytic [i.e., a calculating] expression. In the course of

time gradually more operations were permitted. Here, analytic expressions or
formulas included transcendental functions or implicit functions or functions
arrived at in the integral calculus. Euler classified functions; for example,
he distinguished algebraic, transcendental, odd, even, single or multivalued
functions, etc. This simple characterisation of functions by Euler makes clear
63
Juschkevitsch 1983, 161. “Quantitas quomodocunque ex una vel pluribus quan-
titatibus composita appelatur ejus unius vel plurium functio.”
64
Lausanne: Bousquet 1748. Also in: EO I, 8–9. Vol. I, chap. 1, §4: “Functio quan-
titatis variabilis est expressio analytica quomodocunque composita ex illa quantitate
variabili et numeris seu quantitabus constantibus.”
Fig. 5.29. Title page and frontispiece of the first edition of Euler’s Introductio in
analysin infinitorum (1748, E 101), vol. 1. Courtesy the library of Mathematische
Institut of Universität Leipzig.
that functions were now not only a tool for mathematics but had become
an object itself in mathematics. However, we should not overinterpret Euler’s
classification in the modern understanding as function spaces, i.e. as sets of
functions with an algebraic and topological structure (because such concepts
were not developed in Euler’s time).
In analytic formulas variable quantities represent numbers. Therefore, the
idea of variable quantity assumed — at least implicitly — a concept of any
domain of numbers. When Euler explained the difference between constant
[a] and variable quantities [x], he took into account also such variables as
infinitesimal quantities [dx] having the strange algebraic property that
a + dx = a; (5.1)
in the words of his teacher John Bernoulli in 1691/92:
A quantity, which is increased or decreased only by an infinitely small

quantity, may be considered as remaining the same.65
65
Bernoulli 1922, 3: “Quantitas diminuta vel aucta quantitate infinitis minore
neque diminuitur neque augetur.”
Fig. 5.30. The beginning of the contents of G.F.A. de l’Hospital’s Analyse des
infiniment petit (Analysis of the Infinitely Small), Paris 1696 (on the left), and
L. Euler’s Introductio, vol. 1, Lausanne 1748 (on the right). Obviously, the con-
cept of a function is totally missed in l’Hospital. Courtesy Landes- und Univer-
sitätsbibliothekbibliothek Halle and the library of Mathematische Institut of Uni-
versität Leipzig.
Consequently, Euler even wrote “dx revera = 0” (indeed equal zero)!66 In the
differential calculus Institutiones calculi differentialis (published in 1755 but
prepared about 1748; E 212; EO I, 10) he elaborated the calculus of zeros,
in which he distinguished between arithmetical and geometrical proportions
of zeros. There are various interpretations of this concept. Three examples:
the “classical” understanding of infinitesimals as variables tending to zero,
the view of non-standard analysis regarding infinitesimals as elements of a
non-Archimedean number field, and the “physical” intuition. Euler himself
wrote:
If a [non-negative] quantity is so small that it is smaller than any given

one, then it certainly could not be anything but zero. [. . . ] To those
who ask what the infinitely small quantity [differential] in mathematics
is, we answer that it is actually zero. Hence there are not so many
mysteries hidden in this concept as there are usually believed to be.
The supposed mysteries have rendered the calculus of the infinitely
small quite suspect to many people. Those doubts that remain we
66
Euler 1755, Caput III, p. 78. Also in: EO I, 10. Leipzig, B.G. Teubner, Leipzig
1913, cap. III, §83, p. 69.
shall thoroughly remove in the following pages, where we shall explain

this calculus [which is only a special case of the calculus of finite
differences].67
What before in the calculus of finite differences was assumed arbitrary is

now taken to be infinitely small in the differential calculus. Taking into ac-
count the specific properties of infinitesimal quantities, Euler transformed the
established rules of the calculus of finite differences into that of the differen-
tial calculus. A general principle is that in arithmetical comparison with finite
quantities infinitesimal quantities can be neglected ( = zero), i.e., (1) is indeed
valid.68 On the other hand, geometrical proportions of infinitesimal quantities
led to the rules for differentiating functions. Euler’s pragmatism here is most
remarkable. Already in the early Latin manuscript he pointed out:
It is evident that the differential calculus is a special case of the cal-

culus which I have enunciated above [that is, the calculus of finite
differences].69
Furthermore, in the beginning Euler believed that any analytic expression

could be expressed as a power series. This idea vaguely anticipates the theorem
of Weierstrass: the polynomials are dense in the continuous functions on a
closed and bounded interval. His robust pragmatism may be illustrated by his
remark:
If anyone doubts that every function can be so expressed the doubt
will be set aside by actually expanding functions.70
Obviously, this concept describes analytic functions. However, Euler extended

the concepts of analytical functions and regarded other power series including
so-called Puiseux and Laurent series:71
67
Euler 1755, Caput III, §83, p. 78 (and EO I, 10). “Si enim quantitas tam fuerit
parua, ut omni quantitate assignabili sit minor, ea certe non poterit non ess nulla;
namque nisi esset = 0. [. . . ] Quaerenti ergo, quid sit quantitas infinite parva in
mathesi, respondemus eam ess revera = 0; neque ergo in hac idea tanta mysteria
latent, quanta volgo putantur et quae pluribus calculum infinite parvorum admodum
suspectum reddiderunt. Interim tamen dubia, si quae supererunt, in sequentibus, ubi
hunc calculum sumus tradituri, funditus tollentur.”
68
Euler 1755, Caput III, §87, p. 80. “Hinc sequitur canon ille maxime receptus,
quod infinite parva prae finitis evanescant, atque adeo horum respectu reiici queant.”
69
Cf. Juskewisch 1983, 164. “Perspicuum est Calculum differentialem, ejus, quem
ante exposui, calculi esse casum specialem.”
70
Euler 1748, cap. 4, §59 (and in: EO I, 8). “Si quis dubitet, hoc dubium per
ipsam evolutionem cujusque Functionis tolletur.”
71
Victor Alexandre Puiseux (1820–1883), expansion in “power” series with a fixed
rational exponent (1850); Pierre Alphonse Laurent (1813–1854), (complex) power
series with integral exponents (1843). Such expansions of a function are useful in
the neighborhood of a pole or a singularity.
Fig. 5.31. The use of the universally adopted sign f (x) to stand for a function
of one given variable x is found firstly in an additamentum to Euler’s memoir “De
infinitis curvis (On infinitesimal curves)” of 1734, published in 1740 (E 44). Courtesy
the library of Academy Leopoldina, Halle.
Az α + Bz β + Cz γ + Dz δ + etc.
(where α, β, γ, δ, etc. are arbitrary rational and integer numbers for Puiseux
and Laurent series respectively; Cap. 4, §59). In 1905 in a paper on func-
tions which can be represented analytically, Lebesgue showed the range of
Euler’s conception: when infinite expressions such as infinite series and prod-
ucts and continued fractions are allowed, then the class of functions is equal
to that of measurable Borel functions, which for Rn coincide with the class
of all Baire functions. In later years Euler laid less stress on the need for any
particular kind of the analytical form. He had noticed that the known func-
tions (analytic expressions) were insufficient for the requirements of analysis,
especially in the debate on the vibrating string (1747). The boundary condi-
tions of the problem lead to nonanalytic functions. A most natural shape is
one with a non-differentiable point: the plucked string. If the initial state of
the string is represented by an arbitrary hand-drawn curve, then we cannot
expect the solution of the differential equation of the vibrating string to be
an analytic expression. The debate on the new class of functions continued
for another 20 years and involved prominent mathematicians, among them
Euler, d’Alembert, and Daniel Bernoulli. In fact, in the discussion all the par-
ticipants advanced incorrect arguments, and everybody attacked everybody.
It led to ugly polemics; for instance, Jean d’Alembert (1717–1783) criticised
Euler who for his part wrote to Lagrange in 1759:
I doubt whether he [d’Alembert] is serious, unless perhaps he is thor-
oughly blinded by self-love.72
Incidentally, these were not at all the true colours of d’Alembert that the
annoyed Euler saw.
To summarise the complicated debate, what is the general solution of the
problem of the vibrating string? I mention the central points of disagreement.
72
EO IVA5. Basel: Birkhäuser 1980, p. 420. “Je doute qu’il joue se rôle
sérieusement, à moins qu’il ne soit profondément aveuglé par l’armour-propre.”
Euler and d’Alembert disagreed on which kinds of functions were admitted.

Euler gave up the idea of one law (formula) describing the whole function and
allowed what he called discontinuous functions (that is, piecewise analytic
expressions, whereas discontinuous in the sense of Euler refers to the sev-
eral different analytic expressions that represent the law).73 Daniel Bernoulli
suggested trigonometric series to solve the wave equation. Euler rejected this
possibility because the character of the trigonometric functions imposes cer-
tain restrictions on the form of solution that therefore cannot be fully general.
This is an important question playing a crucial role in further development:
can an arbitrary function be given by a particular type of series representa-
tion? Incidentally, Euler gave cursory attention to trigonometric series and on
this occasion, curiously, he unconsciously did reasonable work to determine
the coefficients in a trigonometric expansion.
In his Institutiones calculi differentialis (1755) one important result of
the controversy appeared. Euler gave a general definition of a function as a
quantity whose values somehow change with the changes of the independent
variables:
If, therefore, x denotes a variable quantity, all quantities which depend

in some way on x or are determined by it, are called functions of this
variable.74
The crucial words of an interpretation are “depend in some way” and “are
determined.” In the actual forming of such functions Euler cannot help but
use the known kinds of determining; i.e., he must use the standard algebraic
and transcendental operations. In other words, in his Introductio (1748) he
had already dealt with the same concept; more or less his definitions given
in 1748 and 1755 are to be regarded as being equivalent. Functionality was
for Euler a matter of formal representation by calculable expressions and not
so much as a description of relations by concepts. Therefore any attempt to
interpret the 1755 definition in modern terms does not at all meet Euler’s
sight. Moreover, in a paper written in 1765 Euler remarked that the known
calculus “can only be applied to curves whose nature can be contained in one
analytic equation.” In 1829 it was Dirichlet who showed Fourier series can
represent a broad class of “arbitrary” functions, among them the functions
of classical physics. In a lecture in Berlin in 1899, Herman Amandus Schwarz
rightly spoke of “empirical functions” which with the help of Fourier series
can be made computable.
Such modern interpretation is also often falsely attributed to the defin-
itions of Jean Baptiste Fourier (1768–1830), Nikolai Ivanovich Lobachevsky
73
Cf. the above definition in the Introductio (Euler 1748, §4), in which Euler
demands one analytic expression for the law.
74
Euler 1755, preface, p. VI. “Si igitur x denotet quantitatem variabilem, omnes
quantitates, quae utcunque ab x pendent, seu per eam determinantur, eius functiones
vocantur.”
Fig. 5.32. On vibrating strings and the extension of the concept of functions.
(left) Euler, “Sur la vibration des cordes” (1749, E 140). There is a Latin fore-
runner (E 140). Like J. le Rond d’Alembert, Euler had the general solution
y = y(x, t) = f (x + ct) + g(x − ct). D’Alembert regarded the arbitrary functions g
and f as represented by analytic expressions, whereas Euler, extending the concept,
did not demand such restricting representations.
(right) D. Bernoulli, “Réflexions et éclairissement sur les nouvelles vibrations des
cordes” (1753). Daniel Bernoulli used the eigenfunctions yn = an sin πx cos nπt and
got the general solution by superposition; Euler did not regard this composition as
the general solution. Courtesy the library of Academy Leopoldina, Halle.
Fig. 5.33. Euler’s general definition of a function in the Institutiones (1755, E 212).
Courtesy the library of Mathematische Institut of Universität Leipzig.
(1792–1856) and Johann Peter Gustav Dirichlet (1805–1859) given in 1834

and 1837, respectively. However, like Euler in his analytic expression, both
regarded above all continuous changes of functions only, and therefore dealt
not with modern functionality, but were guided by “mechanical” motions.
Discontinuities appear in singular points only. Dirichlet’s famous everywhere
discontinuous function (in the modern sense) served in the first place as an
example of a non-integrable function and not as an extension of the func-
tion concept. A one-to-one correspondence between arbitrary sets (= systems
in Dedekind) appears 1887 in the famous paper Was sind und was sollen die
Zahlen? (What are Numbers and what is their Meaning?) by Richard Dedekind
(1831–1916).75 Whereas for numbers a topology is given by intuition, it was
to be developed for arbitrary sets in order to define a continuous function on
such sets.76
Furthermore, Euler stated that there was no need for the relation between
the quantities to be given by the same law throughout an interval, nor was
it necessary that the relation be given by mathematical formulas (and such
functions he called discontinuous). He therefore regarded curves freely drawn
by hand (cum libero manus ductu) as functions — these are the so-called
mechanical (or transcendental) curves. Moreover, Euler even discussed the
graph of y(x) = (−1)x .
Euler did all things as easily as he could. Therefore, he fitted his con-
cepts to the problems (not the other way round), and so he did in the case of
functions too. When he built up analysis, the analytic expression was perma-
nently extended. It is openness that characterizes Euler’s concept of function.
Finally, Euler classified functions as continuous and discontinuous. Continu-
ous functions in the sense of Euler are identical with the functions he used
in the Introductio (1748) and the Institutiones (1755) whereas their nature
can be contained in one analytic expression (i.e., Euler’s notion differs from
the modern concept). Discontinuous functions, on the other hand, cannot be
expressed by such a single analytic expression, but they can be piecewise com-
posed by finitely many continuous functions and they can even be represented
by hand-drawn curves.77
Four years after Euler’s death, in 1787, the Petersburg Academy proposed
the question on the nature of arbitrary functions for a prize competition.
The paper “Sur la nature des fonctions arbitraire (On the nature of arbitrary
functions)” by Antoine Arbogast (1759–1803) was crowned in 1790. In this
prize paper Arbogast summarized Euler’s view of an extended concept of
functions in this way:
Euler had the daring idea not to subject these curves to any laws [i.e.,
to regard arbitrary curves], and it was he who said for the first time
that curves may be any line, that is, irregular and discontinuous, or
composed of different parts of curves [functiones mixtas] and drawn
by hand in a free movement [cum libero manus ductu], which goes
with no spatial restrictions.78
75
Dedekind 1887, §3.
76
Cf. Thiele 2000, 128–181, here especially 170–179.
77
Euler 1765.
78
Arbogast 1790, 4. “M. Euler [. . . ] eut l’idée hardie de n’assujettir ces courbes
à aucune loi, & il a dit le premier, qu’elles pouvoient être quelconques, irrégulières
& discontinues, c’est-à-dire, ou formées de l’assemblage de plusieurs portions de
courbes différentes, ou tracées par le mouvement libre de la main qui se meut sans
loi dans l’espace.”
Fig. 5.34. Nova Acta Petropolitanae, vol. 5 (1787), with the prize problem on the
nature of functions. Courtesy Deutsche Akademie der Naturforscher, Leopoldina,
Halle.
Fig. 5.35. Figures from L.F. Arbogast’s paper Sur le nature des fonctions arbi-
traire (On the Nature of Arbitrary Functions), crowned prize paper of the Peters-
burg Academy 1791, in which composed functions were regarded. Both examples
are discontinuous in Euler’s sense; in the modern understanding only the example
on the right shows a discontinuous curve (function). Courtesy Landes- und Univer-
sitätsbibliothek Halle.
The tendency to render mathematics in arithmetical terms has continued

since the days of Euler. I need only remind you of Lagrange’s book Théorie
des fonctions analytiques (Theory of analytic functions, 1797) or of the book
Vorlesungen über reelle Funktionen (Lectures on real functions, 1918) by Con-
stantin Carathéodory (1873–1950). Felix Klein (1849–1925) spoke of the arith-
metization; however, he was not pleased by these abstract arithmetizing ten-
dencies. It was the isoperimetric problems of the calculus of variations posed
in 1697 and their solutions in which the analytic expression arose that shifted
the setting of analysis from geometry to arithmetic. This tendency, however,
was not straightforward. In 1878 Karl Weierstrass (1815–1897) delivered the
lectures “Einleitung in die Theorie der analytischen Functionen” (Introduc-
tion to the theory of analytic functions). In the lecture notes which were taken
up by Adolf Hurwitz (1859-1919) we read:
John Bernoulli first gave another and seemingly very general definition
of function: if among two variable quantities there is a relation which
determines along with the values of one quantity a certain number
of definite values of the other one, then these quantities are called
functions of each other.79
And some lines later:
First of all, this definition is only valid for real numbers. But it is
completely untenable and completely infertile.80
Why did Weierstrass come to this strange view? For him power series were
the heart of analysis. Weierstrass’s ultimate aim was the representation of a
function.81 To this definition he objected above all that one cannot deduce
some general properties such as differentiability. Incidentally, this definition
was not Bernoulli’s. He neither spoke of “values of quantities” nor of “multi-
valued functions.” The latter functions we find in Euler before 1730, but we
do not find such a concept in the famous controversy on the logarithms of
negative numbers between John Bernoulli and Leibniz in 1712. Using a one-
infinite relation, Euler clarified the meaning of such functions in his paper
“De la controverse entre Mrs. Leibnitz et Bernoulli sur les logarithmes des
nombres négatifs et imaginaires” (Controversy between Mr. Leibniz and Mr.
79
Weierstraß 1878/1988, 48.
80
Weierstraß 1878/1988, 48. “Eine andere und scheinbar sehr allgemeine Defini-
tion einer Function gab zuerst J[ohann]. Bernoulli: Wenn zwei veränderliche Größen
so miteinander zusammenhängen, daßjedem Werth der einen eine gewisse Anzahl
bestimmter Werte der anderen entsprechen, so nennt man jede der Größen eine
Function der anderen. . . . Dieselbe gilt jedoch zunächst nur für reelle Zahlen. Sie ist
aber überhaupt vollkommen unhaltbar und unfruchtbar.”
81
“Das letzte Ziel bleibt doch immer die Darstellung einer Function.” Lecture
notes (Mitschrift) of his cours Ausgewählte Kapitel der Functionenlehre, winter
term 1885/86. Institute Mittag-Leffler, Djursholm, p. 262.
Bernoulli on the logarithms of negative and imaginary numbers)82 in 1749.

Moreover, a numerical interpretation (“value of quantity”) is probably first
stated by Augustin-Louis Cauchy (1789–1857) in 1821 (Cours d’Analyse). In
comparison with Weierstrass we see the advantage of Euler’s pragmatic atti-
tude.
Only a few mathematicians have invented more than two or three symbols
which are universally accepted in modern mathematics. Euler is among them
(because of his influential writings). Indeed he was a great notation builder
and establisher. Some examples in analysis (with the year of print):
1734 f (x)
1734 π
1736 e
1748 sin x, cos x, log x
2
1755 Σ, ∆,
√∆
1794 i = −1
Let me end with a few words on Leonhard Euler himself. One of his most
admirable qualities was a willingness to explain how he did mathematics,
how he made discoveries. His extraordinary memory enabled him to make de-
tailed calculations in his head (like the Austrian composer Wolfgang Amadeus
Mozart (1756–1791), a younger contemporary, who had every composition in
his head before he started writing). He never wanted to have the last word; on
the contrary, in his papers he let the readers have many things to complete in
order to encourage and to involve them. Dirk Struik (1894–2000) once said he
would not like to have a cup of coffee together with the quarrelsome and en-
vious mathematicians of the 18th century. No doubt, Euler is to be excluded.
On a memorial tablet in Swiss Riehen we find a concise characterisation of
nine words by Otto Spiess (1878–1966): “He was a great scholar and a kind
person.”83
Euler’s enormous productivity (from 1725 to 1783 with an average about
800 pages a year) is accompanied by quality and depth of the discoveries. His
prodigious output has been collected in the Leonhardi Euleri Opera omnia,
with more than 70 volumes up to now. We have about 900 items, including
20 books, and in the correspondence there are more than 3,000 letters and
300 addresses. An entire volume is necessary to present the list of Euler’s
publications.84 An evaluation of this volume supplies us with some statistics
about Euler’s work:85
82
Euler 1749.
83
“Er war ein großer Gelehrter und ein gütiger Mensch.” Spiess’s inscription on
a memorial plaque at Riehen, Kirchstrasse 8, erected in 1960. Spiess was a Swiss
historian of mathematics and a biographer of Euler.
84
Eneström 1910–1913.
85
Yushkevitch 1972, III, 37.
Years publications their percentage

in Euler’s lifetime
(approximate)
1725-34 35 5%
1735-44 50 7%
1745-54 150 20%
1755-64 110 14%
1765-74 145 19%
1775-83 270 35%
Opera omnia
(Series I-III)
algebra, number theory, analysis 40%
mechanics, physics 28%
geometry, including trigonometry 18%
astronomy 11%
architecture, ballistics, philosophy,
theory of music, theology, etc. 3%
Mathematics
(Series I)
integral calculus 20%
geometry, including differential geometry 17%
differential equations 13%
series 13%
number theory 13%
algebra, theory of probability 10%
foundation of analysis 7%
calculus of variations 7%
Statistics concerning Euler’s publications and fields of interest (due to

Eneström)
The output during his working life averaged about 800 pages a year or
15 12 pages a week. Adolf Pavlovich Youschkevich (1906–1993) noted that the
19 papers written in 1751 contain about 1,000 pages, as do the 56 papers of
1776 a quarter of a century later.86
Each volume of Euler’s Opera omnia provides important texts. Thirty en-
tries in the index of a Japanese Encyclopedic Dictionary of Mathematics and
even 53 articles of the German Mathematisches Wörterbuch (Mathematical
Dictionary) confirm Euler’s influence over the more than two centuries since
his death in 1783 until modern times.87 Indeed, the esteem in which Leon-
86
Jouschkevich 1971, 2, 741.
87
Encyclopedic Dictionary, K. Ito, ed. Tokyo: Iwanami Shoten 1954, revised edi-
tion 1960, English translation Cambridge: MIT Press 1987, 2 vols.; Mathematisches
Wörterbuch, 2 vols. J. Naas and H. Schmid, eds. Berlin: Akademie-Verlag 1961.
hard Euler was held from the beginning through our days has not diminished.
Johann Heinrich Lambert (1728–1777), Euler’s colleague during the Berlin
period, regarded Euler and d’Alembert as the first mathematicians among his
contemporaries.88 Euler was “the most prolific mathematician in history” and
the “major figure in the development of analysis in the eighteenth century”
(Victor Katz, born 1942).89 One and a half centuries before the princeps math-
ematicorum, Carl Friedrich Gauss (1777–1855) said of Euler’s mathematical
lifework:
The study of Euler’s work will remain the best school for the different
fields of mathematics and nothing else can replace it.90
In his Disquisitiones arithmeticae (1801) Gauss spoke of “summus Euler” us-

ing an epithet he attributed nowhere else to a scholar with the only exception
of Isaac Newton.91
It has been impossible to summarize all of Euler’s important contributions
to mathematics, and I did not try to do so. At best one can merely present
his work in a qualified sense. In this spirit, let me quote in conclusion from
Euler’s younger contemporary Jonathan Swift (1667–1745):
Elephants are always drawn smaller than life.
88
Lambert’s esteem is reported in the memories of D. Thiébault Mes souvenirs
de vingt ans de séjour à Berlin ou Frédéric le Grand (My memoirs of 20 years of a
sojourn in Berlin or Frederick the Great) (Paris: Buission 1804), vol. 5, p. 31f. “Le
premier géomètre vivant, me répondit-il, c’est M. Euler et M. d’Alembert, ou M.
d’Alembert et M. Euler: je les place au même rang.” (The first living mathematician,
he answered [to Thiébault’s question for the most famous living mathematicians],
are Mr. Euler and Mr. d’Alembert or Mr. d’Alembert and Mr. Euler. I put them
on the same rank.) Incidentally, Lambert added proudly: “Le troisième c’est moi.”
(The third, that is me.)
89
Katz 1993, 495.
90
Letter to P.H. Fuss from September, 11, 1849, Niedersächsische Staats- und
Universitätsbibliothek Göttingen, Handschriftenabteilung, Gauss-Nachlass. “Das
Studium der Werke Eulers bleibt die beste Schule in den verschiedenen Gebieten
der Mathematik und knn durch nichts ersetzt werden.” Cf. our motto by Laplace:
Read Euler, he is the master of us all.
91
§151 “vir summus”, §56 “vir sagacissimus”. Furthermore, in his Disquisitione
arithmeticae Gauss quoted not less than 28 papers of Euler (14 of Lagrange). In-
serted in Gauss’s own copy of Euler’s Methodus inveniendi (Niedersächsische Staats-
und Universitätsbibliothek Göttingen, Gauss collection) there is a traced portrait of
Euler made by Gauss (communication by Prof. Dr. K. Reich, Hamburg).
References
Sources and Bibliographies
Akademien im Briefwechsel, 1961. Die Berliner und die Petersburger Akademie

der Wissenschaften im Briefwechsel Leonhard Eulers (Juschkewitsch, A. and
E. Winter, Eds.). Akademie-Verlag, Berlin vol. 2.
Arbogast, A., 1790. Sur la nature des fonctions arbitraire. Academy, St. Pe-
tersburg.
Bernoulli, Johann, 1691/1922. Johannis Bernoulli Lectiones de calculo dif-
ferentialium (Schafheitlin, P., Ed.). Verhandlungen der Naturforschenden
Gesellschaft in Basel 34 (1922) 1–32.
Bernoulli, Johann, 1696. Problema novum. Acta eruditorum, Junii 1696, p.
269. Also in: Struik 1968, 391, and in: Bernoulli 1991, 212, 258–262.
Bernoulli, Johann 1706. Solutio Du Problême proposé par M. Jacques Ber-
noulli. Mémoires de l’Academie Royale des Sciences, Paris.
Bernoulli, Johann 1718. Remarques. Mémoires de l’Academie Royale des Sci-
ences, Paris, pp. 100–138.
Bernoulli, Jakob 1697. Solutio Problematum Fraternorum. Acta eruditorum,
Maji 1697, pp. 211–217. Also in: Bernoulli 1991, 271–282.
Bernoulli, Johann and Jacob, 1991. Die Streitschriften von Jacob und Johann
Bernoulli (Goldstine, H.H., Ed.). Birkhäuser, Basel.
Condorcet, A. de, 1783. Eloge de Euler. Histoire de l’Académie royale des
sciences pour l’année 1783 (1786). Also in: Opera omnia Euleri III, 12, Zürich
1960, and Oeuvres de Condorcet, vol. 3. Paris, Didot 1847, pp. 1–42, p. 41.
Euler, L., 1839. Œuvres complètes en français de L. Euler (Dubois et al., eds.),
5 vols. Etablissement Géographique, Brussels (E 786).
Euler, L., 1849. Opera minora collecta. Commentationes arithmeticae collectae
(P.H. and N. Fuss, Eds.), 2 vols. Academy, St. Petersburg 1849 (E 791).
Euler, L. 1911-, (EO I-IV). Leonhardi Euleri Opera omnia (Collected works),
divided in four series: I Mathematics (29 vols.), II Mechanics and Astronomy
(31 vols., 29 publ.), III Physics (12 vols., 11 publ.), IV A+B Correspondence
and Manuscripts (8 + 7 vols., 4 publ.), edited from the Euler Committee of
the Swiss Academy of Sciences. Basel: Birkhäuser, formerly Leipzig: B. G.
Teubner, and Zürich (Turici): Orell-Füssli, since 1911.
[All references to Euler’s collected works are given by EO followed by series
and volume number in Roman and Hindu-Arabic numerals respectively: “EO
ser., vol.”]
Euler, L., 1737. Solutio problematis ad geometriam situs pertinentis (E 53).
Commentarii Academia Scientiarum Petropolitanae 8 (1736), 128–140, actu-
ally published in 1741. Also in: EO I, 7, and a complete English translation
in Struik 1968, 183–187.
Euler, L., 1744. Methodus inveniendi (E 65). Bousquet, Lausanne. Also in:
EO I, 24. Zürich: Orell-Füssli 1952.
Euler, L., 1748. Introductio in analysin infinitorum, vol. 1. Bousquet, Lau-
sanne 1748. Also in: EO I, 8-9. B.G. Teubner, Leipzig 1922. English translation
by John Blanton, Euler 1988.
Euler, L. 1749. De la controverse entre Mrs. Leibnitz et Bernoulli sur les log-
arithmes des nombres négatifs et imaginaires. Mémoires de l’Académie royale
des sciences et de belles lettres (1749). Also in: EO I, 17. B. G. Teubner,
Leipzig 1914. German translation in: Zur Theorie komplexer Funktionen, in:
Ostwalds Klassiker, No. 261 (P. Juschkewitsch, Ed.). Leipzig: Geest & Portig
1983.
Euler, L., 1755. Institutiones calculi differentialis (E 212). Academy, St. Pe-
tersburg. Also in: EO I, 10. Leipzig, B.G. Teubner 1913. English translation
by John Blanton, Euler 2000.
Euler, L., 1765. De usu functionum (E 322). Nova Commentarii Academia
Scientiarum Petropolitanae 11, 67–102. Also in: EO II, 25.
Euler, L., 1768-1772. Lettres à une Princesse d’Allemagne. St. Petersburg.
Also in: EO III, 11-12; English translation Letter of Euler to a German
Princess, 2 vols., Thoemmes Press 1977.
Euler, L., 1983. Zur Theorie komplexer Funktionen (A.P. Juschkewitsch, Ed.).
Ostwalds Klassiker der exakten Wissenschaften, No. 261. Akademische Ver-
lagsgesellschaft Geest & Portig, Leipzig. [Comments to and German transla-
tions of E 101 (partly), E 168, E 390, E 490, E 656, E 675, E 694.]
Euler, J.A., 1767/1995. Meines Vaters Lebens-Lauf, dictated to his son Johann
Albrecht by L. Euler. Archive of the Russian Academy of Sciences in St.
Petersburg. Also in: Fellmann, Euler. Rowohlt, Hamburg 1995, pp. 11–13.
Euler, K., 1955. Das Geschlecht Euler-Schölpi. Geschichte einer alten Familie
[Genealogy]. W. Schmitz, Gießen.
Eneström, G., 1910–1913. Verzeichnis der Schriften Leonhard Eulers. Jahres-
berichte der Deutschen Mathematiker-Vereinigung, Ergänzungsband 4. B. G.
Teubner, Leipzig. [The bibliography is organised in three parts in view of the
date of publication, of the date of composition, and of the subjects. Each pa-
per has a so-called Eneström number (denoted by E no.). Every volume of
EO IV has a helpful table in which Euler’s paper can be located in Series I to
III by the Eneström number.]
Eneström, G., 1904–1905. Der Briefwechsel zwischen Leonhard Euler und Jo-
hann I Bernoulli. Bibliotheca mathematica (3) 4, 344–388, (3) 5, 248–291; (3)
6, 16-87; incomplete.
Eneström, G., 1906. Der Briefwechsel zwischen Leonhard Euler und Daniel
Bernoulli. Bibliotheca mathematica (3) 7 126–156; incomplete.
Euler, L. and C. Goldbach, 1965. Briefwechsel 1729–1764. (A.P. Juskevic and
E. Winter, Eds.). Akademie-Verlag, Berlin.
Frederick II, 1849, 1852, 1853. Oeuvres de Frédéric le Grand (Preuss, J.D.E.,
Ed.), t. 11; t. 20; t. 23. Decker, Berlin.
Fuss, P.H., 1843. Correspondance mathématique et physique de quelques célè-
bres géomètres du XVIIIème siècle, 2 vols. St.-Pétersburg (Reprint: New York
1968).
Knobloch, W., 1984 (Ed.). Leonhard Eulers Wirken an der Berliner Akademie
der Wissenschaften. 1741-1766. Spezialinventar. Regesten der Euler-Doku-
mente aus dem Zentralen Archiv der Akademie der Wissenschaften. Akademie-
Verlag, Berlin.
Koenig, S., 1751. De universalio principio. Acta eruditorum, March 1751.
Maupertuis, M. de, 1744. Accord de différentes Loix. Mémoires de l’Academie
Royale des Sciences, Paris (shortened Reprint Amsterdam 1751). Also in:
Opera omnia Euleri ser. II, 5. Orell Füssli, Zürich 1957 (includes also some of
Maupertuis’ writings).
Maupertuis, P. de, 1746. Les loix du mouvement et du repos déduites d’un
principe métaphysique. Histoire de l’Académie Royale des Sciences et Belles-
lettres de Berlin, pp. 267–294. Also in: Opera omnia Euleri ser. II, vol. 5.
Weierstraß, K., 1878/1988. Einleitung in die Theorie der analytischen Funk-
tionen. Lecture note by A. Hurwitz in 1878, edited by P. Ullrich. Vieweg,
Braunschweig 1988.
Winter, E., 1957 (Ed.), Die Registres der Berliner Akademie der Wissen-
schaften. 1746–1766. Dokumente für das Wirken Leonhard Eulers in Berlin.
Akademie-Verlag, Berlin.
Youshkevitch, A. and V.I. Smirnov, 1967. Eyler. Perepiska, annoturovany
ukazatel (Annoted register of Euler’s correspondence). In Russian. Nauka,
Leningrad.
English Translations
Euler, L., 1988-1990. Introduction to Analysis of the Infinite, 2 vols. (E 101-

102) Transl. and Ed. J. Blanton. Springer, New York.
Euler, L., 1754. On the general and fundamental principle of all mechanics
whereon all other principles relative to the motion of solids or fluids should be
established. Gentleman’s Magazine 24, 6–7 (E 177). Also in: Fauvel, J., and J.
Gray (Eds.), The History of Mathematics. Macmillan, London 1988, 460-462
(reprints).
Euler, L., 2000. Foundations of Differential Calculus (E 212) (J. Blanton, Ed.).
Springer, New York.
Euler, L., 1828. Elements of Algebra (E 387) (J. Hewlett, Ed.). Longman,
London; Reprint of the 5th edition with an introduction by C.A. Truesdell.
Springer, New York 1984.
Euler. L., 1833. Letters of Euler on Different Subjects in Natural Philosophy (E

343, 344, 417). Trans. David Brewster with additional notes of John Griscom.
Harper, New York 1833; Arno Press, New York 1975.
Euler, L., 1737. Solutio problematis ad geometriam situs pertinentis (E 53).
Commentarii Academia Scientiarum Petropolitanae 8 (1736), 128–140, Trans.
in Struik 1968, 183–187.
Proceedings of Conferences Held in Honour of Leonhard Euler
Festschrift 1907. Festschrift zur Feier des 200. Geburtstages von Leonhard
Euler. Herausgegeben von der Berliner Mathematischen Gesellschaft (P. Schaf-
heitlin, Ed.). B.G. Teubner; Leipzig 1907.
Sbornik 1958. Leonard Eyler. Sbornik statey v tshest 250-letiya so dnya rosh-
deniya, predstablennych Akademy Nauk SSR. In Russian. Moscow. [The 1957
conferences held in Berlin and St. Petersburg were coordinated; therefore, in
the proceedings we find Russian and German abstracts respectively.]
Sammelband 1959. Sammelband der zu Ehren des 250. Geburtstages Leonhard
Eulers der Deutschen Akademie der Wissenschaften zu Berlin vorgelegten Ab-
handlungen (K. Schröder, Ed.). Akademie-Verlag, Berlin.
Beiträge 1983. Leonhard Euler. Beiträge zu Leben und Werk. Gedenkband des
Kantons Basel-Stadt (J.J. Burckhardt et al., Eds.). Birkhäuser, Basel [Very
extensive bibliography by J.J. Burckhardt, pp. 511–552].
A tribute to Leonhard Euler. Special issue of Mathematics Magazine 56, 5
(1983).
Euler-Kolloquium 1983. Zum Werk Leonhard Eulers. Vorträge des Euler-
Kolloquiums 1983 in Berlin. (E. Knobloch et al., eds.). Birkhäuser, Basel.
Wissenschaftliche Konferenz 1985. Festakt und Wissenschaftliche Konferenz
aus Anlaß des 200. Todestages von Leonhard Euler. (W. Engel, ed.). Akademie-
Verlag, Berlin.
Idey Eylera 1988. Rasvitiye idey Leonardo Eylera i sovremennaya nauka.
(N.N. Bogolyubov et al., eds.) In Russian. Nauka, Moscow.
Biographies
Fuß, N., 1783. Éloge de M. Léonard Euler. Lu 23 Octobre 1783. Avec une
liste [in]complette de ouvrages de M. Euler (pp. 74–124). St. Péterbourg:
Academy. German translation by Fuß in: Opera omnia Euleri (= EO) ser. I,
1. B.G. Teubner, Leipzig 1911.
de Condorcet, Marquis, 1783. Éloge de M. Euler. Histoire de l’Academie
Royale des Sciences Paris 1783 (printed 1786), 37–68. Also in: EO III, 11.
Wolf, R., 1862. “Euler.” In: Biographien zur Kulturgeschichte der Schweiz. 4.
Cyclus. Orell-Füssli, Zürich, pp. 87–134.
du Pasquier, L.-G., 1927. Léonard Euler et ses amis. Hermann, Paris.

Spieß, O., 1929. Leonhard Euler. Huber, Frauenfeld-Leipzig.
Fueter, R., 1948. Leonhard Euler. Elemente der Mathematik, Beiheft 3.
Birkhäuser, Basel (Reprints3 1978).
Youschkevitch, A.P., 1971. “Leonhard Euler.” In: Dictionary of Scientific Bi-
ography, vol. 4, pp. 467–484. Scribner’s & Sons, New York. Reprinted in: Bio-
graphical Dictionary of Mathematics, vol. 2, pp. 736–753. New York: Scribner’s
& Sons, 1991.
Thiele, R., 1982. Leonhard Euler. B.G. Teubner, Leipzig. Bulgarian edition
Nauka i Iskustvo, Sofia 1985; Ukrainian edition Visha shkola, Kiev 1983; Ital-
ian edition Pitagora Editrice, Bologna 2000.
Fellmann, E.A., 1995. Leonhard Euler. Rowohlt, Hamburg.
Selected References
Arago, F., 1854. Oeuvres complètes. J.A. Barral, ed. Tom 2: Notices bi-
ographiques. Paris: Gide.
Bernoulli, R., 1983. Leonhard Eulers Augenkrankheiten. In: Leonhard Euler.
Gedenkband des Kantons Basel-Stadt. (J.J. Burckhardt et al., Eds.). Birk-
häuser, Basel, pp. 471–488.
Biermann, K.-R., 1985. Wurde Leonhard Euler durch J.H. Lambert aus Berlin
vertrieben? In: Festakt und Wissenschaftliche Konferenz 1983 in Berlin (En-
gel, W., Ed.). Akademie-Verlag, Berlin, pp. 91–99.
Breidert, W., 1983. Leonhard Euler und die Philosophie. In: Leonhard Euler.
Gedenkband des Kantons Basel-Stadt. (J.J. Burckhardt et al., Eds.). Birk-
häuser, Basel, pp. 447–458.
Calinger, R., 1975–1976. Euler’s letters to a princess of Germany as an ex-
pression of his mature scientific outlook. Archives of the History of the Exact
Sciences 15, 211–233.
Calinger, R., 1996. Leonhard Euler. The first St. Peterburg years (1727-1741).
Historia Mathematica 23, 121–166.
Cotes, R., 1714. Logometria. Philosophical Transactions 1714 (1717); reprinted
in: Harmonia mensurarum. London 1722.
d’Arcy, P., 1749, 1752. Réflexions sur le principe de la moindre action,
and Réplique à une mémoire de M. de Maupertuis, both in Mèmoires de
l’Academie Royale des Sciences, 1749 (published in 1753) and 1752.
Dedekind, R., 1887. Was sind und was sollen die Zahlen? Vieweg, Braun-
schweig. Reprints. English translation in: Essays on the theory of numbers
(Transl. W.W. Beman). Dover, New York 1963.
Descartes, R., 1637/1954. La Géométrie. La Haye, 1637. Cf. D. E. Smith’s
bilingual edition of Dover, New York 1954.
Dunham, W., 1999. Euler. The Master of Us All. Dolciani Mathematical Ex-
positions No, 22. MAA, Washington (DC).
Eckert, M., 2002. Euler and the fountains of Sanssouci. Archive for History of
Exact Science 56, 451–468.
Engelsman, S.B., 1990. What you should know about Euler’s Opera omnia.
Nieuw Archief voor Wiskunde 8,1, 67–79.
Fraser, C., 1991. Mathematical technique and physical conception in Euler’s
investigation of the elastica. Centaurus 34, 211–246.
Fraser, C., 1994. The origin of Euler’s variational calculus. Archives of the
History of the Exact Sciences 47, 103–141.
Fraser, C., 1997. The background to and early emergence of Euler’s analysis.
In: Analysis and Synthesis in Mathematics (M. Otte and. M. Panza, eds.).
Kluwer, Dordrecht.
Gauss, C.F., 1801/1973. Disquisitiones arithmeticae. Göttingen 1801. Also in:
Werke, vol. 1. Reprint Hildesheim, Olms 1973.
Hakfoort, C., 1995. Optics in the age of Euler. University Press, Cambridge.
Hankel, H., 1882. Untersuchungen über die unendlich oft oszillierenden und
unstetigen Fuktionen. Mathematische Annalen 20, 63–112.
Hartweg, F.G., 1979. Leonhard Eulers Tätigkeit in der französisch-reformierten
Kirche von Berlin. Die Hugenottenkirche 32, 4, 14–15; 32, 5, 17–18.
Hult, J., 1985. Eulers Briefe an eine deutsche Prinzessin. Populärwissenschaft
höchster Vollendung. In: Festakt und Wissenschaftliche Konferenz aus Anlaß
des 200. Todestages von Leonhard Euler (W. Engel, Ed.). Akademie-Verlag,
Berlin, pp. 83–90.
Juschkewitsch, A.P., 1959. Euler und Lagrange über die Grundlagen der
Analysis. In: Sammelband (K. Schröder, ed.). Akademie-Verlag, Berlin, pp.
224–244.
Juschkewitsch, A.P., 1983. Euler’s unpublished manuscript Calculus Differen-
tialis. In: Leonhard Euler. Gedenkband des Kantons Basel-Stadt (J. J. Bur-
ckhardt et al., Eds.). Birkhäuser, Basel, pp. 161–170.
Katharina II, 1986. Memoiren, 2 vols. Insel, Leipzig (German translations of
the French and Russian memoirs).
Katz, V., 1993. A History of Mathematics. Harper Collins, New York.
Kline, M., 1983. Euler and infinite series. Mathematics Magazine 56, 5, 307–
315.
von Krockow, C. Graf, 1996. Die preußischen Brüder. Dtv, München.
Laugwitz, D., 1983. Die Nichtstandard-Analysis: Eine Wiederaufnahme der
Ideen und Methoden von Leibniz und Euler. In: Leonhard Euler. Gedenkband
des Kantons Basel-Stadt (J. J. Burckhardt et al., Eds.). Birkhäuser, Basel,
pp. 185–198.
Lichtenberg, G.C., 1972. Vermischte Gedanken über die aërostatischen Maschi-

nen. Schriften und Briefe (Promis, W., Ed.), vol. 3, pp. 62–63. C. Hanser,
München.
Lützen, J., 1983. Euler’s vision of a general partial differential calculus for a
generalized kind of function. Mathematics Magazine 56, 5, 299–306.
Maltese, G., 2000. On the relativity of motion in Leonhard Euler’s science.
Archive for History of Exact Sciences 54, 319–348.
Outram, D., 1996. The Enlightenment. Cambridge 1996 (second edition).
Pulte, H., 1989. Das Prinzip der kleinsten Wirkung und die Kraftkonzeptio-
nen der rationalen Mechanik. Sonderheft 19 der Studia Leibnitiana. Steiner,
Stuttgart.
Soreth, M., 2000. Die Eulerkreise in Eulers Briefen an eine deutsche Prinzessin.
In: Mathesis (R. Thiele, Ed.). GNT-Verlag, Berlin, pp. 55–81.
Raith, M., 1983. Der Vater Paulus Euler. Beiträge zum Verständnis der geisti-
gen Herkunft Leonhard Eulers. In: Leonhard Euler. Gedenkband des Kantons
Basel-Stadt. (J. J. Burckhardt et al., Eds.). Birkhäuser, Basel, pp. 459–470.
Snow, C.P., 1959. The Two Cultures and the Scientific Revolution. University
Press, Cambridge (Eng.).
Stieda, W., 1931. Die Übersiedlung Leonhard Eulers von Berlin nach St.
Petersburg. Berichte über die Verhandlungen der Sächsischen Akademie der
Wissenschaften in Leipzig. Bd. 83, Heft 3. Hirzel, Leipzig.
Struik, D.J., 1968, A Source Book in Mathematics. University Press, Prince-
ton.
Szabó, A., 1987. Geschichte der mechanischen Prinzipien (3rd ed.). Wis-
senschaft und Kultur, Band 32. Birkhäuser, Basel.
Thièbault, D.: Mes souvenirs de vingt ans de séjour à Berlin. Paris: Buisson
1803-04.
Thiele, R., 1996. Euler und Maupertuis vor dem Horizont des teleologis-
chen Denkens. Über die Begründung des Prinzips der kleinsten Aktion. In:
Schweizer im Berlin des 18. Jahrhunderts (Fontius, M., and H. Holzhey, Eds.).
Akademie-Verlag, Berlin, pp. 373–390.
Thiele, R., 1997. Das Zerwürfnis Johann Bernoullis mit seinem Bruder Jakob.
Acta historica Leopoldina 27, 257–276.
Thiele, R., 1999. Ist die Natur sparsam? Betrachtungen zum Prinzip von Mau-
pertuis aus mathematikhistorischer Sicht. In: Pierre Louis Moreau de Mauper-
tuis. Eine Bilanz nach 300 Jahren. Berlin-Verlag/Nomos Verlagsgesellschaft,
Berlin/Baden-Baden, pp. 437–503.
Thiele, R., 2000. Frühe Variationsrechnung und Funktionsbegriff. In: Mathe-
sis. Festschrift zum siebzigsten Geburtstag von Matthias Schramm (R. Thiele,
Ed.). GNT-Verlag, Berlin, pp. 128–181.
Thiele, R., 2002. 300 Jahre Brachistochronenproblem. In: Medium Mathematik

(Löffladt, G., and M. Toepell, Eds.) Franzbecker, Hildesheim, pp. 76–99.
Thiele, R., 2005a. Leonhard Euler. In: Mathematics and the Divine (L.
Bergmans and T. Koetsier, Eds.). Elsevier, Amsterdam.
Thiele, R., 2005b. Von der Bernoullischen Brachistochrone zum Kalibrator-
Konzept. Brepols, Turnhout.
Thiele, R., in print. Leonhard Euler. In: [Ueberwegs] Grundriss der Geschichte
der Philosophie des 18. Jahrhunderts. (H. Holzhey et al., Eds.) Schwabe,
Basel, forthcoming.
Truesdell, C.A., 1968. Essays in the History of Mechanics. Springer; Berlin.
Truesdell, C.A. 1984. An Idiot’s Fugitive Essays on Science. Springer, New
York.
Weil, A., 1984. Number Theory. An Approach through History. Birkhäuser,
Basel.
Weyl, H. 1928/1949. Philosophie der Mathematik und Naturwissenschaft. Old-
enbourg, München 1928; Philosophy of Mathematics and Natural Science. En-
larged American edition. University Press, Princeton 1949.
Weyl, H., 1968. Obituary: David Hilbert, 1862-1943, no. 131. In: Gesammelte
Abhandlungen. J. Springer, Berlin.
Yushkevitch, A.P. 1972. Istoriya matematika, vol. 3. Nauka, Moscow.
Acknowledgments
I take the opportunity of expressing my gratitude to those whose support

has assisted me. In 1997 I spent some weeks at the Institute for the History
and Philosophy of Science and Technology, Victoria College, University of
Toronto as guest of Craig Fraser. It is a pleasure to thank him for the kind
invitation and to acknowledge his excellent hospitality. Especially I would like
to express my thanks to Craig Fraser for many discussions and for his help in
improving this text as also did the editors Glen Van Brummelen and Michael
Kinyon for this revised reprint. Furthermore, I would like to extend my thanks
to my colleague Gerhard Betsch, Tübingen, Germany, and for her technical
assistance to Mrs. Ulrike Rau, Leipzig.
I acknowledge the involved libraries and archives for their readiness to
support my work and to give permission to reprint illustrations. In detail I
thank
Universitätsbibliothek Basel, Handschriftenabteilung und Portraitsammlung,
Berlin-Brandenburgische Akademie, Archiv,
Landes- und Universitätsbibliothek Halle,
Deutsche Akademie der Naturforscher, Leopoldina, Halle, Bibliothek,
Universitätsbibliothek Leipzig,
Bibliothek des Sudhoff-Instituts der Universität Leipzig,

Michael Raith, Riehen.
The final word belongs to the organiser Tom Drucker of the 23rd Annual Meet-
ing of the SCHPM/CSHPM. It would be remiss of me if I did not acknowledge
to them the honour to be an invited speaker.
Fig. 5.36. Leonhard Euler. Bust in form of a hermes statue, about the end of 18th
century. Courtesy Sudhoff-Institut, Universität Leipzig.
6
Mathematics in Canada before 1945: A
Preliminary Survey∗
Thomas Archibald1 and Louis Charbonneau2

1
Department of Mathematics and Statistics, Acadia University
2
Département de Mathématiques, Université de Québec a Montréal
In 1932, Professor J. C. Fields of the University of Toronto summarized the

achievements of Canadian mathematics up to that point:
progress in mathematics in Canada up to the present has not been
all that might have been hoped for, things look more promising for
the future. There is a small but increasing group of the younger men
who are interested in mathematical research, and some of the later
appointments have been encouraging.3
For once, this rather bleak assessment is not the product of Canadian un-
derstatement. Unlike the situation in the U.S., Fields noted, the Canadian
mathematical community was so tiny that virtually all those conducting re-
search were Fellows of the Royal Society of Canada. Since section III of that
society was devoted to Mathematics, Chemistry and the Physical Sciences,
and since the total membership in 1932 of that section was about 100, we see
that we are looking at quite a small group indeed. This situation was to alter
in a few years, with the founding of the Canadian Mathematical Congress and
the Canadian Journal of Mathematics, both called for by Fields in this 1932
retrospective view.
Mathematics has an old history in Canada, however, and like the country
itself it represents two cultures based on and evolving from distinct national
traditions. French-Canadian mathematics is much older, stemming from the
earliest years of the colony of Nouvelle-France. The breaking of contact with
metropolitan France following the conquest curtailed French development af-
ter a healthy start, however. From that time onward, the parallels in math-
ematical development within the two cultural contexts are quite striking. In
order to have a mathematical culture, the first requirement is a population
or regime which recognizes that a portion of the population must have basic
∗
First published in Peter Fillmore, ed., Mathematics in Canada, vol. I, Ottawa,
ON: Canadian Mathematical Society, pp. 1–90.
3
Fields (1932), p. 112
142 Thomas Archibald, Louis Charbonneau
or advanced mathematical skills. These skills are then promoted via elemen-
tary or higher-level schools, by teachers in these schools, and by the materials
these teachers employ. Hence an essential part of the history of mathematics
in Canada is closely bound up with the development of the educational in-
frastructure. Exactly who is doing the teaching is also of interest, since the
single individual can have a very strong influence in a situation where few
individuals are involved and when the system is in a nascent state. The con-
tent of curriculum at all levels is of course important, as are the origins and
the goals of that curriculum. An additional aspect of the local setting is the
book trade: publishers and printers played an important role in the diffusion
of basic mathematical knowledge.
As with most aspects of Canadian culture, and indeed Canadian nation-
hood itself, mathematics in Canada manifested itself rather gradually, emerg-
ing from French, British and U.S. antecedents between the late eighteenth
and the early twentieth centuries. The phases of this emergence into a mathe-
matical maturity are quite similar to what we see elsewhere in the Americas,
chronology apart: an initial phase when all necessary mathematical skill was
imported, or the property of first-generation immigrants educated elsewhere;
a recognition of a requirement for some mathematical training in an elite of
bankers, accountants, merchants, surveyors, navigators, etc., and the corre-
sponding development of an infrastructure for providing such training using
materials and teachers from abroad; the broadening of this infrastructure to
include a wider portion of the population, the development of secondary and
post-secondary education, the local training of teachers, and the creation of
texts which suited local conditions and educational objectives; and finally the
gradual development of a fully articulated mathematical community, engaged
in teaching and research at all levels, publishing, and integrated into an in-
ternational mathematical community. These phases can and do overlap in the
Canadian context, as in most places.
In this paper we will attempt to survey briefly these developments. The
authors apologize for the fact that many individual stories of interest could not
be told in a short space. In particular, materials on the more recent history,
while abundant, were in most cases difficult to access, and the result is a
rather unbalanced account which will as usual favour central Canada. A little
less usually, we will spend more space on the earlier portion of the history.
Little has been written about these events. Apart from Charbonneau’s
article in the Canadian Encyclopedia, and Yves Gingras’ Les origines de la
recherche scientifique au Canada: le cas des physiciens, which has some points
of contact with our story, most of the information presented here comes from
archival research, biographical compilations, memorial volumes, and institu-
tional histories of Departments of Education and universities. We are fortu-
nate to have Karpinski’s Bibliography of Mathematics in the Americas through
1850, which, though incomplete, gives a helpful picture of the situation in
Canada in the first half of the century. In addition, R.S. Harris’s History of
Higher Education in Canada, 1663–1960 provides a good deal of important
6 Mathematics in Canada before 1945: A Preliminary Survey 143
background information. In general, French Québec has been better served by

historians than English Canada, and the following account reflects this.
Nouvelle-France (1635–1760)
Under the French regime the economy of the French colony in America de-
pended to a great extent on navigation of the Saint Lawrence river, and on
exploration of new lands. The population of the colony grew slowly in the sev-
enteenth century: around 11,000 in 1685, 18,500 in 1713, and 55,000 in 1754.
In 1754 Quebec City had only about 8000 people.4 In this setting, mathemat-
ical requirements were restricted to a few practical applications, notably in
surveying, cartography, and navigation. However, the increase in the popula-
tion, especially in Quebec City, led the Jesuits to begin offering a complete
classical course at the Collège de Québec beginning in 1659. The old sys-
tem had required five years, to which they now added two more, devoted to
“philosophy.”5 In keeping with the curriculum in French Jesuit Colleges of
the period (which not so long before had produced Descartes) mathematics
teaching was concentrated in these last two years.
The royal chair of mathematics and hydrography (1660-1760)6 Problems in-
volving property and the fixing of land boundaries are common in any seden-
tary society, and Nouvelle-France was no exception. Champlain (1567-1635),
on his return to Québec in 1632, declared himself “engineer in chief of the
colony,” and concerned himself with such issues. At Champlain’s death, Jean
Bourbon, an engineer in the Compagnie des Cents-Associés, took on the job
and served as judge of land boundary questions until his own death in 1668. In
1674, the Sovereign Council of Nouvelle-France required that all surveying in-
struments be approved by Martin Boutet, Sieur de Saint-Martin (1616-1683),
who was at the time the professor of mathematics at the Collège de Québec.
Jean Talon, Talon, Jean the intendant or chief administrator of the
colony from 1665 to 1681, grasped the fact that the economic future of the
colony would require a better knowledge of the country’s geography, as well as
depending on pilots that were competent to navigate on the Saint Lawrence.
Since 1661, Boutet had given mathematics courses at the Collège which were
oriented towards surveying and navigation. Soon after taking up his position,
Talon requested that these courses be extended to include the training of pi-
lots. The need was urgent: the census of the same year, 1666, shows only 22
“marins” (sailors) in the entire colony. Besides the chronic shortage of naviga-
tors, there was also a need for accurate maps. Once again the natural choice
to provide such training was Boutet. In 1671, Talon named him Professor of
4
Kerr (1966), p. 24.
5
Audet (1971), t. 1, p. 174.
6
The information in this section comes from Audet (1971), t. 1, pp. 192–202,
and from Chartrand, Duchesne, Gingras (1987), pp. 20–34.
Hydrography. Despite repeated efforts by the colonial administration, it was

only in 1678 that he was awarded the official title — and the corresponding
salary of royal engineer, with a mandate to “teach hydrography, piloting, and
other parts of mathematics.” This appointment occurred in a flood of creation
of such royal chairs of mathematics and hydrography in France in the 1670s.
Even though Boutet’s teaching was undertaken at the express request of
Talon, no remuneration accompanied it until 1678. This, it turns out, was
symptomatic of future events. Until the end of the seventeenth century, the
colony had difficulty ensuring continuity in the teaching of practical mathe-
matics applied to surveying and hydrography. Jean-Batiste Louis Franquelin
(1652-1718) received the title of “Hydrographe du roi à Québec” in 1687,
three years after the death of Martin Boutet. Prior to this appointment, he
had given hydrography courses privately and had prepared a map of New
France. After getting the appointment, he spent a number of years in France,
leaving the hydrography courses without a teacher. In 1697, Louis Jolliet
(1565-1700) took up the post. Jolliet, famous for his explorations, had trained
with the Jesuits in Québec, and thus became the first Canadian-born Royal
Hydrographer. Unfortunately he died only three years after taking up the
post. A renewed search led to the appointment of Jean Deshayes (d. 1706), a
French astronomer and cartographer, in 1703. Deshayes had visited Québec
in 1685, during which time he had observed an eclipse of the moon which
allowed a determination of the longitude of Québec by the French astronomer
Jean-Dominique Cassini.7 At his death, Deshayes left what was probably the
first scientific library in New France: it contained about fifteen volumes, in-
cluding the Marquis de l’Hôpital’s 1696 Analyse des infiniment petits pour
l’intelligence des lignes courbes.
These recurring difficulties in staffing the post of hydrographer led the
authorities to petition the crown to hand the chair over to the Jesuits in per-
petuity. The request was a reasonable one: since 1700, the Jesuits had given
a hydrography course at the Collège de Québec; they may have done so as
well at Montréal. The request was successful, and from 1708 until 1759 the
king granted a Chair in Hydrography to the Jesuits at Québec. The priests
who held the chair had an average tenure of five years in the period up to
1741. During the winter the students resided in Québec, where they took
courses in geometry, trigonometry, and physics as well as courses in naval
theory. In the summers, they apprenticed as pilots under the direction of the
second captain of the port. The best-known of the occupants of the chair is
Father Joseph-Pierre de Bonnécamps (1707-1790) , who taught hydrography
from 1741 until the fall of Québec in 1759. In addition to his teaching duties,
Bonnécamps took part in expeditions as cartographer, keeping up a corre-
spondence with astronomers and men of science in the mother country. Thus
he published a memoir on the aurora borealis in the Mémoires de Trévoux, a
scientific periodical published by the Jesuits. Another memoir, presented to
7
Chartrand, Duchesne, Gingras (1987), pp. 26–27.
the Paris Academy of Sciences by the astronomer Jean-Nicolas Delisle (1688-

1768), reported calculations of the longitude of Québec made by Bonnécamps
and de Lotbinière. The cumulative precision of observations resulted in an
improvement over the earlier calculations of Deshayes and Cassini.8
In the final years of the French regime, a distinguished scientist lived in
Québec as an officer under Montcalm, namely Louis-Antoine de Bougainville
(1729-1811). When he arrived in Canada in 1756, he had just completed the
second part of his Traité du calcul intégral pour servir de suite a l’analyse des
infiniments-petits de M. le marquis de l’Hôpital, the first part of which had
appeared in 1754. His relations with Father Bonnécamps seem to have been
good, since he wrote a letter of recommendation for the priest on the latter’s
return to France in 1757.9
The Collège de Québec10 Founded in 1635 by the Jesuits, the Collège de
Québec was the main seat of intellectual training in the colony for the en-
tire French regime, its importance increasing with the growth of the colony.
In 1651, the entire responsibility for teaching a course of study with a nor-
mal duration of five years rested on two of the ten priests, together with six
brothers residing at the college who assisted. One of the two priests taught
mathematics, doubtless mostly elementary and commercial mathematics. At
that time there were only a dozen students, including the “petite école” where
the preparatory teaching was done. As mentioned earlier, the classes of phi-
losophy (two more years) were added in 1659, and mathematics was placed
in one of these years, as was the custom in France. Both Talon and the first
bishop of Québec, Mgr François de Montmorency Laval, concerned themselves
with the curriculum.11 During the second half of the eighteenth century, in
French colleges, the mathematics program was the following:12
8
The longitude of Québec is 73◦ 33 , while Deshayes and Cassim obtained 72◦ 13
and Bonnécamps and de Lotbinière, 72◦ 30 . Chartrand, Duchesne, Gingras (1987),
p. 34.
9
For more about Bougainvile and his time in America, see Struik, Dirk, J. (1956).
Among the French engaged in mathematical activity in Nouvelle-France, we may also
mention Joseph Bernard Chabert who, for geodesic purposes, made astronomical
observations along the Atlantic coast near Louisbourg in 1750-1751. See Struik,
Dirk J. (1976), p. 102.
10
Unless otherwise mentioned, the information in this section is from Audet.
11
In most of the French colleges, the philosophy course was spread out over three
years. Audet (1971), t. 1, pp. 172–189. Note that the 1775 program of the philosophy
class of the Séminaire de Québec corresponds to the French program except that
“mathématiques mixtes” was reduced to practical geometry with some commercial
mathematics.
12
Dainville, Francois de, (1964), p. 52.
Pure mathematics:
arithmetic, algebra, geometry, plane trigonometry
“Mixed” mathematics:
practical geometry: measurement of length, area and volume
mechanics: the science of forces and the actions of bodies
hydrostatics
spherical astronomy
use of the gnomon (i.e. solution of triangles in practical circumstances)
optics: perspective, mirrors and lenses
fortification
pyrotechnics (sometimes)
This is an impressive-sounding program. However, it should be remembered
that mathematics at this time was taught in connection with only one of
the four parts of philosophy (logic, metaphysics, ethics and physics), namely
physics, and that physics was done first.13 The instruction had virtually noth-
ing in common with how mathematics is taught now: there were no exercises,
and practical applications were limited to those things which were dictated in
the notes. In France, students entering the philosophy class had often received
no mathematical training at all. The Abbé Sauri, author of an introductory
textbook called Institutions mathématiques, noted as much in his introduc-
tion:14
I would advise messieurs the philosophy teachers to teach my institu-

tions at the beginning of the course, or at least, to teach arithmetic,
the first four rules of algebra, and the notions of geometry contained
in No. 1, p. 141 to No. 7, p. 147 inclusive. This will place their stu-
dents in a better position to understand logic and metaphysics. In
logic itself one often talks of triangles, circles, etc. How can we expect
that young people who have no acquaintance with these figures will
understand any of the professor’s explanations?
There is no reason to think the situation would be otherwise in Québec.
Still, we should distinguish between the Collège courses and those given
in connection with the Chair of mathematics and hydrography. Although the
history of the Chair is closely associated with the college, it seems that the
professors of hydrography did not simply teach the same material as was
given in the college philosophy courses.15 It should also be noted that, on
this side of the Atlantic anyway, the classical course given by the college
led to only one end: the priesthood. The liberal professions were, so to speak,
13
14
Sauri (1786), p. xvii. Even if the book appeared after 1760, teaching in the
French colleges did not alter significantly between 1750 and 1780.
15
So it seems from the short list of mathematics professors at the college given
by Audet. Audet (1971), t. 1, p. 185.
closed to Canadians. There were no lawyers in the colony, and notaries needed
only a minimal training. Physicians came almost uniquely from France. This
could not help but push the teaching in a direction that was not particularly
favourable to independent teaching in the sciences.
By the time of the defeat of Montcalm, Bougainville, Bonnécamps, and
their associates had left the colony, and the college closed.
English Canada from 1760 to the Union: Creating an

Infrastructure
Reflecting their later start as colonists, English mathematical production and
education start later than in the French colony. The only 18th century Eng-
lish work mentioned by Karpinski is a ready reckoner printed at Quebec in
1790 by Major Williams. It ran to a second edition, is described in the Que-
bec Herald of 1789-91, and appears lost. It deals with rates of exchange, an
important subject since local currencies were in pounds, shillings and pence,
and exchange rates varied.
A second early paramathematical book deserves mention as showing the
state of affairs in the early nineteenth century. This is the 1822 work by Arthur
Fessenden, titled Tables, showing the interest at six per cent of any sum from
1 pound to 1000 pounds, from one day to one hundred days, and from one
month to twelve months. Fessenden was an accountant for the Bank of Canada,
then at Montreal. His tables were published in Montreal by Nahum Mower,
a newspaper printer/publisher, and later editions appeared in Montreal in
1830 and Halifax in 1832. Subsequent editions of 1837, 1841, and 1847 were
extended to 365 days. Of course several things appear odd to us about this.
Only one rate is tabulated, though the existence of reprints suggest that this
was “the going rate” for the entire 25-year period. The tabulation of many
principal amounts shows that while addition was a staple of the counting-
house, multiplication was not. The work bears a recommendation from the
President and Cashier of the Bank of Canada, and from the President of the
Montreal Bank.
In the early years of the nineteenth century the main emphasis was on
elementary education (whether for children or not), and the first years of
mathematical study were devoted almost exclusively to arithmetic. The ear-
liest English Canadian arithmetic book we have identified is the 1809 treatise
of John Strachan, later Bishop of York, which he wrote in order to have an ap-
propriate book for his classes at the Cornwall Grammar School. The work was
titled Concise introduction to practical arithmetic: for the use of schools, and
published in Montreal. Strachan was educated at King’s College, Aberdeen
and at St. Andrews; he had a strong interest in science, and the Scotland
of his day offered in general a better scientific education than was available
in England. Strachan was instrumental in the founding of both McGill and
King’s College (the earliest component of the University of Toronto), and was
an ardent advocate of proper grammar schools (i.e. secondary schools). He was

also the first Inspector of Schools in Upper Canada, and an ardent proponent
of the standardized education that would require the creation of such a post.
Strachan was also important in setting the stage for a domestic production of
textbooks through his ardent opposition to U.S. “democratic” influences, an
opposition widely shared in loyalist British North America.
Both before and after Strachan, British and American texts were often
used by Canadian teachers, a pattern which continued until the 1850s. An ex-
ample of a widely approved English book is Francis Walkingame’s The tutor’s
assistant, being a compendium of arithmetic and complete question-book. A
Canadian edition of the work appeared in Montreal in 1818, though the Eng-
lish original appeared first in 1751, and passed through an enormous number
of editions over the next century. In fact, the Dictionary of National Biography
states the following:
A so-called 71st edition appeared in 1831 ... Except for the section
dealing with the rule of three, which needed improvement, the work
remained little altered down to 1854.16
Little wonder then that Karpinski lists nine Canadian editions prior to
1850, printed at places such as Toronto, Picton, and St. John. A quick look at
the book, however, suggests why it would be of limited use for Strachan and
his contemporaries. It is essentially a book directed at adults, covering a great
deal of ground (e.g. cube root, single-entry bookkeeping, and basic algebra)
in a short span. As the title suggests, it appears to have been intended as an
aid for teachers, structuring the curriculum and providing worked examples.
Answer keys were published soon after its original appearance, and may well
have existed for the Canadian editions as well.
An American competitor for such British arithmetics was provided by
Daniel Adams (1773-1864), a New England physician and educator, who pub-
lished The Scholar’s Arithmetic in 1801 expressly to provide a suitable school
text.17 The work ran to many editions, and was revised in 1827 as Adams’ New
Arithmetic, under which title it appeared in a Canadian version in 1833. The
title, given the American origin of the book, is a bit misleading: Adams’ new
arithmetic, suited to Halifax currency, in which the principles of operating by
numbers are analytically explained and synthetically applied; thus combining
the advantages of the inductive and synthetic mode of instructing. The whole
made familiar by a great variety of useful and interesting examples, calculated
at once to engage the pupil in the study, and to give him a full knowledge of
figures in their application to all the practical purposes of life. Designed for
16
DNB, v. 20, p. 548. It is described as the “most popular arithmetic both in
England and America down to the time of Colenso” that is, to the late nineteenth
century.
17
Adams also published works on grammar, oratory, and geography. See the Dic-
tionary of American Biography, 1, pp. 54–55.
use in the schools and academies in the British provinces. The first Canadian
edition was printed at Stanstead, in the Eastern Townships, with or without
the approval of the author. A second edition was published at Sherbrooke in
1849. We may note the methods of analysis and synthesis being touted here as
pedagogical strategies. It seems likely that this reflects some awareness of con-
temporary British debates on analysis versus synthesis as appropriate means
of procedure in learning higher mathematics, but we have no other evidence
of this at present.
In 1832, hence shortly before the publication of Adams’ book, appeared a
second work written specifically for Canadian schools, with William Phillips
as author. Phillips, apparently based in York, is described as a teacher “in
Ladies’ Schools”, as well as a private tutor. In abbreviated form, the title is:
A new and concise system of arithmetic, calculated to facilitate the improve-
ment of youth in Upper Canada. The work was published by subscription
under the patronage of Sir John Colborne, then Lt. Governor of the Province
of Upper Canada. Other listed subscribers included Strachan, by then the
Archdeacon of York, and Dr. Harris, Principal of the newly-founded Upper
Canada College.
This list of patrons sheds light on some historical issues that require a bit
of background to be completely understood, and which bring us back to John
Strachan. Strachan was a very political creature, one who saw his teaching
as a means of extending his personal influence; in 1817 he had declared with
satisfaction that “all my pupils [are] now the leading characters in many parts
of the province.”18 Strachan had long argued for the establishment of a Uni-
versity in Canada on the “Scottish or German” model. He viewed this as a
Christian institution, though Methodists and other dissenting sects should be
excluded from governance of the institution, which should be firmly in the
hands of the established church (i.e. the Church of England). This stance fit
in well with his generally Tory outlook, one which had enabled him to suc-
cessfully cultivate Lord Maitland, Colborne’s predecessor. With Colborne’s
accession to the Lieutenant-Governor’s post, Strachan’s influence waned, as
did his efforts to establish the newly chartered King’s College. Colborne felt
that Grammar schools of high quality were essential, so that the children of
the appropriate classes in the colonies could return to England for university
education when desirable. Upper Canada College was founded on Colborne’s
initiative, and somewhat against Strachan’s wishes, as such a preparatory
school.19
18
Craig (1986) p. 755.
19
Similar doctrinal battles formed the background for the establishment of in-
stitutions of higher education in Nova Scotia where King’s College (Windsor) was
founded in 1789 and was restricted to members of the Church of England, then
about 20% of the population. This eventually led to the founding of Pictou Acad-
emy by Thomas McCulloch. The strength of mathematics and science there was one
of McCulloch’s key arguments in attempting to obtain provincial funding. Eventu-
ally McCulloch became the first president of Dalhousie (1838).
The Methodists, disinclined to be excluded by Strachan from control of

higher education, had meanwhile been moved to action, establishing Upper
Canada Academy at Cobourg, Ontario. Though this school originally ex-
perienced financial difficulty, Egerton Ryerson, a Canadian-born Methodist,
worked to obtain a royal charter for this school, which eventually became Vic-
toria College in Toronto. Ryerson was the College’s first principal, inducted
in 1841. Ryerson became Superintendent of Schools for Canada West in 1844,
an appointment which had eventual repercussions for the world of elementary
mathematics in Canada. He was also the major architect of the system of
Normal Schools (or education colleges) in Upper Canada.
Despite his “democratic” leanings, Ryerson was a staunch believer in the
importance of a firm adherence to the British Empire and of a close association
between the colonies. This contrasts somewhat with the position of Strachan,
who in his later years became an advocate of Upper Canadian colonial auton-
omy. The imperial focus is reflected in a contemporary arithmetic by G. and J.
Gouinlock. A complete system of practical arithmetic, for the use of schools in
British America, to which are added, a set of book-keeping by single entry, and
a practical illustration of mental arithmetic, federal money, receipts, bills of
exchange, inland and foreign, explanation of commercial terms, etc. The whole
adapted to the business of real life, to the circumstances of the Country, and to
the present improved state of commerce was printed in Hamilton in 1842. The
Gouinlocks describe themselves on the title page as “formerly British teachers
of long experience and extensive practice”. In addition to standard operations
with whole numbers, fractions and decimals, it includes proportion, simple
and compound interest, and the following advanced topics apart from those
mentioned in the title: British exchange of moneys, with a number of coun-
tries (incl. Europe, W. Indies, E. Indies and Canton in China); Alligation,
Involution, Evolution, Square root and Cube root, duodecimal multiplication,
tonnage of ships, and permutation.20
French Canada as an English Colony, from 1760 to the

Union
The first arithmetic books. The period from the conquest to the Union of
Upper and Lower Canada, in 1840, is characterized by a desire on the part
of the authorities to establish a system of primary public education. This
idea of training a large number of people to a minimal level entailed certain
changes in both curriculum and pedagogy. As in English Canada, arithmetic
was emphasized.
The publication of the first arithmetic books in Quebec should be seen
against this shifting backdrop. The first such work, written by Jean-Antoine
Bouthillier (1782-1835), appeared in Quebec in 1809 and was titled Traité
20
Karpinski, p. 438
d’arithmétique pour l’usage des écoles. Bouthillier had studied at the Collège
Saint-Raphaël in Montreal before apprenticing as a surveyor.21 He never
taught, though he worked at a variety of professions, among them journal-
ist, translator, inspector of highways, and justice of the peace.22 His book
dealt with the elementary operations on whole numbers and fractions, as well
as rules useful for merchants (rules of three, rule of false position, simple and
compound interest, and exchange). Despite being intended for schools, its
success was limited at first. This was perhaps in part because it was oriented
toward the memorization of rules, rather than comprehension. One might also
ask whether, at the time of its appearance, the number of schools with instruc-
tion going beyond the four elementary operations justified the scope of the
work. However, after 1830 the use of this book became more widespread, and
it appeared in numerous editions until 1864.
Another student from the same school in Montreal was Michel Bibaud, who
published a manual of arithmetic in 1816. This work was titled: L’arithmétique
en quatre parties, savoir: l’arithmétique vulgaire, l’arithmétique marchande,
l’arithmétique scientifique, l’arithmétique curieuse, suivie d’un précis sur la
tenue des livres de comptes, principalement pour ceux qui veulent apprendre
l’Arithmétique d’eux-mêmes et sans Maı̂tre, ou s’y perfectionner. As the title
indicates, this was intended not for the schools but for an audience of auto-
didacts. Nevertheless it was used in many schools. The “four parts” of the
title are: common arithmetic (the four elementary operations on numbers and
fractions, calculation of areas and volumes); commercial arithmetic (rule of
three, exchange); scientific arithmetic (decimals, powers, roots, proportions
and logarithms) and recreational arithmetic (riddles, games, puzzles, etc.).
Apart from the recreational portion, Bibaud’s content corresponds to that in
Bouthillier’s book. Bibaud’s work is in fact a compilation, and he indicates
his sources. The common and scientific arithmetic were inspired by the 1786
text of Abbé Sauri, which Bibaud had doubtless used as a student at Collège
St. Raphaël.23 The commercial arithmetic comes from Walkingame, a popu-
lar English text of the time. Finally, the recreational material was based on a
book by a M. Despiau called Choix d’amusements physiques mathématiques.24
Bibaud was to rework the contents of his book, republishing it in 1832 un-
der the title L’Arithmétique a l’usage des écoles élémentaires du Bas-Canada
without the recreational portion.25
21
This became the Collège de Montréal in 1806.
22
See Lavoie, Paul (1994), chap. 5, section 5.5.
23
Sauri (1786). One finds in the Archives du Collège de Montréal a Compendium
des institutions mathématiques de l’Abbé Sauri copied at Québec the 7 of August
1785. A copy of the fourth edition (1786) of the Sauri book belonged to the Collège
de Montréal. It is now located at the Bibliothèque Nationale in Montréal.
24
This book seems to have been published in London in 1800, with a translation
in 1801 also at London. See Lavoie, Paul (1994), p. 276, note 1.
25
Lavoie, Paul (1994), pp. 274–282.
A third book deserves mention, written by Casimir Ladreyt (1797-1877)

and published in 1836. As with many books of the period, its title tells us
a great deal: Nouvelle arithmétique raisonnée ou cours complet de calcul
théorique et pratique, a l’usage des collèges et des maisons d’éducation de
l’un et de l’autre sexe, des personnes qui veulent apprendre cette science en
peu de temps et sans le secours d’un maı̂tre, et de celles qui veulent se livrer
au commerce; suivi de quelques leçons sur la plannimétrie et la stéréométrie
(arpentage et cubage), ou Toisé des surfaces et des volumes. We know little
about the author beyond what is on the title page: a “former French trader,
now a teacher”. Though Ladreyt covers much the same ground as Bibaud and
Bouthillier, he distinguishes himself by his pedagogical care. His page layout is
more adapted to the logical hierarchy of the ideas, and reasoning is important
because of its importance in training the faculty of judgement. For this reason
he does not cite the rule of three as such, preferring that the reader reason
out the problem rather than blindly applying a rule. Despite these qualities,
or perhaps because of them, Ladreyt’s book had a rather limited success.26
Finally we note a book which had a limited distribution in the region
near Québec, Joseph Laurin’s Traité d’arithmétique: contenant une claire et
familière explication de ses principes: et suivi d’un traité d’Algèbre, which
corresponded closely in its content to the books of Bouthillier and Bibaud.
The popularity of the books of Bouthillier and Bibaud around 1840 and
even over the next two decades indicates that much mathematics teaching was
individually based. This state of affairs was to evolve following the arrival in
1837 of the Frères des Écoles chrétiennes (FÉC), to which we return below.
Secondary teaching: the classical Collèges. In 1757, the state of war between
France and England had led to the closing of the Collège de Québec, which
subsequently remained closed.27 England did not permit the recruitment of
Jesuits any longer, and besides many returned to France. Those who stayed
tried to take up teaching again, but with little success. In 1765, on the order of
Mgr. Jean-Olivier Briand, the Séminaire de Québec took up the task.28 The
Séminaire organized its teaching directly in the tradition of the Collège de
Québec; from 1765 to 1770, since there were no priests, the course was limited
to the classes of Letters and Humanities concentrating on Latin, French, Eng-
lish and Greek.29 In 1770, the philosophy classes began again, though there
was no mathematics at all for another three years. In Montréal, the Collège
26
Lavoie, Paul (1994), pp. 282–292.
27
Galarneau, Claude (1978), p. 16.
28
Founded in 1668, the Séminaire de Québec had never provided secondary ed-
ucation in a continuous way prior to the arrival of the English. The seminarians
attended courses given by the Jesuits at the Collège de Québec. At times when they
were dissatisfied with the Jesuit courses (as in 1732) they organized their own phi-
losophy courses, but this was never done on a regular basis. Audet, Louis-Philippe
(1971), t. 1, p. 373.
29
The Letters classes, often called Humanities, consist of the first six years of the
course.
Saint-Raphaël, already mentioned in connection with its illustrious students

Bouthillier and Bibaud, began to offer courses in the heart of the city in 1773.
This began by offering only humanities (the introductory years of the course),
with those interested in completing the philosophy course being obliged to
go to Québec. Not until 1790, following pressure from parishioners, did the
Bishop of Québec appoint a philosophy professor. In fact, the parishioners also
argued for a professor who could teach arithmetic and mathematics as well as
a course in writing, complaining that such instruction was available only in
the Protestant schools.30 Such courses began in 1791.31 We may assume that
they revolved around commercial arithmetic, in accordance with the changing
economic nature of Montréal at the time.
The content of the mathematics courses in the philosophy classes of the
Séminaire de Québec are known to us in part because of two presentations of
“theses” in mathematics, one in 1775 by the “physics students of Mr. Thomas
Bédard Diacre”, the other between 1786 and 1790 by the “students of Mr. Ed-
mund Burke, priest”.32 These are thesis defences in the medieval sense: public
academic exercises during which physics students of the philosophy class dis-
cussed a few mathematical propositions. On these occasions, the Séminaire
printed a pamphlet which summarized the propositions debated. The pam-
phlet of 1775 has nine pages, and we find there material from elementary
arithmetic, algebra, and the calculation of proportions. This is followed by
the solution of equations in one to four unknowns, theorems on arithmetic
and geometric progressions, and quadratic equations. There are also propo-
sitions of elementary geometry, practical geometry, and trigonometry. The
1790 pamphlet consists of ten pages of statements of theorems, and goes well
beyond the earlier work in including conic sections, spherical trigonometry,
eleven propositions on differential and integral calculus, and a large number
on various aspects of mechanics. The content of the 1790 pamphlet corre-
sponds much more to the program in French colleges of the time than did
that of 1775. This evolution may well have begun with the arrival in Québec
in 1775 of M. J.-B. Lahaille, a French Jesuit from Bordeaux, who succeeded
Thomas Bédard for one year at that time. It appears to have continued in
the hands of Charles Chauveaux, who taught physics and mathematics from
1776 to 1786, and was one of the students who participated in the “defence”
of 1775.33 His course notes contain many of the subjects treated in the 1790
work, though several of the mechanical topics, as well as conic sections, are
missing.
Subsequently the purely mathematical content of the philosophy classes
ceased to evolve. Nevertheless, there was a change in the attitude to science.
30
Charbonneau, Louis, 1984, p. 43.
31
Galarneau, Claude, 1978, p. 18.
32
Bédard, Thomas (1775), Burke (n.d.).
33
These notebooks are held at the Archives of the Collège de Montréal, of the
Séminaire de Saint-Hyacinthe, and of the Séminaire de Québec. Galarneau, Claude
(1977), pp. 86–87.
This change may be seen in the lesson plans of the philosophy class and even
in the names given to each of the two years.34 In 1790, the first year was
called “Logic”, and logic, metaphysics and morals were covered. The second
year was called “Physics”, and covered physics and mathematics. This divi-
sion corresponds faithfully to that in the French Collèges and doubtless to
that in the Collège de Québec. In 1816, the earlier counsel of the Abbé Sauri
was at last heeded and we see in the course of study the addition of “a part
of mathematics” to the program in the Logic year. Finally, from 1838, the
name of the first year is no longer Logic, but Mathematics. The program of
study consisted of: algebra, geometry, differential and integral calculus, and
conic sections. The second year was still called physics, and covered physics
and chemistry. Despite appearances, logic, metaphysics and ethics were still
taught, but clearly mathematics and science were considered of greater impor-
tance than at the beginning of the century. It is worth noting that in England
at this time differential and integral calculus were part of the curriculum at
Cambridge, but for the most part not elsewhere, and even Cambridge had
provided them only for about twenty years. Hence it is not surprising that we
do not find courses of corresponding sophistication in English Canada at this
time.
Prior to 1835, it was also the case that the philosophy courses alternated:
that is, there was one professor, who taught the two courses of Logic and
Physics in alternate years. Thus the students did not necessarily follow the
course in order. From 1835, because of the growth of the student population,
three teachers divided the task by discipline. Abbé Jérôme Demers taught
“intellectual” philosophy (logic, metaphysics and morals), while mathematics
was taught by Abbé Normandin, and physics by Abbé J. L. Casault. At about
the same time, we see mathematics enter the lower school, under the impul-
sion of Demers and his colleague Abbé John Holmes, a transplanted American
convert. Thus 8e and 7e (the preparatory years) contained arithmetic; frac-
tions and decimals in 6e and 5e ; bookkeeping and the metric system in 4e ; and
then algebra, and elementary geometry in 3e , 2e , and 1re . The other colleges
and seminaries of Lower Canada followed suit, though there were delays in
implementing the program depending on the college.35
What provoked this exodus of mathematics from the philosophy classes?
Certainly we may mention the pressing needs for men outfitted with com-
mercial mathematics. We already mentioned pressures of this kind originating
from the parishioners of Notre-Dame de Montréal at the end of the eighteenth
century. Such demands were to be repeated throughout the nineteenth cen-
tury, the more so because the majority of students left the classical Collèges
without even beginning the two years of philosophy. There is more to it than
this, however: the attitude of the Church had been ambivalent until about
1840. Under the influence of Abbé Demers, however, the Séminaire de Québec
34
Charbonneau, Louis, 1984, p. 43.
35
Lamonde, Yvan (1980) p. 76.
and other Collèges adopted a position that is related to that enunciated by

the Jesuits in their Ratio Studiorum of 1832:
The times require us to give more importance than formerly to physi-

cal sciences and mathematics . . . . For the fact that these sciences have
been abused to oppose our holy religion is not a reason to abandon
them, but on the contrary a reason that our people should devote
themselves to them with all the more ardour, in order to seize the
weapons of the enemy and to employ in defence of the truth the means
which they abused to attack it.36
Given these developments it seems likely that after the mid-1830s stu-
dents entering the philosophy classes were better prepared to undertake study
in intermediate-level mathematics. One may wonder, however, if in fact the
teaching was any different from that which had gone before. Was the training
of mathematics teachers who taught in the philosophy classes of a kind to
improve this teaching? Many of these teachers had traditionally been quite
young. We have already mentioned Thomas Bédard and his student Charles
Chauveaux, the latter having begun teaching the year after completing his
philosophy course. It was traditional in the Collèges for brilliant young semi-
narians to give courses immediately after completing their own studies. Such
a tradition was hardly propitious for an improvement of teaching, since the
young professor would simply repeat courses which he had heard himself a few
months before.37 Nonetheless, there were notable exceptions. Burke, for ex-
ample, who succeeded Chauveaux, had received a solid education in Paris.38
Unfortunately, his career at the Séminaire ended in 1790. Abbé Houdet, a
Sulpician father who immigrated to Canada to escape the anticlerical laws
of the French revolution, was responsible for science and philosophy at the
Collège de Montréal between 1798 and 1826.39 The most remarkable of the
philosophy professors of the first half of the nineteenth century is undoubtedly
Jérôme Demers, who taught both philosophy classes at Québec from 1800 to
1835, then restricted himself to intellectual philosophy from 1835 to 1849. His
influence was felt in all the Collèges of the province. It is interesting to note
that Abbé Demers had worked for some time as a surveyor between the end
of his classical studies and entering the seminary in 1795. His course notes
on mathematics show no special originality, but those in physics indicate that
he was up to date on the latest discoveries, especially in electricity and mag-
netism.40 Demers also encouraged higher study elsewhere, and was involved in
36
Cited in Simard, G., Tradition et Évolution dans l’enseignement classique, Ot-
tawa, 1923, p. 10.
37
Martineau, Armand (1967), pp. 215–216.
38
Galarneau, Claude (1977), p. 87.
39
Galarneau, Claude (1977), p. 90–93.
40
Many student notebooks are held at the Archives of the Collège de Montréal,
the majority dating from 1811.
sending the Abbés Isaac and François Desaulniers to Georgetown University,

a Jesuit institution in Washington, where they obtained M.A.’s in science. On
their return, they too became philosophy professors, Isaac at St.-Hyacinthe
and Francois at Nicolet.41 Despite these examples, young, inexperienced teach-
ers without additional training (such as Michel Racine at Québec) continued
to teach mathematics even after 1835, and the overall quality of these courses
showed little improvement.42
English Canada from the Act of Union to Confederation
Following the Act of Union, the educational system in the newly-established

Canadas was substantially reorganized. Indeed, following a general movement
in Britain, between 1842 and about 1853 all the British North American
colonies restructured education to establish common schools available to all
boys, to encourage teacher training, and to develop uniform textbooks. In
the Canadas, the standardization of texts took the form of a list of approved
books which schools could select, with subsidies being withheld if approved
books were not chosen. This led to a substantial market for publishers to tap;
there were 2500 elementary schools in Canada West in 1844.
The most successful of these publishers was John Lovell (1810–1893) of
Montreal. Originally a printer and newspaper publisher, Lovell had turned to
literary efforts in the 1850s; from the late 1850s he produced a highly suc-
cessful school series of which arithmetic was an important part. His principal
arithmetic author was John Herbert Sangster (1831–1904), a London-born
teacher who had immigrated to Canada at an early age, and who had studied
at Upper Canada College and at Victoria University, Cobourg (M.A. 1861,
and M.D. 1864). Shortly after obtaining the M.D. he became headmaster of
the Normal School at Toronto which Ryerson had founded. In addition to
his arithmetic works, Sangster wrote a treatise on natural philosophy, one on
chemistry, and an introduction to algebra.43
Lovell’s virtual monopoly began to be broken after Confederation, presum-
ably in part because of expansion of the market: in 1871 the Schools Act in
Ontario eliminated tuition for elementary schools and instituted a preliminary
form of compulsory attendance. The main publishers to come onto the scene
41
Science teaching had clearly become important to the directors of the Séminaire
de Nicolet a decade before the Desaulniers had been sent to the U. S. For example, in
1824 and 1829, they had published a work called Nouveau Traité abrégé de la Sphere
d’aprés le systéme de copernic, par demandes et par réponses for the students of the
seminary. It seems likely that this was a reprint of a French textbook. See Lessard,
C. (1980), p.265. Icon Lortie (1955, p. 39) attributes this 24 page work to Isaac
Desaulniers which seems implausible given that he was only 13 years old at the time
of the first printing.
42
Provost, Honorius (1959), p. 667.
43
MacMillan’s Dictionary of Canadian Biography (1963), p. 668.
at this time in Toronto were W. Gage, Copp Clark, and Rose (later Hunter
Rose). English models (and English authors) continued to be of importance
after Confederation, however, though the phrase “for Canadian schools” was
usually tacked on to an adaptation of a British work. Key among the English
authors were Barnard Smith (1810-1876), a fellow of Peterhouse (Cambridge)
and later rector of Glaston and Rutland; and his cousin J. Hamblin Smith
(1829–1901). These two men authored a large number of school textbooks
at a variety of levels. Barnard Smith’s books were adapted in many editions
by Archibald MacMurchy for Copp Clark; James Hamblin Smith’s arithmetic
was likewise adapted, by Thomas Kirkland and William Scott, and appeared
in a number of editions published by Adam Miller and later by his successor
William Gage.
MacMurchy (1832-1912) came to Canada in 1840, and after early education
at Rockwood Academy graduated B.A. (1861) and M.A. (1868) from the
University of Toronto. From 1858 he was the mathematical master at the
Toronto Grammar School (later Jarvis Collegiate), becoming rector in 1872
until his retirement in 1900.44 In addition to his role as an adaptor, he was
editor of the Canadian Educational Monthly for many years; we shall have
occasion to refer to some of his reviews below.
A further textbook author of the 1870s and after is James Alexander
McLellan (1832-1907). Besides his contributions as an author of books on
arithmetic and algebra, McLellan is significant in displaying some of the at-
tributes now associated with professional educators. In particular, he wrote
on educational psychology from 1889 onwards, collaborating with the U.S.
philosopher and psychologist Thomas Dewey on The Psychology of Number
(1903).45
The Beginnings of Higher Mathematics

Sangster, MacMurchy, and others mark a point around the time of Confed-
eration by which Canadian-educated writers were producing materials for el-
ementary education in the Canadian context. These were required to meet
Provincial standards, and included extensive answer keys and examination
practice materials. By the same period, Canadian-authored upper school texts
began to appear, not only in “advanced arithmetic” which included problems
in compound interest, book-keeping, and so on, but also in Euclidean Geom-
etry and in Algebra.46
44
Wallace (1963), p. 481.
45
Wallace (1963), p. 475. McLellan taught at a number of schools, including Upper
Canada College, and became Director of Normal Schools for Ontario in 1875. He
was later Principal of the Ontario Normal College at Hamilton (from 1885).
46
In fact, there is an earlier algebra text by Sangster (1853) though we have
not seen it. Its title, Algebraic formulae: showing the method of deducing the most
During the period in question it is difficult to distinguish works that are

produced for upper secondary and for post-secondary teaching. In some cases
the same book was used for both, the choice depending rather precisely on
the audience and on local conditions. For example, an incoming class of engi-
neering students from a variety of schools might very well require an algebra
course more or less identical to that given in the better high schools, though
the pace might be faster. We will therefore not try to distinguish them too
carefully here, the more so because the same authors wrote for both groups.
This overlap between secondary and post-secondary education extends far
beyond the mathematics courses, however. Virtually all Canadian universi-
ties and colleges of this period began their history as high schools, a gradual
differentiation occurring only when an adequate supply of students became
available to sustain post-secondary education.47 To illustrate this in a math-
ematical context, let us briefly consider the courses of mathematical study at
some Canadian institutions at various periods.
The oldest one known is from King’s College, Windsor, N.S. in 1814 (the
College had been chartered in 1789, but began granting degrees only in 1807).
In this case, Oxford was the model; hence there was a very heavy emphasis
on the classics. There were two professors, one of whom taught “Euclid and
Wood’s algebra”48 in the third year. In Fredericton in 1824 all instruction was
provided by the principal; the amount of mathematics appears to have been
similar.
By 1860, an increased emphasis on mathematics is generally evident. This
is in part due to the Scottish model, urged by Strachan (hence influential
in Ontario, English Quebec, and, via a Royal Commission of 1854, in New
Brunswick). Thomas McCulloch in Nova Scotia also urged its adoption. De-
tails of the mathematical requirements for two B.A. degrees at this time are
provided by Harris:
Queens:
Year One: Euclid 1-6, algebra, plane trigonometry, logarithms
Year Two: Euclid 11, part of 12; plane and spherical trigonometry,
conics, calculus
Year Three: Principia 1-3; hydrostatics
Trinity:
Previous Examination: Euclid 1-4, 6; algebra to the binomial theorem
B.A. Exam. Algebra to the end of the binomial theorem, trigonometry
and solution of triangles, mechanics, hydrostatics
At the same time, Toronto required mathematics in each of the four years,
and McGill in the first three of four. At Toronto mathematics could be avoided
important rules of arithmetic and mensuration suggests something other than a
simple introduction to algebra.
47
For details on the development of Canadian Universities, see Harris (1976).
48
Harris (1976), p. 30. The Wood in question is James Wood of late eighteenth-
century Cambridge.
after first year by obtaining first class honours in the first year. Other insti-
tutions had similar requirements. It is interesting to note that these were
required of all students; as a corollary it follows that the standards in mathe-
matics were doubtless not very high. Honours students at Toronto or McGill,
who wrote specialized examinations in the chosen subject, doubtless had to
meet a higher standard than did general students. Materials for these courses
were originally imported, though as mentioned earlier Canadian texts at a
higher level started to appear in the 1860s. These developments in mathe-
matics education reflected an increased interest in science and engineering in
the universities. Besides an expansion of these programs, this development is
marked by the founding of two scientific societies: the Canadian Institution
(later the Royal Canadian Institute), founded at Toronto in 1849; and the
Royal Society of Canada, established in 1882. Both of these provided a forum
for members to discuss scientific and technological issues of general interest,
and for the presentation and publication of research.
In Ontario the connection between school and university textbook writers
was particularly close in this period, in part owing to the membership of many
of these men in the Royal Canadian Institute. Thomas Kirkland, William
Scott, John H. Sangster and Alexander MacMurchy all were members, as were
several university professors, among them James Loudon and J. B. Cherriman
of the University of Toronto, and Alexander MacKay of MacMaster. The
elementary and secondary teachers among these men were at the top of their
profession by the mid-1880s and early 1890s; all were either Principals of
leading schools, or of one of the Ontario Normal Schools, and many had taught
at Upper Canada College at some point.
Of the authors of the 1870s and 1880s, three stand out because of their
involvement in university teaching, as well as in production of texts for peo-
ple at the university level. These are: John Bradford Cherriman (1823-1908),
professor of mathematics at Toronto from 1850 to 1875; his successor James
Loudon (1841-1916), professor of mathematics and physics at Toronto; and
Nathan Fellowes Dupuis (1836-1917), professor at Queen’s.
Cherriman was born in England, and had graduated from St. John’s Col-
lege, Cambridge, as sixth wrangler in 1845, the year William Thompson was
second. He came to Toronto as an Assistant Professor in 1850, and was pro-
moted to Professor in 1853. Cherriman published a dozen papers in the Cana-
dian Journal, the publication of the Canadian Institute, and three in the first
volume of the Transactions of the Royal Society of Canada after the found-
ing of that Society in 1882. However, as Gingras has pointed out, these are
very much either recreational or teaching-related. They do not build on ear-
lier work, and are at best new proofs of old results.49 Two titles will suffice to
49
Gingras (1991), 17-5 1. In this regard they rather resemble the weaker papers
in the Cambridge and Dublin Mathematical Journal, with which Cherriman was no
doubt well-acquainted.
convey their flavour: Note on the composition of parallel rotations, and Note
on the bishop’s move in chess.
Loudon was a product of the Upper Canadian system to which we have
been devoting our attention. He studied at the Toronto Grammar School, Up-
per Canada College, and the University of Toronto, receiving a B.A. in 1862
and an M.A. in 1864. He then became a tutor in classics, but moved to math-
ematics, eventually becoming the University’s professor of mathematics and
physics in 1875 (and the first Canadian-born professor). In 1887 he became
professor of physics only, and became president of the University in 1892. This
set the stage for the first mathematical research at Toronto. Like Cherriman,
Loudon published a number of articles in the Proceedings of the Royal Cana-
dian Institute; and like Cherriman, these mostly arose from teaching concerns.
Dupuis taught school from 1857 to 1863, presumably following Normal
School training.50 At that point he was able to enter the University of Queen’s
College (Frontenac County was his home), where he worked as an observer
in the Kingston observatory and as librarian. Obtaining a B.A. (with eight
others) in 1866, he went on to obtain an M.A. from the same institution in
1868. At that point he succeeded Robert Bell as Professor of Chemistry and
Natural History, and his writing career began in the same year with a textbook
on geometrical optics. During his time at Queen’s Dupuis taught physics,
geology, mineralogy, biology, mathematics, and various engineering courses.
He was an important institution builder, instrumental in the establishment of
engineering and a medical school on a firm footing.
For reasons that are unclear, Dupuis began to teach mathematics in 1880.
He was at once concerned to provide his students with an up-to-date course
of study, and the two books that he produced in the 1880s attest to this
concern. The first of these, Junior Algebra (1882) eventually became The
Principles of Elementary Algebra, published by MacMillan in 1892. The term
“junior” in the title may refer to the third year, as in U.S. nomenclature,
or it may refer to a preparatory work, presumably for first year students. It
was characterized by Dupuis as an “intermediate algebra”, and owes a good
deal to Chrystal’s Algebra. The content includes a certain amount of formal
algebra, so that terms such as “commutative” are introduced, but the work
is very much concerned with applications, as Dupuis noted in the Preface:
Probably the most distinctive feature of the work is the importance at-
tached to the interpretation of algebraic expressions and results. . . the
results arrived at have little interest and no special meaning until they
are interpreted. This interpretation is either Arithmetical, that is, into
ideas involving numbers and the operations performed upon numbers;
or Geometrical, that is, into ideas concerning magnitudes and their
relations.51
50
Most information about Dupuis is from Varkaris (1980).
51
Dupuis (1892), p. iv.
At the higher end, the book includes such subjects as simultaneous quadratics,
the remainder theorem, approximation of roots, annuities, continued fractions,
series and determinants.
Probably prepared in outline at around the same time as the algebra,
but not in print until 1889, is Geometry of the Point, Line and Circle in
the Plane. This was probably Dupuis’ most successful book, appearing in at
least five editions to 1914. This is at about the same level (again described as
“junior”) as the algebraic work, and is definitely intended as preliminary to
a study of analytic geometry and the calculus. The work shows a good deal
of originality in not simply reorganizing Euclid. Instead, lines and curves are
treated as plane loci, so that triangles are distinguished from the regions that
they bound. One of the inspirations cited is Sylvester:
The principle of motion in the transformation of geometric figures, as

recommend by Dr. Sylvester, and as a consequence the principle of
continuity are freely employed, and an attempt is made to generalize
all theorems which admit of generalization.52
There is a certain naiveté in this last statement, one which is probably quite
genuine. Although Dupuis had worked hard to put together good preliminary
courses for his students, there is little evidence that he had a grasp of higher
geometry as it was practiced in his day. However the book does introduce such
topics as inversion in the circle, pole and polar, homographies and involutions.
These books were well-received in the community for which they were in-
tended, as their success in the U.S. and Canadian markets attests. Favourable
reviews aided in the process. “It is safe to say”, noted the Canada Educa-
tional Journal, “that a student will learn more of the science [of Geometry]
from this book in one year than he can learn from the old-fashioned trans-
lations of a certain ancient Greek treatise in two years”. The same review
urged every mathematical master to study the book “in order to learn the
logical method of presenting the subject to beginners.”53 As for the algebra,
it was described in The Schoolmaster as “one of the most able expositions
of algebraic principles that we have yet met with . . . emphatically a book for
teachers.”54
These books are symptomatic of yet another generational change in Cana-
dian mathematics, one in which students at the universities begin to see rig-
orous mathematics beyond the elementary Euclidean level, and in which a
variety of subjects are treated in an incipiently rigorous way. In addition, the
courses written by Dupuis are clearly prefatory to acquiring a higher level
of expertise, and the debt to more advanced work is acknowledged for the
student to see. These are not simply “everything you need to know” about
a certain subject for the purposes of application. Students prepared under
52
Dupuis (1914), p. vi.
53
Review reprinted at the end of Dupuis (1914).
54
Reprinted at the end of Dupuis (1914).
this regime would clearly be more susceptible of recognizing the existence of

more advanced mathematics, and of beginning to value the subject outside
of its applications (though the latter were clearly very important). Perhaps
significantly, Dupuis’ title for a time was Professor of Pure Mathematics.
By the 1890s, most of the established institutions had one professor of
mathematics, though in some cases duties were still split between mathemat-
ics and physics or engineering. At some schools, notably Toronto and McGill,
there was also an assistant or lecturer. Thus in the course of the nineteenth
century we see that mathematics in English Canada has transformed. An
institutional base for elementary mathematics education was firmly in place
by the time of Confederation, and was closely linked to developing institu-
tions at the secondary and post-secondary levels. University mathematics was
established as a teaching subject in most institutions by the 1890s, though
research was still in a nascent state; and it was to remain in that state at
least until the time of Fields’ comments cited at the beginning of the paper,
despite a general expansion of mathematical education at the universities in
the intervening decades.
Reasons for the lack of research late in the nineteenth century are not
hard to find. While in fields such as physics and chemistry it was still possible
at that time for an relatively inexperienced student to undertake experimen-
tal work of a meaningful sort, participation in mathematical research at an
international level required an extensive exposure to the literature and, ide-
ally, to working research mathematicians. Such literature was difficult to come
by in Canada. Furthermore, there were few rewards for engaging in research
beyond personal satisfaction. The tiny Canadian mathematics community of
the late nineteenth century was instead fully involved with teaching a variety
of largely introductory courses in institutions which were widely separated
geographically.
More than this, the Canadian university was a combination of the English
and Scottish models. Thus liberal and practical education vied for position in
the curriculum, and professors were either British-trained or trained in Canada
according to this model. Research had long taken a secondary position in the
British schools, and where it was undertaken an emphasis on practical results
was emphasized. The education-related, physically oriented papers of Loudon
and Cherriman reflect these values, and several decades were to pass before
the views of Fields, that pure mathematical research was desirable in the
universities, came to find widespread favour in Canada.
Québec from the Union to the Foundation of the

Université de Montréal (1840–1920)
The union of Upper and Lower Canada had followed political disturbances
that were felt in each of the two colonies. In Québec some small changes for
mathematics are detectable, though these changes were mostly felt at the lower
levels (due to a new trend in pedagogy) and in connection with the practical
applications of mathematics. In the classical colleges, mathematics teaching in
the philosophy courses stagnated. We also see in this period the beginning of
the publication of local works which go beyond the elementary level. Among
these only one displays any originality. Later, however, the founding of the
École Polytechnique de Montréal in 1873 marks the beginning of mathematics
teaching beyond the elementary level outside the classical colleges.
Primary teaching: “l’enseignement mutuel”55 The arrival in Canada of the
Frères des Écoles chrétiennes (Brothers of the Christian Schools) marks a turn-
ing point in the history of education in the Province. This French community,
which had been founded in 1684 by Jean-Baptiste de la Salle (1651-1719), had
a long tradition of teaching students in groups. When the first four brothers
arrived in Montreal in November of 1837, they brought with them this tradi-
tion, which permitted them to address the problems of mass education head
on. Their school took in 200 students in the first year, and by 1840 they had
860 students.56 In mathematics, they were innovative, and their approach was
rather more dynamic than that of their predecessors. This is indicated by the
title of their first textbook, published only a year after their arrival: Nouveau
traité d’arithmétique: contenant toutes les operations ordinaires du calcul, les
fractions et les différentes reductions de fractions, les règles de trois, d’intérêt,
de société, d’alliage, l’extraction des racines, les principes pour mesurer les
surfaces et la solidité des corps; enrichi de 400 problèmes a résoudre, pour
servir d’exercice aux élèves: a l’usage des écoles chrétiennes des frères. The
book was a reedition of a recent (1833) French book, hastily modified to take
into account the peculiarities of the British colony (notably currency and the
system of weights and measures).57 Unlike its predecessors, it contains a large
number of problems. Even more indicative of a new approach, most of the
problems were not presented with solutions. This indicates the importance
which the brothers gave to exercises to be completed by the students them-
selves, and it also presupposes the competence of the teacher, at a certain
level at least. Thus the rule-example-rule presentation of earlier manuals was
replaced by a long, careful presentation of each topic, calling on examples,
and concluding with a concise statement of the rule. This was then followed
by exercises and problems. The brothers even attempted to give their book a
deductive structure adapted to the level of the students.
55
Lavoie, Paul (1994), pp. 384–410. Here there are analogies with two U. S. books:
Dilworth, Thomas, Schoolmaster’s Assistant: Being a Compendium of Arithmetic
Both Practical and Theoretical, 1773, and Adams, Daniel, Arithmetic, in which the
Principles of Operating by Numbers are Analytically Explained and Synthetically
applied; thus Combining the Advantages to be Derived both from the Inductive and
Synthetic Mode of Instruction, 1801. The latter had many Canadian editions.
56
Audet, Louis-Philippe (1971), t. 1, p.370.
57
F.É.C (1833).
Another textbook, more properly Québecois than that of the brothers,

appeared around 1843. While emphasizing British currency and weights and
measures, it harks back to its predecessors in presenting few explanations.
The following year, a book of solutions to the exercises was published.58
These books had a large number of editions, and became the paradigmatic
texts for secondary teaching. To follow the developments which follow at this
level would take us too far from our main subject. However, we note that these
works were often used in the Humanities classes of the classical colleges.59
The classical colleges. In 1840, mathematics had solidified its position in the
colleges, particularly in the Humanities classes and in the two years of phi-
losophy. Léon Lortie has argued that 1840-1850 was the golden age of math-
ematics teaching in Québec, a statement which needs some added nuance.60
Certainly mathematics instruction progressed greatly from 1800 to 1840, but
this progress is linked to the construction of a system of elementary education
in the province. The classical colleges participated in these developments by
opening the door to mathematics in the Humanities class. At the upper level,
the progress seems slight, if indeed there was any. The notes for the course
given by Abbé Alexis Pelletier in 1862–1863 in the philosophy class at the
Séminaire de Québec are revealing in this regard.61 It is divided into three
main parts, the first on algebra, the second on geometry, and the third on
plane trigonometry. The first contains, among other things, sections on ratio
and proportion, the rule of three, and questions with a commercial flavour,
for example on interest. This content is exactly what was taught at the end
of the previous century at Québec. The fact that such a notebook exists also
shows that the taking of notes remained an important part of the student’s
activities. This is confirmed by reading the memoirs of another student at the
Séminaire, where we find remarks on the course given by Abbé Théophile-
Étienne Hamel, who replaced Pelletier in 1866–1867.62 Hamel had returned
from Paris ten years previously, with a license in science obtained at the École
des Carmes63 , and was thus the third college professor to have studied physics
or mathematics abroad. His student Gosselin reminisced as follows:
He didn’t have time to rework the course which he had taken in Paris,
and to adapt it to another milieu, to reduce its proportions, to make
it accessible to students of whom many knew almost nothing, even in
58
F.É.C. (1842) and F.É.C. (1843). The date is uncertain. See Lavoie, Paul (1994),
p. 404.
59
For a complete and detailed overview of the teaching of arithmetic in Québec
in the nineteenth century, see Lavoie, Paul (1994).
60
I (L.C.) endorsed this view in (Charbonneau, Louis (1984)).
61
These notes were taken by Jean-Alfred Charlebois and may be found in the
Archives de l’Université Laval, box P 211.
62
Gosselin, D. (1908), pp. 174–175.
63
Gingras, Yves, (1991), p. 34.
arithmetic. The hour of the course was spent copying—at full steam—
what he dictated to us. Under the heading “theory of limits”, incom-
prehensible to the vast majority, I scribbled four hundred lines of tiny,
cramped writing . . . he did not often compel us to grind the mill of
problems, contenting himself to deliver the formulas without which it
could not be moved . . . . Furthermore, arithmetic took the lion’s share
of the time, almost six months out of ten. Only four months remained
for algebra, geometry, and trigonometry. Of this last we got only a
bird’s-eye glimpse.
As we can see, the students often arrived in philosophy with minimal

mathematics, so it is hardly surprising that arithmetic got the lion’s share.
What could the students possibly take away from a mathematics course that
consisted of dictation? While the professors of mathematics were no longer
just out of the philosophy course, the teaching itself did not evolve.
Mathematics had come into the Humanities course of the colleges in the
context of the general reform of the 1830s, but for several reasons this did not
extend to the philosophy classes.64 For one thing, the reform movement was
halted by the events following the rebellion of 1837. As Jarrell has argued,
there appears to have been a detouring of intellectual energy away from the
sciences and toward political concerns.65 As a result, the small number of
French-Canadians who had participated in scientific and literary societies was
not sufficient to orient the colleges towards science teaching after 1837. Grad-
ually an image developed of the French-Canadian who was naturally drawn
toward the “moral and political sciences, history, literature and the arts”, in
contrast with the English, who had an affinity for the “mathematical, physical
and natural sciences”. The sciences became tributary to these values, to the
extent that they could be, and lay society no longer came knocking at the
door of the colleges to insist that the science and mathematics taught there
met the needs of the industrial society developing in Québec. In addition, the
Catholic Church in whose hands this education lay was concerned primarily
at this time with developing its weapons against the new social philosophies;
its priorities did not include the teaching of the sciences, on which a num-
ber of these philosophies claimed to be based. As Abbé J. S. Raymond of the
Séminaire of Saint-Hyacinthe noted in 1872, “Is it not to the deeper knowledge
of the sciences that the materialist movement in which our century so prides
itself is due?” No more was needed to assure that Latin and Greek would be
seen as the most important tools for developing the spirit, to the detriment
of mathematics. The formative qualities of mathematical training were not
contested, but they had the major flaw of abetting the “materialist move-
ment”. As a result, when we compare the program of 1863, as evidenced by
the notes of Abbé Pelletier, with the program of 1921 at Laval for graduates
64
The citations in this paragraph are taken from Charbonneau, Louis (1984), pp.
29–31.
65
Jarrell, R.A. (1977).
of the philosophy class, the similarity of the contents is striking. Whatever

may have happened in the lower schools since 1840, in the upper school the
attitude toward mathematics was, by 1900, one of mere toleration.
The first intermediate level mathematics books in French. In a context
where French-Canadians developed a bit of a complex with regard to the
sciences and mathematics, it isn’t too surprising to find that between 1840
and 1920, apart from school texts, only four mathematical books were pub-
lished. What is perhaps surprising is that the “mathematically-inclined” Eng-
lish Canadians did little better.
The Traité élémentaire de calcul différentiel et de calcul intégral, attributed
to Jean Langevin, (1821-1892) was possibly intended to help in teaching dif-
ferential and integral calculus in the colleges.66 However, the level goes well
beyond what could have been taught there. Langevin was professor of math-
ematics at the Séminaire de Québec in 1838, even before his ordination, and
is best-known as the first principal of the École Normale Laval, and later as
the first bishop of Rimouski.
The book is divided into three parts, followed by three notes. The first
two parts, differential and integral calculus, rest on the idea of the differen-
tial with the Leibnizian notation. Each of the two parts is divided into two
chapters, the first of which gives rules for the differentiation and integration
of the principal functions and the second of which gives “applications” such
as maxima and minima or volumes of solids of revolution. The third part,
“Method of Limits”, uses the difference quotient idea to get at series develop-
ment. The notes deal with further series expansions using Newton’s binomial
series, and with the method of undetermined coefficients. The few “problems”
in the book are generally accompanied by a solution, except in the last note
where only statements are given.
This treatise could well have been written at the beginning of the nine-
teenth century. Allusion is made to the fact that there are several methods
which “all lead to the same result” and that “the difference between them
is more metaphysical than mathematical.” The methods to which the author
refers are those of Newton and Leibniz, but also those of Landen, d’Alembert
and Lagrange. There is no mention of Cauchy. The treatise concludes with a
reference (for additional information) to Lacroix, Hind, and Boucharlat, thus
revealing its mixed ancestry, both English and French.67 While it is not com-
pletely original, it is clearly written, and permits the reader to develop the
basic methods of the calculus.
The Premier livre des elements de géometrie d’Euclide, a l’usage des
étudiants au Collège nautique du Canada (1853) is considerably less original
66
Lortie, Léon (1955, p. 40) gives the origins of this attribution.
67
John Hind (1796-1866) and Jean Louis Boucharlat (1775-1848). The work of
Hind referred to by Langevin remains unidentified. As for Boucharlat, the book is
certainly An elementary treatise on the differential and integral calculus. Translated
from the French by R. Bladelock, Cambridge, 1828.
than the work we have just been speaking of. The author’s name is not given,
but the work is in fact a translation of a brief portion of the well- known book
of Robert Simson, Euclid, Elements, First six books, with the 11th and 12th
and Euclid’s Data.68 An examination of the statements of the propositions
suggests that the anonymous translator was inspired by the French transla-
tion of Euclid’s works made by Peyrard at the beginning of the century.69
The third of these home-grown works is arguably the most interesting, not
only because of its length and its originality, but also because of its author,
Charles Baillargé (1826-1906). The book is Nouveau traité de géométrie et de
trigonométrie rectiligne et sphérique, suivi du toisé des surfaces et des volumes
et accompagné de tables de logarithmes des nombres et sinus, etc. naturels et
logarithmiques et d’autres tables utiles. Ouvrage théorique et pratique illustré
de plus de 600 vignettes, avec un grand nombre d’exemples et de problemes
a l’usage des Arpenteurs, Architectes, Ingenieurs, Professeurs et eleves, Etc.
Weighing in at 900 pages, it appeared in 1866. Baillargé came from a family
of architects and engineers which had come to Canada in 1741, and he con-
tinued the family tradition, working for example as the on-site architect of
the Parliament Buildings in Ottawa from 1863 to 1865.70 Here he was able
to put his mathematical skills to good use, defending himself unsuccessfully
from charges of overspending on construction by calculating the volume of
materials employed in the irregularly-shaped buildings. His book aimed at
introducing higher elementary mathematics to a wide audience, and Baillargé
did not hesitate to redo classical treatments to achieve this end, cutting down
the number of propositions in Euclid’s first six books by half.71 His inspi-
rations included Legendre and Davies, while his trigonometry seems to have
come from Playfair and from the old Institutions mathématiques of Sauri.72 He
was proudest of his chapter on the measurement of areas and volumes, noting
the originality of some of his results, notably his proposition 1521.73 This dis-
68
Simson’s book was first published in 1756. We have consulted the 25th edition,
dating from 1841.
69
Peyrard, F. (1819).
70
Cameron, Christina (1989).
71
Playfair, John, Elements of geometry: containing the first six books of Euclid,
with a supplement on the quadrature of the circle, and the geometry of solids: to
which are added elements of plane and sphericale trigonometry, 8th ed., Edinburgh:
s.n., 1831. First edition, 1795. There was a tenth edition in 1846.
72
Legendre, Adrien-Marie, Elements de géométrie, First published in 1794 with
many editions throughout the nineteenth century. Legendre was also translated
into English: Legendre, Adrien-Marie, Elements of geometry and trigonometry; with
notes, Edinburgh: Olivier & Boyd, 1824. Davies, Charles, Elements of geometry and
trigonometry from works of A.M. Legendre: adapted to the course of mathematical
instruction in the United States, New York: A.S. Barnes & Co., 1862 (reprinted
1871).
73
The statement of the theorem is as follows (p. 662 of the Nouveau traité. . . ): “Of
every prism or right cylinder or oblique – of every pyramid, regular or irregular, or
every cone, whether right or oblique – of every truncated pyramid or cone between
covery was the theoretical basis for his most famous production, the Tableau
stéréométrique, a wooden box on a base five feet by three feet, about five
inches deep which contained 200 wooden models of geometric shapes.74 While
the object could be used as an introduction to solids for classes, its main
use was to assist engineers and architects in the measurement of volumes.
Baillargé promoted the work energetically, and succeeded beyond his wildest
expectations: between 1872 and 1876, the tableau earned him 13 medals and
17 diplomas from eight different countries.75 In Russia, for example, its adop-
tion was urged not only in the primary schools but also in Polytechnics.76 In
Québec, the device was investigated, following Baillargé’s request for official
recognition, by Jean Langevin, who passed it on to Thomas-Étienne Hamel,
now director at the Séminaire de Québec. Together with the Abbé Mainguy
of the Séminaire he declared it satisfactory, and Mainguy even published a
tract on the device. Following this expert examination, the Council of Public
Instruction recommended its use in the schools of the Province.77
In 1882, at the founding of the Royal Society of Canada, Baillargé was a
founding member of Section III on mathematics, physics and chemistry. He
regularly presented papers to the annual meetings of the society, among which
there are three on mathematics. However, as with English-language writers
of the day, his main preoccupations were toward pedagogy and applications,
as his 1882 “Utility of Geometry as applied to the Arts and Sciences” and his
“Hints to Geometers for a new Edition of Euclid” suggest.78
A fourth treatise, the Théorie élémentaire des nombres d’après Buler,
Legendre, Gauss et Cauchy (1870) is a work which falls outside the patterns
we have seen so far. Of course, the appearance of Buler for Euler in the ti-
tle doesn’t exactly inspire confidence in the anonymous author. It is a 22
page pamphlet listing relatively elementary results in number theory, includ-
ing material on Gaussian residues. Without further information, there are
only questions about this work: could such information be found in Québec
at the time? Who could the author have imagined to be the public for such a
work?79
parallel bases – of the sphere . . . [many such figures omitted]: the volume is equal
to the sum of the surface of the base, if there is only one, or of its parallel bases,
if there are two, and four times the surface of a section midway between the bases,
between the base and the vertex, or between opposite vertices, multiplied by one
sixth of the height of the solid.”
74
A complete bibliography of Baillargé’s works is in Cameron, Christina (1989),
pp. 161–166. On the Nouveau Traité. . . , see Chapter 11, pp. 131–138.
75
Prospectus du Tableau Stéréométrique Baillargé, [Quebec]: n.p. [1871].
76
Cameron, Christina (1989), p. 136.
77
Cameron, Christina (1989), pp. 132–133.
78
The complete list of papers given before the Royal Society is given in Cameron,
Christina (1989), p. 191, note 62.
79
The author mentions that the proofs of one of the propositions was given to
him by a Prof. Wantzella. Anonymous (1870), p. 12–13.
The École Polytechnique de Montréal. In 1852, the Université Laval was

founded in Québec, with four faculties: theology, law, medicine and arts.80
The beginnings, as with most institutions of the period, were difficult, with
shortages of both professors and students except possibly in medicine (where
the Faculty was just a restructuring of an existing school).81 The main aim
of the Arts faculty was to control the quality of teaching in the classical
colleges.82 However, the shortage of professors led to the mission of T.-É
Hamel to Paris, from which he returned with his licence in mathematics as
we mentioned above.83 The lukewarm attitude toward science and particu-
larly applied science at, Laval at the time is well-illustrated by the story of
the founding of the École polytechnique de Montréal.84 In 1870, the Chauvau
government had offered Laval a grant to create a school of applied sciences.
After much hesitation, this was refused: the university did not want to see
the government mixed up in its affairs. Such a school remained a political pri-
ority, and in 1873 the new government of Gédéon Ouimet began negotiation
with the Commission of Catholic Schools of Montréal, and in particular with
the director of the Catholic Commercial Academy of Montreal, Urgel-Eugène
Archambault, which led to the introduction of a “Cours scientifique et indus-
triel” at that school. In 1876, the Academy became the École polytechnique
de Montréal. Its beginnings were modest, with 114 diplomas awarded during
the years from 1877 to 1904. Of these graduates, only 54 came from classical
colleges, reflecting the lowly status given to the engineering profession by the
colleges. As Abbé Hamel, himself a science graduate, said in 1876, “We don’t
hesitate to encourage such [engineering] studies for those of our young people
who are not destined for the priesthood.”85
It is hard to assess the level of instruction at the École in its early years.86
The two professors charged with mathematics instruction during the first
decades of its existence, Frédéric André and Emile Balète, had no university
training. André taught for the most part the introductory course from 1875 to
80
81
82
Following Léon Lortie, some authors have argued that, in its efforts toward
uniformization, the Faculty of Arts brought about a decline in the quality of math-
ematics and science teaching at the Séminaire de Québec. We instead support the
view of Chartrand, Duchesne, Gingras (1987), p. 220 who attribute this point of
view to an error.
83
The kind of licence obtained be Abbé Hamel in Paris in unclear. Chartrand,
Duchesne, Gingras (1987), p. 217, suggest a qualification in mathematics while Gin-
gras, Yves (1991), p. 34, speaks of a licence en sciences, for other references in-
dicating a licence en mathématiques, see Charbonneau, Louis (1984), p. 33, note
29.
84
Chartrand, Duchesne, Gingras (1987), pp. 227–230 and Gagnon, Robert (1991),
pp. 39–44.
85
Gagnon, Robert (1991), p. 70.
86
Gagnon, Robert (1991), pp. 64–68.
1911.87 Balète was in charge of the rest of the mathematical program, teaching
from 1875 to 1908 and serving as director of the École from 1882 to 1909. He
had been trained at the French military Collège at Saint-Cyr, and immigrated
to Canada in 1872 following an apparently disappointing military career. De-
spite a lack of advanced mathematical training, Balète also became professor
of mathematics at the Arts faculty of the Montreal Branch of the Université
Laval in 1900-1901.88 A look at the annual programs gives us some idea of the
content of courses at this time, though the testimony of the programs should
be accepted with caution. It appears that differential and integral calculus be-
came somewhat more important during the period in question.89 After 1910,
more profound changes occurred, with two new professors, each more distin-
guished mathematically than their predecessors. Victor Elzéar Beaupré and
Conrad Manseau were both graduates of the school, and both engaged in
scientific activity extramurally. Beaupré, later a professor of mathematics at
the newly-founded Faculty of Science at the Université de Montréal, was an
actuary, indeed the first French-Canadian to become a member of the Society
of Actuaries of America.90 Marseau, on the other hand, was an astronomer,
having obtained a licence from the Sorbonne in 1914.91
Expanding Horizons
By the turn of the century, signs of a new agenda for mathematics in Cana-
dian universities begin to be seen. This agenda both tended away from
applications—though this varied from one school to another—and showed
an awareness of the importance of research and of the diversity of the math-
ematics then being practiced internationally. This awareness, as we shall see,
seems to have been concentrated in a few individuals. It manifests itself by the
beginning of doctoral programs, by an effort to increase exposure of students
to a broader range of mathematical ideas, and, where doctoral programs did
not exist, by preparing students for advanced study in other institutions. We
will take each of these developments in turn.
87
Even in retirement he continued to teach, until just before his death in August
1923. Archives de l’École polytechnique, dossier Frédéric André, dossier 320-300-22.
88
According to the Annuaire of the Université Laval for 1900–1901.
89
In the Bulletin annuel of the École polytechnique for 1878–1879, the “first prin-
ciples of the theory of derivatives” form part of the algebra course in first year. The
“complete theory of derivatives” and series expansions constitute the first part of
the second year algebra course. The booklet Programmes des travaux techniques et
questionnaires des examens généraux of 1896-1897 shows clearly that there was a
calculus course in the third year on differential and integral calculus.
90
La Voix Nationale, Dec. 1935, from a clipping in the dossier Beaupré (dossier
329-300-22), Archives de l’École polytechnique.
91
Building mathematics in the universities. At around the turn of the century,

as was the case in the U.S., young Canadian-trained mathematicians began to
study abroad. This led to the first Canadian Ph.D’s in mathematics, people
with an exposure to a European research tradition who had not only under-
taken research on their own, but who had seen the necessary conditions in
which such research could flourish. The most important of these were those
who returned to Canada to academic positions. Best-known of these is J. C.
Fields (1863-1932), a University of Toronto gold medallist as an undergraduate
in 1884, who took his Ph.D. at Hopkins in 1887. After a short teaching stint
in the U.S., Fields went to Europe, where he spent approximately ten years
in the mathematical centres of Berlin and Paris. Fields was appointed to the
Faculty at Toronto in 1902. At this time, another young Toronto-educated
mathematician was also in the Department. A. T. DeLury had undertaken
graduate work at Clark and in Paris. He had joined the Toronto faculty in
1892. The department at the time thus consisted of Fields, DeLury, Loudon,
Baker, and M. A. Mackenzie, an actuary. Of these, Fields was perhaps the
most important in developing research at Toronto; his student, Samuel Beatty,
was the first mathematics Ph.D. at Toronto, obtaining his degree in 1915. In
the next twenty years, Toronto was to produce eight doctorates in mathemat-
ics, two of them women. This production of Ph.D.’s was however supervised
mostly by imported mathematical talent: W. J. Webber, from Cambridge,
and J. L. Synge from Dublin.92
At other institutions, mathematics was either in a nascent state (as were
the institutions themselves in many cases), or else–as at McGill for example–
it remained firmly subsidiary to other programs, particularly Engineering.
McGill is an interesting case in point. Its faculty included J. Harkness, a Cam-
bridge mathematician with international connections who could have fostered
research interests, though not active in research himself. It is interesting to
look at the situation there and elsewhere through the eyes of Henry Marshall
Tory (1864-1947), a McGill alumnus of 1891 who taught in the department
with Harkness, and was an important figure in the development of scientific
research in Canada.
Tory was from a Nova Scotia Methodist family, and had the good luck to
be “discovered” by a reasonably well-educated teacher who gave him books on
algebra, geometry and trigonometry. His own reminiscences stress the paucity
of books available, then and later. Determining to go to university, he worked
as a clerk to save enough to spend six months completing a teacher’s certifi-
cate. In 1886, having planned to go to Mt. Allison, he heard of McGill, where
the chances of summer employment would be better. Despite his weaknesses
in the classics he graduated at the top of his class in 1890 and was immediately
engaged as a lecturer for the fall of 1891, thus “to be associated as a colleague
with the men by whom I had been taught”, as he put it.93 At the time the
92
See Robinson (1979) for details of the situation at Toronto.
93
Corbett (1954) p. 41.
mathematics faculty included himself, Harkness, George Chandler, who was

the head of the Mathematical Department in the Engineering Faculty, John
Cox, “a Tripos man” who was adept at the style of mathematical physics then
taught at Cambridge, and Alexander Johnson, trained in Dublin, then Dean
of Arts. A student later noted of Tory: “he used to joke about the relative
capacities of himself and his associate Dr. Harkness–a Cambridge Scholar of
very high attainments. Dr. Tory admired and liked him but admitted that he
himself was a better choice for 1st yr work.”94 The student quoted Tory, “The
poor youngsters don’t know what Harkness is talking about.”
The McGill emphasis on engineering and British-style applied mathemat-
ics in those years was all the more influential because of McGill’s activities in
franchising its program. In order to be viable, a good supply of students was
necessary, especially at McGill, which unlike the Ontario and Atlantic schools
received no government subsidy. For this reason, McGill’s admnistration sup-
ported efforts to develop affiliated institutions elsewhere in the country which
would provide instruction equivalent to the first two years at McGill. Students
with satisfactory performance could then travel to Montreal for their final year
or years, completing a degree at a well-equipped institution with a good name.
Such an arrangement had begun in Vancouver and Victoria in 1899 and 1902
respectively, where the leading high schools began to offer McGill first year,
then second year. These arrangements would later lead to the establishment
of the University of British Columbia. Tory himself was instrumental in these
arrangements.
It is clear, then, that what impetus existed to produce research mathemati-
cians in this period came from the German and French models. England was
in the process of changing, but those with English-style training at this time
were inclined to the Cambridge view. While research in other sciences, notably
engineering and physics, was getting under way, mathematics remained as a
service department at most schools.
The role of the First World War in this story is problematic. It must
certainly be true that the general mobilization required in Canada cut into
any efforts to expand graduate studies during the war years, something which
was not the case in the U. S. for example. On the other hand, the war pointed
out the pressing importance of individuals trained to do research, both to the
British and to those in the Dominions, as Canada then was. The war had led
to a general call throughout the Empire for highly qualified researchers. It was
then discovered that there were without exaggeration more trained scientists
in a few of the large German industries than could be found in the whole
Empire.95 Following urging from Britain, the National Research Council was
established in 1916, though the results for mathematics were again slight at
first. The continued emphasis on applications did however begin to produce
some results as far as the use of mathematics in the physical sciences and
94
Corbett (1954), quoting Susan Cameron Vaughan, p. 43.
95
Corbett (1954) p. 154.
engineering, as well as in actuarial work and other applied areas. This is

shown nicely by the International Congress of 1924.
The 1924 International Congress of Mathematicians. In 1920, the Interna-
tional Mathematical Union determined to hold its next quadrennial Congress
in New York. By 1922 New York backed out, and the Union determined to
retain an American location by selecting Toronto, apparently at the urging
of J. C. Fields. Fields thus took on the job of chairing the Organizing Com-
mittee. We can speculate on his motives: by assembling a strong group of
the world’s leading research mathematicians, he would illustrate the diversity
of mathematics and the importance accorded by other nations to research
in mathematics. The meeting would also provide a forum for Canadian re-
searchers. Fields was successful in attracting a rather surprising amount of
outside funding: the federal and provincial governments supplied $27,000 each,
with $6,500 from the Carnegie Corporation and $2000 from the University of
Toronto. There were also contributions from the private sector: Eaton’s gave
$500, for example, as did Imperial Oil and several private citizens.
This sponsored a congress, by all accounts successful, with 444 atten-
dees, among them 107 Canadians and 191 Americans. Following the dic-
tates of the war settlement, Germany was still excluded. Those who have
been involved in organizing conferences recently will doubtless appreciate
the complexity of the task when many participants were arriving by ship,
and where the ancillary travel options included a rail trip to the west
coast. Fifteen of the papers or abstracts were given by Canadians, and as
this should give a fairly complete picture of research mathematics in the
country at the time, we list the mathematicians giving them: J. C. Fields
(Toronto–algebra); F. H. Murray (Dalhousie–partial differential equations);
N. B. MacLean (Manitoba–geometry); C. T. Sullivan (McGill–geometry);
J. L. Synge (Toronto–geometry); Daniel Buchanan (UBC–mechanics); L. V.
King (McGill–numerical analysis); H. B. Dwight (Westinghouse, Hamilton–
electrical engineering); Alan Ferrier (RCAF, Ottawa–aeronautics); T. R. Rose-
brugh (Toronto–electrical engineering); T. R. Wilkins (Brandon–ballistics); R.
W. Angus (Toronto–hydraulics); R. H. Coats (Dominion Statistician, Ottawa–
descriptive statistics); H. H. Wolfenden (Consulting Actuary, Grimsby–ac-
tuarial science). The breadth of topics is particularly interesting, comprising
what would now be called the mathematical sciences broadly conceived. Only
one third of the papers are in pure mathematics as we think of it today, and
these are really not at the leading edge of research. For the most part these
consist of elaborations on the authors’ theses, Synge’s paper constituting an
exception.
The congress undoubtedly brought together a large portion of the Cana-
dian mathematical community, establishing new ties and reaffirming old ones.
Nevertheless, Fields’ assessment, quoted at the beginning of this paper, is
quite accurate: there was very little mathematical research being done in the
country in the period immediately before or after the Congress. This had to
await further developments in the university sector and internationally.
After 1920: The University Period in Québec

With the exception of the École polytechnique, as of 1920 higher level math-
ematics in Québec had not evolved since the mid-nineteenth century. Mathe-
matics was thus out of step with Québec society as a whole in this regard. Par-
ticularly in Montréal, industrialization made evident the absence of French-
Canadians in scientific and engineering areas. Many observers suggested that
it was important to create institutions where a scientific education of high
quality would be offered. Those institutions which did exist had a primarily
practical orientation. As with the École polytechnique, mathematics was seen
as a tool in these schools. For a research community to develop, it was first
necessary that those interested in science and mathematics begin to act in a
concerted fashion. After 1920, a mathematical community gradually formed;
this was the necessary prelude to the more extensive transformation which
followed the second world war.96
The universities: mathematics in the service of the other sciences. On Febru-
ary 14, 1920, the Montreal branch of the Université Laval received its charter
and began to function independently as the Université de Montréal. In the
original plan, the faculty of science would be created at the École polytech-
nique. Mgr Georges Gauthier, the university’s first rector, saw to it that the
science faculty got rolling quickly, not least because a substantial grant from
the Rockefeller Foundation required solid improvements in the level of scien-
tific training provided to medical students. Thus one of the principal raisons-
d’être of the new Faculty of Science was to ensure that students admitted
to the medical faculty had acquired a good level of training in basic science.
Under these circumstances, the university decided to create an independent
Faculty of Science outside the École polytechnique, one which would share
laboratory facilities with the Faculty of Medicine.97 This was probably not
the best thing for mathematics, which would likely have found an affiliation
with the École polytechnique more stimulating.
Courses and teaching in the Faculty were organized on the French model.
Students enrolled in a one-year “certificate” course in a given discipline. The
accumulation of three certificates entitled the student to a licence.98 By far
96
For an overview of this movement, see Chartrand, Duchesne, Gingras (1987),
chap. 8, pp. 239–272.
97
The history of this decision is sketched by Léon Lortie in a note in the Archives
de l’Université de Montréal, box 3523 (18-8-5-1), probably written in 1970. See also
98
The decision to imitate the French system may have been taken, among other
reasons, from the fact that the Baccalaureate in Sciences was given to those who
the most popular certificate was the “PCN”, that is, physics, chemistry and
natural sciences, which was the certificate required for entrance to the medical
faculty. From 1920 to 1945, this program accounted for between 40% and 80%
of the students in the faculty. Between 1920 and 1944, only eight licences in
mathematical sciences were awarded.99 The number of mathematics profes-
sors was very small. At the time of creation of the faculty, Arthur Léveillé,
who had an honours B.A. in mathematics from London, was named professor
of mathematics, permitting him to leave his previous post as clerk in a book-
store. Victor-Elzéar Beaupré of the École polytechnique also became professor
in the Faculty of Science. It was not until 1936 that a graduate continued to
higher study, when Abel Gauthier went to Columbia to undertake a Master’s.
He obtained this in 1939 with a thesis called Theory of Group Representation
by Matrices. In the same year he was hired by the Faculty of Science at the
Université de Montréal. He published several articles in the period from 1936
and 1941, and continued his education with courses at Chicago, Columbia and
Brown in the early forties. This set a new tone, and soon Maurice L’Abbé,
Francois Mumer and Jacques Saint-Pierre went abroad to complete doctor-
ates. On their return, they brought a new mathematical culture with them.
In 1947, after the death of Arthur Léveillé, Abel Gauthier became director of
the Department of Mathematics; during his tenure as head, until 1957, the
Department positioned itself to become the research centre it was to be in the
following decade.100
As for Laval, it founded its École supérieure de chimie in 1921. The first
mathematics courses were given there by Athéod Tremblay, a surveyor and
geometer from Québec. More important for mathematics in Québec was the
appointment the following year of Adrien Pouliot, who was to become the
guiding spirit of Québec mathematics.101 Pouliot had graduated from the
École polytechnique de Montréal, and in 1928 obtained a licence in mathe-
matics from the Sorbonne. From 1929 to 1939 he spent his summers in Chicago
in order to improve his mathematics. He was the only professor of mathemat-
ics at Laval until 1936, when the Abbé Alexandre LaRue joined him.102 In
1923, Pouliot founded the Société mathématique de Québec; in 1929 he be-
came known to a wider public by sparking a lively debate on the quality of
science teaching in the classical Collèges and in secondary schools generally.
University science programs had continual difficulties in finding students who
failed the rhetoric examination for the Baccalaureate in Arts. Rhetoric was the final
year of the classical course prior to the philosophy classes. See the text of Leon
Lortie mentioned in n. 97.
99
Charbonneau, Louis (1988), pp. 8–9.
100
Formerly, following the French model, the group of mathematics professors was
termed the Institut de mathématiques. With the progressive abandonment of the
French system after 1945, the institutes became departments.
101
For a detailed biography of Adrien Pouliot, see Ouellette, Danielle (1986).
102
Althéod Tremblay continued teaching for many years. Richard, Guy W. (1982),
p. 18.
were sufficiently qualified to undertake university-level studies in science. The

controversy initiated by Pouliot’s criticisms started the long process of insert-
ing science properly into secondary education in the province, an end which
wasn’t really achieved until the 1960s.103 In 1939, two years after the founding
of a Faculty of Science at Laval, Pouliot organized a Department of Pure and
Applied Mathematics. This had six members in 1945, including Basil White,
educated abroad, and Paul Lorrain, a physicist whose later career was spent
principally at the Université de Montréal. The first five mathematics degrees
were awarded in 1951.
During the period from 1920 to 1945, then, we see little that resembles
mathematical research in Québec. This was also the case at the École poly-
technique, though that institution remained central to mathematics in the
province under the direction of Augustin Frigon. But even though research
was still not taking place, the formation of a mathematical community was
occurring.
Mathematics in scientific societies, 1923-1945.104 In addition to the Société
mathématique de Québec, founded at Quebec by Pouliot in 1923, the Société
de Mathématiques et d’astronomie du Canada was established at Montreal in
April of the same year.105 Although there are traces of activities from that
date until 1942, this group only met regularly in the period from 1925 to 1932,
during which it organized four to six lectures a year which were intended for
a general audience, notably teachers.106 At the peak of its activity it had
34 members, of whom the most active came from the École polytechnique.
These included André-V. Wendling, Lorenzo Brunotto, and Beaupré, as well
as Léveillé from the Université de Montréal. The lectures don’t seem to have
provoked any real scientific interaction or discussion, with the possible excep-
tion of one by Léveilleé given in November 1927 which provoked responses from
Beaupré and Jules Poivert (another polytechnicien) at the December meeting.
After a decade of inactivity, Brunotto tried in 1942 to reactivate the society,
which was eventually reorganized under the name Société mathématique de
Montréal in 1944 with 15 members. This society, as well as Pouliot’s SMQ,
103
Chartrand, Duchesne, Gingras (1987) pp. 257–260 and Galarneau, Claude
(1978) p. 221-228.
104
For an overview of the activities of mathematical associations in Québec, see
Richard, Guy W. (1982).
105
The name Société de Mathématiques et d’Astronomie du Canada might lead to
confusion. The founders hoped that mathematicians and astronomers from elsewhere
in Canada would eventually join the society. In reality, the society stayed essentially
Montréal-based, and even francophone.
106
Minutes of the Société de Mathématiques et d’Astronomie, in the Archives de
l’École polytechnique, box 999-303-87 (24). The minutes cover the first 32 meetings,
the last of which took place on the 16 of August 1931. Activities continue after that
date, as Guy W. Richard (1982) has stressed. Unfortunately he does not mention
his sources.
both became members of ACFAS, the umbrella group, founded in 1923.107

After a decade of little activity, ACFAS began organizing large numbers of
lectures in the 1930s, and from 1933 has had an annual meeting. Between 1933
and 1945, there were 32 communications in mathematics at these meetings,
the mathematical level of which was higher than that given at the meetings
of the Montreal or Quebec societies. These were nevertheless intended as ex-
pository, and hence did not reach a research level. Pouliot gave 12 of these
papers, and his colleague Althéod Tremblay gave eight, with participation by
others from the universities and the École polytechnique.
Conclusion
In the years 1935–1945 there are distinct signs of research mathematics be-
ginning to come to Canada. Synge returned to Toronto in 1930 as the head
of a new Department of Applied Mathematics, which later included Alexan-
der Weinstein and Leopold Infeld. In addition, the Nuremberg Laws brought
the first refugee mathematician of what would later be a large and produc-
tive group: Richard Brauer came to Toronto in 1935. Brauer’s appointment
was apparently made at the suggestion of Emmy Noether, as Robinson re-
ports.108 However, Robinson also reports that “Our chairman was anxious
to build up the department, and the suggestion was immediately accepted”,
while Morawetz notes “It is hard to imagine today the struggle to make that
appointment”.109 This is just one example of history that this brief article
has not been able to unravel. In addition to the refugee influx–in part un-
willing, as enemy aliens arrested in Britain and transported–we see a general
growth in interest in both pure and applied mathematics across the country
after the war. Young mathematicians began to leave the country for math-
ematical study. Their return to teaching posts, their research activity, and
the founding of the Canadian Mathematical Society/Société mathématique
du Canada brought new perspectives to mathematics in Canada. Mathemat-
ics began to develop in an independent fashion, with new contacts with the
world mathematics community.
Bibliography
Adams, Daniel. Adams’ new arithmetic, suited to Halifax currency, in which
the principles of operating by numbers are analytically explained and synthet-
ically applied; thus combining the advantages of the inductive and synthetic
mode of instructing. The whole made familiar by a great variety of useful and
107
For a history of ACFAS, see Gingras, Yves (1994).
108
Robinson (1979), p. 41.
109
Morawetz (1993). p. 14.
interesting examples, calculated at once to engage the pupil in the study, and
to give him a full knowledge of figures in their application to all the practical
purposes of life. Designed for use in the schools and academies in the British
provinces. Stanstead, L.C. Walton and Gaylord (1883). A second edition was
published at Sherbrooke in 1849.
Anand, Kailash K. “Canadian women Mathematicians from the Early Nine-
teenth Century to 1960–A More Comprehensive Study”. CMS Notes, 21 no.
5, 1989, 31-42.
Anonymous, Nouveau Traité abrégé de la Sphère d’après le système de Coper-
nic, par demandes et par reponses, Nouvelle édition a l’Usage du Séminaire de
Nicolet, Trois-Rivières: Ludger Duvernay Imprimeur, 1824. The 1829 edition
is a l’usage des Ecoles de cette Province)
Anonymous (attributed to Jean Langevin), Traité élémentaire de calcul différ-
entiel et de calcul intégral, Québec: Imprimerie d’Aug. Côté et Cie, 1848.
Anonymous, Théorie élémentaire des nombres d’après Buler, Legendre, Gauss
et Cauchy, 1er fascicule, Montréal: Eugene Senécal, 1870.
Audet, Louis-Philippe, Histoire de l’enseignement au Québec, 2 vol., Montréal,
1971.
Baillairgé, Charles, Nouveau traité de géométrie et de trigonometrie rectiligne
et sphérique. Suivi du toisé des surfaces et des volumes et accompagné de
tables de logarithmes des nombres et sinus, etc. naturels et logarithmiques et
d’autres tables utiles. Ouvrage théorique et pratique illustré de plus de 600
vignettes, avec un grand nombre d’exemples et de problèmes a l’usage des
Arpenteurs, Architectes, Ingénieurs, Professeurs et Élèves, Etc., Québec : C.
Darveau, 1866.
Bibaud, Michel, L’arithmétique en quatre parties, savoir : l’arithmétique vul-
gaire, l’arithmétique marchande, l’arithmétique scientifique, l’arithmétique
curieuse, suivie d’un précis sur la tenue des livres de comptes, principalement
pour ceux qui veulent apprendre l’Arithmétique d’eux-même et sans Maı̂tre,
ou s’y perfectionner, Montréal: Nahum Mower, 1816.
Bibaud, Michel, L’arithmétique a l’usage des écoles élémentaires du Bas-
Canada, Montréal Workman & Bowman, 1832. Nouvelle impression en 1847.
Burke, Edmund, Thèses de mathématique et de physique qui seront soutenues
au Séminaire de Québec sans lieu d’édition, sans date, mais entre 1786 et 1790.
Bouthillier, Jean Antoine, Traité d’arithmétique pour l’usage des écoles, Qué-
bec: John Neil-son, 1809. Une seconde édition en 1829 et une neuvième et
dernière édition en 1864. (Voir Lavoie, Paul (1994), la Bibliographie)
Cameron, Christina, Charles Baillairgé, Architect & Engineer, Montréal,
Kingston, 1989.
Charbonneau, Louis, “L’abbé Jérôme Demers (1774-1853), ‘L’homme qui lit
dans les astres’,” Bulletin AMQ (Association mathématiques du Québec),
décembre 1983, pp. 4–6.
Charbonneau, Louis, “L’enseignement des mathématiques dans les collèges

classiques du Québec au XIXe siécle”, Bulletin AMQ (Association mathéma-
tiques du Québec), mai 1984, pp. 41–44 et octobre 1984, pp. 29–34.
Charbonneau, Louis, “Les mathématiques a Montréal, 1920-1960. Bulletin
AMQ May 1988, pp. 8–13.
Chartrand, Luc, Duchesne Raymond, Gingras, Yves, Histoire des sciences au
Québec, Montréal, 1987.
Corbett, E. A. Henry Marshall Tory, Beloved Canadian. Ryerson Press, 1954.
Craig, G. M. “John Strachan” in Dictionary of Canadian Biography, v. IX
(1861 to 1870), pp. 751–766. Toronto, University of Toronto Press,
Dainville, Francois de, “L’enseignement scientifique dans les collèges des
Jésuites”, in Taton, R., Enseignement et diffusion des sciences en France au
XVIIIe siècle, Paris, 1964, pp. 27–65.
F.É.C., Nouveau traité d’Arithmétique décimale, contenant toutes les opé ra-
tions ordinaires du calcul, les fractions, la racine carrée, etc. enrichie de 1316
problémes a résoudre, pour servir d’exercice aux élèves, Paris, Lille, 1833.
F.É.C., Nouveau traité d’arithmétique: contenant toutes les opérations ordi-
naires du calcul, les fractions et les différentes reductions de fractions, les
règles de trois, d’intérêt, de société, d’alliage, l’extraction des racines, les
principes pour mesurer les surfaces et la solidité des corps; enrichi de 400
problèmes a résoudre, pour servir d’exercice aux éléves: a l’usage des écoles
chrétiennes des frères, Montréal,: C.P. Leprehon, 1838.
F.É.C., Traite d’arithmétique contenant toutes les opérations ordinaires du
calcul, les fractions, l’extraction des racines, les principes pour mesurer les
surfaces et la solidité des corps, enrichi d’un grand nombre de problèmes a
résoudre, pour servir d’exercice aux éléves: a l’usage des écoles chrétiennes,
Montréal, 1842.
F.É.C., Solutions des problèmes avec leurs réponses du Traité d’arithmétique
a l’usage des écoles chrétiennes, Montréal, 1843.
Fessenden, Arthur. Tables, showing the interest at six per cent, of any sum
from 1 pound to 1000 pounds, from one day to one hundred days and from
one month to twelve months. Montreal, Nahum Mower, 1822. Editions: 1830,
Montreal, printed by Workman and Bowen for E.C. Tuttle; 1832, Halifax, C.S.
Belcher, 1837, extended to 365 days; 1841.
Fields, J.C. “The Royal Society of Canada and Canadian Mathematics”, in
Fifty Years Retrospect, Royal Society of Canada, 1932.
Gagnon, Robert. Histoire de l’École polytechnique de Montréal, 1873-1990 La
Montée des lngémieurs, Montreal, 1991.
Galarneau, Claude, Les Collèges classiques au Canada français, Montréal,
1978.
Galarneau, Claude, “L’enseignement des sciences au Québec et Jérôme De-

mers (1765-1835)”, Revue de l’Université d’Ottawa, vol. 47, no. 1-2, 1977, pp.
84–94.
Gingras, Yves. Les Origines de la recherche scientifique au Canada: le cas des
physiciens. Montréal, Borèal, 1991.
Gingras, Yves. Pour l’avancement des sciences: histoire de l’ACFAS, 1923-
1993, Montréal, 1994.
Gosselin, D., Les étapes d’une classe au Petit Séminaire de Québec, 1859-1868,
Québec, 1908.
Gouinlock, G. and J. A complete system of practical arithmetic, for the use of
schools in British America, to which are added, a set of book-keeping by single
entry, and a practical illustration of mental arithmetic, federal money, re-
ceipts, bills of exchange, inland and foreign, explanation of commerical terms,
etc. The whole adapted to the business of real life, to the circumstances of the
Country, and to the present improved state of commerce. Hamilton, Ontario.
J. Ruthven, 1842, 215pp.
Harris, Robin S. A History of Higher Education in Canada, 1663-1960.
Toronto, University of Toronto Press, 1976.
Jarrell, R.A., “The Rise and Decline of Science at Quebec, 1824-1844”, Social
History / Histoire sociale, vol. 10, no. 1, 1977, pp. 77–91.
Karpinski, Louis. Bibliography of Mathematics in America through 1850, Ann
Arbor, University of Michigan Press and London, Humphrey Milford, Oxford
University Press, 1940.
Kerr, D.G.G., A Historical Atlas of Canada, Don Mills, 1966.
Ladreyt, Casimir, Nouvelle arithmétique raisonnée ou cours complet de calcul
théorique et pratique, a l’usage des collèges et des maisons d’éducation de
l’un et de l’autre sexe, des personnes qui veulent apprendre cette science en
peu de temps et sans le secours d’un maı̂tre, et de celles qui veulent se livrer
au commerce; suivi de quelques leçons sur la plannimétrie et la stéréométrie
(arpentage et cubage), ou toisé des surfaces et des volumes, Montréal: s.d.,
1836.
Lamonde, Yvan, La philosophie et son enseignement au Québec (1665-1920),
Montréal: HMH, 1980.
Lavoie, Paul, Contribution a une histoire des mathématiques scolaires au
Québec: L’arithmétique dans les écoles primaires (1800-1920), Thèse, Uni-
versité Laval, Québec, 1994.
Lessard, C., Le Séminaire de Nicolet, 1803-1969, Trois-Rivières: Editions du
Bien Public, 1980.
Lortie, Léon, “Les mathématiques de nos ancêtres”, Mémoire de la Société
royale du Canada, t. XLIX, troisiéme série, Section 1, juin 1955.
Martineau, Armand, “Programme des études au Canada durant la période
1760-1790”, Revue de l’Université d’Ottawa, vol. 37, no. 2, 1967, pp. 206–230.
Morawetz, Cathleen. “The early history of applied mathematics in Canada”.

CMS Notes, 25, 12–15, 1993.
Ouellet, Damelle. Adrien Pouliot, un homme en avance sur son temps,
Montréal, 1986.
Peyrard, F., Les oeuvres d’Euclide, traduites littéralement d’après un manu-
scrit grec très-ancien, resté inconnu jusqu à nos jours, Paris: Chez C.-F. Pa-
tris, 1819.
Phillips, William. A new and concise system of arithmetic, calculated to fa-
cilitate the improvement of youth in upper Canada. Published by subscription
under the patronage of his excellency Sir John Colborne, K.C.B., Lt. Governor
of the Province of Upper Canada, Inc. Under the partronage of the honourable
and venerable the Archdeacon of York, the reverend Dr. Harris, Principal of
the Collège, & c. York, [1832] Eastwood and Skinner.
Provost, Honorius, “Documents pour une histoire du Séminaire de Québec
(Suite), CXXXII–Renseignement sur le Séminaire, 28 décembre 1836,” Revue
de l’Université Laval, vol. XIII, n 7, mars 1959, pp. 665–669.
Robinson, Gilbert de B. The Mathematics Department in the University of
Toronto, Toronto, University of Toronto Press, 1979.
Sauri, Institutions mathématiques, servant d’introduction a un cours de philo-
sophie, a l’usage des universités de France, Paris, quatrième edition, 1786.
Simson, Robert, Euclid, Elements, First 6 books, with the 11th and 12th,
Euclid’s Data, 25th edition, London, 1841.
Strachan, John. Concise introduction to practical arithmetic: for the use of
schools. Montreal, 1809.
Struik, Dirk J., “Mathematicians at Ticonderoga,” The Scientific Monthly,
vol. 82, no. 5. May, 1956, pp. 236–240.
Struik, Dirk J., “Mathematics in Colonial and Early Republican America”,
in Men and Institutions in American Mathematics, Graduate Studies, Texas
Tech. university, No. 13, 1976.
The University of Toronto and Its Collèges, Toronto, The University Library,
1906.
Varkaris, Jane and Costas, Nathan Fellowes Dupuis, Professor and Clock-
maker of Queen’s University, and his Family. Toronto, Ontario Genealogical
Society
Walkingame, Francis, The Tutor’s Assistant: Being a Compendium of Arith-
metic, and Complete Question-book . . . to Which is Added A compendium of
Book-keeping, 51st edition, Montréal: Nahum Mower, 1818.
Wallace, William Stewart. The MacMillan Dictionary of Canadian Biography,
Toronto, MacMillan, 1963.
Wallace, William Stewart, et al. The Royal Canadian Institute Centenary
Volume, 1849-1949. Toronto, Royal Canadian Institute, 1949.
Acknowledgments
L.C. would like to thank the personnel of the Archives de l’Université Laval,
the Archives de l’École polytechnique and the Archives de l’Université de
Montréal for their speed and assistance in searching their respective collec-
tions.
T.A. gratefully acknowledges the support of SSHRC during the conduct of
the research which led to this paper, and thanks the National Museum of
American History, Smithsonian Institution, for kind hospitality. He also wishes
to thank Peter Fillmore for provoking this work, and Louis Charbonneau
for agreeing to make available his expertise and the results of many years
of research. Thanks are also due to Jan Marontate for encouragement and
assistance, and to the editors of this volume [8] for the opportunity to republish
this still-provisional work.
7
The Emergence of the American Mathematical
Research Community
∗
Karen Hunger Parshall
Departments of History and Mathematics, University of Virginia
Various factors affected the emergence of a community of mathematical re-

searchers in the United States in the closing quarter of the nineteenth cen-
tury. In our book on this subject, David Rowe and I focus principally on the
roles of three men—James Joseph Sylvester, Felix Klein, and Eliakim Hastings
Moore—the institutions in which they worked—the Johns Hopkins University,
Göttingen University, and the University of Chicago—and their productions—
both mathematical and organizational—in the formation of this community.
Largely through their work and efforts, we argue, American mathematicians
united in common cause to create and disseminate research-level mathematics.
James Joseph Sylvester, a sixty-two-year-old British algebraist seemingly
well past his prime in 1876; Felix Klein, a rising German geometer recover-
ing from nervous exhaustion in the mid-1880s; and Eliakim Hastings Moore,
a young and unproven American mathematician in the early 1890s—what
could these three men, separated by generation, mathematical training, and
cultural background, possibly have to do with the emergence of a mathe-
matical research community in the United States between 1876 and 1900?1
∗
This text was originally given as a talk in June of 1994 and drew extensively
from the manuscript then in press of Karen Hunger Parshall and David E. Rowe,
The Emergence of the American Mathematical Research Community, 1876–1900:
J. J. Sylvester, Felix Klein, and E. H. Moore, HMATH, vol. 8 (Providence: American
Mathematical Society and London: London Mathematical Society, 1994). It presents
an overview of the argument of that book.
The bibliography given here is necessarily abbreviated. For the complete list of
the sources upon which the book was based, see Parshall and Rowe, pp. 455-485.
Much additional work on the American mathematical scene in the nineteenth and
early twentieth centuries has been done in the decade since 1994; it is not reflected
in the bibliography presented here.
I thank the American Mathematical Society for permission to publish the present
text.
1
The substance of this and the next eight paragraphs closely follows the argument
in the preface of ibid., pp. ix-xv. Here, and in what follows, for the full range of
184 Karen Hunger Parshall
Sylvester, after all, worked as an actuary from 1844 to 1855 and taught at the
Royal Military Academy in Woolwich from 1855 until his forced retirement
in 1870. In neither of these posts did he have an opportunity to train his own
countrymen, much less aspiring Americans, in research-level mathematics.2
Indeed, training at the research level did not even form part of the univer-
sity mission in nineteenth-century Britain. Felix Klein taught and conducted
his mathematical research in Germany: first at Erlangen, next at Munich,
then at Leipzig, and finally at Göttingen. These institutions and their respec-
tive mathematical traditions were not only geographically remote from late
nineteenth-century America but also intellectually far-removed from a country
where institutions of higher education functioned primarily at a collegiate—as
opposed to a university—level and where basic mathematical research received
little encouragement.3 Finally, Eliakim Hastings Moore, thirty years old when
the University of Chicago opened its doors in 1892 with him as acting head
of its Department of Mathematics, had received reasonably solid training at
Yale, had studied abroad, and had even done a bit of original—if unexciting—
research. His own development into a major researcher was, however, by no
means assured. He had never taught students at the graduate level, and his
youth, inexperience, and Midwestern vantage point all seemed to militate
against his chances of becoming a major voice in a community of mathe-
maticians which, insofar as it was discernible at all, had begun to coalesce
around the fledgling East Coast undertaking—the New York Mathematical
Society—founded with six members in 1888.4
On the surface, then, these three men would appear to be unlikely protag-
onists in the story of the emergence of an American mathematical research
community. Moreover, the period in which their impact on American math-
ematics was most immediate and decisive, the last quarter of the nineteenth
century, would seem too early for the detection of significant contributions to
higher mathematics from a country known more for its “Yankee ingenuity”
than for the cultivation of abstract ideas. Nevertheless, a confluence of his-
torical trends and events made this disparate trio the formative figures in the
creation of a community of mathematical researchers on American shores in
the years from 1876 to 1900. Moreover, the analysis of these trends and events
fills a conspicuously large gap in the literature on the history of American sci-
ence.
In 1986 the book, Historical Writing on American Science: Perspectives
and Prospects, appeared, owing to an initiative taken by the History of Science
Society. Representing the collective effort of over a dozen specialists, this work
aimed both to survey the various areas of the history of American science
sources upon which the argument and material draw, see the notes on the cited
pages of ibid.
2
On Sylvester’s early career, see ibid., pp. 59-75.
3
On Klein’s career trajectory, see ibid., pp. 167-187.
4
For an account of Moore’s biography, see ibid., pp. 279-285.
7 The Emergence of the American Mathematical Research Community 185
and to suggest fruitful avenues for further research.5 In the volume’s pref-
ace, editors Sally Gregory Kohlstedt and Margaret Rossiter explained that
by the mid-1980s “[s]everal Americanists were ready to assess the current
state of various specialties and to indicate what ‘needs and opportunities’ re-
mained after more than a decade of significant activity.”6 In their collective
assessment, institutional history along with science in medicine, religion, and
the federal government constituted the four so-called “classical themes”; the
“newer areas” of native American scientific knowledge in addition to science
and technology, war, and public policy received special attention; and the
history of the scientific specialties of geology, astronomy, chemistry, biology,
physics, and the social sciences were singled out for analysis.7 Notably absent
from the specialties treated? Mathematics.
To be sure, the history of mathematics in general and the history of Ameri-
can mathematics in particular have been relatively neglected in the last several
decades by the American community of historians of science. The “glory days”
of the 1930s and 1940s when George Sarton declared the primacy of mathe-
matics within the history of science from his lofty positivist heights have long
since passed. “[T]he history of mathematics should really be the kernel of the
history of culture,” he wrote in 1937. “Take the mathematical developments
out of the history of science, and you suppress the skeleton which supported
and kept together all the rest. Mathematics gives to science its innermost
unity and cohesion, which can never be entirely replaced with props and but-
tresses or with roundabout connections, no matter how many of these may
be introduced.”8 Yet despite such pronouncements, even during Sarton’s era,
the history of American mathematics and, in fact, the history of American
science failed to satisfy the prescripts of a generation of internalist historians
of science which largely adhered to a “great name” approach to the discipline.
Since the 1950s and in response initially to the diverse points of view re-
flected in the work of Alexandre Koyré and Thomas Kuhn, among others,
historians of science have increasingly broadened their purview to embrace
issues like the impact of philosophical and religious ideas on science, the role
of external, social factors in the development of scientific thought, and the
interrelations between science and society at large.9 This changed historio-
5
Sally Gregory Kohlstedt and Margaret W. Rossiter, Historical Writing on
American Science: Perspectives and Prospects (Baltimore: The Johns Hopkins Uni-
versity Press, 1985).
6
Ibid., p. 7.
7
Ibid., pp. 9-15.
8
George Sarton, The Study of the History of Mathematics (Cambridge, MA:
Harvard University Press, 1937; reprint ed., New York: Dover Publications, Inc.,
1957), p. 4.
9
See, for example, Alexandre Koyré, Études galiléennes, 3 parts, 1935–1939;
reprinted in one volume (Paris: Hermann, 1939); and From the Closed World to the
Infinite Universe (Baltimore: The Johns Hopkins University Press, 1957); as well
as Thomas Kuhn, The Structure of Scientific Revolutions (Chicago: The University
graphical climate has encouraged a new interest in American science, and

especially in those sciences perceived as impinging most upon society. Medi-
cine, biology, and physics (particularly in the twentieth century with relativity
theory and the atomic bomb)—sciences like these lend themselves naturally
to this sort of analysis. Mathematics, however, with its abstruse language
and arcane symbolism, with its seemingly insulated practitioners, and with
its unapparent impact on public policy or daily life, seems uncongenial—and
so uninteresting—within such an historiographical framework.10 While this
explanation may shed some light on the relative indifference toward the his-
tory of American mathematics from the 1970s through the 1980s, a period
characterized by Kohlstedt and Rossiter as one of “significant activity” in the
history of American science, it fails to illuminate the deeper aspects of the
relation of the subdiscipline of the history of mathematics to the history of
science as a whole.
In a 1990 position paper published in History of Science, Ivor Grattan-
Guinness stated the case bluntly: the history of mathematics has been largely
ignored, making it one of the least developed subdisciplines within the his-
tory of science.11 In his view, historians of science have failed to treat the
historical development both of mathematics per se and of mathematics as re-
lated to other sciences due to a fundamental fear of the subject, their mostly
empty discussion of notions like “mathematization” aside. As he put it, “His-
torians of Science, like most of the population, do not like mathematics, or
at least find nothing particularly interesting or appealing in it.”12 As for the
other constituency that might have advanced the subdiscipline, the math-
ematicians, Grattan-Guinness contended that “[e]ven those . . . who become
somewhat interested in history usually assert its importance only for trivial
reasons of anecdote and general heuristic without consideration of basic ques-
tions of historiography. Further and more importantly,” he continued, “they
usually view history as the record of a ‘royal road to me’—that is, an account
of how a particular modern theory arose out of older theories instead of an ac-
count of those older theories in their own right.”13 While, as he fully admitted,
counterexamples to this general assessment of the recent historiography of the
of Chicago Press, 1962). For a concise and cogent discussion of these developments
within the history of science, see Allen G. Debus, Science and History: A Chemist’s
Appraisal (Coimbra: Serviço de Documentação e Publicações da Universidade de
Coimbra, 1984), pp. 17-33.
10
This refers to the naı̈ve perception of pure mathematics. It goes without saying
that applied probability and statistics have affected public policy in key and obvious
ways. Perhaps due to their relatively short histories in the United States, not even
these areas of the history of American mathematics have received much attention
from historians of science.
11
Ivor Grattan-Guinness, “Does the History of Science Treat of the History of
Science? The Case of Mathematics,” History of Science 28 (1990):149-173.
12
Ibid., p. 155.
13
Ibid., p. 157.
history of mathematics certainly exist, they are indeed rare. As he claimed,

the neglect of the historians of science coupled with the mostly ahistorical
approach of the mathematicians have rendered the history of mathematics “a
classical example of a ghetto subject: too mathematical for historians and too
historical for mathematicians.”14
Grattan-Guinness’s analysis of the state of the history of mathematics
within the history of science in particular, and within history in general, goes
far to explain the absence of a discussion of the history of mathematics in
Kohlstedt and Rossiter’s book. In light of the book’s prefatory remarks, the
omission of mathematics from this otherwise competent study of the histori-
ography of American science thus raises several obvious questions. Was there
no one “ready to assess the current state” of the history of American math-
ematics? Were there no “needs and opportunities” for historical research in
this scientific discipline? Was there simply no history of American mathemat-
ics to survey in the mid-1980s? Or, perhaps, were the historians of American
science merely ignoring the case of mathematics?
Like the history of American astronomy, which did receive treatment in
Historical Writing on American Science, the history of American mathemat-
ics—although just as clearly “underdeveloped”—was not nonexistent in the
mid-1980s.15 In the massive, three-volume collection of original and reprinted
essays published in 1988 and 1989 to commemorate the centenary of the Amer-
ican Mathematical Society, Uta Merzbach provided much of Kohlstedt and
Rossiter’s “missing chapter” on mathematics.16 Although she treated the pe-
riod from 1969 to the late 1980s cursorily, Merzbach discussed in some detail
the work done from 1890, when Florian Cajori published his book, The Teach-
ing and History of Mathematics in the United States, through the 1930s, when
David Eugene Smith and Jekuthiel Ginsburg produced A History of Mathe-
matics in America before 1900, and up to the 1960s, when Clifford Truesdell
and Kenneth O. May conceived journals to promote and encourage high-level
research in the history of mathematics irrespective of geographical and po-
litical boundaries.17 While much of the research Merzbach surveyed did not
deal exclusively with the history of American mathematics, her article made
clear that there was a small but extant body of research which, if surveyed,
14
Ibid., p. 158.
15
Marc Rothenberg, “History of Astronomy,” pp. 117-131 on p. 131 in Kohlstedt
and Rossiter.
16
Uta C. Merzbach, “The Study of the History of Mathematics in America: A
Centennial Sketch,” in A Century of Mathematics in America–Part III, ed. Peter
Duren et al. (Providence: American Mathematical Society, 1989), pp. 639-666.
17
Florian Cajori, The Teaching and History of Mathematics in the United States
(Washington, D. C.: Government Printing Office, 1890); and David Eugene Smith
and Jekuthiel Ginsburg, A History of Mathematics in America before 1900 (Chicago:
Mathematical Association of America, 1934; reprint ed., New York: Arno Press Inc.,
1980). Truesdell founded the Archive for History of Exact Sciences while May began
Historia Mathematica.
would have highlighted precisely those questions and areas begging for fur-
ther study and analysis. Moreover, a closer look at the works she cited—such
as the volume by Smith and Ginsburg—underscored the fact that the exist-
ing historical literature on American mathematics suffered from a paucity of
archival sources, a near total absence of substantive discussions of the math-
ematics actually produced by the Americans, and a failure to situate this
research within the broader context of the history either of mathematics or
of American science. The book, The Emergence of the American Mathemati-
cal Research Community, 1876–1900, thus aims to lay a solid foundation for
further research by defining and documenting one crucial process and one
key period in the history of American mathematics, namely, the emergence of
a mathematical research community in the United States between 1876 and
1900.
The notion of periodization inherent here is central to the argument. To
focus on the period from 1876 to 1900 explicitly draws the boundaries of two
other periods in the historical development of American mathematics. In the
first period, the century from 1776 to 1876, mathematics evolved not as a sep-
arate discipline but rather within the context of the general structure-building
of American—as opposed to colonial—science.18 The colleges formed a pri-
mary locus of scientific activity, but, by and large, they did little to encourage
the pursuit of research for the advancement of science. At the same time,
however, the concept of research in American science—as in other academic
disciplines—emerged as scientists looked toward Europe as their model and
measured themselves against the yardstick of European scientific achievement.
Indeed, the Americans could point to the work of Nathaniel Bowditch in ce-
lestial mechanics; the scientific accomplishments at the United States Coast
Survey of Superintendents Ferdinand Hassler, Alexander Dallas Bache, and
Benjamin Peirce; the influence of engineering mathematics at West Point with
its curriculum modeled on that of the École polytechnique; and the astronomi-
cal research of Simon Newcomb and George William Hill. However, few of the
achievements associated with these institutions and individuals had any last-
ing effect on the generation that followed in the last quarter of the nineteenth
century.
Quite simply, prior to 1876, nothing even remotely resembling a math-
ematical research community existed in the United States, nor did the time
appear ripe for its imminent emergence. Rather, the century from 1776 to 1876
witnessed the formation of an American scientific community, which, loosely
characterized, earned its living primarily through undergraduate teaching (al-
though to some extent also through federal-governmentally-supported jobs)
but which defined itself in terms of its extracurricular research.19 General sci-
18
For a fuller discussion of this period, see Parshall and Rowe, pp. 1-51.
19
For more on the issues discussed in this and the next paragraph, see Parshall
and Rowe, pp. 33-49. See also John C. Greene, American Science in the Age of Jef-
ferson (Ames: The Iowa State University Press, 1984); George H. Daniels, American
Science in the Age of Jackson (New York: Columbia University Press, 1968); and
entific societies and their publications, like the American Association for the
Advancement of Science and the National Academy of Sciences, provided the
communications outlets for the scientific community, since critical numbers
of practitioners of the individual sciences did not yet exist to sustain spe-
cialized societies or journals.20 On the educational front, colleges broke from
the confines of the colonial era by expanding their faculties with scientists
and their curricula with the sciences.21 Concomitantly, the traditional math-
ematics curriculum, which had largely been restricted to Euclid’s Elements,
incorporated pedagogical innovations issuing mostly from France and began
to include the calculus, among other topics. These changes within higher edu-
cation, however, did not imply the existence of institutional support for or an
encouragement of basic scientific research. In fact, the lack of support for re-
search within the institutions of higher education fundamentally distinguishes
the periods before and immediately after 1876.
Before that time, when it was fostered at all, research was promoted pri-
marily within the federal government—in agencies like the Coast Survey and
the Nautical Almanac Office—but only exceptionally within the colleges.22 As
a result, the research done had, by and large, an applied flavor. The accom-
plishments of Hill and others notwithstanding, American mathematics, as it
unfolded after 1876, had little in common with the research “tradition” of the
previous era. The next generation—associated with institutions of higher ed-
ucation in which departmental structures discouraged the kind of cooperation
generally needed for applied research—focused its attention almost exclusively
on the pure side of the mathematical spectrum, rather than pursuing areas like
celestial mechanics as had Bowditch, Benjamin Peirce, and Hill. Moreover, its
leading figures reinforced their mathematical predilections by forging a viable
community during this period, which successfully incorporated research-level
mathematics into the intellectual fabric of the country.
As noted, three men and the institutions within which they worked largely
shaped this second period: the Englishman James Joseph Sylvester, at The
Johns Hopkins University; the German Felix Klein, first from Leipzig but
more crucially from Göttingen; and the American Eliakim Hastings Moore,
at the University of Chicago.23 Indeed, their respective periods of involve-
Robert V. Bruce, The Launching of Modern American Science: 1846–1876 (New

York: Alfred A. Knopf, 1987).
20
See, for example, Sally Gregory Kohlstedt, The Formation of the American
Scientific Community: The American Association for the Advancement of Science
1848–60 (Urbana: University of Illinois Press, 1976).
21
Stanley Guralnick made this argument in “The American Scientist in Higher Ed-
ucation: 1820–1910,” pp. 99-141 in The Sciences in the American Context: New Per-
spectives, ed. Nathan Reingold (Washington: Smithsonian Institution Press, 1979).
22
On the government’s role in the support of science, see A. Hunter Dupree,
Science in the Federal Government: A History of Policies and Activities (Baltimore:
The Johns Hopkins University Press, 1986).
23
This paragraph follows Parshall and Rowe, pp. xiv-xv.
ment in American mathematics clearly define three distinct phases centered

around three separate programs. An analysis of these programs and of the
nature of the influence of their progenitors reflects not only the progressive
deepening of research standards and output in the United States throughout
the twenty-five-year period but also the generational differences separating
Sylvester, Klein, and Moore. It also highlights, by focusing on the students
and research issuing from these institutions, the process of maturation of an
American mathematical research community which had fully emerged by 1900.
A British import, Sylvester assumed the first professorship in mathematics
in 1876 at The Johns Hopkins University, an institution pivotal in the history
of higher education in America and, by extension, in the history of American
mathematics.24 Unlike other, older colleges and universities in the United
States, Hopkins emphasized graduate education, although it did not neglect
the important function of the undergraduate college as a sort of feeder into its
graduate programs. Having as two of its primary goals the training of future
researchers and the maintenance of high levels of research productivity among
its faculty, this institution departed radically from the traditional model of
the undergraduate teaching college.
At Hopkins, Sylvester was able to continue his own lines of inquiry unfet-
tered by the heavy teaching loads and burdensome duties that encumbered
most of his colleagues around the country.25 Motivated—and often talented—
students accepted graduate fellowships and came to Baltimore determined to
do original work under his guidance. In mathematics, that work tended to re-
flect Sylvester’s interests in the theories of invariants, partitions, and algebras,
but geometry and mathematical logic also found their respective proponents
in William Story, whom Sylvester had stolen from Harvard to be his Teach-
ing Associate, and in the mathematician-logician-philosopher Charles Peirce.
Likewise, the work tended to reflect Sylvester’s own spontaneous approach to
mathematical research. In conducting his so-called Mathematical Seminarium,
Sylvester challenged his students to fill in the details—big and small—in his
presentations and then to present their findings before the group in polished
form. Not all of the research produced in this environment was exciting; not
all of it was important; but most of it was solid; and some of it was genuinely
remarkable. Of real importance, however, is not so much the quality of the
research as the evidence of the conviction, shared by the members of the De-
partment of Mathematics (and by the University as a whole for that matter),
of the importance and primacy of the production of new knowledge. Thanks
24
On The Johns Hopkins University and its importance for the development of
research-level mathematics in the United States, see Parshall and Rowe, pp. 53-
58 and 75-94. See also Hugh Hawkins, Pioneer: A History of the Johns Hopkins
University, 1874–1889 (Ithaca: Cornell University Press, 1960).
25
On the research done at The Johns Hopkins during Sylvester’s tenure there, see
Karen Hunger Parshall, “America’s First School of Mathematical Research: James
Joseph Sylvester at The Johns Hopkins University, 1876–1883,” Archive for History
of Exact Sciences 38 (1988):153-196, and Parshall and Rowe, pp. 99-146.
to a singular intellectual environment and a common drive towards an ideal

of research, the Department of Mathematics at The Johns Hopkins defined
America’s first school of research mathematics.
After Sylvester returned to Britain in 1883 to take up the Savilian Chair
of Mathematics at Oxford’s New College, prospective American mathemati-
cians suffered from the absence, on American shores, of mathematical in-
struction comparable to that available on the Continent.26 While a dozen or
so schools—among them, Harvard, Yale, Princeton, Cornell, and the Univer-
sities of Virginia, North Carolina, Texas, Michigan, and Wisconsin—offered
programs ostensibly at the graduate level, none of these institutions yet had a
staff of researchers teaching or working at the frontiers of mathematics.27 The
Hopkins-trained students emerged from their mathematically charged environ-
ment only to find themselves unable to maintain their research momentum in
other home institutions. Sylvester himself recognized and commented on this
deplorable state of affairs in his farewell address to the Hopkins community.
Rhetorically questioning his audience, he asked:
What happens to them? They are absorbed by inferior though valu-
able colleges and institutions, and their work droops. They write to
me or to their friends, “We miss the stimulus of the Johns Hopkins.”
What a great thing it would be if means were found for providing trav-
eling scholarships or Fellowships for a year or two, that they might
prolong their studies, and come in contact with scientific men and
science in England and on the Continent of Europe.28
Of course, Sylvester offered only a stopgap solution to the problem. His stu-
dents would have returned from their trips abroad only to face the same math-
ematical isolation. For example, one of his students, William Durfee, took a
professorship of mathematics at teaching-intensive Hobart College in Geneva,
New York in 1884, became Dean in 1888, and dropped from the research
ranks. A similar fate befell Sylvester’s two number-theoretically-oriented stu-
dents. Oscar Mitchell accepted a position at another small college, Marietta
College in Ohio, and George Ely became an examiner at the United States
Patent Office. As late as 1888, Sylvester’s last student, Ellery William Davis,
was the Mathematics Department at the University of South Carolina and
26
This and the next paragraph follow Parshall and Rowe, pp. 144-145.
27
See Cajori for descriptions of these various programs, and Karen Hunger Par-
shall, “A Century-Old Snapshot of American Mathematics,” The Mathematical In-
telligencer 12 (3) (1990):7-11.
28
Remarks of Prof. Sylvester at a Farewell Reception Tendered to him by the
Johns Hopkins University, Dec. 20, 1833 (Reported by Arthur S. Hathaway),” 24
typescript pages, Daniel Coit Gilman Papers Ms. 1, Special Collections Division,
The Milton S. Eisenhower Library, The Johns Hopkins University. For the quote,
see p. 12.
tried, but failed, to institute a graduate program there before moving on to

somewhat greener pastures at the University of Nebraska in 1893.29
To get a sense of what it was like for these students, consider the case
of Sylvester’s student Arthur Hathaway at Cornell. By the standards of the
day, Cornell supported a huge mathematics faculty at seven strong, counting
all ranks from professor to instructor.30 The leader of this group, James Ed-
ward Oliver, had graduated in 1849 from a midcentury Harvard dominated
mathematically by Benjamin Peirce and had returned in the mid-1850s to pur-
sue Peirce’s advanced course in mathematics through the Lawrence Scientific
School.31 Earning his living in Cambridge at the federally supported Nautical
Almanac Office, Oliver entered the academic ranks in 1871 when he accepted
the assistant professorship of mathematics at Cornell. Only two years later,
he assumed the mathematical chair and directed Cornell mathematics from
this vantage point until his death in 1895. A modest and unassuming man,
Oliver pursued mathematics more for his own pleasure than for the reputa-
tion publication might have brought, yet he was fully attuned to the growing
importance and desirability of publication within the emergent mathematical
community. In Oliver’s words,
[w]e [at Cornell] are not unmindful of the fact that by publishing more,
we could help to strengthen the university, and that we ought to do so
if it were possible. Indeed, every one of us five is now preparing work
for publication or expects to be doing so this summer, but such work
progresses very slowly because the more immediate duties of each day
leave us so little of that freshness without which good theoretical work
can not be done.32
The duty that sapped their energies most completely was teaching.
During the 1886–1887 academic year, Oliver, together with his four
colleagues, Associate Professor Lucien Augustus Wait, Assistant Professor
George William Jones, and Instructors James McMahon and Arthur Hath-
away, taught an average of seventeen to twenty hours each week. The next
year two more Instructors, Duane Studley and George Egbert Fisher, joined
the staff, but this 40% increase in the teaching faculty hardly lessened the
burden. According to Oliver, “our department’s whole teaching force, com-
posed of only about one-eleventh of all resident professors, has to do about
one-ninth of all the teaching in the University.”33
In spite of this load, which Oliver clearly viewed as inequitable, he and
his colleagues managed to make a fair showing publication-wise in 1887–1888.
Oliver seemed quite proud to report that
29
Parshall, “A Century-Old Snapshot of American Mathematics,” pp. 9-10.
30
On the situation at Cornell, see ibid., and Parshall and Rowe, pp. 269-271.
31
For an idea of the advanced nature of Peirce’s curriculum, see Cajori, pp. 137-
138, reproduced as Table 1.1 in Parshall and Rowe, p. 50.
32
Cajori, p. 180.
33
Ibid., p. 186.
Professor Oliver has sent two or three short articles to the [Annals of
Mathematics], and has read, at the National Academy [of Sciences’]
meeting in Washington, a preliminary paper on the Sun’s rotation,
which will appear in the Astronomical Journal. Professor Jones and
Mr. Hathaway have lithographed a little Treatise on Projective Geom-
etry. Mr. McMahon has sent to the [Annals] a note on the circular
points at infinity, and has also sent to the Educational Times, Lon-
don, solutions (with extensions) of various problems. Other work by
members of the department is likely to appear during the summer,
including a new edition of the Treatise on Trigonometry.34
The latter work comprised part of the popular series of textbooks by Oliver,
Wait, and Jones designed primarily for use in the college classroom. Thus,
the Cornell faculty, although perhaps more active in textbook writing than in
original research, was nonetheless alive mathematically. In Oliver’s view, only
a sufficiently high level of vitality would successfully attract that increasingly
desirable entity—the graduate student—to the department. Apparently, he
and his colleagues attained the necessary level, for their program attracted
eleven graduate students in the 1887–1888 academic year.
Relatively speaking, then, the situation was positively rosy at Cornell, but
Cornell was the exception rather than the rule. In the 1880s, the United States
simply did not yet support the critical mass of mathematicians necessary
for the sustenance of a specialized research community, and with Sylvester’s
departure to England, it no longer had the means to train such a commu-
nity’s membership effectively. With unencouraging educational prospects at
home, Americans turned to Europe, and particularly to the lecture halls of
Felix Klein, for their mathematical training. For roughly a decade following
Sylvester’s departure, Klein actively served as the mathematical standard-
bearer for the United States. Why was Klein the main conduit for the sudden
transfusion of abstract mathematics in the German style that so decisively
enlivened the fledgling community of American mathematicians? An under-
standing of this requires penetrating beyond domestic factors and external
causes to the man himself and the unusually rich sources that defined and
shaped his career.35
Klein’s mathematics embodied many of the ideals characteristic of Ger-
man scholarship in the nineteenth century.36 Even from his youth, he sought
to attain a unified conception of mathematical knowledge that embraced the
34
Ibid., p. 181. See James Oliver, Lucien Wait, and George Jones, A Treatise on
Trigonometry, 4th ed. (Ithaca: G. W. Jones, 1890).
35
On Klein and his influence in training American mathematicians, see Parshall
and Rowe, pp. 147-259.
36
This and the next paragraph follow Parshall and Rowe, pp. 147-148. See also
David E. Rowe, “The Early Geometrical Works of Sophus Lie and Felix Klein,” in
The History of Modern Mathematics, ed. David E. Rowe and John McCleary, 2 vols.
(Boston: Academic Press, 1989), 1:209-273.
achievements of his predecessors. To this end, he strove for and attained an

extraordinary breadth of knowledge, much of which he acquired in discussions
with colleagues and friends. A master of give-and-take, he cultivated scien-
tific relations with many of the leading mathematicians of his day and then
imparted the ideas so gained to the students in his lecture courses. Further-
more, he freely shared the hard-won insights which allowed him to capture the
essence of a mathematical theory. Klein’s approach clearly met with success,
for by the age of thirty, he had already begun to attract talented students
from outside of Germany, many of whom—like Maxime Bôcher and William
Fogg Osgood at Harvard—went on to prominent positions as scholars and
leaders within their respective scientific communities.
In contrast to Sylvester, whose mathematical style was dominated by com-
plex computations and directed towards fairly restricted problems within spe-
cialized branches of algebra, Klein tended to soar above the terrain that oc-
cupied ordinary workaday mathematicians to take in vast expanses of math-
ematical knowledge. His one glaring weakness, as Richard Courant once put
it, was that he often found it difficult to land his plane.37 Klein had no pa-
tience for thorny problems that required abstruse technical arguments. For
him, what counted was the “big picture,” and he drew it largely from the
work of his forerunners—Gauss, Abel, Galois, Riemann, and Weierstrass—
as well as from leading contemporary figures, including Lie, Schwarz, and
Dedekind. Throughout the course of his career, Klein made important contri-
butions to geometry, group theory, Riemannian function theory, Galois theory,
rigid-body mechanics, and even general relativity theory. In his own mind at
least, all of this seemingly disparate work was of a piece, and, what is more,
he ultimately viewed it as largely embedded within the mathematical tradi-
tion associated with Göttingen University. Thus, to learn mathematics from
Felix Klein meant gaining an overview of a substantial part of the deepest
parts of nineteenth-century mathematics as put forth by some of its foremost
proponents. Furthermore, it meant participating in the challenging and highly
competitive seminar setting.
Klein generally organized his seminars by choosing a subject closely related
to the content of his lecture courses.38 In sharp contrast to Sylvester, Klein
tended to orchestrate every detail of his seminar in advance and to unveil his
overall plan during the seminar’s initial session. At that time, too, he discussed
his goals and assigned the specific topics that he wanted his individual students
to research and present in a series of formal lectures. Before the student could
step to the podium, however, he or she had to meet with Klein in order
to discuss the content of the lecture. Although himself a born pedagogue,
Klein believed that good teaching was an acquired characteristic that could
be inculcated in properly receptive minds. Motivated by this belief as well
37
Richard Courant, “Felix Klein,” Die Naturwissenschaften 37 (1925):765-772 on
p. 772.
38
This paragraph follows Parshall and Rowe, p. 191.
as by his deeply ingrained sympathy for holistic solutions, he conducted his

seminars in accordance with the principle that the activities of teaching and
learning must be viewed as integrally related, and his students, both American
and otherwise, imbibed this philosophy.
Unlike Sylvester, however, who trained American students in the United
States, Klein taught in Germany.39 His influence on the emergent mathe-
matical research community in the United States was nevertheless far more
pervasive than that of Sylvester. No fewer than six of Klein’s students went
on to become President of the American Mathematical Society, and thirteen
served as Vice President.40 None of these Americans, though, perpetuated
the distinctly Kleinian approach to mathematics. In later years, in fact, Klein
expressed some bitterness that the research he and his contemporaries had
cultivated, and which had captivated the attention of his American students
in Göttingen, had fallen out of fashion. This had partly to do with the in-
tensely personal nature of Klein’s mathematical vision. As one commenta-
tor remarked, “few of Klein’s . . . contemporaries were willing to assimilate
his singularly personal methods. . . . Klein’s mathematics demanded too much
knowledge of too many things for mastery in a reasonable time, and in addi-
tion it frequently presupposed a facility in spacial [sic] linguistics beyond the
capacities of most mathematicians.”41
In sum, Klein’s influence on his American students had less to do with his
specific research program than with his general ability to inspire them and to
train them to do mathematical research. Indeed, to many Americans, Klein
represented an emissary of mathematical culture at large, and it was that cul-
ture that they very much wanted to transplant to the United States. In 1893,
E. H. Moore and his Chicago colleagues helped secure Klein as the keynote
speaker at the Mathematical Congress associated with the World’s Columbian
Exposition in order to bring that avatar of mathematical culture directly to
American shores. Klein’s appearance at this event, however, foreshadowed not
only his desire to step back from his involvement in the American scene but
also the appearance of E. H. Moore as the third formative figure in research
mathematics in the United States.42
Together with his colleagues, Oskar Bolza and Heinrich Maschke, Moore
established an environment at Chicago conducive to the training of research-
level mathematicians.43 Like Sylvester at Hopkins, Moore worked within an
39
This and the next paragraph follow Parshall and Rowe, p. 253-254.
40
See Table 5.3 in Parshall and Rowe, p. 259.
41
Eric Temple Bell, The Development of Mathematics (New York: McGraw-Hill,
1945), pp. 511-512.
42
On the Mathematical Congress at the World’s Columbian Exposition, see Par-
shall and Rowe, pp. 295-330. See also Karen Hunger Parshall and David E. Rowe,
“Embedded in the Culture: Mathematics at the World’s Columbian Exposition of
1893,” The Mathematical Intelligencer 15 (2) (1993):40-45.
43
On Moore and the Chicago program in mathematics, see Karen Hunger Par-
shall, “Eliakim Hastings Moore and the Founding of a Mathematical Community in
institutional setting which greatly facilitated this achievement. At the under-

graduate level, he helped fashion a program which would not only expose
the student body to mathematics as part of general culture but also prepare
the mathematically inclined for more advanced work at the graduate level.
For those already at this higher stage, he worked to put together a broad
range of courses designed to bridge the gap between studying mathematics
and doing mathematical research. As researchers, the Chicago mathematicians
embraced many of the leading areas of late nineteenth- and early twentieth-
century mathematics.44 Moore, who had started his career working in alge-
braic geometry, had switched into group theory by the 1890s, and had moved
into axiomatics and the foundations of analysis by the turn of the century. His
colleague, Bolza, initially focused on the theories of hyperelliptic and elliptic
integrals and functions, spending the 1890s in such pursuits before shifting
into the calculus of variations after 1901. Finally, Maschke had first devoted
his energies to the theory of finite linear groups before taking up the invariant
theory of differential forms from 1900 until his death in 1908. All three math-
ematicians shared their work and their new ideas with their advanced stu-
dents and guided them to open problems. Yet, it was E. H. Moore who, sniff-
ing the changing mathematical winds, uncannily and successfully shifted his
mathematical course and brought generations of students—including Leonard
E. Dickson, Oswald Veblen, Robert L. Moore, and George D. Birkhoff—along
in his wake. Finally, at the post-doctoral level, he and his colleagues strove to
heighten research activity both within the department, through the regular
meetings of their Mathematical Club (an institution reminiscent of Klein’s
seminar), and nationally, through their involvement particularly in the Amer-
ican Mathematical Society.45
Sylvester, Klein, and Moore, looking at these men within their respective
institutional settings allows for an examination of the environmental factors
that encouraged them to train students at the research level as well as to pur-
sue their own research objectives. Furthermore, an analysis of the training of
this first research-oriented generation of American mathematicians provides
an illuminating look at the different styles these men employed in training
young men and women to do mathematical research as well as at the re-
search so produced. Viewing the students’ research output against the back-
drop of late-nineteenth-century mathematics highlights the great strides made
in mathematical research in the United States between 1876 and 1900. In a
key shift that partially delineates the first and second periods in the history
of American mathematics, academic institutions in the United States, led by
America, 1892–1902,” Annals of Science 41 (1984):313-333, and Parshall and Rowe,

pp. 275-294 and 363-372.
44
On the research that issued from the Chicago department in its early years, see
Parshall and Rowe, pp. 372-401.
45
On the broader community activiism of the Chicago department, see Parshall,
“Eliakim Hastings Moore and the Founding of a Mathematical Community in Amer-
ica,” and Parshall and Rowe, pp. 401-419.
The Hopkins, slowly adopted and adapted a research ethic that had become
firmly entrenched in German higher education early in the nineteenth century.
This translated into an increasing emphasis on—and an increasing production
of—research as an officially sanctioned and supported endeavor in the emer-
gent university setting and, by intimate association, in the emergent American
mathematical community.
Although the training of mathematicians at the research level represented
a critical ingredient in the emergence of this community, the formation of
a community—an interacting group of people linked by common interests—
required more than just advanced training in mathematics.46 Various organi-
zational activities proved crucial in forging the requisite communications links.
In 1878, Sylvester founded, under the auspices of The Johns Hopkins Univer-
sity, the American Journal of Mathematics, the oldest research-level mathe-
matics journal in the United States; and in 1899, E. H. Moore established the
Transactions of the American Mathematical Society. Neither of these journals
suffered for lack of high-quality material for publication, in contradistinction
to the numerous failed attempts at research-level mathematical periodicals be-
fore 1876.47 As already mentioned, the American Mathematical Society was
founded as the New York Mathematical Society in 1888 and grew from a mem-
bership of six to over two hundred in three years. The first major mathematics
meeting was organized by E. H. Moore and his Chicago colleagues in 1893 as
the Mathematical Congress of the Chicago World’s Columbian Exposition.
On that occasion, Felix Klein served not only as the keynote speaker but also
as an official cultural emissary of the Prussian government at the fair. He
brought with him contributed papers from some of the most influential math-
ematicians in Germany, and the meeting attracted an audience totaling some
forty-five Americans. Following the Congress, Klein proceeded to Northwest-
ern University, where the same Chicago contingent had organized the so-called
Evanston Colloquium, the first research-level colloquium on American shores,
again with Klein as the featured speaker.48 There, he gave a two-week-long
series of ten lectures attended by some two dozen specialists or specialists-to-
be, in which he surveyed the mathematical landscape of the late nineteenth
century and propagandized for his own unique vision of and approach to the
subject. As a result of the late-nineteenth-century conjunction of these and
other innovations and their innovators, of changed attitudes as to the value
and desirability of research at both an individual and institutional level, and
of the existence of a critical mass of practitioners, an American mathematical
research community had emerged by 1900.
46
For this definition, see Parshall and Rowe, p. xvi.
47
On these journals, see Parshall and Rowe, pp. 88-94 and 411-415, respectively.
A list of America’s failed journals appears in Table 1.2 in ibid., p. 51.
48
For more on the Evanston Colloquium and Klein’s role in it, see Parshall and
Rowe, pp. 331-361.
As this last statement suggests, the community of research-level mathe-

maticians clearly extended beyond the immediate spheres defined by Sylvester,
Klein, and Moore.49 However, using their careers as a point of departure, it is
possible to examine meaningfully a broad spectrum—although by no means
the complete roster—of participants in the field at the turn of the twentieth
century. An identification of their students and, to some extent, their stu-
dents’ students and colleagues penetrates beyond the crème de la crème to
underscore the existence of an extended population of mathematicians—some
talented and some not, some well-known and some obscure—based at colleges
and universities throughout the country who actively shared an interest in
mathematics at the research level.
In summary, three factors that had been altogether absent before 1876
serve to define the emergent period of research-level mathematics in Amer-
ica.50 The first—the founding of research-oriented universities beginning with
The Johns Hopkins in 1876, followed by Clark University in 1889, and cul-
minating with the University of Chicago in 1892—reflected a fundamental
change in the American academic climate. Within these new institutions, re-
search in pure mathematics as well as in other fields attained a degree of
credibility and support previously lacking. Moreover, the existence and ex-
ample of these schools prompted the older institutions—Harvard, Columbia,
Yale, and Princeton—to develop viable graduate programs during this era or
soon thereafter. It had quickly become clear to leading American educators,
many of them inspired by the German university model, that graduate train-
ing and research went hand in hand, and those institutions that pursued this
policy soon established a competitive edge within a fast-changing academic
environment.
As a measure of the rapidity of the transformation that ensued, consider
the following statistics on the number of doctoral degrees in mathematics
earned by Americans between 1875 and 1900.51 Before 1875, American uni-
versities had conferred a total of only six degrees in the field. During the next
fifteen years, thirty-nine Americans took doctorates in the United States, and
another fifteen earned their degrees abroad. These figures were dwarfed again
by those of the final decade of the century, which witnessed a total of 107 new
Ph.D.s in mathematics, eighty-four of them earned at home.
A second factor crucial in the emergence of the American mathematical
research community during the last quarter of the nineteenth century was
the founding of the New York Mathematical Society in 1888. In the wake of
the Chicago Congress and the prototypic Evanston Colloquium, the organi-
zation assumed national dimensions and changed its name to the American
49
Compare Parshall and Rowe, pp. xvi-xvii.
50
This and the next four paragraphs follow Parshall and Rowe, pp. 429-431.
51
R. G. D. Richardson, “The Ph.D. Degree and Mathematical Research,” Ameri-
can Mathematical Monthly 43 (1936):199-215; reprinted in A Century of Mathemat-
ics in America–Part II, ed. Peter Duren et al. (Providence: American Mathematical
Society, 1989), pp. 361-378 on p. 366.
Mathematical Society (AMS) in 1894. Spurred by E. H. Moore and Henry

White, a group of Midwestern mathematicians established the Chicago Sec-
tion of the AMS in 1897. Following the Chicago lead, twenty West Coast
Society members—including Sylvester’s student W. Irving Stringham and the
Klein-trained Mellon W. Haskell—met and founded a San Francisco Section
of the Society in May of 1902. By December of 1906, the Göttingen Ph.D.
Earle Raymond Hedrick, the Russian-born Alexander Chessin, and another
Sylvester student, Ellery Davis, among others, had met in Columbia, Missouri
and had formed a Southwestern Section initially numbering some thirty-five
strong. Thus, from the mid-1890s onward, the AMS served as the principal
organizational vehicle for meetings, colloquia, and other activities of interest
to the budding community of research mathematicians across the country. By
the turn of the century, it had also initiated its Transactions as a complement
both to its Bulletin (begun in 1891) and to the two older, research-oriented
periodicals, the American Journal of Mathematics and the Annals of Mathe-
matics, the latter founded by astronomer Ormond Stone at the University of
Virginia in 1884. Thus, by the end of this pivotal period, the United States had
already developed sufficient research capacity to support four major mathe-
matical journals.
While it would be difficult to overestimate the role the American Math-
ematical Society played in promoting mathematical research in the United
States, its early success depended on a third factor: a generation of Americans
not only interested in mathematics but also possessing the requisite knowl-
edge to educate their successors. The older members of this group necessarily
turned to European scholars, notably J. J. Sylvester and Felix Klein, for their
training and inspiration. Whether they studied in Baltimore, Göttingen, Paris,
or Leipzig, this generation went on to establish a new standard for mathemat-
ics instruction at colleges and universities in every region of the United States.
Led by E. H. Moore at Chicago, a few select institutions—most notably Har-
vard and Princeton—also developed solid doctoral programs, the graduates
of which shaped and directed American mathematics in the opening decades
of the twentieth century.52
These three factors interacted to transform what amounted to a few scat-
tered pockets of mathematical expertise into a cohesive and extensive math-
ematical community. Quantitatively speaking, from 1891 to 1906 or during
the first fifteen years of publication of the American Mathematical Society’s
Bulletin, over 1,000 geographically dispersed participants at various levels
of interest and activity formed a pyramidal community broadly based upon
more than 500 interested—if not active—participants and tapering to an apex
of some sixty highly active and productive researchers.53 Within this commu-
nity, those most active in research spanned the landscape of pure mathematics
52
For a sketch of these developments, see Parshall and Rowe, pp. 432-453.
53
See Della Dumbaugh Fenster and Karen Hunger Parshall, “A Profile of the
American Mathematical Research Community, 1891–1906,” pp. 228-261 in The His-
from algebra to analysis to geometry, while tending to favor the subdisciplines

of group theory, the theory of automorphic functions, the calculus of varia-
tions, real and complex function theory, classical projective, algebraic, and
differential geometry, and foundational studies. (Applied areas, although rep-
resented, held less attraction for turn-of-the-century American research math-
ematicians, due, in part, to the fact that they drew their inspiration from pure
mathematicians, like Sylvester, Klein, and E. H. Moore.) This broad support
of pure mathematics at the research level that solidified from 1876 to 1900
undoubtedly contributed to the sustained activity in American mathemat-
ics that characterized the ensuing period of consolidation and growth from
1900 to 1933. The final quarter of the nineteenth century thus marks a true
watershed in the history of American mathematics. The major institutional
structures and research traditions of American mathematics stood firmly in
place by the end of the first decade of the twentieth century. No dramatic
new qualitative changes affected the community’s overall contours until the
influx of European refugees began in the mid-1930s. This would mark the
beginning of a fourth period in the history of American mathematics—from
1933 to roughly 1960—characterized by this infusion of Europeans as well as
by the large-scale governmental funding of basic research during and after the
Second World War.
Just as internal institutional and external international influences affected
the shift in the historical development of American mathematics from its third
period of quiet consolidation and expansion to the tumultuous era that fol-
lowed, so similar forces had shaped the crucial second period during which a
self-sustaining American mathematical research community emerged.54 Indi-
viduals both at home and abroad, educational institutions both domestic and
foreign, general developments in science and its social and cultural status,
broader philosophies of education, political rivalries, and the encroachment
of modernity in its several guises, these were among the components of the
matrix from which research-level mathematics evolved in the United States
during the last quarter of the nineteenth century.
Selected References
Bell, Eric Temple. The Development of Mathematics. New York: McGraw-
Hill, 1945.
Bruce, Robert V. The Launching of Modern American Science: 1846–1876.
New York: Alfred A. Knopf, 1987.
Cajori, Florian. The Teaching and History of Mathematics in the United
States. Washington, D. C.: Government Printing Office, 1890.
tory of Modern Mathematics, vol. 3, ed. Eberhard Knobloch and David E. Rowe
(Boston: Academic Press, Inc., 1994).
54
This paragraph follows Parshall and Rowe, p. 453.
Courant, Richard. “Felix Klein.” Die Naturwissenschaften 37 (1925):765-

772.
Daniels, George H. American Science in the Age of Jackson. New York:
Columbia University Press, 1968.
Debus, Allen G. Science and History: A Chemist’s Appraisal. Coimbra:
Serviço de Documentação e Publicações da Universidade de Coimbra,
1984.
Dupree, A. Hunter. Science in the Federal Government: A History of Poli-
cies and Activities. Baltimore: The Johns Hopkins University Press, 1986.
Fenster, Della Dumbaugh and Parshall, Karen Hunger. “A Profile of the
American Mathematical Research Community, 1891–1906.” Pp. 228-261.
In The History of Modern Mathematics. Vol. 3. Ed. Eberhard Knobloch
and David E. Rowe. Boston: Academic Press, Inc., 1994.
Grattan-Guinness, Ivor. “Does the History of Science Treat of the History
of Science? The Case of Mathematics.” History of Science 28 (1990):149-
173.
Green, John C. American Science in the Age of Jefferson. Ames: The Iowa
State University Press, 1984.
Guralnick, Stanley M. “The American Scientist in Higher Education:
1820–1910.” Pp. 99-141. In The Sciences in the American Context: New
Perspectives. Ed. Nathan Reingold. Washington, D. C.: Smithsonian In-
stitution Press, 1979.
Hawkins, Hugh. Pioneer: A History of the Johns Hopkins University,
1874–1889. Ithaca: Cornell University Press, 1960.
Kohlstedt, Sally Gregory. The Formation of the American Scientific Com-
munity: The American Association for the Advancement of Science 1848-
60. Urbana: University of Illinois Press, 1976.
Kohlstedt, Sally Gregory and Rossiter, Margaret W. Historical Writing
on American Science: Perspectives and Prospects. Baltimore: The Johns
Hopkins University Press, 1985.
Koyré, Alexandre. Études galiléennes. 3 Parts, 1935–1939; Reprinted in
one volume. Paris: Hermann, 1939.
Koyré, Alexandre. From the Closed World to the Infinite Universe. Balti-
more: The Johns Hopkins University Press, 1957.
Kuhn, Thomas. The Structure of Scientific Revolutions. Chicago: The Uni-
versity of Chicago Press, 1962.
Merzbach, Uta C. “The Study of the History of Mathematics in Amer-
ica: A Centennial Sketch.” Pp. 639-666. In A Century of Mathematics in
America–Part III. Ed. Peter Duren et al. Providence: American Mathe-
matical Society, 1989.
Oliver, James; Wait, Lucien; and Jones, George. A Treatise on Trigonom-

etry. 4th Ed. Ithaca: G. W. Jones, 1890.
Parshall, Karen Hunger. “America’s First School of Mathematical Re-
search: James Joseph Sylvester at The Johns Hopkins University 1876–
1883.” Archive for History of Exact Sciences 38 (1988):153-196.
Parshall, Karen Hunger. “A Century-Old Snapshot of American Mathe-
matics.” The Mathematical Intelligencer 12 (3) (1990):7-11.
Parshall, Karen Hunger, “Eliakim Hastings Moore and the Founding of a
Mathematical Community in America, 1892–1902.” Annals of Science 41
(1984):313-333.
Parshall, Karen Hunger and Rowe, David E. The Emergence of the Amer-
ican Mathematical Research Community, 1876–1900: J. J. Sylvester, Felix
Klein, and E. H. Moore. HMATH. Vol. 8. Providence: American Mathe-
matical Society and London: London Mathematical Society, 1994.
Parshall, Karen Hunger and Rowe, David E. “Embedded in the Culture:
Mathematics at the World’s Columbian Exposition of 1893.” The Mathe-
matical Intelligencer 15 (2) (1993):40-45.
Richardson, R. G. D. “The Ph.D. Degree and Mathematical Research.”
American Mathematical Monthly 43 (1936):199-215; Reprinted in A Cen-
tury of Mathematics in America–Part II. Ed. Peter Duren et al. Provi-
dence: American Mathematical Society, 1989. Pp. 361-378.
Rothenberg, Marc. “History of Astronomy,” Pp. 117-131. In Historical
Writing on American Science: Perspectives and Prospects. Ed. Sally Gre-
gory Kohlstedt and Margaret Rossiter. Baltimore: The Johns Hopkins Uni-
versity Press, 1985.
Rowe, David E. “The Early Geometrical Works of Sophus Lie and Felix
Klein.” In The History of Modern Mathematics. Ed. David E. Rowe and
John McCleary. 2 Vols. Boston: Academic Press, Inc., 1989, 1:209-273.
Sarton, George. The Study of the History of Mathematics. Cambridge,
MA: Harvard University Press, 1937; Reprint ed., New York: Dover Pub-
lications, Inc., 1957.
Smith, David Eugene and Ginsburg, Jekuthiel. A History of Mathemat-
ics in America before 1900. Chicago: The Mathematical Association of
America, 1934; Reprint ed, New York: Arno Press Inc., 1980.
Sylvester, James Joseph. “Remarks of Prof. Sylvester at a Farewell Re-
ception Tendered to him by the Johns Hopkins University, Dec. 20, 1833
(Reported by Arthur S. Hathaway).” 24 typescript pages. Daniel Coit
Gilman Papers Ms. 1. Special Collections Division. The Milton S. Eisen-
hower Library. The Johns Hopkins University.
8
19th Century Logic Between Philosophy and
Mathematics∗
Volker Peckhaus
Universität Paderborn, Fakultät für Kulturwissenschaften – Philosophie
Summary. The history of modern logic is usually written as the history of mathe-
matical or, more general, symbolic logic. As such it was created by mathematicians.
Not regarding its anticipations in Scholastic logic and in the rationalistic era, its
continuous development began with George Boole’s The Mathematical Analysis of
Logic of 1847, and it became a mathematical subdiscipline in the early 20th century.
This style of presentation cuts off one eminent line of development, the philosoph-
ical development of logic, although logic is evidently one of the basic disciplines of
philosophy. One needs only to recall some of the standard 19th century definitions
of logic as, e.g., the art and science of reasoning (Whateley) or as giving the norma-
tive rules of correct reasoning (Herbart). In the paper the relationship between the
philosophical and the mathematical development of logic will be discussed. Answers
to the following questions will be provided:
1. What were the reasons for the philosophers’ lack of interest in formal logic?
2. What were the reasons for the mathematicians’ interest in logic?
3. What did “logic reform” mean in the 19th century? Were the systems of math-
ematical logic initially regarded as contributions to a reform of logic?
4. Was mathematical logic regarded as art, as science or as both?
8.1 Introduction
Most 19th century scholars would have agreed to the opinion that philosophers
are responsible for research on logic. On the other hand, the history of late 19th
century logic indicates clearly a very dynamic development instigated not by
philosophers, but by mathematicians. The central feature of this development
was the emergence of what has been called the “new logic”, “mathematical
logic”, “symbolic logic”, or, since 1904, “logistics”. This new logic came from
Great Britain, and was created by mathematicians in the second half of the
19th century, finally becoming a mathematical subdiscipline in the early 20th
century.
∗
First published in Bulletin of Symbolic Logic vol. 5 (1999), 433-450.
204 Volker Peckhaus
Charles L. Dodgson, better known under his pen name Lewis Carroll
(1832–1898), published two well-known books on logic, The Game of Logic
([13]) of 1887 and Symbolic Logic of 1896 ([14]) of which a fourth edition ap-
peared already in 1897. These books were written “to be of real service to
the young, and to be taken up, in High Schools and in private families, as a
valuable addition of their stock of healthful mental recreations” ([14, p. xiv]).
They were meant “to popularize this fascinating subject,” as Carroll wrote in
the preface of the fourth edition of Symbolic Logic ([14, p. xiv]). But, astonish-
ingly enough, in both books there is no definition of the term “logic”. Given
the broad scope of these books the title “Symbolic Logic” of the second book
should at least have been explained.
Maybe the idea of symbolic logic was so widely spread at the end of the
19th century in Great Britain that Carroll regarded a definition as simply
unnecessary. Some further observations support this thesis. They concern a
remarkable interest by the general public in symbolic logic, after the death of
the creator of the algebra of logic, George Boole, in 1864.
Recalling some standard 19th century definitions of logic as, e.g., the art
and science of reasoning (Whately) or the doctrine giving the normative rules
of correct reasoning (Herbart), it should not be forgotten that mathematical or
symbolic logic was not set up from nothing. It arose from the old philosophical
collective discipline logic. The standard presentations of the history of logic
ignore the relationship between the philosophical and mathematical side of its
development; they sometimes even deny that there has been any development
of philosophical logic at all. Take for example William and Martha Kneale’s
programme in their eminent The Development of Logic. They wrote ([32, p.
iii]): “But our primary purpose has been to record the first appearances of
these ideas which seem to us most important in the logic of our own day,”
and these are the ideas leading to mathematical logic.
Another example is J. M. Bocheński’s assessment of “modern classical
logic” which he dated between the 16th and the 19th century. It was for him
a noncreative period in logic which can therefore justly be ignored in a problem
history of logic ([7, p. 14]). According to Bocheński classical logic was only a
decadent form of this science, a dead period in its development ([7, p. 20]).
Such assessments show that the authors adhered to the predominant views
on logic of our time, i.e., actual systems of mathematical or symbolic logic. As
a consequence, they have not been able to give reasons for the final divorce
between philosophical and mathematical logic, because they have ignored the
seed from which mathematical logic has emerged. Following Bocheński’s view
Carl B. Boyer presented a consistent periodization of the development of logic
([11, p. 633]): “The history of logic may be divided, with some slight degree of
oversimplification, into three stages: (1) Greek logic, (2) Scholastic logic, and
(3) mathematical logic.” Note Boyer’s “slight degree of oversimplification”
which enabled him to skip 400 years of logical development and ignore the
fact that Kant’s transcendental logic, Hegel’s metaphysics and Mill’s inductive
logic were called “logic”, too.
8 19th Century Logic Between Philosophy and Mathematics 205
In discussing the relationship between the philosophical and the mathe-

matical development of logic, at least the following questions will be answered:
1. What were the reasons for the philosophers’ lack of interest in formal
logic?
3. What did “logic reform” mean in the 19th century? Were the systems of
mathematical logic initially regarded as contributions to a reform of logic?
4. Was mathematical logic regarded as art, as science or as both?
This paper focuses not only on the situation in Great Britain, but also on the
development in Germany. British logicians of that time regarded Germany as
the logical paragon. John Venn can be regarded as a chief witness. He deplored,
in the second edition of his Symbolic Logic of 1894, the lack of a tradition in
logic in Great Britain which caused problems in creating the collection of
books on logic for the Cambridge University Library ([67, p. 533]):
At the time when I commenced the serious study of Symbolic Logic
many of the most important works which bore on the subject were
not to be found in any of those great libraries in this country to which
one naturally refers in the first place, and could therefore only be
obtained by purchase from abroad. [. . . ] I suppose that the almost
entire abandonment of Logic as a serious academic study, for so many
years in this country at least, had prevented the formation of those
private professorial libraries, the frequent appearance of which in the
market has kept the secondhand booksellers’ shops in Germany so well
supplied with works on this subject.
It should be stressed, however, that when speaking of German logic Venn
wasn’t referring to contemporary German logical sytems, but to the great
18th century rationalistic precursors of the British algebra of logic beginning
with Gottfried Wilhelm Leibniz and ending with the Swiss, Johann Heinrich
Lambert.
In the following sections surveys are given of the philosophical and math-
ematical contexts in which the new logic emerged in Great Britain and Ger-
many. The strange collaboration of mathematics and philosophy in promoting
the new systems of logic will be discussed, and finally answers to the four ques-
tions already posed will be given.
8.2 Contexts
8.2.1 The philosophical context in Great Britain
The development of the new logic started in 1847, completely independent of

earlier anticipations, e.g., by the German rationalist Gottfried Wilhelm Leib-
niz (1646–1716) and his followers (cf. [45]; [48, ch. 5]). In that year the British
206 Volker Peckhaus
mathematician George Boole (1815–1864) published his pamphlet The Math-

ematical Analysis of Logic ([9]). Boole mentioned that it was the struggle for
priority concerning the quantification of the predicate between the Edinburgh
philosopher William Hamilton (1788–1856) and the London mathematician
Augustus De Morgan (1806–1871) which encouraged this study. Hence, he re-
ferred to a startling philosophical discussion which indicated a vivid interest
in formal logic in Great Britain. This interest was, however, a new interest,
not even 20 years old. One can even say that neglect of formal logic could
be regarded as a characteristic feature of British philosophy up to 1826 when
Richard Whately (1787–1863) published his Elements of Logic.1 In his preface
Whately added an extensive report on the languishing research and educa-
tion in formal logic in England. He complained ([69, p. xv]) that only very
few students of the University of Oxford became good logicians and that
by far the greater part pass through the University without knowing
any thing of all of it; I do not mean that they have not learned by
rote a string of technical terms; but that they understand absolutely
nothing whatever of the principles of the Science.
Thomas Lindsay, the translator of Friedrich Ueberweg’s important System
der Logik und Geschichte der logischen Lehren ([63], translation [64]), was
very critical of the scientific qualities of Whately’s book, but he, nevertheless,
emphasized its outstanding contribution for the renaissance of formal logic in
Great Britain ([38, p. 557]):
Before the appearance of this work, the study of the science had fallen
into universal neglect. It was scarcely taught in the universities, and
there was hardly a text-book of any value whatever to be put into the
hands of the students.
One year after the publication of Whately’s book, George Bentham’s An Out-
line of a New System of Logic appeared ([6]) which was to serve as a com-
mentary to Whately. Bentham’s book was critically discussed by William
Hamilton in a review article published in the Edinburgh Review ([20]). With
the help of this review Hamilton founded his reputation as the “first logical
name in Britain, it may be in the world.”2 Hamilton propagated a revival
of the Aristotelian scholastic formal logic without, however, one-sidedly pre-
ferring the syllogism. His logical conception was focused on a revision of the
standard forms by quantifying the predicates of judgements.3 The controversy
1
[53] lists nine editions up to 1848 and 28 further printings to 1908. Van Evra
([66, p. 2]) mentions 64 printings in the USA to 1913.
2
This opinion can be found in a letter of De Morgan’s to Spalding of 26th June,
1857 (quoted in [23, p. xii]) which was, however, not sent. Hamilton was for George
Boole one of the “two greatest authorities in logic, modern and ancient” ([9, p. 81]).
The other authority is Aristotle. This reverence to Hamilton might not be without
irony because of Hamilton’s disregard of mathematics.
3
Cf. [22, vol. 4, p. 287].
about priority arose, when De Morgan, in a lecture “On the Structure of the
Syllogism” ([16]) given to the Cambridge Philosophical Society on 9th Novem-
ber 1846, also proposed quantifying predicates. Neither had any priority, of
course. Application of the diagrammatic methods of the syllogism proposed,
e.g., by the 18th century mathematicians and philosophers Leonard Euler,
Gottfried Ploucquet, and Johann Heinrich Lambert, presupposed quantifica-
tion of the predicate. The German psychologistic logician Friedrich Eduard
Beneke (1798–1854) suggested quantifying the predicate in his books on logic
[4] and [5], the latter of which he sent to Hamilton. In the context of this
paper it is irrelevant to solve the priority question. It is, however, important
that a dispute of this extent arose at all. It indicates, there was new interest
in research on formal logic.
This interest represented only one side of the effect released by Whately’s
book. Another line of research stood in the direct tradition of Humean empiri-
cism and the philosophy of inductive sciences: the inductive logic of John Stu-
art Mill (1806–1873), Alexander Bain (1818–1903) and others. Boole’s logic
was in clear opposition to inductive logic. It was Boole’s follower William
Stanley Jevons (1835–1882; cf. [29]) who made this opposition explicit.
Boole referred to the controversy between Hamilton and De Morgan, but
this influence should not be overemphasized. In his main work on the Laws
of Thought ([10]) Boole went back to the logic of Aristotle by quoting from
the Greek original. This can be interpreted as indicating that the influence of
contemporary philosophical discussion was not as important as his own words
might suggest. In writing a book on logic he was doing philosophy, and it
was thus a matter of course that he related his results to the philosophical
discussion of his time. This does not mean, of course, that his thoughts were
really influenced by this discussion.
8.2.2 The philosophical context in Germany
It seems clear that, in regard to the 18th century dichotomy between German
and British philosophy represented by the philosophies of Kant and Hume,
Hamilton and Boole stood on the Kantian side. There are some analogies
between the situations in Great Britain and Germany, where philosophical
discussion on logic after Hegel’s death was determined by the Kantian influ-
ence. In the preface to the second edition of his Kritik der reinen Vernunft of
1787 ([31]), Immanuel Kant (1723–1804) wrote that logic has followed the safe
course of a science since earliest times. For Kant this was evident because of
the fact that logic had been prohibited from taking any step backwards from
the time of Aristotle. But he regarded it as curious that logic hadn’t taken a
step forward either ([31, B VIII]). Thus, logic seemed to be closed and com-
plete. Formal logic, in Kant’s terminology the analytical part of general logic,
did not play a prominent rôle in Kant’s system of transcendental philosophy.
In any case, it was a negative touchstone of truth, as he stressed ([31, B
84]). Georg Wilhelm Friedrich Hegel (1770–1831) went further in denying any
208 Volker Peckhaus
relevance of formal logic for philosophy ([24, I, Introduction, pp. XV-XVII]).

Referring to Kant, he maintained that from the fact that logic hadn’t changed
since Aristotle one could infer that it needed a complete rebuilding ([24, p.
XV]). Hegel created a variant of logic as the foundational science of his philo-
sophical system, defining it as “the science of the pure idea, i.e., the idea in
the abstract element of reasoning” ([25, p. 27]. Hegelian logic thus coincides
with metaphysics ([25, p. 34]).
This was the situation when after Hegel’s death philosophical discussion
on formal logic in Germany started again. This discussion on logic reform
stood under the label of “the logical question”, a term coined by the Neo-
Aristotelian Adolf Trendelenburg (1802–1872). In 1842 he published a paper
entitled “Zur Geschichte von Hegel’s Logik und dialektischer Methode” with
the subtitle “Die logische Frage in Hegel’s Systeme”. But what is the logical
question according to Trendelenburg? He formulated this question explicitly
towards the end of his article: “Is Hegel’s dialectical method of pure reasoning
a scientific procedure?” ([62, p. 414]). In answering this question in the nega-
tive, he provided the occasion of rethinking the status of formal logic within
a theory of human knowledge without, however, proposing a return to the
old (scholastic) formal logic. In consequence the term “the logical question”
was subsequently used in a less specific way. Georg Leonard Rabus, the early
chronicler of the discussion on logic reform, wrote that the logical question
emerged from doubts concerning the justification of formal logic ([49, p. 1]).
Although this discussion was clearly connected to formal logic, the so-
called reform did not concern formal logic. The reason was provided by the
Neo-Kantian Wilhelm Windelband who wrote in a brilliant survey on 19th
century logic ([70, p. 164]):
It is in the nature of things that in this enterprise [i.e., the reform of
logic] the lower degree of fruitfulness and developability power was on
the side of formal logic. Reflection on the rules of the correct progress
of thinking, the technique of correct thinking, had indeed been brought
to perfection by former philosophy, presupposing a naive world view.
What Aristotle had created in a stroke of genius, was decorated with
the finest filigree work in Antiquity and the Middle Ages: an art of
proving and disproving which culminated in a theory of reasoning,
and after this constructing the doctrines of judgements and concepts.
Once one has accepted the foundations, the safely assembled building
cannot be shaken: it can only be refined here and there and perhaps
adapted to new scientific requirements.
Windelband was very critical of English mathematical logic. Its quantification
of the predicate allows the correct presentation of extensions in judgements,
but it “drops hopelessly” the vivid sense of all judgements, which tend to
claim or deny a material relationship between subject or predicate. It is “a
logic of the conference table”, which cannot be used in the vivid life of science,
a “logical sport” which has, however, its merits in exercising the final acumen
([70, pp. 166–167]).
The philosophical reform efforts concerned primarily two areas:
1. the problem of a foundation of logic which itself was approached by psy-
chological and physiological means, leading to new discussion on the ques-
tion of priority between logic and psychology, and to various forms of
psychologism and anti-psychologism (cf. [50], [33]);
2. the problem of logical applications focusing interest on the methodological
part of traditional logic. The reform of applied logic attempted to bring
philosophy in touch with the stormy development of mathematics and
sciences of the time.
Both reform procedures had a destructive effect on the shape of logic and
philosophy. The struggle with psychologism led to the departure of psychol-
ogy (especially in its new, experimental form) from the body of philosophy
at the beginning of the 20th century. Psychology became a new, autonomous
scientific discipline. The debate on methodology resulted in the creation of
the philosophy of science which was separated from the body of logic. The
philosopher’s ignorance of the development of formal logic caused a third de-
parture: Part of formal logic was taken from the domain of the competence of
philosophy and incorporated into mathematics where it was instrumentalized
for foundational tasks.
8.2.3 The mathematical context in Great Britain
As mentioned earlier, the influence of the philosophical discussion on logic in

Great Britain on the emergence of the new logic should not be overemphasized.
Of greater importance were mathematical influences. Most of the new logicians
can be related to the so-called “Cambridge Network” ([12, pp. 29–71]), i.e., the
movement which aimed at reforming British science and mathematics which
started at Cambridge. One of the roots of this movement was the foundation
of the Analytical Society in 1812 (cf. [17]) by Charles Babbage (1791–1871),
George Peacock (1791–1858) and John Herschel (1792–1871). In regard to
mathematics Joan L. Richards called this act a “convenient starting date for
the nineteenth-century chapter of British mathematical development” ([51, p.
13]). One of the first achievements of the Analytical Society was a revision
of the Cambridge Tripos by adopting the Leibnizian notation for the calculus
and abandoning the customary Newtonian theory of fluxions: “the principles
of pure D-ism in opposition to the Dot-age of the University” as Babbage
wrote in his memoirs ([1, p. 29]). It may be assumed that this successful
movement triggered by a change in notation might have stimulated a new
or at least revived interest in operating with symbols. This new research on
the calculus had parallels in innovative approaches to algebra which were
motivated by the reception of Laplacian analysis. Firstly the development of
symbolical algebra has to be mentioned. It was codified by George Peacock in
210 Volker Peckhaus
his Treatise on Algebra ([42]) and further propagated in his famous report for
the British Association for the Advancement of Science ([43], especially pp.
198–207). Peacock started by drawing a distinction between arithmetical and
symbolical algebra, which was, however, still based on the common restrictive
understanding of arithmetic as the doctrine of quantity. A generalization of
Peacock’s concept can be seen in Duncan F. Gregory’s (1813–1844) “calculus
of operations”. Gregory was most interested in operations with symbols. He
defined symbolical algebra as “the science which treats of the combination of
operations defined not by their nature, that is by what they are or what they
do, but by the laws of combinations to which they are subject” ([18, p. 208]). In
his much praised paper “On a General Method in Analysis” ([8]) Boole made
the calculus of operations the basic methodological tool for analysis. However
in following Gregory, he went further, proposing more applications. He cited
Gregory who wrote that a symbol is defined algebraically “when its laws of
combination are given; and that a symbol represents a given operation when
the laws of combination of the latter are the same as those of the former”
([19, pp. 153–154]). It is possible that a symbol for an arbitrary operation
can be applied to the same operation ([19, p. 154]). It is thus necessary to
distinguish between arithmetical algebra and symbolical algebra which has to
take into account symbolical, but non-arithmetical fields of application. As
an example Gregory mentioned the symbols a and +a. They are isomorphic
in arithmetic, but in geometry they need to be interpreted differently. a can
refer to a point marked by a line whereas the combination of the signs +
and a additionally expresses the direction of the line. Therefore symbolical
algebra has to distinguish between the symbols a and +a. Gregory deplored
the fact that the unequivocity of notation didn’t prevail as a result of the
persistence of mathematical practice. Clear notation was only advantageous,
and Gregory thought that our minds would be “more free from prejudice, if
we never used in the general science symbols to which definite meanings had
been appropriated in the particular science” ([19, p. 158]).
Boole adopted this criticism almost word for word. In his Mathematical
Analysis of Logic of 1847 he claimed that the reception of symbolic algebra
and its principles was delayed by the fact that in most interpretations of
mathematical symbols the idea of quantity was involved. He felt that these
connotations of quantitative relationships were the result of the context of
the emergence of mathematical symbolism, and not of a universal principle
of mathematics ([9, pp. 3–4]). Boole read the principle of the permanence
of equivalent forms as a principle of independence from interpretation in an
“algebra of symbols”. In order to obtain further affirmation, he tried to free
the principle from the idea of quantity by applying the algebra of symbols to
another field, the field of logic. As far as logic is concerned this implied that
only the principles of a “true Calculus” should be presupposed. This calculus is
characterized as a “method resting upon the employment of Symbols, whose
laws of combination are known and general, and whose results admit of a
consistent interpretation” ([9, p. 4]). He stressed (ibid.):
It is upon the foundation of this general principle, that I purpose to

establish the Calculus of Logic, and that I claim for it a place among
the acknowledged forms of Mathematical Analysis, regardless that in
its objects and in its instruments it must at present stand alone.
Boole expressed logical propositions in symbols whose laws of combination
are based on the mental acts represented by them. Thus he attempted to
establish a psychological foundation of logic, mediated, however, by language.
The central mental act in Boole’s early logic is the act of election used for
building classes. Man is able to separate objects from an arbitrary collection
which belong to given classes, in order to distinguish them from others. The
symbolic representation of these mental operations follows certain laws of
combination which are similar to those of symbolic algebra. Logical theorems
can thus be proven like mathematical theorems. Boole’s opinion has of course
consequences for the place of logic in philosophy: “On the principle of a true
classification, we ought no longer to associate Logic and Metaphysics, but
Logic and Mathematics” ([9, p. 13]).
Although Boole’s logical considerations became increasingly philosophical
with time, aiming at the psychological and epistemological foundations of logic
itself, his initial interest was not to reform logic but to reform mathematics.
He wanted to establish an abstract view on mathematical operations without
regard to the objects of these operations. When claiming “a place among the
acknowledged forms of Mathematical Analysis” ([9, p. 4]) for the calculus of
logic, he didn’t simply want to include logic in traditional mathematics. The
superordinate discipline was a new mathematics. This is expressed in Boole’s
writing: “It is not of the essence of mathematics to be conversant with the
ideas of number and quantity” ([10, p. 12]).
8.2.4 The mathematical context in Germany

The results of this examination of the British situation at the time when the
new logic emerged–a reform of mathematics, with initially a lack of interest
in a reform of logic, by establishing an abstract view of mathematics which
focused not on mathematical objects, but on symbolic operations with arbi-
trary objects–these results could be transferred to the situation in Germany
without any problem. The most important representative of the German al-
gebra of logic was the mathematician Ernst Schröder (1841–1902) who was
regarded as having completed the Boolean period in logic (cf. [7, p. 314]).
In his first pamphlet on logic, Der Operationskreis des Logikkalkuls ([55]), he
presented a critical revision of Boole’s logic of classes, stressing the idea of
the duality between logical addition and logical multiplication introduced by
William Stanley Jevons in 1864. In 1890 Schröder started on the large project,
his monumental Vorlesungen über die Algebra der Logik ([56], [57], [58], [59])
which remained unfinished although it increased to three volumes with four
parts, of which one appeared only posthumously. Contemporaries regarded the
first two volumes alone as completing the algebra of logic (cf. [68, p. 196]).
212 Volker Peckhaus
Schröder’s opinion concerning the question as to the end to which logic is

studied (cf. [44], [46]) can be drawn from an autobiographical note, published
in 1901 (and written in the third person), the year before his death. It contains
Schröder’s own survey of his scientific aims and results. Schröder divided his
scientific production into three fields:
1. A number of papers dealing with some of the current problems of his
science.
2. Studies concerned with creating an “absolute algebra,” i.e., a general the-
ory of connections. Schröder stressed that such studies represent his “very
own object of research” of which only little was published at that time.
3. Work on the reform and development of logic.
Schröder wrote ([60]) that his aim was
to design logic as a calculating discipline, especially to give access to
the exact handling of relative concepts, and, from then on, by eman-
cipation from the routine claims of spoken language, to withdraw any
fertile soil from “cliché” in the field of philosophy as well. This should
prepare the ground for a scientific universal language that, widely dif-
fering from linguistic efforts like Volapük [a universal language like
Esperanto, very popular in Germany at that time], looks more like a
sign language than like a sound language.
Schröder’s own division of his fields of research shows that he didn’t consider
himself a logician: His “very own object of research” was “absolute algebra,”
which was similar to modern abstract or universal algebra in respect to its
basic problems and fundamental assumptions. What was the connection be-
tween logic and algebra in Schröder’s research? From the passages quoted one
could assume that these fields belong to two separate fields of research, but
this is not the case. They were intertwined in the framework of his heuristic
idea of a general science. In his autobiographical note he stressed ([60]):
The disposition for schematizing, and the aspiration to condense
practice to theory advised Schröder to prepare physics by perfecting
mathematics. This required deepening–as of mechanics and geometry–
above all of arithmetic, and subsequently he became by the time aware
of the necessity for a reform of the source of all these disciplines, logic.
Schröder’s universal claim becomes obvious. His scientific efforts served to
provide the requirements to found physics as the science of material nature
by “deepening the foundations,” to quote a famous metaphor later used by
David Hilbert ([26, p. 407]) in order to illustrate the objectives of his axiomatic
programme. Schröder regarded the formal part of logic that can be formed as
a “calculating logic,” using a symbolical notation, as a model of formal algebra
that is called “absolute” in its last state of development.
But what is “formal algebra”? The theory of formal algebra “in the narrow-
est sense of the word” includes “those investigations on the laws of algebraic
operations [. . . ] that refer to nothing but general numbers in an unlimited

number field without making any presuppositions concerning its nature” ([54,
p. 233]). Formal algebra therefore prepares “studies on the most varied num-
ber systems and calculating operations that might be invented for particular
purposes” ([54, p. 233]).
It has to be stressed that Schröder wrote his early considerations on formal
algebra and logic without any knowledge of the results of his British prede-
cessors. His sources were the textbooks of Martin Ohm, Hermann Günther
Graßmann, Hermann Hankel and Robert Graßmann. These sources show that
Schröder was a representative of the tradition of German combinatorial alge-
bra and algebraic analysis (cf. [48, ch. 6]).
Like the British tradition, but independent of it, the German algebra of
logic was connected to new trends in algebra. It differed from its British
counterpart in its combinatorial approach. In both traditions, algebra of logic
was invented within the enterprise to reform basic notions of mathematics
which led to the emergence of structural abstract mathematics. The alge-
braists wanted to design algebra as “pan-mathematics”, i.e., as a general dis-
cipline embracing all mathematical disciplines as special cases. The indepen-
dent attempts in Great Britain and Germany were combined when Schröder
learned about the existence of Boole’s logic in late 1873, early 1874. Finally
he enriched the Boolean class logic by adopting Charles S. Peirce’s theory of
quantification and adding a logic of relatives according to the model of Peirce
and De Morgan.
The main interest of the new logicians was to utilize logic for mathematical
and scientific purposes, and it was only in a second step, but nevertheless an
indispensable consequence of the attempted applications, that the reform of
logic came into the view. What has been said of the representatives of the
algebra of logic also holds for the proponents of competing logical systems such
as Gottlob Frege or Giuseppe Peano. They wanted to use logic in their quest
for mathematical rigor, something questioned by the stormy development in
mathematics.
8.3 Accepting the New Logic
Although created by mathematicians, the new logic was widely ignored by fel-
low mathematicians. In Germany Schröder was only known as the algebraist
of logic, and regarded as rather exotic. George Boole was respected by British
mathematicians, but his ideas concerning an algebraical representation of the
laws of thought received very little published reaction. He shared this fate
with Augustus De Morgan, the second major figure of symbolic logic at that
time. In 1864, Samuel Neil, the early chronicler of British mid 19th century
logic, expressed his thoughts about the reasons for this negligible reception:
“De Morgan is esteemed crotchety, and perhaps formalizes too much. Boole
demands high mathematic culture to follow and to profit from” ([41, p. 161]).
214 Volker Peckhaus
One should add that the ones who had this culture were usually not interested
in logic. The situation changed after George Boole’s death in 1864. In the fol-
lowing comments only some ideas concerning the reasons for this new interest
are hinted at. In particular the rôles of William Stanley Jevons and Alexander
Bain are stressed which exemplify “the strange collaboration of mathemat-
ics and philosophy in promoting the new systems of logic” mentioned in the
introduction.
8.3.1 William Stanley Jevons
A broader international reception of Boole’s logic began when William Stanley

Jevons made it the starting point for his influential Principles of Science ([28]).
He used his own version of the Boolean calculus introduced in his Pure Logic
of 1864. Among his revisions were the introduction of a simple symbolical
representation of negation and the definition of logical addition as inclusive
“or”. He also changed the philosophy of symbolism ([27, p. 5]):
The forms of my system may, in fact, be reached by divesting his
[Boole’s] of a mathematical dress, which, to say the least, is not es-
sential to it. The system being restored to its proper simplicity, it
may be inferred, not that Logic is a part of Mathematics, as is almost
implied in Professor Boole’s writings, but that the Mathematics are
rather derivatives of Logic. All the interesting analogies or samenesses
of logical and mathematical reasoning which may be pointed out, are
surely reversed by making Logic the dependent of Mathematics.
Jevons’ interesting considerations on the relationship between mathematics
and logic representing an early logicistic attitude will not be discussed. Similar
ideas can be found not only in Gottlob Frege’s work, but also in that of
Hermann Rudolf Lotze and Ernst Schröder. In the context of this paper,
it is relevant that Jevons abandoned mathematical symbolism in logic, an
attitude which was later taken up by John Venn. Jevons attempted to free logic
from the semblance of being a special mathematical discipline. He used the
symbolic notation only as a means of expressing general truths. Logic became
a tool for studying science, a new language providing symbols and structures.
The change in notation brought the new logic closer to the philosophical
discourse of the time. The reconciliation was supported by the fact that Jevons
formulated his Principles of Science as a rejoinder to John Stuart Mill’s A
System of Logic of 1843, at that time the dominating work on logic and the
philosophy of science in Great Britain. Although Mill called his logic A System
of Logic Ratiocinative and Inductive, the deductive parts played only a minor
rôle, used only to show that all inferences, all proofs and the discovery of
truths consisted of inductions and their interpretations. Mill claimed to have
shown “that all our knowledge, not intuitive, comes to us exclusively from
that source” ([40, Bk. II, ch. I, §1]). Mill concluded that the question as to
what induction is, is the most important question of the science of logic, “the
question which includes all others.” As a result the logic of induction covers
by far the largest part of this work, a subject which we would today regard
as belonging to the philosophy of science.
Jevons defined induction as a simple inverse application of deduction. He
began a direct argument with Mill in a series of papers entitled “John Stuart
Mill’s Philosophy Tested” ([29]). This discourse proved that symbolic logic
could be of importance not only for mathematics, but also for philosophy.
Another effect of the attention caused by Jevons was that British algebra
of logic was able to cross the Channel. In 1877, Louis Liard (1846–1917),
at that time professor at the Faculté de lettres at Bordeaux and a friend of
Jevons, published two papers on the logical systems of Jevons and Boole ([34],
[35]). In [36] he added a booklet entitled Les logiciens anglais contemporaines
which had five editions until 1907, and was translated into German in [37].
Although Herman Ulrici had published a first German review of Boole’s Laws
of Thought as early as 1855 ([65]; cf. [47]), the knowledge of British symbolic
logic was conveyed primarily by Alois Riehl, then professor at the University
of Graz, in Austria. He published a widely read paper “Die englische Logik
der Gegenwart” (“English contemporary logic”) in 1877 ([52]) which reported
mainly Jevons’ logic and utilized it in a current German controversy on the
possibility of scientific philosophy.
8.3.2 Alexander Bain
Finally a few words on Alexander Bain (1818–1903): This Scottish philoso-

pher was an adherent of Mill’s logic. Bain’s Logic, first published in 1870,
had two parts, the first on deduction and the second on induction. He made
explicit that “Mr. Mill’s view of the relation of Deduction and Induction is
fully adopted” ([2, I, p. iii]). Obviously he shared the “[. . . ] general conviction
that the utility of the purely Formal Logic is but small; and that the rules
of Induction should be exemplified even in the most limited course of logical
discipline” ([2, p. v]). The minor rôle of deduction showed up in Bain’s defi-
nition “Deduction is the application or extension of Induction to new cases”
([2, p. 40]).
Despite his reservations about deduction, Bain’s Logic was quite important
for the reception of symbolic logic because of a chapter of 30 pages entitled
“Recent Additions to the Syllogism.” In this chapter the contributions of
William Hamilton, Augustus De Morgan and George Boole were introduced.
Presumably many more people became acquainted with Boole’s algebra of
logic through Bain’s report than through Boole’s own writings. One exam-
ple is Hugh MacColl (1837–1909), the pioneer of the calculus of propositions
(statements) and of modal logic. He created his ideas independently of Boole,
eventually realizing the existence of the Boolean calculus by means of Bain’s
report. Even in the early parts of his series of papers “The Calculus of Equiva-
lent Statements” he quoted from Bain’s presentation when discussing Boole’s
216 Volker Peckhaus
logic ([39]). In 1875 Bain’s logic was translated into French, in 1878 into Pol-
ish. Tadeusz Batóg and Roman Murawski ([3]) have shown that it was Bain’s
presentation which motivated the first Polish algebraist of logic, Stanislaw
Piatkiewicz
(1848–?) to begin his research on symbolic logic.
The remarkable collaboration of mathematics and philosophy can be seen
in the fact that a broader reception of symbolic logic commenced only when
its relevance for the philosophical discussion of the time came to the fore.
8.4 Conclusions
Finally, these are the answers to the initial questions:

1. What were the reasons for the philosophers’ lack of interest in formal
logic?
In Germany philosophers shared Kant’s opinion that formal logic was a
completed field of knowledge. They were interested primarily in the founda-
tions and application of logic. In Great Britain there was hardly any vivid
logical tradition. Philosophy was dominated by empiricist conceptions. New
systems of formal logic therefore had difficulties in gaining a footing in the
philosophical discussion.
Foundational problems and problems in grasping new mathematical ob-
jects forced some mathematicians to look intuitively at the logical foundations
of their subject. The interest in formal logic was thus a result of the dynamic
development of late 19th century mathematics. One should not assume, how-
ever, that this was a general interest. Most mathematicians did not (and still
do not) care about foundations.
3. How did the mathematicians’ logical activities fit into the reform of logic
conceptions of the time?
In Germany in the second half of the 19th century, logic reform meant
overcoming the Hegelian identification of logic and metaphysics. In Great
Britain it meant enlarging the scope of the syllogism or elaborating the phi-
losophy of science. Mathematicians were initially interested in utilizing logic
for mathematical means, or they used it as a language for structuring and sym-
bolizing extra-mathematical fields. Applications were, e.g., the foundation of
mathematics (Boole, Schröder, Frege), the foundation of physics (Schröder),
the preservation of rigour in mathematics (Peano), the theory of probabili-
ties (Boole, Venn), the philosophy of science (Jevons), the theory of human
relationships (Alexander Macfarlane), and juridical questions. The mathe-
maticians’ preference for the organon aspect of formal logic seems to be the
point of deviation between mathematicians and philosophers who were not
interested in elaborating logic as a tool.
4. Was mathematical logic regarded as art or as science?
From the applicational interest it follows that it was mainly regarded as

an art. The scientific aspect grew, however, with the insight into the power
of logical calculi. Nevertheless, in an institutional sense the new logic was
established only in the beginning of the 20th century as an academic subject,
i.e., as an institutionalized domain of science.
References
1. Charles Babbage, Passages from the life of a philosopher, Longman, Green,
Longman, Roberts, & Green, London 1864; reprinted Greeg, Westmead, 1969.
2. Alexander Bain, Logic, 2 volumes, pt. 1: Deduction, pt. 2: Induction, Longmans,
Green, & Co., London, 1870.
3. Tadeusz Batóg and Roman Murawski, Stanislaw Piatkiewicz
and the beginnings
of mathematical logic in Poland, Historia Mathematica, vol. 23 (1996), pp. 68–
73.
4. Friedrich Eduard Beneke, Syllogismorum analyticorum origines et ordinem nat-
uralem, Mittler, Berlin, 1839.
5. , System der Logik als Kunstlehre des Denkens, 2 volumes, F. Dümmler,
Berlin, 1842.
6. George Bentham, An outline of a new system of logic: With a critical exami-
nation of Dr. Whately’s “Elements of logic”, Hunt and Clark, London, 1827;
reprinted Thoemmes, Bristol, 1990.
7. Joseph Maria Bocheński, Formale Logik, Alber, Freiburg/München, 1956 (=
Orbis Academicus, vol. III, no. 2), 4 1978.
8. George Boole, On a general method in analysis, Philosophical Transactions of
the Royal Society of London for the Year MDCCCXLIV, pt. 1 (1844), pp. 225–
282.
9. , The mathematical analysis of logic: Being an essay towards a calculus
of deductive reasoning, Macmillan, Barclay, and Macmillan, Cambridge/George
Bell, London, 1847; reprinted Basil Blackwell, Oxford, 1951.
10. , An investigation of the laws of thought, on which are founded the
mathematical theories of logic and probabilities, Walton & Maberly, London,
1854; reprinted Dover, New York, ca. 1951.
11. Carl B. Boyer, A history of mathematics, John Wiley & Sons, New York/Lon-
don/Sydney, 1968.
12. Susan Faye Cannon, Science in culture: The early Victorian period, Dawson
Scientific History Publications, New York, 1978.
13. Lewis Carroll, The game of logic, MacMillan, London, 1887; reprinted in [15].
14. , Symbolic logic, MacMillan, London, 1896; reprinted in [15].
15. , Mathematical recreations of Lewis Carroll: Symbolic logic and The
game of logic (both books bound as one), Dover Publications and Berkeley En-
terprises, New York, 1958.
16. Augustus De Morgan, On the syllogism I: On the structure of the syllogism, and
on the applications of the theory of probabilities to questions of arguments and
authority, Transactions of the Cambridge Philosophical Society, vol. 8 (1846),
pp. 379–408.
17. Philip C. Enros, The analytical society (1812–1813): Precursor of the renewal
of Cambridge Mathematics, Historia Mathematica, vol. 10 (1983), pp. 24–47.
218 Volker Peckhaus
18. Duncan Farquharson Gregory, On the real nature of symbolical algebra, Trans-
actions of the Royal Society of Edinburgh, vol. 14 (1840), pp. 208–216.
19. , On a diffculty in the theory of algebra, The Cambridge Mathematical
Journal, vol. 3 (1842), pp. 153–159.
20. William Hamilton, Logic. In reference of the recent English treatises on that
science, Edinburgh Review, vol. 66 (April 1833), pp. 194–238; again in [21], pp.
16–174.
21. , Discussions on philosophy and literature, education and univer-
sity reform. Chiefly from the Edinburgh Review; corrected, vindicated, en-
larged, in notes and appendices, Longman, Brown, Green and Longmans, Lon-
don/Maclachlan and Stewart, Edinburgh, 1851.
22. , Lectures on metaphysics and logic, 4 volumes (H. L. Mansel and J.
Veitch, editors), William Blackwood and Sons, Edinburgh/London, 1859–1866.
23. Peter Heath, Introduction, in Augustus De Morgan, On the syllogism and other
logical writings (Peter Heath, editor), Routledge & Kegan Paul, London, 1966
(= Rare masterpieces of philosophy and science), pp. vii–xxxi.
24. Georg Wilhelm Friedrich Hegel, Wissenschaft der Logik, vol. 1: Die objective
Logik, Johann Leonhard Schrag, Nürnberg, 1812/1813; critical edition: Hegel,
Wissenschaft der Logik. Erster Band. Die objektive Logik (Friedrich Hegemann
and Walter Jaeschke, editors), Felix Meiner, Hamburg, 1978 (= Hegel, Gesam-
melte Werke, vol. 11).
25. , Encyclopädie der philosophischen Wissenschaften im Grundrisse.
Zum Gebrauch seiner Vorlesungen. Dritte Ausgabe, Oßwald’scher Verlag, Hei-
delberg, 1830. Critical edition: Hegel, Enzyklopädie der philosophischen Wis-
senschaften im Grundrisse (1830) (Wolfgang Bonsiepen and Hans-Christian
Lucas, editors), Felix Meiner, Hamburg, 1992 (= Hegel, Gesammelte Werke,
vol. 20).
26. David Hilbert, Axiomatisches Denken, Mathematische Annalen, vol. 78 (1918),
pp. 405–415.
27. William Stanley Jevons, Pure logic or the logic of quality apart from quantity
with remarks on Boole’s system and the relation of logic and mathematics, E.
Stanford, Londone, 1864; reprinted in [30], pp. 3–77.
28. , The principles of science. A treatise on logic and scientific method, 2
volumes, Macmillan and Co., London, 1874 [New York, 1875].
29. , John Stuart Mill’s philosophy tested, The Contemporary Review, vol.
31 (1877/78), pp. 167–182 and 256–275; vol. 32 (1878), pp. 88–99; again in [30],
pp. 137–172.
30. , Pure logic and other minor works (Robert Adamson and Harriet
A. Jevons, editors), Macmilland and Co., London/NewYork, 1890; reprinted
Thoemmes Press, Bristol, 1991.
31. Immanuel Kant, Critik der reinen Vernunft, 2nd ed., Johann Friedrich Hart-
knoch, Riga, 1787; again in Kant’s gesammelte Schriften, vol. 3 (Königlich
Preußische Akademie der Wissenschaften, editor), Reimer, Berlin, 1911.
32. William Kneale and Martha Kneale, The development of logic, Clarendon Press,
Oxford, 1968; reprinted 1986.
33. Martin Kusch, Psychologism. A case study in the sociology of philosophical
knowledge, Routledge, London/New York, 1995 (= Psychological issues in sci-
ence).
34. Louis Liard, Un nouveau système de logique formelle. M. Stanley Jevons, Revue
philosophique de la France et de l’Étranger, vol. 3 (1877), pp. 277–293.
35. , La logique algébrique de Boole, Revue philosophique de la France et

de l’Étranger, vol. 4 (1877), pp. 285–317.
36. , Les logiciens anglais contemporains, Germer Baillière, Paris, 1878,
5
1907.
37. , Die neuere englische Logik (J[ohannes] Imelmann, editor), Denicke’s
Verlag, Berlin, 1880, 2 1883.
38. Thomas M. Lindsay, On recent logical speculation in England, in [64], pp. 557–
590.
39. Hugh MacColl, The calculus of equivalent statements (second paper), Proceed-
ings of the London Mathematical Society, vol. 9 (1877–1878), pp. 177–186.
40. John Stuart Mill, A system of logic, ratiocinative and inductive: Being a connec-
tive view of the principles of evidence and the methods of scientific investigation,
2 volumes, J. W. Parker, London, 1843.
41. Samuel Neil, John Stuart Mill, The British Controversialist and Literary Mag-
azine n.s. (1864), pp. 161–173 and 241–256.
42. George Peacock, A treatise on algebra, J. & J. J. Deighton, Cambridge/G. F. &
J. Rivington, London, 1830.
43. , Report on the recent progress and present state of certain branches
of analysis, in Report of the third meeting of the British Association for the
Advancement of Science held at Cambridge in 1833, John Murrary, London,
1834, pp. 185–352.
44. Volker Peckhaus, Ernst Schröder und die ‘pasigraphischen Systeme’ von Peano
und Peirce, Modern Logic, vol. 1 (Winter 1990/1991), no. 2/3, pp. 174–205.
45. , Leibniz als Identifikationsfigur der britischen Logiker des 19. Jahrhun-
derts, in VI. Internationaler Leibniz-Kongreß. Vorträge I. Teil, Hannover, 18.-
22.7.1994, Gottfried-Wilhelm-Leibniz-Gesellschaft, Hannover, 1994, pp. 589–
596.
46. , Wozu Algebra der Logik? Ernst Schröders Suche nach einer univer-
salen Theorie der Verknüpfungen, Modern Logic, vol. 4 (1994), pp. 357–381.
47. , Hermann Ulrici (1806–1884): Der Hallesche Philosoph und die en-
glische Algebra der Logik. Mit einer Auswahl von Texten Ulricis und einer Bibli-
ographie seiner Schriften, Hallescher Verlag, Halle a. S., 1995 (= Schriftenreihe
zur Geistes- und Kulturgeschichte. Texte und Dokumente).
48. , Logik, Mathesis universalis und allgemeine Wissenschaft. Leibniz und
die Wiederentdeckung der formalen Logik im 19. Jahrhundert, Akademie Verlag,
Berlin, 1997 (= Logica Nova).
49. Georg Leonhard Rabus, De neuesten Bestrebungen auf dem Gebiete der Logik
bei den Deutschen und Die logische Frage, Deichert, Erlangen, 1880.
50. Matthias Rath, Der Psychologismus in der deutschen Philosophie, Karl Alber
Verlag, Freiburg/München, 1994.
51. Joan L. Richards, Mathematical visions: The pursuit of geometry in Victorian
England, Academic Press, Boston, etc., 1988.
52. Alois Riehl, Die englische Logik der Gegenwart, Vierteljahrsschrift für wis-
senschaftliche Philosophie, vol. 1 (1877), pp. 51–80.
53. Wilhelm Risse, Bibliographica Logica. Verzeichnis der Druckschriften zur Logik
mit Angabe ihrer Fundorte, vol. 2: 1801–1969, Olms, Hildesheim/New York, 1973
(= Studien und Materialien zur Geschichte der Philosophie; 1).
54. Ernst Schröder, Lehrbuch der Arithmetik und Algebra für Lehrer und Studirende,
vol. 1 [no further volumes published]: Die sieben algebraischen Operationen, B.
G. Teubner, Leibzig, 1873.
220 Volker Peckhaus
55. , Der Operationskreis des Logikkalkuls, Teubner, Leipzig, 1877; re-

printed as “special edition”, Wissenschaftliche Buchgesellschaft, Darmstadt,
1966.
56. , Vorlesungen über die Algebra der Logik (exakte Logik), vol. 1, B. G.
Teubner, Leipzig, 1890; reprinted in [61].
57. , Vorlesungen über die Algebra der Logik (exakte Logik), vol. 2, B. G.
Teubner, Leipzig, 1891; reprinted in [61].
58. , Vorlesungen über die Algebra der Logik (exakte Logik), vol. 3, pt. 1:
Algebra und Logik der Relative, B. G. Teubner, Leipzig, 1895; reprinted in [61].
59. , Vorlesungen über die Algebra der Logik (exakte Logik), vol. 2, pt. 2
(Karl Eugen Müller, editor), B. G. Teubner, Leipzig, 1905; reprinted in [61].
60. , Grossberzoglich Badischer Hofrat Dr. phil. Ernst Schröder[,] ord. Pro-
fessor für Mathematik an der Technischen Hochschule in Karlsruhe i. Baden,
in Geistiges Deutschland. Deutsche Zeitgenossen auf dem Gebiete der Liter-
atur,Wissenschaften und Musik, Adolf Eckstein, Berlin-Charlottenburg, 1901.
61. , Vorlesungen über die Algebra der Logik (exakte Logik), [“second edi-
tion”], 3 volumes, Chelsea, Bronx, New York, 1966.
62. Friedrich Adolf Trendelenburg, Zur Geschichte von Hegel’s Logik und dialektis-
cher Methode. Die logische Frage in Hegel’s Systeme. Eine Auffoderung [sic!]
zu ihrer wissenschaftlichen Erledigung, Neue Jenaische Allgemeine Literatur-
Zeitung, vol. 1 (23 April 1842), no. 97, pp. 405–408; (25 April 1842) no. 98, pp.
409–412; (26 April 1842) no. 99, pp. 413–414; separately published in Trende-
lenburg, Die logische Frage in Hegel’s System. Zwei Streitschriften, Brockhaus,
Leipzig, 1843.
63. Friedrich Ueberweg, System der Logik und Geschichte der logischen Lehren,
Adolph Marcus, Bonn, 1857.
64. , System of logic and history of logical doctrines, translated from the
German with notes and appendices by Thomas M. Lindsay, Longmans, Green,
and Co., London, 1871; reprinted Thoemmes Press, Bristol, 1993.
65. Hermann Ulrici, Review of [10], Zeitschrift für Philosophie und philosophische
Kritik, vol. 27 (1855), pp. 273–291.
66. James Van Evra, Richard Whately and the rise of modern logic, History and
Philosophy of Logic, vol. 5 (1894), pp. 1–18.
67. John Venn, Symbolic logic, 2nd ed., “revised and rewritten,” Macmillan & Co.,
London, 1894; reprinted Chelsea Publishing, Bronx, New York, 1971.
68. Alexander Wernicke, Review of [57], Deutsche Litteraturzeitung, vol. 12 (1891),
col. 196–197.
69. Richard Whately, Elements of logic: Comprising the substance of the article
in the Encyclopaedia Metropolitana: with additions &c., J. Mawman, London,
1826.
70. Wilhelm Windelband, Logik, in Windelband (ed.), Die Philosophie im Beginn
des zwanzigsten Jahrhunderts. Festschrift für Kuno Fischer, vol. 1, Carl Winter,
Heidelberg, 1904, pp. 163–186.
9
The Battle for Cantorian Set Theory
Joseph W. Dauben∗
Department of History, Herbert H. Lehman College, CUNY

Ph.D. Program in History, The Graduate Center, City University of New York
The substance of Georg Cantor’s revolutionary mathematics of the infinite is

well-known: in developing what he called the arithmetic of transfinite num-
bers, he gave mathematical content to the idea of actual infinity. In so doing he
laid the groundwork for abstract set theory and made significant contributions
to the foundations of the calculus and to the analysis of the continuum of real
numbers. Cantor’s most remarkable achievement was to show, in a mathemat-
ically rigorous way, that the concept of infinity is not an undifferentiated one.
Not all infinite sets are the same size, and consequently, infinite sets can be
compared with one another. But so shocking and counter-intuitive were Can-
tor’s ideas at first that the eminent French mathematician, Henri Poincaré,
condemned Cantor’s theory of transfinite numbers as a “disease” from which
he was certain mathematics would one day be cured.1 Leopold Kronecker, one
of Cantor’s teachers and among the most prominent members of the German
mathematics establishment, even attacked Cantor personally, calling him a
“scientific charlatan,” a “renegade,” and a “corrupter of youth.”2
∗
This paper was first presented as the invited centennial address for history of
mathematics on the occasion of the 100th anniversary of the American Mathemati-
cal Society, held jointly with the Mathematical Association of America in Atlanta,
Georgia, January, 1988. A revised version of that lecture, “Cantor and the Episte-
mology of Set Theory,” was presented in Montreal on June 3, 1995, as one of the
Kenneth O. May lectures of the Canadian Society for History and Philosophy of
Mathematics, at its Annual Meeting. Although a videotape of the AMS centennial
lecture is distributed by the American Mathematical Society, the lecture itself has
not been published until now, and I am grateful to the Canadian Society for History
and Philosophy of Mathematics for including it in this memorial volume of May
lectures. I am especially pleased to dedicate this paper to the memory of Kenneth
O. May, from whom I learned so much in the course of working with him from
1977-1979 as Managing Editor of Historia Mathematica, the journal he founded in
1974.
1
Poincaré 1908, p. 182.
2
For Kronecker’s criticism of Cantor, see Schoenflies 1927, p. 2.
222 Joseph W. Dauben
It is also well-known that Cantor suffered throughout his life from a series
of “nervous breakdowns” which became increasingly frequent and debilitating
as he got older. Some have tried to link this to his dangerous flirtations with
the infinite, but in the opinion of Karl Pönitz, who treated Cantor at the Halle
Nervenklinik:
Cantor’s illness was basically endogenous, and probably showed some
form of manic-depression: exogenous factors, such as the difficulties of
his researches and the controversies in Halle University, are likely to
have played only a small part in the genesis of his attacks, little more
than the clap that starts the avalanche. Thus he would have suffered
his attacks if he had pursued only an ordinary mundane career.3
Nevertheless it was all too easy for his early biographers to present Can-
tor, who was trying to defend his complex theory, as the hapless victim of the
infinite, due to his increasingly long periods of mental breakdown that began
in the 1880s, all of which were exacerbated by the persecutions of his con-
temporaries.4 But such accounts distort the truth by trivializing the genuine
intellectual concerns that motivated some of the most thoughtful contempo-
rary opposition to Cantor’s theory. They also fail to credit the power and
scope of the defense he offered for his ideas in the battle to win acceptance
for transfinite set theory.
At first Cantor himself resisted the implications of his research–because he
had always believed that the idea of the actual infinite could not be consis-
tently formulated, and so had no place in rigorous mathematics. Nevertheless,
by his own account, he soon overcame his “prejudice” regarding the transfinite
numbers because he found they were indispensable for the further develop-
ment of his mathematics.5 Because of his own early doubts he was able to
anticipate opposition from diverse quarters, which he attempted to meet with
philosophical and theological arguments as well as mathematical ones. More-
over, when he was called upon to respond to his critics, he was able to muster
3
The description is from Grattan-Guinness 1971, pp. 268-69. Among the doctors
who were responsible for treating Cantor at the Universitätsnervenklinik Halle were
Karl Pönitz and a Dr. Mekus. See Purkert and Ilgauds 1985, pp. 52-59, and pp. 118-
119; Purkert and Ilgauds 1987, pp. 79-82, and pp. 193-95. Cantor’s mental condition
has been analyzed in detail by the French Lacanian psychiatrist, Nathalie Charraud.
See especially her chapter, “La maladie,” in Charraud 1994, pp. 193-216.
4
This is the view, among others, of Schoenflies 1927; and Bell 1937, chapter 29.
Schoenflies’s account, it should be noted, was concerned exclusively with Cantor’s
first major breakdown in 1884, and it was not difficult for him to draw explicit
lines between Cantor’s illness and specific anxieties which the climate of Cantor’s
research had produced. But his later bouts of manic-depression seem to reflect no
such concerns or connections. According to the doctor who treated his next major
episode of manic-depression in 1899, it was most likely triggered by the broken
engagement of one of his daughters; see Purkert and Ilgauds 1985, p. 118; and
Purkert and Ilgauds 1987, p. 193.
5
Cantor 1883, p. 175.
9 The Battle for Cantorian Set Theory 223
his ideas with considerable force. His mental illness, far from playing an en-
tirely negative role, in its manic phases may well have contributed to the
energy and single-mindedness with which he promoted and defended his the-
ory, just as the theological dimension of Cantor’s understanding of the infinite
also reassured him–in fact convinced him–of its absolute truth, regardless of
what opponents like Kronecker might say against the theory.
Before it is possible to appreciate the origins, scope and significance of
Cantor’s battle to win acceptance for his transfinite numbers–the alephs–it will
be helpful to say something, briefly, about his life and the early development
of set theory.
Cantor’s Early History
Georg Ferdinand Ludwig Philip Cantor was born on March 3, 1845, in St.
Petersburg.6 His mother, a Roman Catholic, came from a family of notable
musicians; his father, a successful tradesman, was the son of a Jewish business-
man, but a devout Lutheran, having been raised in a Lutheran mission in St.
Petersburg. Cantor’s father passed on his own deep religious convictions to his
son. According to Eric Temple Bell’s widely-read book, Men of Mathematics,
first published in 1937, Georg Cantor’s insecurities in later life stemmed from
a ruinous Freudian conflict with his father, but surviving letters and other
evidence concerning their relationship indicate quite the contrary. Georg’s fa-
ther appears to have been a sensitive man who was attentive to his children
and took a special but not coercive interest in the welfare and education of
his eldest son.7
When the young Cantor was still a child the family moved from Russia to
Germany, and it was there that he began to study mathematics. After receiv-
ing his doctorate from the University of Berlin in 1868 for a dissertation on
the theory of numbers, two years later he accepted a position as Privatdocent,
or instructor, at the University of Halle, a respected institution but not as
prestigious for mathematics as the universities at Göttingen or Berlin. One
of his colleagues at Halle was Heinrich Eduard Heine, who was then working
on the theory of trigonometric series, and he encouraged Cantor to take up
the difficult problem of the uniqueness of such series. In 1872, when he was
twenty-seven, Cantor published a paper that included a very general solu-
tion to the problem, along with his theory of real numbers, which contained
6
The details provided here of Cantor’s life and early career are drawn largely
from my study, Dauben 1979, especially pp. 271-299. Other biographical studies
worth consulting include Meschkowski 1967, Grattan-Guinness 1971, Purkert and
Ilgauds 1985, and Purkert and Ilgauds 1987.
7
Bell 1937, chapter 29. For alternative evaluations of Cantor’s relationship with
his father, see Dauben 1979, esp. pp. 272-280, and Charraud 1994.
the seeds of what would develop later into his theory of transfinite sets and
numbers.8
Cantor in the 1870s Eduard Heine Richard Dedekind

(1821–1881) (1831–1916)
Cantor’s Discovery that the Real Numbers are

Uncountably Infinite
Cantor was not alone in studying the properties of the continuum of real
numbers in rigorous detail. In 1872, the same year that Cantor’s paper on
8
Cantor 1872. Dauben 1971 provides a detailed analysis of this paper and four
others that preceded it, which led Cantor from an early version of his representation
theorem in 1870 proving that if a function f (x) is represented by a trigonometric
series convergent for every value of x, then the series is unique, to the theorem
of 1872 which established the uniqueness of the representation even for an infinite
number of “exceptional” points. Remarkably, the trick Cantor needed to establish a
limited version of this theorem (published in 1871), namely that for a certain finite
number of values of x either the representation of the function or the convergence of
the trigonometric series could be given up, was provided by Kronecker (see Dauben
1979, p. 34). Within the year Cantor realized that he could establish the uniqueness
theorem even if an infinite number of exceptional points were permitted, provided
that they were distributed in a particular way. This was to lead eventually to the
first stages of Cantor’s development of set theory in terms of his theory of point sets.
To describe how the exceptional point sets were distributed over the continuum in
the case of the representation theorem, he also found that he needed first to develop
a rigorous theory of real numbers. Thus from a series of papers on the representation
of functions by trigonometric series, Cantor was led to consider a rich concatenation
of ideas that would prove especially fertile for his thinking about sets, the structure
and nature of the continuum, and eventually, his theory of transfinite numbers over
the next few decades.
trigonometric series appeared, the German mathematician Richard Dedekind

also published an analysis of the continuum that was based on infinite sets.
In his monograph on the subject of continuity and the irrational numbers,
Dedekind articulated an idea that Cantor later made more precise:
The line L is infinitely richer in point-individuals than is the domain
R of rational numbers in number-individuals.9
Dedekind’s statement, however, conceals a serious weakness. If anyone had
asked Dedekind how much richer the infinite set of points in the continuum was
than the infinite set of rational numbers, he could not have replied. Cantor’s
major contribution to this question was published in 1874, in Crelle’s Journal
für die reine und angewandte Mathematik. What Cantor showed was the non-
denumerability of the real numbers, a discovery that was soon to transform
much of modern mathematics. Cantor’s paper was short, three pages, and
bore a very strange title:
Über eine Eigenschaft des Inbegriffes aller reellen algebraischen Zahlen.
On a Property of the Collection of All Real Algebraic Numbers.10
No one scanning the title of this short paper would have guessed that
this was the paper that disclosed Cantor’s revolutionary discovery of the non-
denumerability of the continuum of real numbers, which established that some
infinite sets were larger than others, in particular that the set of natural
numbers N was of a lower magnitude of infinity than the set of real numbers
R. Instead, the article bore a deliberately misleading title suggesting that its
major result was a theorem about algebraic numbers, thus failing even to hint
at the more significant point that the paper actually contained. What could
possibly have prompted Cantor to choose such an inappropriate title for what
now, in retrospect, strikes any mathematician as one of the most important
discoveries in modern mathematics?
The answer hinges on one of Cantor’s teachers at Berlin, Leopold Kro-
necker. Having studied with Kronecker, Cantor was well-acquainted with Kro-
necker’s work in number theory and algebra, and with his highly conservative
philosophical views with respect to mathematics. By the early 1870s, Kro-
necker was already vocal in his opposition to any infinitary arguments, in-
cluding the Bolzano-Weierstrass theorem and upper and lower limits, as well
as to irrational numbers in general. Kronecker’s later pronouncements against
analysis and set theory, as well as his adamant and well-known insistence
upon using the natural numbers to provide the only satisfactory foundation
for mathematics, were simply extensions of these early views.
With this in mind, it is not unreasonable to suspect that Cantor had
good reason to anticipate Kronecker’s opposition to his proof of the non-
denumerability of the real numbers, which proved that they comprised a set
9
Dedekind 1872, p. 9.
10
Cantor 1874.
Cantor’s paper in which he proves the Leopold Kronecker

nondenumerability of the real numbers. (1823–1891)
infinitely larger than the set of integers. Certainly any result (such as Can-
tor’s) which confirmed the existence of transcendental numbers–against which
Kronecker’s opinions were well-known–would have been subject to criticism
from Kronecker.
Worse yet, Kronecker was on the editorial board of the journal to which
Cantor submitted his paper. Had Cantor been more direct with a title like
“The Set of Real Numbers is Non-Denumerably Infinite,” or “A New and In-
dependent Proof of the Existence of Transcendental Numbers,” he could have
counted on a strongly negative reaction from Kronecker. In fact, when Fer-
dinand Lindemann later established the transcendence of π in 1882, meaning
that it was not only irrational but also not an algebraic number, Kronecker
asked what value the result could possibly have, since irrational numbers did
not exist.11 As Cantor contemplated publishing his paper on the nondenumer-
ability of the real numbers in 1874, an innocuous title was clearly a strategic
choice. Reference only to algebraic numbers would have had a much better
chance of passing Kronecker’s eye unnoticed, for there was nothing to excite
either immediate interest or censure.
If the idea that Cantor may have harbored fears about Kronecker’s op-
position to his work at such an early date seems unwarranted, it is worth
noting that Kronecker had already tried to dissuade Cantor’s colleague at
Halle, namely Heine, from publishing an article on trigonometric series in
11
Kronecker made this remark in a lecture at the Berliner Naturforscher-
Versammlung in 1886; see Weber 1893, p. 15; Kneser 1925, p. 221; and Pierpoint
1928, p. 39.
Crelle’s Journal. Although Heine’s article eventually appeared, Kronecker was

at least successful in delaying its appearance, about which Heine was partic-
ularly vocal in letters to Schwarz, also a friend of Cantor’s. Heine explained
the circumstances of the situation in a letter to Schwarz on May 26, 1870:
My little work “On Trigonometric Series,” of which I now have the
page-proofs in hand, and which at present is still under debate with
Kronecker, who wanted to persuade me to retract it (the particulars
below), appears in the current volume (71) of the journal p. 353, and
has made me very happy; I had sent it to Borchardt in February, where
Kr[onecker] saw it and to whom it was given and he kept it without
my knowing it, until I came to Berlin.12
Heine to Schwarz, May 26, 1870. Cantor and his wife, Vally, about 1880.
Doubtless Schwarz and Heine would both have brought Kronecker’s readi-
ness (and ability) to block ideas with which he disagreed to Cantor’s attention.
Indeed, several years later Kronecker also delayed publication of a paper Can-
tor had written on the invariance of dimension.13 This so angered Cantor that
12
The complete letter is transcribed in Dauben 1979, pp. 308-09.
13
Cantor 1878; Cantor wrote a bitter letter to Dedekind about the incident and
even planned to withdraw the paper from Crelle’s Journal,a step Dedekind per-
suaded him not to take. See Cantor’s letter to Dedekind of October 23, 1887, in
Cantor/Dedekind 1937, p. 40. Cantor gave a lecture in Braunschweig in 1897 in
which he recalled the incident and said that it was Weierstrass who had interceded
on his behalf, thanks to which the paper was eventually published. See Fraenkel
1930, p. 10; full details are given in Dauben 1979, pp. 66-70.
he never submitted anything to Crelle’s Journal again. A decade later he re-

garded Kronecker as both a private and a public menace–not only because he
was condemning set theory openly, but Weierstrassian analysis as well.
There was however a positive side to Kronecker’s opposition to Cantor’s
work, for it forced Cantor to evaluate the foundations of set theory as he was in
the process of creating it. This concern prompted long philosophical passages
in Cantor’s major publication of the 1880s on set theory, his Grundlagen einer
allgemeinen Mannigfaltigkeitslehre of 1883. It was here that Cantor issued
one of his most famous pronouncements about mathematics, namely that
the essence of mathematics is exactly its freedom.14 This was not simply an
academic or philosophical message to his colleagues, for it also carried a hidden
and deeply personal subtext. It was, as he later admitted to David Hilbert,
a plea for objectivity and openness among mathematicians. This, he said,
was inspired directly by the oppression and authoritarian closed-mindedness
that he felt Kronecker represented, and worse, had wielded in a flagrant and
damaging way against those he opposed.
Thus at the very beginning of his career, even before he had begun to
develop any of his more provocative ideas about transfinite set theory, Cantor
had experienced his first bitter taste of Kronecker’s opposition to his work.
Doubtless Cantor knew that he could expect more trouble in future.
Set Theory Begins to Develop

Meanwhile, Cantor devoted himself to developing further the ideas about point
sets which he had first investigated in the context of representing functions
by trigonometric series in the 1870s. By the end of the decade he had mar-
ried Vally Guttman, and shortly thereafter he began to publish an important
series of papers on linear point sets. These eventually led to his Grundlagen
einer allgemeinen Mannigfaltigkeitslehre. Ein mathematisch-philosophischer
Versuch in der Lehre des Unendlichen, which emphasized transfinite ordinal
numbers introduced rather vaguely in terms of what Cantor termed “prin-
ciples of generation.” The Grundlagen also offered a detailed philosophical
defense of his new ideas on the infinite.15 It was in 1883 that Cantor first
tried vigorously to establish his Continuum Hypothesis in the version that
the set of real numbers was the next largest after the denumerable set of
natural numbers.
Despite his vigorous efforts to prove the correctness of the Continuum
Hypothesis, he was greatly frustrated by his inability to do so. Early in 1884
he thought he had found a proof, but a few days later he reversed himself
completely and thought he could actually disprove the hypothesis. Finally he
14
Cantor 1883, p. 182.
15
Cantor 1883, pp. 131-132.
realized that he had made no progress at all, as he reported in letters to his

friend and editor of Acta Mathematica, Gösta Mittag-Leffler in Stockholm.16
Cantor in the 1880s.
All the while Cantor was facing mounting opposition and threats from
Kronecker, who said he was preparing an article to show that “the results of
modern function theory and set theory are of no real significance.”17
Soon thereafter, in May of 1884, Cantor suffered the first of his serious ner-
vous breakdowns. Although his lack of progress on the Continuum Hypothesis
or stress from Kronecker’s attacks may have helped to trigger the breakdown,
it now seems clear that such events had little to do with its underlying cause.
The illness took over with startling speed and lasted somewhat longer than a
month. At the time, only the manic phase of manic-depressive psychosis was
recognized as a symptom. When Cantor “recovered” at the end of June, 1884,
and entered the depressive phase of his illness, he complained that he lacked
the energy and interest to return to rigorous mathematical thinking. He was
content to take care of trifling administrative matters at the university, but
felt capable of little more.
Although Cantor eventually returned to mathematics, he also became in-
creasingly absorbed in other interests. He undertook a study of English history
and literature and became engrossed in a scholarly diversion that was taken
16
See especially letters Cantor wrote to Mittag-Leffler between August and No-
vember of 1884, in Meschkowski 1967, pp. 240-241 and p. 243; and in Schoenflies
1927, p. 12 and pp. 17-18.
17
Quoted by Cantor in a letter to Mittag-Leffler dated January 26, 1884, in
Schoenflies (1927), p. 5.
with remarkable seriousness by many people at the time: namely, the suppo-
sition that Francis Bacon was the true author of Shakespeare’s plays. Cantor
also tried his hand without success at teaching philosophy instead of mathe-
matics, and he began to correspond with several theologians who had taken
an interest in the philosophical implications of his theories about the infinite.
This correspondence was of special significance to Cantor because he was con-
vinced that the transfinite numbers had come to him as a message from God.
But more about the significance of this, as already promised, in a moment.
Transfinite Cardinal Numbers: Cantor’s Alephs

There is still one last element of Cantor’s technical development of transfi-
nite set theory that needs to be mentioned as part of his continuing efforts
to mount a convincing and satisfactory mathematical defense of his ideas,
namely, the nature and status of the transfinite cardinal numbers. The evo-
lution of Cantor’s thinking about the transfinite cardinals is curious, because
although the alephs are probably the best-known legacy of Cantor’s creation,
they were the last part of his theory to be given either rigorous definition or a
special symbol. Indeed, it is difficult in the clarity of hindsight to reconstruct
the obscurity within which Cantor must have been groping, and up to now
his work has been discussed here largely as if he had already recognized that
the power of an infinite set could be understood as a cardinal number. In fact,
beginning in the early 1880s, Cantor first introduced notation for his infinite
(actually transfinite) sequence of derived sets P ν , extending them well beyond
the limitation he had earlier set himself to sets of the first species.18 At this
time he spoke of the indexes only as “infinite symbols” with no reference to
them in any way as numbers.
By the time he wrote the Grundlagen in 1883, the transfinite ordinal num-
bers had finally achieved independent status as numbers and were given the
familiar omega notation, ω being the first transfinite ordinal number following
the entire sequence of finite ordinal numbers, i.e., 1, 2, 3, . . . , ω. However, there
was no mention whatsoever of transfinite cardinal numbers, although Cantor
clearly understood that it is the power of a set that establishes its equivalence
(or lack thereof) with any other set; from this he would eventually develop his
concept of transfinite cardinal numbers. But in the Grundlagen, he carefully
avoided any suggestion that the power of an infinite set could be interpreted
as a number.
Soon, however, he began to equate the two concepts and in September of
1883, did so in a lecture to mathematicians at a meeting in Freiburg. Even
18
An infinite set P was said to be of the first species if there were some finite
number ν for which the νth derived set of limit points of P was empty, i.e., P ν = ∅.
Infinite sets for which there was no such ν were sets of the second species. See Cantor
1872, and Dauben 1979, pp. 43-45.
so, no symbol was as yet provided for distinguishing one transfinite cardinal
number from another. Since he had already adopted the symbol ω to designate
the least transfinite ordinal number, when Cantor finally introduced a symbol
for the first transfinite cardinal number, it was borrowed from the symbols
already in service for the transfinite ordinals. By 1886, in correspondence,
∗
Cantor had begun to represent the first transfinite cardinal as ω ; the next
∗
larger he denoted Ω . This notation was not very flexible, and within months
Cantor realized the need for a more general notation capable of representing
the entire ascending hierarchy of transfinite cardinals. Temporarily he used
fraktur o’s, obviously derivatives from his omegas, to represent his sequence of
cardinal numbers. For a time Cantor actually used superscripted stars, bars,
and his fraktur o’s interchangeably for transfinite cardinals, without feeling
any need to decide upon one or the other notation as preferable.19
The Paradoxes of Set Theory

However, in 1893 the Italian mathematician Giulio Vivanti was preparing a
general account of set theory, and Cantor realized it was time to adopt a
standard notation. Only then did he choose the Hebrew aleph (ℵ) for the
transfinite cardinals, because he thought the familiar Greek and Roman al-
phabets were too common and already widely employed in mathematics for
other purposes. His new numbers deserved something distinct and unique–and
the Hebrew alphabet was readily available among the type fonts of German
printers. The choice was particularly clever, as Cantor was pleased to admit,
because the Hebrew aleph was also the symbol for the number one. Since the
transfinite cardinal numbers were themselves infinite unities, the aleph could
be taken to represent a new beginning for mathematics. Cantor designated
the cardinal number of the first number class ℵ1 in 1893, but in 1895 changed
his mind; from then on, he used ℵ0 to represent the first and least transfinite
∗
cardinal number, the number he had previously designated by ω. From ℵ0 ,
he went on to designate the cardinal number of the second number class as
ℵ1 , after which there followed an unending sequence of transfinite cardinal
numbers.
Cantor made his last major contributions to set theory in 1895 and 1897.
He had already used his famous method of diagonalization in 1891 to show
that the cardinal number of any set P is always less than the cardinal number
of its power-set, the set of all subsets of P . A few years later he presented a
corollary to this result, namely that the cardinal number of the continuum is
equal to 2ℵ0 , and he hoped this result would soon lead to a solution of the
Continuum Hypothesis–which he could now express as 2ℵ0 = ℵ1 .
19
For a detailed discussion of the evolution of Cantor’s notation for the transfinite
cardinal numbers, see Dauben 1979, pp. 179-183.
The arguments Cantor used in his proof about the cardinal number of the
power-set of all subsets of any given set, however, led to far different conclu-
sions. Rather than leading to a resolution of the Continuum Hypothesis, they
led directly to the discovery of the paradoxes of set theory, for the fact that
there could be no “largest” transfinite cardinal number immediately raised
the question of the cardinality of the set of “all” transfinite cardinal num-
bers. Cantor resolved the problem by excluding this possibility entirely; the
aggregate of “all” transfinite numbers was what he called an “inconsistent”
aggregate, and therefore was simply not to be considered as a “set.” Bertrand
Russell, in contemplating this problem, drew far more problematic conclu-
sions, for what he discovered was that a paradox can be derived in set theory
by considering those sets that do not include themselves as members.20
Russell’s paradox suggested that there was something fundamentally
wrong with Cantor’s definition of a set, and the consequences of this realiza-
tion immediately became an important problem in 20th-century mathemat-
ics. Even before Bertrand Russell, however, Cantor had already come upon
his own version of the paradoxes of set theory in the form of contradictions
he associated with the idea of a largest ordinal or cardinal number. This was
all explained in letters first to Hilbert in 1897, and then to Dedekind in 1899.
As Cantor wrote to Dedekind on August 3, 1899, if one considers the collec-
tion of all transfinite ordinal numbers Ω, “the system Ω of all numbers is an
inconsistent, absolutely infinite aggregate.”21
But it is possible that Cantor may have been aware of the paradoxes of
set theory much earlier, perhaps as early as the 1880s when his difficulties
with Kronecker were weighing on his mind and as he was just beginning to
experience his first serious technical problems with set theory. For example,
in his Grundlagen of 1883, Cantor referred to collections that are too large
to be comprehended as a well-defined, completed, unified entity. At the time
he wrote obscurely, with references to absolute sets in explicitly theological
terms, explaining that “the true infinite or Absolute, which is in God, permits
no determination.”22 Was this a hint that he already understood that the
collection of all transfinite ordinal numbers was inconsistent, and therefore
not to be regarded as a set? Later, Cantor said that it was–that he meant this
to be a veiled sign, even then, that he was aware of the paradoxical results
that followed from trying to determine what transfinite ordinal number should
correspond to the well-ordered set of all transfinite ordinal numbers.
20
In contemplating the result of Cantor’s diagonalization proof, Russell consid-
ered the implications of the fact that there could be no one-to-one correspondence
between the elements of a set P and the elements of its power set. In asking himself
what elements of the power set were left out of such a correspondence, Russell was
led to the discovery of his paradox of sets which are not members of themselves. For
details see Russell 1907; Dauben 1979, pp. 261-263; and Chapter IV of Garciadiego
1992: “Russell’s discovery of the ‘paradoxes’,” pp. 81-130.
21
Cantor to Dedekind, August 3, 1899, in Cantor 1932, p. 445.
22
Cantor 1883, Note 2 to Section 4 of the Grundlagen, p. 205.
By the mid-1890s, Cantor could no longer be so vague about absolute enti-

ties, and was forced to be much more explicit about the paradoxes that arose
from contemplating the sets of all transfinite ordinal or cardinal numbers.
The solution Cantor chose for dealing with such mathematical paradoxes was
simply to exclude them from set theory. Anything that was too large to be
comprehended as a well-defined, unified, consistent set was declared incon-
sistent. These were “absolute” collections, and lay beyond the possibility of
mathematical determination. This, in essence, is what Cantor communicated
first to Hilbert in 1897, and somewhat later to Dedekind in Cantor’s letters
of 1899.23
Foundations and Philosophy of Mathematics
Whatever the extent of Cantor’s awareness of the paradoxes may have been
in the early 1880s, he was certainly sensitive to Kronecker’s, growing and
increasingly vocal opposition. Above all, it is clear that explicitly philosophical
concerns expressed in his Grundlagen were in Cantor’s opinion strategically
crucial for a comprehensive defense of his new theory. This was unusual at the
time; it still is. When Mittag-Leffler arranged to publish French translations
of Cantor’s papers on set theory for Mittag-Leffler’s newly-founded journal,
Acta Mathematica, he persuaded Cantor that it would be best to omit all of
the philosophical portions of the Grundlagen as unnecessary (and possibly
repugnant) to mathematicians who might find the theory of interest but the
philosophy unacceptable.24
The philosophical arguments, however, were essential to Cantor, if not to
Mittag-Leffler. They were essential because they were part of the elaborate
defense he had begun to construct to subvert opposition from any quarter, but
especially from Kronecker. One particularly important ploy was to advance a
justification of transfinite set theory based upon the freedom of mathematics
to admit any self-consistent theory. Applications might eventually determine
which mathematical theories were useful, but for mathematicians, Cantor in-
sisted that the only real question was consistency. This of course was just
the interpretation he needed to challenge an established mathematician like
Kronecker. Cantor clearly felt obliged, early in his career, to plead as best he
could for a fair hearing of his work. So long as it was self-consistent it should
23
Cantor’s letters to Hilbert about the “absolute” character of the collection of
all transfinite numbers were long thought to be lost (or nonexistent), but two of
Cantor’s letters on this subject to Hilbert dated September 26 and October 2, 1887,
have recently been published by Walter Purkert and Hans Joachim Ilgauds; see
their Georg Cantor 1845-1918 (1987), pp. 224-227. These same two letters are also
reproduced in Meschkowski 1991, pp. 388-390.
24
Mittag-Leffler, in a letter to Cantor dated March 11, 1883, in the archives of
the Institut Mittag-Leffler, Djursholm, Sweden; cited in Dauben 1979, p. 297.
be taken as mathematically legitimate, and the constructivist, finitist criti-

cisms of Kronecker might be disregarded by most mathematicians, for whom
consistency alone should be the viable touchstone.
Gösta Mittag-Leffler (1846-1927)
The Freedom of Mathematics

Cantor put his philosophy about the freedom of mathematics into action
early in the 1890s, when his career had reached the point where he could
do more than simply write about it. During the 1880s he had already begun
to lay the strategic foundations for an independent union of mathematicians
in Germany. The specific goal of such a union, as he often made clear in his
correspondence, was to provide an open forum, especially for young mathe-
maticians. The union (as Cantor envisaged it) would guarantee that anyone
could expect free and open discussion of mathematical results without preju-
dicial censure from members of the older establishment, whose conservatism
might easily ruin the career of an aspiring mathematician. This was primarily
needed in cases where the ideas in question were at all new, revolutionary or
controversial.
Cantor labored intensively for the creation of the Deutsche Mathematiker-
Vereinigung.25 Eventually, agreement was reached and the Union of German
Mathematicians held its first meeting in conjunction with the annual meet-
ing of the Gesellschaft Deutscher Naturforscher und Ärzte at Halle in 1891.
25
For details, see Dauben 1979, pp, 160-163.
Cantor was elected the Union’s first president, and at its inaugural meeting
he presented his now famous proof that the real numbers are nondenumerable
using his new method of diagonalization.26
The German Union was not the end of Cantor’s vision. He also recognized
the need to promote international forums, and thus began lobbying for inter-
national congresses shortly after formation of the DMV. These were eventually
organized through the cooperative efforts of many mathematicians, and not
directly as a result of Cantor’s exclusive efforts by any means. The first of
these was held in Zürich in 1897, the second in Paris in 1900.27
Promoting new avenues for discussion of mathematics was one way Can-
tor reacted to opposition of the sort his own research had provoked. Despite
criticisms, especially from Kronecker, Cantor persevered, even in the face of
his own repeated failure to resolve some of the most basic questions about
set theory (notably his Continuum Hypothesis), and even though he began
to suffer increasingly serious cycles of manic-depression. Ironically, like his
conflicts with Kronecker, Cantor’s manic-depression may have served a use-
ful purpose. In his own mind it was closely linked to the infallible support
set theory drew from his strongly-held religious convictions. Letters (and the
testimony of colleagues who knew him) reveal that Cantor believed he was
chosen by God to bring the truths of set theory to a wider audience. He also
regarded the successive waves of manic-depression that began to plague him
in the 1880s–peaks of intense activity followed by increasingly prolonged inter-
vals of introspection–as divinely inspired. Long periods of isolation in hospital
provided opportunities for uninterrupted reflection during which Cantor en-
visioned visits from a muse whose voice reassured him of the absolute truth
of set theory, whatever others might say about it.
In promoting set theory among mathematicians, philosophers, and theolo-
gians (he even wrote to Pope Leo XIII at one point on the subject of the
infinite), Cantor was convinced he would succeed in securing the recognition
that transfinite set theory deserved.28 By stressing self-consistency and the in-
trinsic freedom of mathematics, he also advanced an essential element of any
intellectual inquiry, namely that the mind must be free to pursue the truth
wherever it may lead. Inspiration should be encouraged, not confounded by
arbitrary prejudice, and for Cantor this meant that theories should be judged
upon standards of consistency and utility.
26
Cantor 1891.
27
For details, see Dauben 1979, pp. 163-165.
28
Cantor wrote in Latin to Pope Leo XIII, February 13, 1896; transcribed in
Purkert and Ilgauds 1987, pp. 198-199; published in Latin with German translation
in Meschkowski 1991, p. 383.
Transfinite Mathematics and Cantor’s Manic-Depression

To understand Cantor’s tenacious promotion and defense of set theory, es-
pecially in his later years after the publication of the Beiträge of 1895/1897,
it is important to appreciate the connection between Cantor’s faith in God,
his mental illness, and his mathematics. Certain documents suggest that in
addition to enforcing periodic intervals of contemplation and withdrawal from
daily affairs, Cantor’s periods of depression were productive in other ways. In
fact, he was often able to pursue his mathematical ideas in the solitude of
the hospital or quietly at home. This may have supported his belief that the
transfinite numbers had been communicated to him from God. In fact, as he
noted in the third motto he chose to head his last publication, the Beiträge
of 1895:
Veniet tempus, quo ista quae nunc latent, in lucem dies extrahat et
longioris aevi diligentia.
The time will come when these things which are now hidden from you
will be brought into the light.29
Cantor’s Beiträge of 1895.
This is a familiar passage from the Bible and reflects Cantor’s belief that he
was an intermediary serving as the means of revelation. It may also be taken
29
The Bible, I Corinthians 4:5.
to reflect Cantor’s faith that despite any prevailing resistance to his work, it
would one day enjoy recognition and praise from mathematicians everywhere.
It is easy, of course, to misinterpret the religious element in Cantor’s think-
ing, as popularizers often do. This was certainly the case in an article that
appeared in 1977 in the French magazine La Recherche, which supplied the
following caricatures to illustrate an expository article about Cantor, his re-
ligious convictions, psychological illness, and transfinite set theory. The first
drawing depicts Cantor in ecstasy, as it were, receiving the divine message:
Cantor in ecstasy! The precarious balance.
In the second illustration, the figure with the gun is meant to be Leopold
Kronecker–with God helping Cantor to maintain his balance–all of which rests
precariously on a transfinite aleph.30
But there is a very serious side to all of this and it deserves to be empha-
sized. For example, following a long period of hospitalization in 1908, Cantor
wrote to a friend in Göttingen, the British mathematician Grace Chisholm
Young. As he described it, his manic-depression took on a strikingly creative
quality:
30
The two drawings above from the article on Cantor by Pierre Thuillier [Thuillier
1977] are reproduced here by kind permission of the artist, André Barbe, and the
editors of La Recherche.
A peculiar fate, which thank goodness has in no way broken me, but
in fact has made me stronger inwardly. . . has kept me far from home–
I can say also far from the world. . . In my lengthy isolation neither
mathematics nor in particular the theory of transfinite numbers has
slept or lain fallow in me.31
Elsewhere, Cantor actually described his conviction about the truth of his
theory explicitly in quasi-religious terms:
My theory stands as firm as a rock; every arrow directed against it

will return quickly to its archer. How do I know this? Because I have
studied it from all sides for many years; because I have examined all
objections which have ever been made against the infinite numbers;
and above all, because I have followed its roots, so to speak, to the
first infallible cause of all created things. 32
Later generations might dismiss the philosophy, look askance at Cantor’s

abundant references to St. Thomas or to the Church Fathers, overlook his
metaphysical pronouncements, and miss entirely the deeply religious roots of
Cantor’s later faith in the absolute truth of his theory. But all these com-
mitments contributed to his resolve not to abandon the transfinite numbers.
31
Cantor to Grace Chisholm Young, June 20, 1908; transcribed in Meschkowski
1971, pp. 30–34; translated in Dauben 1979, p. 290.
32
Cantor to K. F. Heman, June 21, 1888; quoted from Dauben 1979, p. 298.
Opposition seems to have strengthened his determination. His forbearance,

as much as anything else Georg Cantor might have contributed, ensured that
set theory would survive the early years of doubt and denunciation to flourish
eventually as a vigorous, revolutionary force in 20th-century mathematics.
Bibliography
Charraud, Nathalie, Infini et Inconscient. Essai sur Georg Cantor (Paris: An-
thropos, 1994).
Cantor, Georg, “Über die Ausdehnung eines Satzes aus der Theorie der
trigonometrischen Reihen,” Mathematische Annalen, 5 (1872), pp. 123–132; in
Cantor 1932, pp. 92–102; translated as “Extension d’un théorème de la théorie
des series trigonométriques,” Acta Mathematica, 2 (1883), pp. 336-3-48.
, “Über eine Eigenschaft des Inbegriffes aller reellen algebraischen
Zahlen,” Journal für die reine und angewandte Mathematik, 77 (1874), pp.
258–262; in Cantor (1932), pp. 115–118; translated as “Sur une propriété du
système de tous les nombres algébriques reels,” Acta Mathematica, 2 (1883),
pp. 205–310.
, Grundlagen einer allgemeinen Mannigfaltigkeitslehre. Ein mathema-
tisch-philosophischer Versuch in der Lehre des Unendlichen (Leipzig: B.G.
Teubner, 1883). In Cantor (1932), pp. 165–208. Translated, in part, into
French as “Fondaments d’une Théorie générale des ensembles,” Acta Mathe-
matica, 2 (1883), pp. 381-408; and into English as “Foundations of the Theory
of Manifolds” (trans. U. Parpart), The Campaigner (The Theoretical Jour-
nal of the National Caucus of Labor Committees), 9 (January and February,
1976), pp. 69–96; a better translation into English is by W. B. Ewald, “Foun-
dations of a General Theory of Manifolds: A Mathematico-Philosophical In-
vestigation into the Theory of the Infinite,” in From Kant to Hilbert: a Source
Book in the Foundations of Mathematics, ed. W. B. Ewald (New York: Oxford
University Press, 1996), vol. 2, pp. 878–920.
, “Über eine elementare Frage der Mannigfaltigkeitslehre,” Jahres-
bericht der Deutschen Mathematiker-Vereinigung, 1 (1891), pp. 75–78; in Can-
tor 1932, pp. 278–280.
, Gesammelte Abhandlungen mathematischen und philosophischen In-
halts, ed. E. Zermelo (Berlin: J. Springer; rep. Hildesheim: Olms, 1966, and
Berlin: Springer, 1990).
Cantor, Georg, and Richard Dedekind, Briefwechsel Cantor-Dedekind, eds. E.
Noether and J. Cavaillès (Paris: Hermann, 1937).
Dauben, Joseph W., “The Trigonometric Background to Georg Cantor’s The-
ory of Sets,” Archive for History of Exact Sciences, 7 (1971), pp. 181–216.
, Georg Cantor: His Mathematics and Philosophy of the Infinite

(Cambridge, MA: Harvard University Press, 1979; rep. Princeton: Princeton
University Press, 1990).
Dedekind, Richard, Stetigkeit und irrationale Zahlen (2nd ed. Braunschweig:
Vieweg, 1892); English translation: Dedekind 1901/1963.
, Essays on the Theory of Numbers, Continuity of Irrational Numbers,
the Nature and Meaning of Numbers, trans. W.W. Beman (Chicago: Open
Court, 1901; repr. New York: Dover, 1963).
Fraenkel, A., “Georg Cantor,” Jahresbericht der Deutschen Mathematiker-
Vereinigung, 39 (1930), pp. 189–266.
Garciadiego, Alejandro R., Bertrand Russell and the Origins of the Set-
theoretic ‘Paradoxes’ (Basel: Birkhäuser Verlag, 1992).
Grattan-Guinness, Ivor, “Towards a Biography of Georg Cantor,” Annals of
Science, 27 (1971), pp. 345–391.
Kneser, A., “Leopold Kronecker,” Jahresbericht der Deutschen Mathematiker-
Vereinigung, 33 (1925), pp. 210–228.
Meschkowski, Herbert, Probleme des Unendlichen. Werk und Leben Georg
Cantors (Braunschweig: Vieweg, 1967).
, “Zwei unveröffentlichte Briefe Georg Cantors,” Der Mathematikun-
terricht, 4 (1971), pp. 30-34.
, Georg Cantor–Briefe, eds. Herbert Meschkowski and Winfried Nil-
son (Berlin: Springer-Verlag, 1991).
Moore, Gregory H., Zermelo’s Axiom of Choice (New York: Springer-Verlag,
1982).
Pierpont, J., “Mathematical Rigor, Past and Present,” Bulletin of the Amer-
ican Mathematical Society, 34 (1928), pp. 23–53.
Poincaré, Henri, “L’avenir des mathématiques,” Atti del IV Congresso Inter-
nazionale dei Matematici, Rome, 6-11 April, 1908 (Rome: Tipografia della R.
Accademia dei Lincei, C.V. Salviucci, 1909), pp. 167–182.
Purkert, Walter, and Hans Joachim Ilgauds, Georg Cantor (Biographien her-
vorragender Naturwissenschaftler, Techniker und Mediziner, vol. 79, Leipzig:
B.G. Teubner, 1985).
, Georg Cantor 1845-1918 (Vita Mathematica, vol. 1, Basel: Birkhäuser
Verlag, 1987).
Russell, Bertrand, “On Some Difficulties in the Theory of Transfinite Numbers
and Order Types,” Proceedings of the London Mathematical Society, 4 (1907),
pp. 29-53.
Schoenflies, A., “Die Krisis in Cantor’s mathematischem Schaffen,” Acta

Mathematica 50 (1927), pp. 1–23.
Thuillier, Pierre, “Dieu, Cantor et l’infini,” La Recherche, 84 (December,
1977), pp. 1110–1116.
Weber, A., “Leopold Kronecker,” Mathematische Annalen, 43 (1893), pp. 1–
25.
10
Hilbert and his Twenty-Four Problems
Rüdiger Thiele
Karl-Sudhoff-Institut für Geschichte der Medizin und der Naturwissenschaften,

University of Leipzig
If you can look into the seeds of time,

And say which grain will grow, and which will not,
Speak then to me.
William Shakespeare, Macbeth (I, 3)
At the turn of the 19th century David Hilbert (1862-1943) was well known
for fundamental results in invariant theory, for his profound Zahlbericht (Re-
port on the Theory of Numbers), and the far-reaching Grundlagen der Geome-
trie (Foundations of Geometry; 14 German editions) which opened the way for
the axiomatic method in mathematics. A substantial part of Hilbert’s fame,
however, rests on his address “Mathematical Problems”, delivered at the sec-
ond International Congress of Mathematicians (ICM) in Paris. Although the
discussion which followed Hilbert’s lecture on that summer morning of August
8, 1900 was desultory, “it became quite clear” [by the printed versions of the
lecture in some languages], as Constance Reid (born 1917) remarked, “that
David Hilbert had captured the imagination of the mathematical world with
his list for the 20th century. His practical experience seemed to guarantee that
they met the criteria which he had set up in his lecture, his judgment, that
they could actually be solved in the years to come.”1
Elie Cartan (1869–1951) emphasized the importance of this speech in a
letter to Constantin Carathéodory (1873–1950) which was written shortly
after Hilbert’s death: “We will never hear the like of such a talk at con-
gresses.”2 On the other hand, it was Carathéodory who pointed out that
these 23 problems divided Hilbert’s career into two parts. In Germany up to
this speech Hilbert was a respected mathematician; after the talk his interna-
tional fame grew rapidly (partly because the American Mathematical Society
very quickly supplied English readers with both a report and a translation
1
Reid, 1996, p. 84.
2
Carathéodory 1943, p. 350. “On n’entendra plus dans les Congrès de conférence
pareille.”
David Hilbert
1862-1943
Courtesy Niedersächsische Staats- und Universitätsbibliothek Göttingen,
Handschriftenabteilung, Voit Collection.
of Hilbert’s speech).3 In the second period David Hilbert gathered more and
more pupils and his seed began to grow. Hilbert had become one of the most
famous mathematicians of the day, his fame possibly exceeded only by that of
Henri Poincaré (1854–1912). Anyone able to solve one of the problems could
instantly make a reputation for himself.
Immediately mathematicians set about their work and the mathemati-
cal community watched each contribution attentively. Hermann Weyl (1885–
1955) once remarked:
3
Hilbert 1902, 2000; G.B. Halsted, 1900; Scott 1900.
10 Hilbert and his Twenty-Four Problems 245
Fig. 10.1. Hilbert’s Paris lecture “Mathematische Probleme (Mathematical Prob-

lems)” was first published in Göttinger Nachrichten 1900. Courtesy Mathematisches
Institut, Universität Leipzig.
Fig. 10.2. English translation of Hilbert’s Paris lecture by M.W. Newson, Bulletin of
the American Mathematical Society 8 (1901/02). Courtesy Mathematisches Institut,
Universität Leipzig.
Fig. 10.3. French translation of Hilbert’s Paris lecture by L. Laugel in Compte

Rendu Deuxième Congrès International des Mathématiciens. Paris: Gauthier 1902.
Courtesy Mathematisches Institut, Universität Leipzig.
We mathematicians have often measured our progress by checking

which of Hilbert’s questions have been settled in the meantime.4
Elsewhere he added:
The problems of mathematics are not isolated problems in a vacuum;

there pulses in them the life of ideas which realize themselves in con-
creto [in practice] through our human endeavors in our historical ex-
istence, but forming an indissoluble whole transcend any particular
science.5
Hilbert, best known for his axiomatic foundations of mathematics and his
formalist viewpoint, knew the value of important problems. As his disciple
and biographer Otto Blumenthal (1876–1944) put it: “Hilbert is the man
of problems. He collects and solves existing problems; he poses new ones.”6
Indeed, it is just by the solution of concrete problems that mathematics will
be developed; in the end, problem solving and theory building go hand in
hand. That’s why Hilbert risked offering a list of unsolved problems instead
of presenting new methods or results, as was usually done at meetings. “He
who seeks for methods without having a definite problem in mind seeks for
the most part in vain,”7 Hilbert told his Paris audience.
Let us examine Hilbert’s career from the first stages up to the Paris lec-
ture. From 1880 to 1884 Hilbert studied at the University of Königsberg in
East Prussia, far from European scientific centers. At that time, Adolf Hur-
witz (1859–1919) (three years older than Hilbert) was appointed professor in
Königsberg, and Hermann Minkowski (1864–1909) (two years younger) was a
brilliant student and close friend of Hilbert’s at Königsberg. It was this mathe-
matical community that had great influence on Hilbert the student. Ferdinand
Lindemann (1852–1939), famous for his 1882 proof that π is not an algebraic
but a transcendental number, had been professor at Königsberg since 1883.
Under Lindemann’s influence Hilbert became interested in the then-flourishing
theory of invariants, the area of research in which he wrote his “Inauguraldis-
sertation” (his Ph. D.) Über die invarianten Eigenschaften specieller binärer
Formen (On the Invariant Properties of Special Binary Forms) in 1884. Most
of Hilbert’s work was devoted to the theory of algebraic invariants during his
Königsberg period.
This theory of invariants appears also as an example in the canceled 24th
problem, and therefore we will go into some details. In the last decades of the
19th century, besides group theory, above all invariants show how structural
4
Weyl 1968, p. 466.
5
Weyl 1944, p. 615.
6
Blumenthal 1922, p. 67. “Hilbert ist der Mann der Probleme. Er sammelt und
löst vorhandene, er weist neue.”
7
Hilbert 1900/1901, p. 32/444. “Denn wer, ohne ein bestimmtes Problem vor
Augen zu haben, nach Methoden sucht, dessen Suchen ist meist vergeblich.”
Fig. 10.4. Hilbert’s lecture notes for his first lecture course in winter term 1886 in
Königsberg. Title page and the facing page with an inscription.
Left: Colleg Invariantentheorie, Dreistündig nebst einer Uebungsstunde (Lecture
Course: Theory of Invariants, three hours a week and one hour for exercising).
Right: The large handwriting is a typical example of Hilbert for corrections and
insertions, usually done with a thick pencil in blue. The meaning is:
Slow development, step by step,
Small and clear handwriting [on the desk],
Watch the audience to see if they understood.
Courtesy Niedersächsische Staats- und Universitätsbibliothek Göttingen, Hand-
schriftenabteilung, Cod. Ms. D. Hilbert 521.
and abstract thinking was developed in mathematics. Invariants are quanti-

ties, entities, relationships, properties, etc., that are unaltered by particular
transformations. For example, all geometric facts that are independent of the
coordinate system are invariants in Euclidean geometry, such as magnitudes
that are not altered by geometric transformations like rotation, dilatation,
and reflection. A more general geometric invariant property is that of a cross
ratio in projective geometry (with no metrical properties) which is invariant
under projectivities. In each geometry every invariant property is based on a
certain group of transformations. The task of a geometry in question consists
in setting up the invariants of this group. In the language of modern algebra,
the theory of invariants deals with linear groups G which act on an n-space
(field) K and the polynomials p(x) = p(x1 , x2 , . . . , xn ) on K that are invariant
under G.
On the other hand – analytically expressed – such invariants are invariants
of tensors: in the terminology of Hilbert’s time, invariants of an n-ary form F
of degree m under transformations. For example, this is a binary form in x1
and x2 of degree three:
ax31 + bx21 x2 + cx1 x22 + dx32 .
The central questions in the theory of invariants are whether there is a system
of invariants in which each invariant can be represented rationally and whether
there is a finite system of this type.
Fig. 10.5. David and Käthe (née Jerosch) Hilbert, in the year of their marriage,
1892. Courtesy Niedersächsische Staats- und Universitätsbibliothek Göttingen,
Handschriftenabteilung, Voit Collection.
Invariant theory was the main research field of Paul Gordan (1837–1912),
who was regarded as the greatest expert in the field (“the king of invariants”).
Gordan developed many constructive techniques for the representation and
generation of invariants, among them his finite basis theorem saying that the
invariants of a system of binary forms (with arbitrary many variables) possess
a finite basis (1868). Despite the efforts of Gordan it was an open question
whether such finite bases exist for forms of arbitrary order. In 1888 the solution
came with Hilbert.
From his summer resort on August 28, 1888, Hilbert wrote a letter to
Hurwitz.8 After reporting on a boating trip with four young ladies, among
them Miss Käthe Jerosch (1864–1945) (later Mrs. Hilbert), David Hilbert said:
“Nevertheless my algebraic-arithmetic questions don’t sleep.”9 He informed
Hurwitz that when he was finishing a paper on invariant theory, he had found
a new approach to Gordan’s problem, considering its pure algebraic kernel.
Already in the first days of September Felix Klein (1849–1925) received the
improved paper, and Hilbert asked for a speedy publication in the Göttinger
Nachrichten to secure his priority of the new “powerful methods”.10 For any
form F with arbitrary degree Hilbert showed that its invariants form a ring
R = K[x] with a finite basis i1 , i2 , . . . , ik ; i.e., the ring is generated by this
basis or, in other words, any invariant i belonging to the ring R is expressible
as a polynomial in this basis: i = P1 i1 + · · · + Pk ik , where Ps (s = 1, . . . , k)
are polynomials with a degree lower than i (Hilbert’s basis theorem).
The numerous papers on invariant theory of the time consisted of masses
of endless algorithmic calculations, and the authors concerned with these cal-
culations had not recognized the general law. They were, so to speak, unable
to see the wood for the trees. Through Hilbert’s insight this presentation of
the theory of invariants was modified in an essential way. He proved its general
theorems in a few pages, especially the most important theorem that every
invariant of a given configuration can be expressed by a rational combination
of a finite number of them. Hilbert’s approach to invariant theory was quite
different; the theory was transformed from what it had been in the hand of
Gordan. His proof was based on existence procedures and therefore provided
no method for constructing the basis in a given case. In resolving the prin-
cipal problems of invariant theory in his own way Hilbert “had dealt it11 a
mortal blow”, as Jean Dieudonné (1906–1992) said. However, in doing so he
had laid the foundations of polynomial ideals and, moreover, prepared the
way for modern algebra as developed later by, for instance, Emmy Noether
(1882–1935), who was Gordan’s only doctoral student, and Emil Artin (1898–
1962), who studied with Gustav Herglotz (1881–1953) in Leipzig. Gian-Carlo
8
Niedersächsische Staats- und Universitätsbibliothek Göttingen, Handschriften-
abteilung, Cod. Ms. Math.-Archiv 76, no. 229.
9
Letter to Hurwitz from August 28, 1888. Ibid., no. 229. “Trotzdem schlafen
meine mathematischen Ideen nicht.”
10
Hilbert 1888, 1889; English translations Hilbert 1970. Göttinger Nachrichten =
Nachrichten der königlichen Akademie der Wissenschaften in Göttingen.
11
“It” means the symbolic methods of invariant theory, i.e. proofs by calculation.
Dieudonné 1971 begins the preface with the line: “Invariant theory has already
been pronounced dead several times, and like the phoenix it has been again and
again rising from its ashes.” Already Weierstraß told Hilbert in 1888: “In invariant
theory many will go to ruin but not of it alone (Untergehen werde auch vieles
in der Invariantentheorie, aber nicht von ihr allein)”, Niedersächsische Staats- und
Universitätsbibliothek Göttingen, Handschriftenabteilung, Bericht über meine Reise
(1888), Cod. Ms. D. Hilbert 741, p. 1/7.
Rota (1932–1999) even regarded Lie theory and algebraic geometry as off-
springs of invariant theory.12 From that time on, about the mid-1880’s, in
Hilbert’s research discoveries of the first order were to follow one another in
rapid succession for more than twenty years.
Hilbert concluded a paper, a brief résumé of his papers in invariant theory,
read in his absence at the International Mathematical Congress in Chicago13
on August 22, 1893, with historical remarks:
In the history of a mathematical theory three periods can easily and
clearly be distinguished: the naive, the formal, and the critical. As to
the theory of algebraic invariants, its founders Cayley and Sylvester
are both representatives of the naive period. [. . . ] The discoverers
and perfecters of the symbolic calculus Clebsch and Gordan are the
representatives of the second period, whereas the critical period has
found its expression in the above mentioned theorems.14
The theorems Hilbert referred to were his own. Until then the criterion of
mathematical existence had been constructibility. Hilbert’s revolutionary ap-
proach, consisting of pure existence proof, ignored this criterion and perplexed
his colleagues, above all the great algorithmician Paul Gordan, who exclaimed:
“That is not mathematics, that is theology (Das ist nicht Mathematik, das
ist Theologie!).” Felix Klein, however, appreciated Hilbert’s approach at once:
“the matter is obviously very important.”15
Hilbert, from 1892 an appointed professor at Königsberg, was offered a
chair at the University of Göttingen in 1895, exactly one hundred years after
Carl Friedrich Gauss (1777–1855) had enrolled in Göttingen, mainly because
of his results in the theory of invariants. Hilbert accepted and remained there
12
Rota, 1999. Two Turning Points in Invariant Theory. The Mathematical In-
telligencer 21, 1 (1999) 22-28; a modern presentation for example is V.L. Popov
1992
13
This Congress was part of the World’s Columbian Exposition which was held on
the occasion of the 400th anniversary of Columbus’s discovery of America in 1893.
It was attended by 45 mathematicians; four were from abroad, among them Felix
Klein, who brought a number of European contributions. The Congress in Zurich in
1897 is regarded as the first International Mathematical Congress; the last but one
took place in Berlin 1998, the last in Beijing 2002.
14
Hilbert 1893, p. 124 (the given translation is based on that of Reid 1996,
p. 34). “In der Geschichte einer mathematischen Theorie lassen sich meist 3 En-
twicklungsperioden leicht und deutlich unterschieden: Die naive, die formale und
die kritische. Was die Theorie der algebraischen Invarianten anbetrifft so sind die
ersten Begründer derselben, Cayley und Sylvester, zugleich auch als die Vertreter
der naiven Periode anzusehen. [. . . ] Die Erfinder und Vervollkommener der sym-
bolischen Rechnung Clebsch und Gordan sind die Vertreter der zweiten Periode,
während die kritische Periode in den oben genannten Sätzen [. . . ] ihren Ausdruck
findet.”
15
Frei 1985. Letter to Hilbert from October 1, 1888., p. 43 (no. 32). “Die Sache
ist offenbar sehr wesentlich.”
Fig. 10.6. Hilbert’s home in Göttingen, 29 Wilhelm-Weber-Straße, from 1897. The

lecture halls and the Mathematical Institute were in walking distance of the house.
Photo R. Thiele.
until his death in 1943 – almost half a century. Before Klein and Hilbert there
had been brilliant mathematicians at Göttingen, for instance Carl Friedrich
Gauss, Johann Peter Dirichlet (1805–1859), and Bernhard Riemann (1826–
1866); but after Riemann’s death in 1866 Göttingen had become a backwater
compared to Prussian Berlin, which housed such luminaries as Jakob Steiner
(1796–1863), Eduard Kummer (1810–1893), Leopold Kronecker (1823–1891),
and Karl Weierstrass (1815–1897). However, Göttingen’s bygone mathemati-
cal tradition was restored. The second flower prepared by Felix Klein achieved
even greater eminence, largely because of Hilbert, who made Göttingen the
leading center of mathematics in Germany.
In his Obituary on Hilbert, Hermann Weyl remarked that Hilbert concen-
trated his energies and focused them on a new area.16 According to the years
of publication of Hilbert’s research we have six (almost) sharply distinguished
periods. We give some examples:
16
Weyl 1944, p. 617.
until 1893 algebraic forms:

Über die Theorie der algebraischen Formen.
Mathematische Annalen 36 (1890), 473–531.
Über die vollen Invariantensysteme.
Über die Theorie der algebraischen Invarianten, paper read
at the International Mathematical Congress in Chicago 1893.
New York: Macmillan 1896, pp. 116–124.
1894–1899 algebraic number theory:
Die Theorie der algebraischen Zahlkörper [=Zahlbericht].
Jahresbericht der Deutschen Mathematiker Vereinigung
4 (1897), 157–546.
1899–1903 foundations of geometry (axiomatic method):
Grundlagen der Geometrie.
Leipzig: Teubner 1899, 14th ed. 1999.
1903–1912 analysis:
Grundzüge einer allgemeinen Theorie der linearen
Integralgleichungen, 6 Mitteilungen, in: Nachrichten der
königlichen Akademie der Wissenschaften in Göttingen
1904–1910; as a book Leipzig: Teubner 1912.
1912–1928 mathematical physics:
Begründung der kinetischen Gastheorie.
Die Grundlagen der Physik.
Nachrichten der Akademie der Wissenschaften in Göttingen
1915 and 1916, again in Mathematische Annalen 92 (1924), 1–32.
R. Courant/D. Hilbert [Hilbert only pro forma],
Methoden der mathematischen Physik. 2 vols.
Berlin: Springer 1924 and 1937.
Über die Grundlagen der Quantenmechanik
(with J. v. Neumann, L. Nordheim),
after 1918 foundations of mathematics:

Axiomatisches Denken.
Neubegründung der Mathematik [1. Mitteilung].
Abhandlungen aus dem Mathematischen Seminar
der Universität Hamburg 1 (1922), 157–177.
Die Grundlagen der Mathematik [2. Mitteilung].
Abhandlungen aus dem Mathematischen Seminar
der Universität Hamburg 6 (1928), 64–85.
Die Grundlegung der elementaren Zahlenlehre.
Mathematische Annalen 104 (1931), 485-494.
Grundzüge der theoretischen Logik
(with W. Ackermann). Berlin: Springer 1928.
Grundlagen der Mathematik
(with P. Bernays). 2 vols. Berlin: Springer 1934 and 1939.
However, if we take into account that Hilbert’s choice in his famous speech
surveyed nearly all the mathematics of his day, our division into certain special
fields simply shows the printed results of Hilbert’s scientific activities during
these periods, not the enormous diversity of his actual interests. A tolerable
overview of Hilbert’s interests cannot be obtained from the published sources
alone; we must study his drafts of lectures, corresponding lecture notes, his
notebooks, and correspondence. Fortunately the Hilbert Nachlass is almost
complete. The Nachlass and many other sources are held in the Library of the
University of Göttingen (Niedersächsische Staats- und Universitätsbibliothek
Göttingen, Handschriftenabteilung), in the Mathematical Institute of the Uni-
versity of Göttingen, and in the Staatsbibliothek Berlin (Nachlässe Born and
Hückel). Among the items of the Nachlass there are three mathematical note-
books17 containing his handwritten notices from 1886. From our point of view,
however, rather surprisingly we find (almost) no references to any of the 23
problems in these notebooks.
10.1 How did Hilbert’s Paris Talk Come About?

Like any other science, mathematics is an international matter. To put it in
Hilbert’s own words: “Mathematics knows no races or geographic boundaries;
for mathematics, the whole cultural world is a single country.”18 However, it
was necessary to wait for the development of means of transportation in the
second half of the 19th century in order to be able to organize international
meetings where mathematicians were brought together to communicate in
17
abteilung. Cod. Ms. D. Hilbert 600:1-3.
18
Reid, 1996, p. 188. “Die Mathematik kennt keine Rassen oder geographische
Grenzen, denn für sie ist die gesamte kulturelle Welt ein einziges Land.”
person rather than writing letters (which indeed had been effective in the
days of Father Mersenne (1588–1647)).
Fig. 10.7. Map of Europe before World War I. (In this projection the three towns
Königsberg, Göttingen, and Paris related to Hilbert are collinear.) Based on a map
of the Karl-Sudhoff-Institut, Universität Leipzig.
In August 1897, 209 mathematicians gathered in Zurich for the first Inter-
national Congress of Mathematicians; the second ICM met in Paris in 1900,
with 262 participants attending. For each ICM it has been customary to invite
some mathematicians to deliver lectures on special topics. At Zurich Poincaré
delivered his speech “Sur les rapports de l’analyse pure et de la physique
mathématique (On the relations between pure mathematics and mathematical
physics)”.19 In winter 1899-1900 Hilbert, one of the most respected German
mathematicians of the day and nearly 38 years old, was invited to make one of
the major addresses in the opening session of the coming ICM in Paris. Hilbert
hesitated whether he should reply to Poincaré’s Zurich lecture or choose an-
other subject and he asked for Minkowski’s opinion on a report on individual
problems. On January 5, 1900, his friend wrote:
19
Rudio 1898, pp. 81–90.
Fig. 10.8. Report on the Paris Congress by A.S. Scott, Bulletin of the Ameri-
can Mathematical Society 7 (1900). Courtesy Mathematisches Institut, Universität
Leipzig.
Most alluring would be the attempt at a look into the future and a
listing of the problems which mathematicians should try themselves
during the coming century. With such a subject you could have people
talking about your lecture decades later.20
In the end Minkowski was right, but Hilbert was still wavering, so he consulted
Hurwitz on March 29:
I must start preparing for a major talk at Paris, and I am hesitating
about a subject. [. . . ] The best would be a view into the future. What
20
Letter to Hilbert from January 5, 1900. Niedersächsische Staats- und Univer-
sitätsbibliothek Göttingen, Handschriftenabteilung, Cod. Ms. D. Hilbert 258: 76;
English translation in Reid 1996, p. 69. “Am anziehendsten würde der Versuch
eines Vorblicks auf die Zukunft sein, also eine Bezeichnung der Probleme, an welche
sich die künftigen Mathematiker machen sollten. Hier könntest Du unter Umständen
erreichen, dass man von Deiner Rede noch nach Jahrzehnten spricht.”
Fig. 10.9. Letter from Minkowski (Zurich) to Hilbert from January 5, 1900. Cour-
tesy Niedersächsische Staats- und Universitätsbibliothek Göttingen, Handschriften-
abteilung, Cod. Ms. D. 258, no. 76.
do you think about the likely direction in which mathematics will

develop during the next century? It would be very interesting and
instructive to hear your opinion about that.21
We have no record of Hurwitz’s reply. In fact Hilbert hesitated rather long

and did not decide before the deadline. The mailed program for the Congress
included neither an announcement of a major lecture nor any other contri-
bution from Hilbert. Minkowski was disappointed: “Without your lecture the
program of the Paris Congress was a great disappointment. The desire on my
part to travel to the congress is now almost gone.”22
In the middle of July, Hilbert surprised his friends Minkowski and Hurwitz
(both professors at Zurich) with proofs of a paper entitled “Mathematische
Probleme” – the complete version of his Paris talk worked out for publication
in the Göttinger Nachrichten. Both friends read the proofs carefully and made
21
I did not find this letter to Hurwitz in the libraries of either the University of
Göttingen or of ETH Zürich; I quote the English translation in Reid 1996, p. 70.
22
Letter to Hilbert, June 22, 1900. Niedersächsische Staats- und Universi-
t”atsbibliothek Göttingen, Handschriftenabteilung, Cod. Ms. D. Hilbert 258: 81.
English translation partly by Reid 1996, p. 70. “Das Programm des Pariser Con-
gresses ohne Deinen Vortrag war für mich eine grosse Enttäuschung. Fast ist mir
überhaupt die Lust, zum Congress hinzugehen, vergangen.”
suggestions for the presentation. Later in a letter to Hurwitz in view of the

invariants Hilbert noticed that in his haste he had failed to give credit to
Hurwitz (in problem 14) in the paper and begged Hurwitz for pardon. He
then mentioned, amused, that Minkowski had even used the opportunity of
proof-reading to insert his own results (probably into problem 5).23
Fig. 10.10. Hermann Minkowski (1883) in the year he was awarded the prize of
the Paris Academy. Courtesy Niedersächsische Staats- und Universitätsbibliothek
Göttingen, Handschriftenabteilung, Voit Collection.
On one point both colleagues agreed: the lecture was too long. Minkowski
wrote on July 17 and 28 respectively:
The section on the calculus of variations, to wit: the formulas might
be better placed in a note at the end of the lecture.
23
Letter to Hurwitz from November 21, 1900. Niedersächsische Staats- und Uni-
versitätsbibliothek Göttingen, Handschriftenabteilung, Cod. Ms. Math.-Archiv 76,
no. 278. “Minkowski hat sich sogar selbst beim Correcturlesen an einer Stelle hinein-
verbessert.”
Hurwitz and I started under the erroneous impression that a stop

was to be put to Ignorabimus, with the calculus of variations neatly
dealt with. And now you go and take a lease on mathematics for the
twentieth century, and will go down as Managing Director.
Actually I believe that through this lecture, which indeed every math-
ematician in the world without exception will be sure to read, your
attractiveness for young mathematicians will increase – if that is pos-
sible.24
A week later Hilbert met Minkowski in Paris; the disappointed Hurwitz

did not attend. In the morning of August 8 Hilbert’s lecture was delivered at
a joint session of two sections “Bibliographie et Histoire” and “Enseignement
et Méthodes” chaired by the German historian of mathematics Moritz Cantor
(1829-1920).25 Hilbert introduced his 23 problems with a longer essay, but
he followed the advice to shorten the lecture, and presented a selected list
of only 10 problems. Moreover, he canceled a 24th problem, which has not
been published until now. In September 1900, Hilbert’s talk, including the
complete list of 23 problems, was published in the Göttinger Nachrichten and
later in slightly revised versions (mainly for the 23rd problem), and in French
and English translations in 1902.26
After the Congress Hilbert took a holiday and traveled to the Baltic Sea
near Königsberg. He felt somewhat unsatisfied with the Paris Congress for two
reasons. The discussion after his lecture was rather disappointing: Giuseppe
Peano (1858–1932) declared that a future paper of Alessandro Padoa (1868–
1937) would give an answer to the second problem (consistency of arithmetic
axioms), and Rudolf Mehmke (1857–1944) obscurely insisted he had already
proposed some monographs concerning the solution of equations of the seventh
degree (problem 13).
24
Letters to Hilbert. Niedersächsische Staats- und Universitätsbibliothek Göt-
tingen, Handschriftenabteilung, Cod. Ms. D. Hilbert 258, nos. 83 & 84. “Der Ab-
schnitt über Variationsrechnung, namentlich die Formeln sind wohl besser in eine
Anmerkung hinter den Vortrag zu verweisen.” (17. 7. 1900) – “Hurwitz und ich
hatten uns zunächst ein ganz falsches Bild gemacht, indem wir dachten, mit dem
Ignorabimus sollte Schluss gemacht werden, namentlich da die Variationsrechnung
schon so genau abgehandelt war. Nunmehr hast Du wirklich die Mathematik für das
20te Jahrhundert in Generalpacht genommen und man wird Dich allgemein gern als
Generaldirector anerkennen. - Namentlich glaube ich, dass Deine Anziehungskraft
auf junge Mathematiker durch diese Rede, die wohl jeder Mathematiker ohne Aus-
nahme lesen wird, wenn überhaupt möglich noch wachsen wird.” (28. 7. 1900) Eng-
lish translation of the last quotation in Reid 1996, p. 72.
25
Moritz Cantor, Germany’s leading historian of mathematics and no relative of
Georg Cantor (the creator of set theory), is best known for his History of Mathe-
matics (Leipzig: B.G. Teubner 1880-1908) in four volumes.
26
Cf. Grattan-Guinness 2000.
Fig. 10.11. The program of Paris Congress for the sections “Bibliographie et His-
toire” and “Enseignement et Méthodes” on August 8, 1900 with the announcement
of Hilbert’s lecture at 9 a.m. In: Compte Rendu deuxième Congrès International des
Mathématiciens. Paris: Gauthier 1902. Courtesy Mathematisches Institut, Univer-
sität Leipzig.
Hilbert believed that science is also to be propagated orally; books alone

are infertile.27 This is why in a letter28 to Hurwitz written immediately after
27
Mathematische Notizhefte. Niedersächsische Staats- und Universitätsbibliothek
Göttingen, Handschriftenabteilung, Cod. Ms. D. Hilbert 600:2, p. 99. “Die Wis-
senschaft wird auch mündlich übertragen, nur aus Büchern ist unfruchtbar – so
etwa.”
28
Letter to Hurwitz from August 25, 1900. Niedersächsische Staats- und Univer-
sitätsbibliothek Göttingen, Handschriftenabteilung, Cod. Ms. Math.-Archiv 76, no.
272.
Fig. 10.12. Letter from Hilbert to Hurwitz (Zurich) written from Hilbert’s hol-
iday place at the Baltic Sea near Königsberg on August 25, 1900. The Paris
Congress ended on August 11, 1900. Courtesy Niedersächsische Staats- und Uni-
versitätsbibliothek Göttingen, Handschriftenabteilung, Math-Archiv 76, no. 272.
the Conference, Hilbert regretted the attendance was poor in quantity and
in quality because for some reason important French mathematicians were
absent. One evening Poincaré even disappeared instead of chairing a banquet.
Hilbert also complained that there were not enough rooms for informal meet-
ings. As for the upcoming Congress at Heidelberg, he remarked, at any rate
the Germans must make a more efficient organization.29
The Paris address intertwined concrete but important problems with the
theoretical context. It is this interlocking character which made the collection
so fruitful. Hilbert’s “leitstern” (lode-star) in research was to find that spe-
cial case which contains all the germs of generality. His leitmotif was to start
investigations (or lectures) with an elementary but instructive example. One
outstanding feature of Hilbert’s works is that he liked to explain general meth-
ods through examples, leaving enough questions for other researches. These
are the deeper reasons why Hilbert built up an important school and why
Hilbert’s list of problems charted the course of mathematics during the 20th
century.
Noteworthy, however, is the fact that neither Hilbert himself nor any of
his disciples did work exclusively on the 23 problems. Among those who
contributed to solutions of the 23 problems we find the following students
of Hilbert: Max Dehn (1878–1952, Ph. D. with Hilbert 1899), Teiji Takagi
(1875–1960), Georg Hamel (1877–1954, Ph. D. with Hilbert 1901), Paul Funk
(1886–1969, Ph. D. with Hilbert 1911), Erich Hecke (1887–1947, Ph. D. with
Hilbert 1910), Richard Courant (1888–1972, Ph. D. with Hilbert 1910), Emil
Artin (1898–1962), Herbert Busemann (1905–1994), Gerhard Gentzen (1909–
29
Letter to Hurwitz from August 25, 1900. Ibid., no. 272.“Wir müssen uns mit
der Vorbereitung jedenfalls mehr Mühe geben und eine bessere und einheitlichere
Organisation ins Werk setzen.”
Fig. 10.13. Lecture hall (Auditorium) of the University of Göttingen (at the
cross roads Weender Straße and Berliner Straße) in which H.A. Schwarz, F.
Klein, D. Hilbert, H. Minkowski, C. Runge, E. Landau, and others lectured. The
new Mathematical Institute, Bunsenstraße, was completed in 1929 with the sup-
port of the Rockefeller Foundation. Courtesy Niedersächsische Staats- und Univer-
sitätsbibliothek Göttingen, Handschriftenabteilung.
1945), and others. Three months after the Paris lecture Hilbert reported the
first solution of one problem (number 3) by Max Dehn who—so he wrote to
Hurwitz—is “one of my best disciples” and “with his results [in his Ph. D. in
1899] I am totally delighted”.30 Dehn was the first to solve a Hilbert prob-
lem. Already one year later in Münster he wrote his Habilitationsschrift and
became a Privatdozent.
10.2 On the Problems

Hilbert had not had time to deal with all 23 problems, nor did his list cover all
branches of mathematics.31 Moreover, in each survey, complete or incomplete,
we can regard the problems from different points of view. So it becomes quite
clear that it will be impossible to give an adequate summary or an appreciation
of the problems and their sequels here. Hilbert told his audience:
The problems mentioned are merely samples of problems; yet they are
sufficient to show how rich, how manifold and how extensive mathe-
matical science is today, and the question is urged upon us whether
30
Letter to Hurwitz from November 5–12, 1899. Ibid. no. 275 (sheet 588); “einer
meiner besten Schüler”, “über dessen Resultate ich ganz entzückt bin”. In the letter
from November 21, 1900 this appreciation is repeated (ibid., no. 278).
31
See Grattan-Guinness 2000; for the sequels see Yandell 2002.
Fig. 10.14. David Hilbert about 1900. Courtesy Niedersächsische Staats- und Uni-
versitätsbibliothek Göttingen, Handschriftenabteilung, Voit Collection.
mathematics is doomed to the fate of those other sciences that have

split up into separate branches. [. . . ] I do not believe this nor wish it.
Mathematical science is in my opinion an indivisible whole. [. . . ] The
organic unity of mathematics is inherent in the nature of this science,
for mathematics is the foundation of all exact knowledge of natural
phenomena. [. . . ] May the new century bring it gifted prophets and
many zealous and enthusiastic disciples.32
32
The last sentences of the Paris lecture, English translation Newson 1902, p. 479.
“Die genannten Probleme sind nur Proben von Problemen; sie genügen jedoch, um
uns vor Augen zu führen, wie reich, wie mannigfach und wie ausgedehnt die math-
ematische Wissenschaft schon heute ist, und es drängt sich uns die Frage auf, ob
der Mathematik einst bevorsteht, was anderen Wissenschaften schon längst wider-
fahren ist, nämlich daß sie in einzelne Teilwissenschaften zerfällt. [. . . ] Ich glaube
und wünsche dies nicht; die mathematische Wissenschaft ist meiner Ansicht nach
ein unteilbares Ganzes. [. . . ] Der einheitliche Charakter der Mathematik liegt im
inneren Wesen dieser Wissenschaft begründet; denn die Mathematik ist die Grund-
lage alles exacten naturwissenschaftlichen Erkennens. [. . . ] Mögen ihr [der Mathe-
Furthermore, Hilbert told his audience of what nature problems should be.
Instead of the lecture I quote a similar remark from his notebook:
The problems must be difficult and clear – but not easy and com-
plicated, because confronted with them we would be helpless or we
would need some exertion of our memory to bear all the assumptions
and conditions in mind.33
Hilbert told a related story:
An old French mathematician [probably Hermite] said: “A mathemat-

ical theory is not to be considered complete until you have made it so
clear that you can explain it to the first man whom you meet on the
street.”34
I would like to update this statement: you should be content with the fact
there is a person on the same floor of your institute who possibly understands
you.
Today, with insight, we know that no problem was trivial; all were inter-
esting and fertile. Even so, despite the great impact of Hilbert’s problems, we
should not regard him as a prophet for the future of mathematics. Hilbert
himself regarded the statement “absolutely accurate prophecies are impossi-
ble”35 as an axiom. Indeed, the problems do not indicate that Hilbert would
have foreseen the rapid development of functional analysis in the following
decade, to which he himself contributed the theory of integral equations.
The difficulty as well as the length of the problems varies. The shortest
consists of only six lines, but it took six decades to find its solution; some of
the problems are as yet unsolved (nos. 8, 12, 13; no. 8 includes the Riemann
matik] im neuen Jahrhundert geniale Meister erstehen und zahlreiche in edlem Eifer
erglühende Jünger!”
33
Göttingen, Handschriftenabteilung, Cod. Ms. D. Hilbert 600:1, p. 55. “Die Prob-
leme müssen schwierig und einfach - nicht leicht und kompliziert sein, so dass man
zunächst rathlos vor ihnen steht - nicht so, dass man schon das Gedächtnis anstren-
gen muss, um bloss alle Voraussetzungen und Bedingungen zu behalten.”
34
At the beginning of his Paris lecture “Mathematische Probleme”, Hilbert 1900,
p. 254. English translation by Newson 1901, p. 479. “Ein alter französischer Math-
ematiker hat gesagt: Eine mathematische Theorie ist nicht eher als vollkommen
anzusehen, als bis du sie so klar gemacht hast, daß du sie dem ersten Mann erklären
könntest, den du auf der Straße triffst.”
35
Mathematische Notizhefte, Niedersächsische Staats- und Universitätsbibliothek
Göttingen, Handschriftenabteilung, Cod. Ms. D. Hilbert 600:3, inserted pages.
“Nimm das Axiom: Absolut richtige Prophezeiungen sind unmöglich (auch für den
Laplaceschen Weltgeist?) (Postulate the axiom: absolutely accurate prophecies are
impossible (also for Laplace’s demon?)).”
Fig. 10.15. List of the lectures Hilbert delivered from winter term 1897 until
winter term 1899, written by Hilbert’s wife Käthe. The numbers before the lec-
tures indicate the number of hours per week. Part of Hilbert’s “Verzeichnis meiner
Vorlesungen, 1886-1930 (List of my lectures)”; Hilbert’s list is incomplete. Cour-
tesy Niedersächsische Staats- und Universitätsbibliothek Göttingen, Handschriften-
abteilung, Cod. Ms. D. Hilbert 520.
hypothesis); and at least two problems are so general that they do not have
an ultimate solution:36
36
Hilbert 1900, pp. 253-297, and 1935, vol. 3, pp. 290-329; cf. Bieberbach 1930,
Aleksandrov 1969, Fang 1970, Browder 1976, Gray 2000, Grattan-Guinness 2000,
Yandell 2002.
no. 6: the mathematical treatment of the axioms of physics, and

no. 23: the further development of the methods of the calculus of variations.
We can (roughly) divide the problems into four groups:

Actual Lecture Published Version Canceled
Problems of foundation 1, 2, 6 24
Problems of analysis 19, 21, 22 20, 23
Problems of geometry 3, 4, 15, 18
Problems of arithmetic
and algebra 7, 8, 13, 16 5, 9, 10, 11,
12, 14, 17
Once again, Hilbert’s rapidly growing international fame was caused
largely by his lucky choice of mathematical problems, not by the more usual
practice of presenting new solutions, methods, or results. Of course, problems
are the lifeblood of science, but in general answers and not questions are more
estimated, expected and above all honored. That is our attitude not only in
science but also in everyday life. Let me give as a simple example a fictitious
but typical dialogue between a father and his child:
Dad, why is the grass green?

Mm, no idea.
Dad, why is the sky blue?
I don’t know that either.
Dad, do I disturb you? You don’t mind me asking all those questions,
do you?
Son, if you don’t ask questions how are you going to learn anything?37
You will probably be amused by Dad’s last answer because you have not
expected a logical, but a psychological response like: “Be quiet.” We are,
however, in a logical context, not in real life.
At the fourth ICM in Rome in 1908 Poincaré said in his lecture “L’avenir
des mathématiques (The future of mathematics)”: “At one time there were
prophets of misfortune; they reiterated that all problems had been solved, that
after them there would be nothing but gleanings left. [. . . ] But,” he added,
“the pessimists have always been compelled to retreat, so that I believe there
are none left today.”38 Here Poincaré echoes precisely Hilbert’s conviction: “A
branch of science is full of life as long as it offers an abundance of problems;
37
English saying, cf. Vollmer 1993, p. 192.
38
Poincaré 1908, p. 167. “Il y a eu autrefois des prophètes de malheur. Il répétaient
volontiers que tous les problèmes susceptibles d’être résolus l’avaient été déjà, et
qu’après eux il n’y aurait plus qu’à glaner. [. . . ] Les pessimistes se trouvaient ainsi
toujours débordés, toujours forcés de reculer, de sort qu’à présent je crois bien qu’il
n’y en a plus.”. English translation in A. Weil 1971, p. 321.
a lack of problems is a sign of death.”39 And yet in his swan song at the
Königsberg Meeting in 1930 Hilbert mentioned with satisfaction the failure of
Auguste Comte (1798-1857) to pose an unsolvable problem (see next section).
André Weil (1906–1998) explained: “If logic is the hygiene of the mathemati-
cian, it is not his source of food; the great problems furnish the daily bread
on which he thrives.”40 At any rate to pose a problem is no small feat, above
all to pose an interesting and important one. In a speech “Die Naturgesetze
und die Struktur der Materie (Laws of nature and structure of matter)”41 in
1961, Werner Heisenberg (1901–1976) said of the ancient Greek philosophers
that, disregarding their insufficient and highly speculative answers, above all
their attainments in posing the right questions were incredible. Among the
philosophers above all it was Karl Popper (1902–1994) who emphasized that
progress depends on questions.
There is almost nothing to be compared with that which Hilbert had
undertaken: Hilbert’s choice of problems is unique, at least as the product of
a single mind. There have since been other compilations of problems in books
and papers, especially in single branches of mathematics, and some problem
columns in journals, among them:
H. T. Croft, Unsolved Problems in Geometry. New York: Springer 1991,
R. K. Guy, Unsolved Problems in Number Theory. New York: Springer
1981,
D. Mauldin, The Scottish Book. Basel: Birkhäuser 1989,
C. S. Ogilvy, Tomorrow’s Math. Unsolved Problems for the Amateur.
New York: OUP 1962,
G. Pólya and G. Szegö, Aufgaben und Lehrsätze aus der Analysis, 2 vols.
Berlin: Springer 1925; English translation: Problems and Theorems of
Analysis, 2 vols. New York: Springer 1998.
P. de Souza, Berkeley Problems in Mathematics. New York: Springer 1998,
W. Sierpinski, A Selection of Problems in the Theory of Numbers.
New York: Macmillan 1964,
D. Shanks, Solved and Unsolved Problems in Number Theory. Washington:
Spartan Books 1962.
H. Tietze, Gelöste und ungelöste mathematische Probleme (Solved and
Unsolved Problems). München: Beck 1949,
S. Ulam, A Collection of Mathematical Problems. New York: Interscience
1960.
In 1976 Jean Dieudonné inspired Felix Earl Browder (born 1927) to ask a
number of mathematicians to describe some unsolved problems in their fields.
39
Hilbert 1900/1901, p. 254/438. “Solange ein Wissenszweig überfluß an Prob-
lemen bietet, ist er lebenskräftig; Mangel an Problemen bedeutet Absterben oder
Aufhören der selbständigen Entwicklung.”
40
Weil 1971, p. 324.
41
Heisenberg 1971, p. 237.
The result, published under the title “Problems of Present Day Mathemat-
ics” in the two-volume book Mathematical Developments Arising from Hilbert
Problems,42 is likely the largest collection of important problems, at least
with respect to the number of represented branches. There is also an earlier
Russian edition, Problemy Gil’berta (Hilbert’s Problems),43 edited by Pavel
Sergeevich Aleksandrov (1896–1982) with comments by competent Russian
mathematicians. Both editions show the influence of Hilbert’s problems even
after seven decades. I restrict myself to quoting James Serrin (born 1926)
on the “The solvability of boundary value problems” in the AMS collection
Mathematical Developments Arising from Hilbert Problems (on problem 19,
which was omitted in the actual lecture):
Among the prophetic problems in Hilbert’s famous list one must surely
include the 20th, the general problem of boundary values for elliptic
partial differential equations. This subject, only a seedling in the year
1900, has burst into flower during our century, has developed in di-
rections Hilbert never imagined, and today encompasses a vast area
of work which to a mathematician of 75 years ago would seem little
short of astonishing.44
In the same collection Enrico Bombieri (born 1940) concluded his paper
“Variational problems and elliptic equations” in this way: “In this sense, it
can be said that Hilbert’s 19th problem has opened one of the most interesting
chapters in mathematics” (p. 434). James Serrin further remarked: “The 20th,
like so many of the others in Hilbert’s list, consisted as much in a program
as in a specific problem requiring some definite answer, and in just this fact
we can see one facet of Hilbert’s genius and breadth” (p. 507). Indeed, it
was the proof of Dirichlet’s principle for a specific problem that led Hilbert
to the general questions finally arising in the 19th and 20th problems. By
the concepts of existence in a generalized sense and of regularity he pointed
out two very important issues in the modern theory of partial differential
equations.
Inspired in part by Hilbert’s list and on behalf of the International Math-
ematical Union, Vladimir Igorovic Arnol’d (born 1937) recently wrote a letter
to some mathematicians asking for the description of great problems for the
next century. In response, on the occasion of Arnol’d’s 60th birthday, Steve
Smale (born 1930) gave a lecture on a “Conference in Honor of Arnol’d” at
the Fields Institute in Toronto in June 1997.45 His talk “Great Problems”
is published under the promising title “Mathematical Problems for the Next
42
Browder 1976.
43
Aleksandrov 1969.
44
Browder 1976, p. 507.
45
Cf. also Atiyah’s Fields lecture “Mathematics in the 20th Century”, delivered
in Toronto, Ont., in 2000. Video Tape of the Fields Institute, Toronto; printed in
N.T.M (N.S.) 10 (2002), 25-39.
Fig. 10.16. First introduction of Hilbert’s independent integral in the lecture

“Flächentheorie, II (Theory of Surfaces, II)”, summer term 1900, only a few weeks
before the Paris talk was delivered in which the integral is part of the 23rd problem.
Hilbert’s lecture note. Courtesy Niedersächsische Staats- und Universitätsbibliothek
Göttingen, Handschriftenabteilung, Cod. Ms. D. Hilbert 552, p. 12.
Century”46 and lists 18 problems, among them two of the Hilbert problems.
In Smale’s opinion the Riemann hypothesis, the Poincaré Conjecture, and the
question “Does P = NP?” are the three greatest open problems. The first two
belong completely to classical mathematics; the last also concerns computer
science.
Fig. 10.17. Kneser’s letter to Hilbert probably from September 2, 1900 (no exact
date is available) in which Kneser pointed out his own introduction of the indepen-
dent integral and desired to be mentioned in an appropriate way in the final version
of Hilbert’s paper “Mathematical Problems”. Courtesy Niedersächsische Staats- und
Universitätsbibliothek Göttingen, Handschriftenabteilung, Cod. Ms. D. Hilbert 180,
no. 4.
46
Smale 1998.
Incidentally, the problem of the zeroes of Riemann’s zeta function is usu-

ally considered to be the most important unsolved problem in mathematics,
and so thought Hilbert. Moreover, Hilbert went so far as to maintain that
it was absolutely the most important question for humanity. I cannot resist
telling you a story I found somewhere in the writings of George Pólya (1887–
1985), although the underlying German myth needs some initial explanation.
Frederick I (1152–1190), named Barbarossa (i.e. Redbeard), was a German
king and Roman emperor of the 12th century who died while on Crusade to
the Holy Land. Since the 14th century popular legend has it that he is only
sleeping in an imperial castle near Göttingen and every five hundred years
looks to see what is going on in Germany, emerging in case Germany needs
him.47 Returning to Hilbert, somebody allegedly asked him: “If you came
back, like Barbarossa, after five hundred years, what would you do?” Hilbert
replied at once: “I would ask, has somebody proved the Riemann hypothesis?”
Furthermore, consider the role of problems in the work of the four extra-
ordinary Hungarian mathematicians John von Neumann (1903–1957), Paul
Erdös (1913–1999), György Pólya, and Gabor Szegö (1895–1985), who at the
end of their careers taught in the USA. In 1954 Von Neumann lectured at the
Amsterdam ICM on “Unsolved Problems”, but he pointed out:
The total subject of mathematics is clearly too broad for any of us. I
do not think that any mathematician since Gauss has covered it uni-
formly and fully; even Hilbert did not and all of us are of considerably
lesser width quite apart from the question of depth than Hilbert.48
That is why he restricted himself to a particular area of mathematics:

operator theory viewed in its connections with other subjects. For the out-
standing mathematician Erdös, problems were a downright passion. He solved
problems, posed problems, and in the end he paid for problems. He paid for
solutions or refutations – sometimes, for a solution or a refutation of the same
problem, he even offered different amounts. Problem books are relatively re-
cent in mathematics. Among them an eminent and early example is the book
Aufgaben und Lehrsätze aus der Analysis (Springer’s Yellow Series, vols. 19
and 20; 1925) of Pólya and Szegö included in the above list. Furthermore,
Pólya wrote books on problem solving, among them How to Solve It?, and
there is also a film Let Us Teach It on a class he gave on this topic.49
Recent works on problems include Jean-Michel Kantor’s (born 1946)
“Hilbert problems and their sequels” and Shiing-Shen Chern’s (born 1911)
complement “Remarks on Hilbert’s 23rd Problem”50 , as well as Jeremy Gray’s
47
The story is like many others, among them that of England’s hero Francis Drake
(1540?–1596) and the legendary fame of Drake’s drum, described in a poem of Henry
Newbolt (1862–1938).
48
Redei 1999.
49
Pólya 1957; the film is distributed by the MAA.
50
Kantor 1996, Chern 1996.
Fig. 10.18. The hills of the Kyffhäuser Mountains are crowned by a ruined castle
which was one of the largest German castles, destroyed in 1178. The ruins are sur-
mounted by a monument (total height 64 m) showing Kaiser Wilhelm I (1797–1888),
Prussian King since 1861 and German Emperor since 1871, as well as the German
Emperor Frederick I (1122–1190) asleep within the mountain. The Kyffhäuser lies in
Thuringia on the South-East side of the Harz Mountains, whereas Göttingen lies on
the North-West side of the Harz Mountains, about 100 km in air distance. Photos
R. Thiele.
(born 1947) The Hilbert Challenge and Benjamin Yandell’s (born 1951) The
Honors Class.51 As Victor Katz (born 1943) wrote in 1993: “Hilbert’s prob-
lems have in fact proved to be central in twentieth-century mathematics. Many
have been solved, and significant progress has been achieved in the remain-
der. Perhaps a late-twentieth-century mathematician will present a new list of
problems at the International Congress of Mathematicians in 1988 in Berlin.”
In Katz’s last sentence resound the concluding words of Hilbert’s Paris lecture
which expressed the hope that well-posed problems will evoke enthusiasm and
inspiration among mathematicians.52 Of course, in Berlin there was no new
list presented.53 In 2000, however, the Clay Institute posed seven problems
for our century; for each solution the Institute will award $1,000,000.54
In concluding this report on collections of problems let us look back at
the 17th and 18th centuries with their rich heritage of important problems.
Famous problems of this period are Kepler’s sphere problem, Fermat’s last
theorem, Pascal’s cycloid problem, de Beaune’s problem, Viviani’s Floren-
tine enigma, Goldbach’s conjecture, Mascheroni’s constructions, etc. In these
centuries it was quite common to pose problems to the mathematical commu-
nity in public, so-called “provocationes” (provocations). Each mathematician
who solved such a provocation was authorized to pose another one, and so
on, bringing a cascade of questions into existence. Some of these problems
became famous: Leibniz’s problem of isochronous curves, Jakob Bernoulli’s
isoperimetric problems, Johann Bernoulli’s problem of the shortest line on a
surface, and others.
In 1696 John (Johann) Bernoulli (1667–1748) challenged “the most inge-
nious mathematicians of the whole terrestrial globe”55 with a new problem,
the problem of quickest descent, or, the Brachistochrone Problem.56 Guil-
laume François Antoine de l’Hôpital (1661–1701) declared the problem to be
one of “the most curious and most beautiful that has ever been proposed.”57
In another announcement John Bernoulli pointed out that nothing encour-
ages noble minds more than the praise of later ages, and that fame and glory
is all that a noble expects for his efforts.58 At this time John Bernoulli and
51
Gray 2000; Yandell 2002.
52
Hilbert 1900, p. 297. “Mögen ihr [Mathematik] im neuen Jahrhundert geniale
Meister erstehen und zahlreiche in edlem Eifer erglühende Jünger.” (cf. footnote 32).
53
Katz 1993/1998, p. 729/808.
54
The Clay Mathematics Institute, Millennium Prize Problems, announced May
4, 2000 at the Collège de France, Paris. Presented by J. Tate and Sir Michael Atiyah.
Cf. http://www.claymath.org/millennium/
55
“Acutissimis qui toto Orbe florent Mathematicus”, headline of a broadsheet
(Programm Editum Groningae), distributed in 1697; also in: Speiser 1991, p. 259.
56
“Problema novum”, added at the end of a paper in Acta eruditorum, June 1696,
p. 269; Speiser 1991, p. 212. Cf. Thiele 2002.
57
Spiess 1955, p. 319. “Ce probleme [me] paroist des plus curieux et des plus jolis
[que] l’on ait encore proposé.” Letter to Joh. Bernoulli from June 15, 1696.
58
“Programm Editum Groningae 1697”, also in Speiser 1991, p. 259.
Fig. 10.19. Figures illustrating the influential Brachistochrone Problem (1696) and
the Isoperimetric Problem (1697) of John and James Bernoulli respectively. Courtesy
Deutsche Akademie der Naturforscher. Leopoldina. Halle. Acta eruditorum, June
1696 and May 1697, Journal de Sçavans, August 1698.
his brother James (Jacob) (1654–1705) began to quarrel. The challenge of

the Brachistochrone Problem and above all the statement about reward by
the argumentative John Bernoulli were aimed especially at his brother. Also
quarrelsome, James (who solved the problem by a fundamental technique)
took his revenge for the provocation by introducing three further challenges,
among them the famous “isoperimetric problems.”59 In order to humble his
brother he publicly offered John the amount of 50 imperial ducats which—in
case John would solve the problems—were to be paid by an unnamed gen-
tleman.60 John for his part goaded and sneered at his brother. The situation
escalated to the well-known quarrel, ended in 1705 by James’s death. John,
thirsting for honor and glory, regarded offering payment to a mathematician
59
Speiser 1991. Solutio Problematum Fraternorum . . . cum Propositione reciproca
aliorum. Acta eruditorum, May 1697, pp. 211–217; also Speiser 1991, pp. 271–282.
60
Ibid., Speiser 1991, p. 276; cf. Thiele 1997a.
for the solution of a problem as a slander and wicked defamation.61 Bygone

times, tempi passati.
Just as the problems posed by the Bernoullis were significant for later
developments, although it is very difficult to select such problems, Hilbert
expected his chosen problems to play an important role in the future devel-
opment of mathematics. Hermann Weyl aptly compared Hilbert’s insight into
the future of mathematics with that of politics:
How much better he predicted the future of mathematics than any

politician foresaw the gifts of war and terror that the new century
was about to lavish upon mankind.62
10.3 Remarks on Hilbert’s Philosophy of Mathematics

On one hand Hilbert demanded certainty by formalizing (axiomatic theory),
but on the other hand he believed that mathematics advances by solving
problems. For this reason it is all too easy to regard Hilbert exclusively as
a pure formalist. Some of his remarks taken from his unpublished notebooks
reinforce this: “Where does mathematics begin? As soon as concepts are fit-
ted together, and only those facts that are contained in the concepts may be
employed further.”63 For that reason axiomatics play an essential role: “The
criterion for scientific method (truth) is axiomatizability. The axiomatic is
the rhythm that makes music of the method, the magic wand that directs all
the individual efforts to a common goal.”64 Incidentally, Hilbert answered the
old enigma “Why can the world be described by mathematics? Why is math-
ematical science possible?” with: “Between thought and action there is no
fundamental and no quantitative difference. This explains the pre-established
harmony [understood in the sense of Leibniz (1646–1716)] and that simple
61
In a letter to Legendre, July 2, 1830 C.G. Jacobi held a similar view: “Mais un
philosophe comme lui [Fourier] aurait dû savoir que le but unique de la science, c’est
l’honneur de l’esprit humain, et que sous ce titre, une question des nombres vaut
autant qu’une question du système du monde.” (But a philosopher like him ought
to have known that the sole aim of knowledge is the honor of the human mind and
that from this viewpoint a problem of number theory is as valuable as a problem of
the system of the universe.) Jacobi 1830/1875, p. 272f.; also in Jacobi’s Werke, vol.
1, p. 454f. Hilbert quoted this letter in his 1930 speech “The Knowledge of Nature”
(Hilbert 1930, in the last paragraph).
62
Weyl 1951/1968, p. 466.
63
Mathematische Notizhefte. Library of the University of Göttingen. Cod. Ms. D.
Hilbert 600:3, p. 116. “Wo fängt Mathematik an? Sobald Begriffe zusammengefügt
werden, und nur das in den Begriffen liegende weiterhin benutzt werden darf.”
64
Ibid., Cod. Ms. D. Hilbert 600:2, p. 45. “Kriterium für Wissenschaftlichkeit
(Wahrheit) ist die Axiomatisierbarkeit. Axiomatik ist der Rhythmus, der die Meth-
ode zur Musik macht - ist der Zauberstab, der alle die Einzelbestrebungen auf ein
gemeinsames Ziel richtet.”
experimental laws generate ever simpler theories.”65 So we live in the best

of all possible worlds, at least in the sense that the world has the simplest
(mathematical) description. Hilbert was convinced of this metaphysical prin-
ciple: there is a realm behind phenomena, and the universe is governed in such
a way that a maximum of simplicity and perfection is realized. In a notice we
read: “Thus pre-established harmony also, because nature does not make such
complicated things that cannot solved by mathematicians.”66
We have many sources for Hilbert’s deep belief that each well-posed prob-
lem can be solved. In opposition to the pessimism of the fin de siècle about
1900, Hilbert maintained his inspiring optimism which played a crucial role
in his research and philosophy. Hilbert proudly noticed in 1930 that Auguste
Comte’s effort to pose an unsolvable celestial problem was met with a solution
a few years later. And in conclusion he added the self-confident words:
The true reason, according to my thinking, why Comte could not find
an unsolvable problem lies in the fact that there is no such thing as
an unsolvable problem.67
In 1922 he concluded his lecture “Wissen und mathematisches Denken (Knowl-

edge and Mathematical Thinking)” with the famous words taken from his
Paris lecture in 1900: “Wir müssen wissen, wir werden wissen (We must know,
we shall know)” and again in a speech at Königsberg in 1930, partly repeated
on air by the local radio station. In the end this optimistic axiom of the
solvability of every problem was engraved on his tombstone in Göttingen.68
To the end of his career Hilbert repeatedly denied the “foolish ignora-
bimus”69 of Emil du Bois-Reymond (1818–1896) and his successors, which he
regarded as the battle-cry of slaves and reactionists, even as an intellectual
sadism.70 Du Bois-Reymond concerned himself with the limits of knowledge
65
Ibid., Cod. Ms. D. Hilbert 600:3, p. 95. “Zwischen Denken und Geschehen ist
kein prinzipieller und kein qualitativer Unterschied! Dadurch erklärt sich die praesta-
bilierte Harmonie und die Tatsache, dass einfache experimentelle Gesetze auch im-
mer einfachere Theorien ermöglichen.”
66
Ibid., Folder Quantentheorie, Cod. Ms. D. Hilbert 666.“Praestabilierte Har-
monie auch darum [,] dass [weil] die Natur so komplizierte Sachen nicht macht, wie
auch der Math. sie nicht lösen kann.” (cryptic German).
67
Reid 1996, p. 196.
68
If you arrive at Göttingen by train from south you pass very close to Hilbert’s
grave, as well as those of Max Born (1882–1970), Otto Hahn (1879–1968), and Max
Planck (1858–1947). The embankment and the cemetery, especially the line of these
graves, are separated only by hedges.
69
Keyword in the famous and widespread speech “Über die Grenzen der Natur-
erkenntnis (On the Limitations of Knowledge in Natural Sciences)” by the physiol-
ogist E. du Bois-Reymond (brother of the mathematician Paul du Bois-Reymond,
1831-1889), delivered in Leipzig in 1872.
70
Göttingen, Handschriftenabteilung, Cod. Ms. D. Hilbert 600:3, inserted pages.
Fig. 10.20. Hilbert’s gravestone in Göttingen, cemetery at the Groner Landstraße.

The optimistic lines “Wir müssen wissen. Wir werden wissen (We must know. We
shall know)” are engraved in the pedestal and barely legible. Photo R. Thiele.
of nature – a widespread question around 1900. In accordance with the pes-

simistic spirit of the “fin de siècle (end of the century)” he maintained that
there are problems we cannot and shall not solve. His well-known catchphrase
was “Ignoramus et ignorabimus (We are ignorant and we shall remain igno-
rant)”. For Hilbert such fruitless scepticism and prophesy of a downfall of
culture was unacceptable. Against the general belief of the time Hilbert’s de-
vice was “Noscemus”,71 which means “We can and shall know.” In his 1930
speech he declared: “For the mathematicians there is no ignorabimus, nor,
in my opinion, for any part of natural science.”72 In his notebook he wrote:
“That there is no ignorabimus in mathematics can probably be proved by my
71
Ibid., Cod. Ms. D. Hilbert, 600:1, p. 72.
72
Hilbert 1930, p. 963; also in Hilbert 1932–1935, vol. 3, p. 387. “Für den Mathe-
matiker gibt es kein Ignorabimus, und meiner Meinung nach auch für die Naturwis-
senschaften nicht.” English translation from Ewald 1996, p. 1165.
theory of logical arithmetic”, and “Perhaps it will turn out that there is no
purport in saying there are insoluble problems.”73
However, in one of life’s little ironies, at almost the same time as Hilbert’s
great optimistic speech in Königsberg, a 25-year-old Kurt Gödel (1906–1978)
proved striking results in 25 pages in a way that Hilbert had not anticipated.
As a consequence of Gödel’s and Paul Cohen’s (born 1924) results, from any
systems of axioms the continuum hypothesis74 can neither be proved nor dis-
proved (assuming the standard axioms of set theory are consistent, Zermelo-
Fraenkel axioms plus the axiom of choice).
Despite Gödel’s incompleteness theorem of 1931, Hilbert, at this stage 69
years old, continued his work to lay the foundations of mathematics. In his
program, proposed in 1905 and more specifically after 1917, Hilbert intended
to justify all of mathematics on the basis of elementary methods of finite rea-
soning. In a lecture on the infinite he declared: “Our thinking is finite; as we
are thinking a finite process is going on. [...] The infinite is nowhere realized;
it does not exist in nature, nor it is an admissible basis of our thinking - a
remarkable harmony between being and thinking.”75 He repeated this convic-
tion in his Königsberg speech in 1930, and in more detail he said: “We must
be clear to ourselves that ‘infinite’ has no intuitive meaning and that without
more detailed investigation it has absolutely no sense. For everywhere there
are only finite things. There is no infinite speed, and no force or effect that
propagates itself infinitely quickly. Moreover the effect itself is of a discrete
nature and exists only in quanta. There is absolutely nothing continuous that
can be divided infinitely often. Even light has atomic structure just like the
quanta of action. I firmly believe that even space is of finite extent. [. . . ] In-
finity, because it is the negation of a condition that prevails everywhere, is a
gigantic abstraction.”76
73
Göttingen, Handschriftenabteilung, Cod. Ms. D. Hilbert, 600:3, pp. 104, 98. “Dass es
kein Ignorabimus in der Mathematik giebt, ist wahrscheinlich durch meine Theorie
der Logik-Arithmetik beweisbar.” – “Vielleicht stellt sich auch heraus: es hat keinen
Sinn zu sagen, es gäbe unlösbare Probleme.”
74
Hilbert’s first problem and described by him as a “very plausible theorem (einen
sehr wahrscheinlichen Satz)”, Hilbert 1900, p. 263.
75
Hilbert 1924, p. 134. “Unser Denken ist finit, indem wir denken, geschieht
ein finiter Prozeß [. . . ] Das Unendliche findet sich nirgends realisiert; es ist weder
in der Natur vorhanden, noch als Grundlage unseres Denkens zulässig - eine be-
merkenswerte Harmonie zwischen Sein und Denken.”
76
Hilbert 1930, “[Wir müssen] uns klarmachen, daß ‘Unendlich’ keine anschauliche
Bedeutung und ohne nähere Untersuchung überhaupt keinen Sinn hat. Denn es gibt
überall nur endliche Dinge. Es gibt keine unendliche Geschwindigkeit und keine sich
unendlich rasch sich fortpflanzende Kraft oder Wirkung. Zudem ist die Wirkung
selbst diskreter Natur und existiert nur quantenhaft. Es gibt überhaupt nichts Kon-
tinuierliches, was unendlich oft geteilt werden könnte. Sogar das Licht hat atom-
istische Struktur, ebenso wie die Wirkungsgröße. Selbst der Weltraum ist, wie ich
sicher glaube, nur von endlicher Ausdehnung. [. . . ] Die Unendlichkeit, weil sie eben
Gödel’s results clarified the reach of the program and marked its limits.
Admittedly, to a certain extent Hilbert reacted to Gödel’s results and accepted
stronger means in his proof theory such as transfinite induction.77 Another
essential roadblock was the 1936 result of Alan Turing (1912–1954) that the
decision problem is unsolvable. To a certain extent Hilbert had dealt with
this difficulty when he admitted that there may be no algorithm to find a
certain element in an infinite set: “In my proof theory I do not maintain
that among infinite objects the discovery of an object can be effected, but
we can imagine that the choice is made.”78 We could continue here with the
names of Ernst Zermelo (1871–1953), Bertrand Russell (1872–1970), Luitzen
Egbert Brouwer (1881–1966), Rudolf Carnap (1891–1970), Alonzo Church
(1903–1995), Stephen Cole Kleene, and others. However, that is another story.
Nevertheless, taken all in all Hilbert’s program proved valuable for the study
of formal systems.
However, Hilbert was not only an old but also a sick man; his career was
almost over. Furthermore, in 1933 the Nazis came to power and started to
rule over Germany and over the German universities. A number of the best-
known German mathematicians, including a great many of Hilbert’s friends,
colleagues and coworkers, became refugees or were murdered. The Hilbert
circle in Göttingen was destroyed by the Nazis. The years to follow became
for Hilbert years of tragic loneliness. We have a report on those days which
shows the old and lonely Hilbert still had his sharp tongue. Asked by the
Nazi minister of education whether mathematics in Göttingen suffered by
eliminating the Jewish influence, Hilbert sadly answered: “There is really none
any more.”79
10.4 The 24th Problem

It is widely believed among mathematicians that simplicity is a reliable guide
to the beauty or elegance of proofs, but like all aesthetic value judgments, such
statements are highly subjective. The physicist Ludwig Boltzmann (1844–
1906) once declared that we should cede elegance to the tailors. Hilbert ad-
mitted that despite the fact that (mathematical) beauty is highly satisfying,
die Negation eines überall herrschenden Zustands ist, eine ungeheuerliche Abstrak-
tion.” Translation by Ewald 1996, p. 1159 (§10).
77
In the Preface of Hilbert 1934.
78
Hilbert 1924, p. 134. “In meiner Beweistheorie wird [. . . ] nicht behauptet,
dass die Auffindung eines Gegenstandes unter den unendlich vielen Dingen stets
[tatsächlich] bewirkt werden kann, wohl aber, dass man [. . . ] stets so tun kann, als
wäre die Auswahl getroffen.”
79
Cf. Reid 1996, p. 205. In oral history there is a report using Hilbert’s East
Prussian dialect: “Jelitten? Dat hat nich jelitten, dat jibt es doch janich mehr”
(Suffered? It [mathematics] did not suffer, it does not exist anymore).
it cannot used in argumentation.80 Can one really say that certain mathemat-
ical proofs are simpler than others? “Mathematical formalism is the base of
all natural laws because with its help it is possible to say what simple is; for
example, addition is necessarily the first simplest mathematical operation,”81
wrote Hilbert in his notebooks. In the field which Hilbert later called proof
theory and metamathematics, as early as 1900 he wanted a detailed investi-
gation of the question of simplicity; of course, the needed formalism was yet
to develop. But he believed that such an investigation carried out by his ax-
iomatic method would be not only promising but also necessarily successful,
especially by means of the reduction of proofs to an algebraic calculus, and
this remained his belief in the 1920’s. Of course, it is an open problem whether
or how such a reduction can be carried out.
In his Mathematische Notizhefte Hilbert made a note of such a question
he intended to include in his Paris lecture as a 24th problem; however, he then
canceled it. As far as I know the canceled 24th problem has remained unpub-
lished until now and I am not aware of any responses or references to the chal-
lenge, with the exception of Hilbert himself. I came across the problem in the
Hilbert Nachlass at the Niedersächsische Staats- und Universitätsbibliothek
Göttingen, Handschriftenabteilung, while I was studying Hilbert’s notices on
the 23rd problem. Let me quote the problem itself:
The 24th problem in my Paris lecture was to be: Criteria of simplicity,

or proof of the greatest simplicity of certain proofs. Develop a theory
of the method of proof in mathematics in general. Under a given set
of conditions there can be but one simplest proof. Quite generally, if
there are two proofs for a theorem, you must keep going until you have
derived each from the other, or until it becomes quite evident what
variant conditions (and aids) have been used in the two proofs. Given
two routes, it is not right to either take these two or look for a third;
it is necessary to investigate the area lying between the two routes.
Attempts at judging the simplicity of a proof are in my examination of
syzygies and syzygies [Hilbert made a slip in writing] between syzygies.
The use or the knowledge of a syzygy essentially simplifies a proof that
a certain identity is true.82
80
abteilung, Cod. Ms. D. Hilbert 657, folder Physik, sheet 37. “Die Schönheit [... ist]
ein wunderbares und den menschlichen Geist hoch befriedigendes Accedenz, aber
kein Beweismittel.”
81
Göttingen, Handschriftenabteilung, Cod. Ms. D. Hilbert 600:3, inserted pages. “Der
mathematische Formalismus ist die Grundlage aller Naturgesetze weil durch ihn
möglich ist, zu sagen, was einfach ist. z.B. Addieren ist notwendig die erste einfachste
mathematische Operation.”
82
Göttingen, Handschriftenabteilung. Cod. Ms. D. Hilbert 600:3, p. 25. “Als 24stes
Fig. 10.21. Entry in Hilbert’s notebook on a canceled 24th problem demanding the
“simplest proof”, beginning with the eighth line from top. Courtesy Niedersächsische
Staats- und Universitätsbibliothek Göttingen, Handschriftenabteilung, Cod. Ms. D.
Hilbert 600:3, p. 25f.
The entry stems from between autumn 1900 and summer 1901. Hilbert
might have been inspired to remember the canceled problem again by the
notice he made just before this entry. In the preceding entry he dealt with
how to transform a geometrical construction into an analytical formalism with
the help of a mechanism (linkage), and he asked for the simplest apparatus.
Moreover, he had the idea to demonstrate the greatest possible simplicity
(“grösste Einfachheit”) by an analytic proof.83
I will not risk more than a few disjointed remarks.84 Firstly, I have re-
marked on the desired proof theory. Secondly, with respect to the mentioned
diversity of proofs, I refer to the history of the Fundamental Theorem of Al-
gebra as an example. There are two ideas: one for an algebraic proof (Euler,
1707–1783) and one for an analytic one (d’Alembert, 1717–1783); furthermore,
there is the research of Gauss between these ideas. Despite his “Pauca, sed
matura (Few but ripe)”, Gauss returned to this matter at various times and
gave four proofs in total.
Thirdly, a proof is the most straightforward way to justify mathematical
reasoning. To quote Godfrey Harold Hardy (1877–1947): “A mathematical
Problem in meinem Pariser Vortrag wollte ich die Frage stellen: Kriterien für die
Einfachheit bez. Beweis der grössten Einfachheit von gewissen Beweisen führen. Ue-
berhaupt eine Theorie der Beweismethoden in der Mathematik entwickeln. Es kann
doch bei gegebenen Voraussetzungen nur einen einfachsten Beweis geben. Ueber-
haupt, wenn man für einen Satz 2 Beweise hat, so muss man nicht eher ruhen,
als bis man sie beide aufeinander zurückgeführt hat oder genau erkannt hat, welche
verschiedenen Voraussetzungen (und Hülfsmittel) bei den Beweisen benutzt werden:
Wenn man 2 Wege hat, so muss man nicht bloss diese Wege gehen oder neue suchen,
sondern das ganze zwischen den beiden Wegen liegende Gebiet erforschen. Ansätze,
die Einfachheit der Beweise zu beurteilen, haben meine Untersuchungen über Syzy-
gien und Syzyzien [sic] zwischen Syzygien. Die Benutzung einer oder Kenntnisse
einer Syzygieen vereinfacht den Beweis, dass eine gewisse Identität wahr ist.” – The
slips and corrections in writing show that Hilbert wrote in haste.
83
Göttingen, Handschriftenabteilung. Cod. Ms. D. Hilbert 600:3, p. 25. “Meine Con-
struktion durch allgemeine Gelenkcirkel ins Analytische übersetzen und fragen,
welcher Apparat jedesmal [-] z.B. um zu bewirken, dass 3 Punkte immer in einer
Geraden, 4 in einem Kreis etc. bleiben - der einfachste ist, den analytische strengen
Nachweis für die grösste Einfachheit führen.” In the lecture Hilbert 1910 (Elements
and Essential Questions in Mathematics) Hilbert explained (p. 42): A Gelenkcirkel
or a Gelenkmechanismus (linkage) is a generalization of a circle. In the xy-plane
any system of rigid rods the endpoints of which are fixed, either in certain points
of the plane or in certain points of the rods, and able to rotate about these points,
forms a linkage (joint mechanism) which is assumed to have one degree of freedom.
A point of such linkage describes an algebraic curve with the equation f(x, y) = 0.
Working backward it is possible (at least in principle) to invent such a mechanism
for each algebraic curve. Cf. Hilbert 1932; Courant 1953, chap. 3, esp. pp. 155-160;
Bieberbach 1952, esp. §§6, 7, and 12.
84
For more details see Thiele 2003.
proof should resemble a simple and clear-cut constellation, not a scattered

cluster in the Milky Way.”85 Allow me to vary another famous bon mot of
Hardy: simplicity is a first test; there is no permanent place for other math-
ematical patterns.86 Hilbert declared: “The mathematician’s function should
be to simplify the intricate; instead they do just the opposite, and complicate
what is simple, and call it ‘generalizing’.”87 Hilbert went on:
Besides, it is an error to believe that rigor in proof is the enemy of

simplicity. On the contrary we find it confirmed . . . that the rigorous
method at the same time is the simpler and the more easily compre-
hended. The very effort for rigor forces us to find simpler methods of
proof. [. . . ] The most striking example is the calculus of variations.88
In general, a mathematical theorem is regarded as “deep” if its proof is

difficult. The opposite of deep is “trivial”. Nevertheless, there is a constant
effort towards simplification, towards the finding of ways of looking at the
matter from an easier, more trivial point of view. There is no question that
simplicity of proof depends on the length of its presentation, on the method
employed, on our familiarity with the used concepts, on its abstract generality,
the novelty of ideas, and so on.
Having all such viewpoints in mind, what kinds of simplicity can we define
at all precisely? Instead of proving mathematical theorems, how does one
examine the deductive systems themselves and prove theses about them?
It is remarkable that, as always, Hilbert started the 24th problem with
examples, and that specific results led him to the general idea. For Hilbert the
simplest mathematical operation is addition, and to each addition there is a
corresponding geometric or logical process. Hilbert investigated the possibility
of such corresponding constructions.89 “The geometrical figures are graphic
formulas”,90 he said in his Paris talk. Without doubt, certain constructions or
proofs rest on countable processes; that is, by an examination of the number of
85
Hardy 1992, p. 113.
86
In Hardy 1992 we find “beauty” instead of “simplicity”, p. 85.
87
Mathematische Notizhefte. Library of the University of Göttingen, Cod. Ms. D.
Hilbert 600:1, p. 45. “Die Thätigkeit der Mathematiker sollte darin bestehen, das
Verwickelte einfach zu machen. Statt dessen machen sie umgekehrt das Einfache ver-
wickelt und nennen das Verallgemeinern.” See also Hardy 1992, p. 105: “Generality
is an ambiguous and rather dangerous word.”
88
Hilbert 1900, pp. 257, 258. “Zudem ist es ein Irrtum zu glauben, daß die Strenge
in der Beweisführung die Feindin der Einfachheit wäre. An zahlreichen Beispielen
finden wir im Gegenteil bestätigt, daß die strenge Methode auch zugleich die ein-
fachere und leicht faßlichere ist. Das Streben nach Strenge zwingt uns eben zur
Auffindung einfacherer Schlußweisen. [. . . ] Das schlagendste Beispiel aber für meine
Behauptung ist die Variationsrechnung.”
89
For example, Hilbert investigated the possibility of such constructions using
only a ruler with a given unit in chapter 7 of Hilbert 1899.
90
Hilbert 1900, p. 259. “Die geometrischen Figuren sind gezeichnete Formeln.”
Fig. 10.22. Notice of Hilbert which considers proofs (third entry). “Watch where
in a proof for the first time an auxiliary concept is introduced which is later elim-
inated in the result. Consider why and whether this auxiliary concept is needed
for the proof. Shortest proof!” Courtesy Niedersächsische Staats- und Univer-
sitätsbibliothek Göttingen, Handschriftenabteilung, Cod. Ms. D. Hilbert 603, sheet
20.
operations involved in the proof we can decide which of two proofs is simpler.
In this view it seems possible to arrange mathematical proofs somehow in
strata. Mathematics appears as a well-ordered stock, as a hierarchical ordering
of formulas, and the task under consideration is to examine the proofs or the
corresponding chains of formulas respectively with respect to simplicity.
The French mathematician Émile Lemoine (1840–1912) proposed a crite-
rion for simplicity of geometric constructions.91 In 1888 in La Géométrographie
ou l’Art des Constructions Géometriques (The Geometrography or the Art of
Geometric Constructions), Lemoine reduced all geometric constructions by
ruler and compass to only five basic constructions. One of them is the plac-
ing of a compass end at a given point. Lemoine called the total number of
times any of these basic operations was used the simplicity of the construction
(“coëfficient de simplicité”, or just “la simplicité”). This measure of the com-
plexity of a geometric construction produced some unexpected and surprising
results with respect to standard constructions. For example: the standard con-
91
É.M.H. Lemoine, La Géométrographie ou l’Art des Constructions Géometriques.
Paris: Naud 1902. There is a summary by Van der Waerden 1937/38.
struction of four tangents to two circles has degree of simplicity 92; Lemoine
gave another construction with a reduced degree of only 34.92
Hilbert similarly wanted to make proofs a measurable object of another
theory by his “logical arithmetic”, in which only finite methods should be
used. Of course, the discovery of such a measure is a delicate business. Invari-
ant theory is a bridge between geometry and algebra. Hilbert’s fundamental
theorem reads that we can pick a finite number of invariants i1 , i2 , . . . , ik by
which we can express any invariant as a polynomial in these basic invariants.
In his aim Hilbert was probably guided by his investigations on invariants,
especially on “syzygies”. What are syzygies? The basic invariants i1 , i2 , . . . , ik
are not algebraic independently but themselves fulfill a homogeneous polyno-
mial relation f (i1 , i2 , . . . , ik ) = 0. Such an identity is called a syzygy.
What is the role of such syzygies in geometry? As an example consider
any geometry with inhomogeneous Cartesian coordinates and all linear ho-
mogeneous transformations (affine transformations). In this affine geometry
geometric magnitudes are invariants which under linear homogeneous trans-
formation are altered only by a factor (such as the area of an triangle, which
can be expressed analytically by certain determinants of the coordinates).
Such invariant magnitudes have a geometric meaning in the affine geome-
try. Moreover, in affine geometry the mentioned determinants make the full
system of invariants. In other words, every invariant can be expressed as a
polynomial in these determinants and because of the geometric meaning of an
invariant, this meaning is expressed analytically by the polynomial relation
(syzygy). Conversely, to each theorem (on invariants) of the affine geome-
try corresponds a syzygy. This means that by determining the full system of
syzygies the theory of invariants allows one to describe all theorems of affine
geometry.
Can we extend such considerations to other geometries? The polynomial
relators, syzygies, can be added and multiplied; they are closed under the
operations of addition and multiplication and form a subring of R, where R
denotes the ring formed by the invariants (see above, p. 250). Moreover, they
even form an ideal I of R which has a finite basis (Hilbert). That means we can
pick a finite number of relators, f1 , f2 , . . . , fk , and express every polynomial
relator f in the form f = Q1 f1 + Q2 f2 + · · · + Qk fk , the Qi being polynomials
(Hilbert’s basis theorem). This finite basis f1 , f2 , . . . , fk is not algebraically
independent. Thus one obtains new relators (second-order syzygies) again
with a finite basis; these relators are not algebraically independent, and so
on. However, Hilbert proved that the cascade of syzygies stops in at most
r + 1 steps, where r = ξ(m) is the number of invariants of the full invariant
system of the n-ary forms F of degree m; ξ(m) is called the characteristic
function and is a rational function of m. It was in the years up to 1892 that
92
Cf. also Lemoine 1893, submitted to the Chicago Congress 1893. Hilbert con-
tributed to the same volume with Hilbert 1893 (On the Theory of Algebraic Invari-
ants), pp. 116–124, cf. footnote 14. See also Hilbert 1890.
Hilbert proved the fundamental finiteness theorem of the theory of invariants

for the full projective group, the Hilbert Syzygy Theorem: “The chain of
syzygies terminates after finitely many steps.”93 Furthermore, Hilbert’s basis
theorem reads: “If every ideal in a commutative ring R is finitely generated,
then so is R[x1 , . . . , xn ].” Taken all in all, we stay in a finite system.
Suppose we can eliminate complex geometric conceptions, i.e., express
them in a finite form. Then it should be possible to exploit this finiteness to
establish the simplicity of geometric proofs. More generally Hilbert pointed
out: “With each mathematical theorem [. . . ] one can ask whether there is any
way to determine how many operations are needed at most to carry out the
assertion of the theorem. Kronecker has particularly emphasized the question
of whether one can carry it out in a finite number of steps.”94
As to the complexity of technical details for proofs consider the well-known
Four-Color Problem.95 Aside from the possibility of a computer-aided proof
which we cannot survey step by step, there remains Hilbert’s practical ques-
tion “whether in mathematics problems exist that cannot be dealt with in a
prescribed short time?”96 His example: calculate the n-th digit in the decimal
expansion of π, where n is equal to (1010 )10 . A simple calculation on an en-
velope shows that an ideal computer of the largest size working since the Big
Bang would have been able to carry out only a finite number of operations
somewhere between 10120 and 10160 . Nevertheless, we read in Hilbert’s note-
book: “All our effort, investigation, and thinking is based on the belief that
there can be but one valid opinion”, or: “The proof of proofs: that it must
93
Hilbert 1897. Lecture Theorie der algebraischen Invarianten nebst Anwendun-
gen auf Geometrie, delivered in Göttingen summer term 1897. Lecture notes taken
by S. Marxsen, Mathematisches Institut, Universität Göttingen. English transla-
tion: Theory of Algebraic Invariants by R. Laubenbach, Cambridge: University Press
1994, p. 173. “Die Kette der Syzygien bricht nach einer endlichen Anzahl von Schrit-
ten ab.” Bl. 773.
94
Ibid., p. 133. “Man kann bei jedem mathematischen Satze [...] fragen, ob man
auf irgendeine Weise darüber Aufschluß geben kann, wie viele Operationen man
höchstens gebraucht, um das im Satze Gesagte auszuführen. Mit besonderem Nach-
druck hat Kronecker die Frage nach der Ausführbarkeit durch ein endliche Anzahl
von Schritten betont.” Bl. 563f.
95
Using this example, Hilbert pointed out that a peculiarity of mathematics con-
sists in the state of affairs that despite the fact that problems can be very sim-
ply formulated, their solutions can be very hard. (“Es ist eine Eigentümlichkeit
der Math.[ematik], dass einfachste spezielle Probleme (z.B. das 4-Farbenproblem)
sehr schwierig zu lösen sind. Erkläre dies Phänomen.” Mathematische Notizhefte.
Niedersächsische Staats- und Universitätsbibliothek Göttingen, Handschriftenab-
teilung, Cod. Ms. D. Hilbert 600: 3, inserted pages.)
96
Göttingen, Handschriftenabteilung. Cod. Ms. D. Hilbert 600:1, p. 53. “Ob es Prob-
leme in der Mathematik giebt, deren Erledigung nicht in einer vorgeschriebenen
kleinen Zeit möglich ist?”
always be possible to arrive at a proof.”97 Thus there must be a simplest proof

for any given proposition. Colloquially he added: “always apply the strictest
proof! Philological-historic import must be wiped out. Given a 15-inch gun,
we don’t shoot with the crossbow.”98
But if we are willing to remain in the sphere of the finite, can we actually
justify all the needed mathematical conclusions? Let us consider the proofs
needed in mathematics and all proofs we can actually execute. Maybe a gap
appears: with the help of finite methods we are only able to deduce a count-
able set of proofs (Gromov’s Hilbert tree), while the set of possible proofs is
probably uncountable.99 Was Hilbert’s belief in the power of thinking some-
what naive? Was it indebted to the widespread belief of progress at the turn
of the century, part of the zeitgeist? Consider this quotation from his Paris
lecture:
Occasionally it happens that we seek the solution under insufficient

presuppositions or in an incorrect sense, and for this reason we do
not succeed. The problem then arises: to show the impossibility of the
solution under the given hypothesis.100
Nevertheless, despite an increasing diversity of mathematical branches and

their problems by the simplification of proofs due to axiomatic methods, very
much in the spirit of Hilbert, mathematics as a whole has become more eco-
nomically designed, has been increasingly unified and has widened our hori-
zons. Hilbert’s spirit and influence has been vivid.
This is a good point to stop. The last word is Hilbert’s: “Mathematics
stalks on earthly ground and at the same time touches the divine firma-
ment.”101
97
Ibid. 600:3, p. 96. “Allem unserem Streben, Forschen und Denken liegt doch die
Meinung zu Grunde, dass es nur eine richtige Meinung geben kann (Maximum)” -
“Beweis aller Beweise: dass man den Beweis immer muss finden können!”
98
Ibid., 600:3, inserted pages. “Immer das schärfste Mittel anwenden! philologisch-
historische Sinn muss ausgerottet werden. Wenn wir 42cm-Kanone haben, schiessen
wir doch nicht mit Armbrust.”
99
For details see Gromov 2000, p. 1213f.
100
Hilbert 1900, p. 261. “Mitunter kommt es vor, daß wir die Beantwortung
unter ungenügenden Voraussetzungen oder in unrichtigem Sinne erstreben und
infolgedessen nicht zum Ziele gelangen. Es entsteht dann die Aufgabe, die Un-
möglichkeit der Lösung des Problems unter den gegebenen Voraussetzungen [. . . ]
nachzuweisen.”
101
Mathematische Notizbücher. Niedersächsische Staats- und Universitätsbib-
liothek Göttingen, Handschriftenabteilung, Cod. Ms. D. Hilbert 600:3, p. 104. “Die
Math.[ematik] schreitet auf der Erde und berührt zugleich den Himmel.”
Hilbert’s Mathematical Problems102

This is the problem, find its solution. You can find it by pure thinking
since in mathematics there is no Ignorabimus.103
1. The cardinality of the continuum, including well-ordering (Cantor’s

problem, 1878),
Gödel 1940, Cohen 1963 (awarded Fields Medal 1966), Vopenka 1975.
2. The consistency of the axioms of arithmetic,
Gödel 1931, Gentzen 1936, Novikov 1941.
3. The equality of the volumes of two tetrahedra of equal bases and equal
altitudes,
Dehn 1902, Kagan 1903.
4. The straight line as shortest connection between two points,
Hamel 1901, Funk 1929, Busemann 1955.
5. Lie’s concept of a continuous group of transformations without the
assumption of the differentiability of the functions defining a group,
Kolmogorov 1930, E. Cartan 1930, Pontrjagin 1932,
von Neumann 1933, Chevalley 1941, Malcev 1946, Gleason 1952,
Montgomery 1952, Zippin 1952.
6. The axioms of physics,
It is not yet clear what axiomatizing physics really means;
not a problem but a program of research.
Particular fields were axiomatized:
Classical mechanics by Hamel 1906,
Thermodynamics by Carathéodory 1909,
and others.
7. Irrationality and transcendence of certain numbers,
Siegel 1921, Gelfond 1929, 1934, Schneider 1934, Baker 1966 et al.
8. Prime number theorems (including the Riemann hypothesis),
Hecke 1917, Schnirelman 1930, Vinogradov 1937.
9. The proof of the most general reciprocity law in arbitrary number fields,
Hilbert 1897, Takagi 1920, Artin 1928, Hasse 1935, Shafarevich 1950.
10. Decision on the solvability of a Diophantine equation,
no solution: Thue 1908, Siegel 1928, Robinson 1969,
Matijasevich 1970.
11. Quadratic forms with any algebraic coefficients,
Hasse 1929, Siegel 1936, 1951, Weil 1964, Ono 1964.
102
Problems with a number in bold print were presented in the historic speech in
Paris. For more details cf. Bieberbach 1930, Aleksandrov 1969, Fang 1970, Browder
1976, Gray 2000, Yandell 2002.
103
“Da ist das Problem, suche die Lösung. Du kannst sie durch reines Denken
finden; denn in der Mathematik gibt es kein Ignorabimus.” End of the introductory
essay in Hilbert 1900, p. 262.
12. The extension of Kronecker’s theorem on Abelian fields to arbitrary

algebraic fields,
unsolved.
13. Impossibility of solving the general seventh degree equation by means of
functions of only two variables,
Arnold 1957; unsolved if analyticity is required.
14. Finiteness of systems of relative integral functions,
no solution: Nagata 1959.
15. A rigorous foundation of Schubert’s enumerative calculus,
van der Waerden 1930f., in general unsolved.
16. Topology of real algebraic curves and surfaces,
partial results.
17. Representation of definite forms by squares,
for real-closed fields Artin 1926,
negative solution in general Du Bois 1967.
18. The building up of space from congruent polyhedra,
Bieberbach 1908, Hajos 1941.
19. The analytic character of solutions of variation problems,
special results.
20. General boundary value problems,
Hilbert dealt with the special case of Dirichlet’s principle 1900, 1904
and Plateau’s problem, Bernstein 1904, Douglas 1939, Courant 1950,
Morrey 1966, and many others.

21. Linear differential equations with a given monodromy group,
Hilbert 1905, negative solution in general Anasov 1994,
Bolibruch 1994.
22. Uniformization of analytic relations by means of automorphic functions,
Koebe 1907, Poincaré 1907.
23. The further development of the methods of the calculus of variations,
Kneser 1900, Hilbert 1905, Carathéodory 1935, Weyl 1936,
Boerner 1936, Lepage 1936, and others.
[24.] The simplicity of proofs,
canceled.
References
Sources
Hilbert, D., Nachlass. Niedersächsische Staats- und Universitätsbibliothek

Göttingen, Handschriftenabteilung (Special Collections). Cod. Ms. D. Hilbert
and Cod. Ms. Math.-Archiv 76 (some letters): among them
Mathematische Notizhefte, 3 vols, some inserted pages.
Cod. Ms. D. Hilbert 600:1–3.
Fimpel, M., 2002. Spezialinventar zur Geschichte der Mathematik und Natur-
wissenschaften an der Universität Göttingen von 1880-1933. Schriften des
Universitätsarchivs Göttingen, Bd. 1 (U. Hunger and H. Wellenreuter, eds.).
Göttingen: Universitätsarchiv.
II. B.4 Nachlaß Hilbert, Cod. Ms. D. Hilbert, pp. 438-485,
II. B. 15 Mathematik-Archiv, Math-Arch., pp. 605-616.
Hilbert, D., 1886. Invariantentheorie. Wintersemester 1886. (Theory of invari-
ants.) Lecture notes from Winter 1886 by Hilbert. Niedersächsische Staats-
und Universitätsbibliothek Göttingen, Handschriftenabteilung, Cod. Ms. D.
Hilbert 521.
Hilbert, D., 1897. Theorie der algebraischen Invarianten nebst Anwendungen
auf Geometrie. Sommersemster 1897. (Theory of algebraic invariants including
applications to geometry.) Lecture notes taken by S. Marxsen. University of
Göttingen, Library of the Mathematical Institut.
Hilbert, D., 1905. Logische Prinzipien des mathematischen Denkens. Som-
mersemester 1905. (Logical principles of mathematical thinking.) Lecture
notes taken by M. Born. Niedersächsische Staats- und Universitätsbibliothek
Göttingen, Handschriftenabteilung, Cod. Ms. D. Hilbert 558; Lecture notes
taken by E. Hellinger. University of Göttingen, Library of the Mathematical
Institut.
Hilbert, D., 1910. Elemente und Principienfragen der Mathematik. Som-
mersemester 1910. (Elements and principal questions of mathematics.) Lec-
ture delivered in the summer term 1910 and recorded by R. Courant. Library
of the Mathematical Institute, University of Göttingen.
Hilbert, D., 1917. Principien der Mathematik. Wintersemester 1917. (Prin-
ciples of Mathematics, winter term 1917/18.) Lecture notes by P. Bernays.
University of Göttingen, Library of the Mathematical Institut.
Hilbert, D., 1919. Natur und mathematisches Erkennen. Wintersemester 1919.
(The mathematical knowledge of nature.) Lecture notes taken by P. Bernays,
winter term 1919/20. University of Göttingen, Library of the Mathematical
Institut.
Hilbert, D., 1921. Geometrie und Physik. Sommersemester 1921. (Geome-
try and Physics.) Type-written manuscript. Staatsbibliothek Berlin, Hand-
schriftenabteilung, Nachlaß Hückel 2.14.
Hilbert, D., 1921a. Kopenhagener Vorträge, 1921. Lectures delivered in Copen-
hagen 1921, 2 copy books. Niedersächsische Staats- und Universitätsbibliothek
Göttingen, Handschriftenabteilung, Cod. Ms. D. Hilbert 589.
Hilbert, D., 1921b. Hamburger Vorträge, 1921. Lectures delivered in Hamburg
1921, 3 copy books with an insertion. Niedersächsische Staats- und Univer-
sitätsbibliothek Göttingen, Handschriftenabteilung, Cod. Ms. D. Hilbert 596.
Hilbert, D., 1924. Über das Unendliche. Wintersemester 1924/25. (On the In-
finite, winter term 1924/25.) Lecture notes, taken by L. Nordheim. University
of Göttingen, Library of the Mathematical Institut.
Hilbert’s printed works

Hilbert, D., 1932-1935. Gesammelte Abhandlungen. 3 vols. Berlin: Springer.
Hilbert, D., 1900/1902. “Mathematische Probleme”. Vortrag, gehalten auf
dem internationalen Mathematiker-Kongreß. Nachrichten der Akademie der
Wissenschaften in Göttingen (1900), 253–297; English translation by Mary
Winston Newson with the permission of Hilbert in: Bulletin of the American
Mathematical Society 8 (1901/02), 437–479; reprinted in the same Bulletin 37
(2000), 407–436.
Hilbert, D., 1888. Zur Theorie der algebraischen Gebilde. Nachrichten der
Akademie der Wissenschaften in Göttingen, p. 450–457.
Hilbert, D., 1889. Über die Endlichkeit des Invariantensystems für binäre
Grundformen. Mathematische Annalen 33, 223–226.
Hilbert, D., 1890. Über die Theorie der algebraischen Formen. Mathematische
Annalen 36, 473–534.
Hilbert, D., 1896. Über die Theorie der algebraischen Invarianten. In: Mathe-
matical Papers read at the International Congress held in Chicago 1893. E.H.
Moore et al., eds. New York: Macmillan 1896, pp. 116–124.
Hilbert, D., 1897/1993. Theory of algebraic invariants. Translation of a lec-
ture delivered in 1897. Eds. R.C. Laubenbacher and B. Sturmfels. New York:
Cambridge University Press 1993.
Hilbert, D., 1899. Grundlagen der Geometrie. Leipzig: B.G. Teubner.
Hilbert, D., 1930. “Naturerkennen und Logik” speech delivered in Königsberg.
Die Naturwissenschaften 18 (1930), 959–963; also in Hilbert 1932–1935, vol.
3, pp. 378–387; English translation in Ewald 1996. pp. 1157–1165.
Hilbert, D. and S. Cohn-Vossen, 1932. Anschauliche Geometrie. Berlin: Sprin-
ger.
Hilbert, D. and P. Bernays, 1934. Grundlagen der Mathematik, I. Berlin:
Springer.
Hilbert, D., 1970. Hilbert’s Invariant Theory Papers. Translated by M. Ack-
erman. Brookline (MA): Math Sci Press.
Frei, G. (ed.), 1985. Der Briefwechsel David Hilbert - Felix Klein (1886-1918).
Göttingen: Vandenhoeck & Ruprecht.
Selected references
Aleksandrov, P.S., 1969/1976. Problemy Gilberta. Moscow: Nauka 1969; Ger-
man translation Die Hilbertschen Probleme. Leipzig: Geest & Portig 1976
(reprints).
Bernays, P., 1922. Über Hilberts Gedanken zur Grundlegung der Arithmetik.
Jahresbericht der Deutschen Mathematiker Vereinigung 31, 65–65.
Bernays, P., 1935. Hilberts Untersuchungen über die Grundlagen der Arith-
metik. In: Hilbert 1932–1935, vol. 3, pp. 196–216.
Bieberbach, L., 1930. Über den Einfluß von Hilberts Pariser Vortrag über
“Mathematische Probleme” auf die Entwicklung der Mathematik in den let-
zten dreißig Jahren. Die Naturwissenschaften 18 (51), 1101–1111. B
Bieberbach, L., 1952. Theorie der geometrischen Konstruktionen. Basel: Birk-
häuser.
Blumenthal, O., 1922. David Hilbert. Die Naturwissenschaften 10 (4), 67–72.
Browder, F.E. ed., 1976. Mathematical Development Arising from Hilbert
Problems. Proceedings of Symposia in Pure Mathematics, vol. 28 (2 parts).
Providence: AMS.
Carathéodory, C., 1943. David Hilbert. Sitzungsberichte der Bayerischen
Akademie der Wissenschaften. Mathematisch-naturwissenschaftliche Klasse,
pp. 350–354.
Chern, S.-S., 1996. Remarks on Hilbert’s 23rd Problem. The Mathematical
Intelligencer 18 (4), 7–9.
Courant, R. and H. Robbins, 1953. What is Mathematics? New York: Oxford
University Press.
Dieudonné, J., 1971. Invariant Theory, Old and New. New York: Academic
Press.
Ewald, W., 1996. From Kant to Hilbert, 2 vols. Oxford: Claredon Press.
Fang, J., 1970. Hilbert. Towards a Philosophy of Modern Mathematics, II.
Hauppauge: Paideai.
Freudenthal, H., 1970. “David Hilbert”. In: Dictionary of Scientific Biography
(C. Gillispie, ed.), vol. VI. New York: Scribner’s, pp. 388–395.
Grattan-Guinness, I., 2000/2001. A Sideway Look at Hilbert’s Twenty-three
Problems of 1900. Notices of the American Mathematical Society 47, 752–757;
corrections by G.H. Moore and response of Grattan-Guinness, in same Notices
48 (2001), 167.
Gray, J., 2000. The Hilbert Challenge. Oxford: Oxford University Press.
Gray, J., 2000a. The Hilbert Problems 1900-2000. Newsletter European Math-
ematical Society 36, 10–13.
Gromov, M., 2000. Interview with Gromov. In: Development of Mathematics,
1950-2000 (J.-P. Pier, ed.). Basel: Birkhäuser, pp. 1213–1227.
Halsted, G.B., 1900. The International Congress of Mathematicians. The
American Mathematical Monthly 7, 188–189.
Hardy, G.H., 1992. A Mathematician’s Apology. Foreword by C.P. Snow. Cam-
bridge: Cambridge University Press. (First edition 1940).
Heisenberg, W., 1971. Talk “Die Naturgesetze und die Struktur der Materie”,
delivered in Athens 1964. In: W. Heisenberg, Schritte über Grenzen. München:
Piper. English translation: Across the Frontiers. Oxbow Press 1990 (Reprint).
Hellinger, E., 1935. Hilberts Arbeiten über Integralgleichungen und unendliche
Gleichungssysteme. In: Hilbert 1932–1935, pp. 94–145.
Jacobi, C.G.J., 1830/1875. Correspondance Mathématique. Journal für die
reine und angewandte Mathematik 80, 205–279.
Kantor, J.-M., 1996. Hilbert Problems and Their Sequels. The Mathematical
Kaplansky, I., 1977. Hilbert’s Problems. Lecture notes. University of Chicago.
Katz, V.J., 1993/1998. A History of Mathematics. New York: Harper Collins
1993; 2nd ed. Reading (MA): Addison-Wesley 1998.
Koetsier, T., 2001. Een verrassende vondst: Duits historicus ontdekt Hilberts
24ste probleem. Nieuw Archief voor Wiskunde 5 (2), 65–67.
Lemoine, É.M.H., 1893. Considérations Générales sur la Mesure de la Sim-
plicité ... , in: Mathematical Papers Submitted to the International Mathe-
matical Congress in Chicago 1893, published by the American Mathematical
Society, vol. 1. E.H. Moore et al., eds. New York: Macmillan 1896, pp. 143–154.
Lemoine, É.M.H., 1902. La Géométrographie ou l’Art des Constructions Géo-
metriques. Paris: Naud.
Parshall, K.V.H., 1990. The One-Hundredth Anniversary of the Death of In-
variant Theory? The Mathematical Intelligencer 12 (4), 10–16.
Pier, J.-P. (ed.), 1994/2000. Development of Mathematics. 1900-1950 and
1950-2000 (2 vols.) Basel: Birkhäuser 1994 and 2000.
Poincaré, H., 1908. L’Avenir des Mathématiques. In: Atti del IV Congresso
Internazionale Mathematici, Roma 6-11 Aprile 1908 (G. Castelnuovo, ed.),
vol. I. Roma: Accademia dei Lincei 1909, pp. 1167–182.
Pólya, G., 1954. Mathematics and Plausible Reasoning, 2 vols. Princeton:
Princeton University Press.
Pólya, G., 1957. How to Solve It? Princeton: University Press.
Popov, V.L., 1992. Groups, Generators, Syzygies, and Orbits in Invariant
Theory. Translated from the Russian. Providence: American Mathematical
Society.
Reid, C., 1970. Hilbert. New York: Springer 1970 and Copernicus 1996.
Redei, M., 1999. Unsolved Problems in Mathematics. John von Neumann’s
Address to the International Congress of Mathematicians, Amsterdam, 1954.
The Mathematical Intelligencer 21 (4), 7–12.
Rota, G.C., 1999. Two Turning Points in Invariant Theory. The Mathematical
Rudio, F. (ed.), 1898. Verhandlungen des ersten Internationalen Mathemati-
ker-Kongresses, Zürich 9.-11. August 1897. B.G. Teubner: Leipzig.
Schmidt, A., 1933. Zu Hilberts Grundlagen der Geometrie. In: Hilbert 1932–
1935, vol. 2, pp. 404–414.
Scott, C.A., 1900. The International Congress of Mathematicians in Paris.
Bulletin of the American Mathematical Society 7, 57–79.
Smale, S., 1998. Mathematical Problems for the Next Century. The Mathe-
matical Intelligencer 20 (2), 7–15.
Speiser, D. (ed.), 1991. Die Streitschriften von Jacob und Johann Bernoulli
(commented by H.-H. Goldstine). Birkhäuser: Basel.
Spiess, O. (ed.), 1955. Der Briefwechsel von Johann Bernoulli, vol. I. Basel:
Birkhäuser.
Thiele, R., 1997. Über die Variationsrechnung in Hilberts Werken zur Analy-
sis. N.T.M. (N.S.) 5, 23–42.
Thiele, R., 1997a. Das Zerwürfnis Johann Bernoullis mit seinem Bruder Jakob.
Acta Historica Leopoldina (Volume in Honor of K.-R. Biermann) 27, 257–276.
Thiele, R., 2002. 300 Jahre Brachistochronenproblem. In: Medium Mathematik
(Proceedings of an International Leibniz Conference at Altdorf in 1996). G.
Löffladt and M. Toepell, eds. Hildesheim: Franzbecker 2002, pp. 76–99.
Thiele, R., 2003. Hilbert’s Twenty-fourth Problem. The American Mathemat-
ical Monthly 110, 1–24.
Thiele, R., 2003. Hilbert und Hamburg. Mitteilungen der Mathematischen
Gesellschaft in Hamburg 22, 99–126.
Thiele, R. 2004. Von der Bernoullischen Brachistochrone zum Kalibratorkon-
zept. Turnhout: Brepols.
Thiele, R., 2004a. Variationsrechnung und Wirkungsprinzipien. In: Formale
Teleologie und Kausalität. Proceedings of a Symposia held in Salzburg 2002,
ed. by M. Stöltzner and P. Weingartner. Paderborn: Mentis Verlag. In print.
Thiele, R., 2004b. Van der Waerdens Leipziger Jahre. Mitteilungen der DMV.
In print.
Van der Waerden, B.L. 1933. Bemerkungen zu Hilberts algebraischen Ar-
beiten. In: Hilbert 1932–1935, vol. 2, pp. 401–403.
Van der Waerden, B.L., 1936/37. Die Geometrographie von Lemoine. Semes-
ter-Berichte zur Pflege des Zusammenhangs von Universität und Schule,
Mathematisches Seminar Münster. 11. Semester 1937/38, pp. 86–93.
Vinnikov, V., 1999. We Shall Know: Hilbert’s Apology. The Mathematical
Vollmer, G., 1993. Gelöste, ungelöste und unlösbare Probleme. In: G. Vollmer,
Wissenschaftstheorie im Einsatz. Stuttgart: S. Hirzel 1993, pp. 183–210.
Weil, A., 1971. The Future of Mathematics. In: Great Currents of Mathemat-
ical Thought (ed. F. le Lionnais), New York: Dover, vol. 1. Translation of the
French edition Les Grands Courants de la Pensée Mathématique, Paris 1962.
Weyl, H., 1944. David Hilbert and his Mathematical Work. Bulletin of the
American Mathematical Society 50, 612–654.
Weyl, H. 1951/1968. A Half-century of Mathematics. The American Mathe-
matical Monthly 58, 523–553; also in: Gesammelte Abhandlungen. Heidelberg:
Springer 1968 (I quote this edition).
Yandell, B.H. 2002. The Honors Class. Hilbert’s Problems and Their Solvers.
Natick: Peters.
11
Turing and the Origins of AI∗
Stuart Shanker
Distinguished Research Professor, Departments of Philosophy and Psychology,

Atkinson College, York University
11.1 Turing’s Two Questions
The publication of Mechanical Intelligence 1 provides us with an invaluable

opportunity to survey Turing’s contributions to the origins of AI. Reprinted
here are all of the major articles which Turing published on the Mechanist
Thesis.2 Reading through these papers gives us a chance to appreciate anew
the sparkling freshness of Turing’s thought, but also, to reassess the profound
and in many ways problematic influence which Turing had on the foundations
of AI. This exercise takes on a special significance in light of the fact that,
in the past few years, some of the most influential figures in the cognitive
revolution have repudiated AI (see Bruner 1990; Shanker 1992). Mechanical
Intelligence enables us to see quite clearly, both why the fledgling cognitive
revolution so quickly adopted AI as its paradigm, and the reasons for this
inevitable rupture.
The book is dominated by Turing’s celebrated 1950 paper, ‘Computing
machinery and intelligence’, which in turn is dominated by the philosophical
question that Turing asks at the outset: ‘Can machines think?’ It is a ques-
tion whose import cannot be divorced from the resounding success of Turing’s
version of Church’s Thesis (viz., that mechanically calculable functions are
Turing-machine computable). Moreover, if judged by the amount of interest
which it has aroused, this question surely stands unrivalled in post-war ana-
lytic philosophy. And yet, as far as the foundations of Cognitive Science are
concerned, there is a definite sense in which this question places the emphasis
on the wrong issue. Indeed, there is even a sense in which it places the empha-
sis on the wrong issue as far as Turing’s own interests and contributions are
∗
First published in Philosophia Mathematica 3 (1) (1995), 52–85.
1
Volume III (D. C. Ince, ed., 1992, ISBN 0-444-88058-5, pp. xix + 227) of the
Collected Works of A. M. Turing, Amsterdam: North-Holland.
2
The two most notable omissions are ‘Intelligent Machinery: A Heretical View’
and Turing’s 1951 BBC lecture ‘Can digital computers think?’.
298 Stuart Shanker
concerned. For over and over again we find him returning to the psychological
question: Do thinkers compute? This is a different, and in many ways, a much
more significant matter.
These two questions belong to very different traditions. The former was
a central concern of English mathematicians in the nineteenth century (e.g.,
Babbage, Jevons and Marquand); the latter a mainstay of empiricist psychol-
ogy in Germany, England, and America. But Turing not only regarded these
two questions as intimately connected: in fact, he thought they were internally
related—that in answering one you would ipso facto be answering the other.
The result was a remarkable synthesis: not only did Turing succeed in merg-
ing recursive function theory and cognitive psychology, but within psychology
itself, he brought together two distinct—and even hostile—schools of thought
under the banner of post-computational mechanism. But while Turing’s con-
ception of automata may have been strikingly original, his approach to the
analysis of thought pursues themes that can be traced back to the Greeks.
Admittedly, it is difficult to view the question ‘Can machines think?’ as
anything other than a modern phenomenon (which, in philosophical terms,
means post-Cartesian). But the question ‘Do thinkers compute?’ is another
matter. The succession of mechanical metaphors of mind—qua hydraulic
pipes, clock, telegraph system, telephone exchange, feedback circuit, serial
and parallel computer—are part of a tradition that stems from a persisting
picture of a mechanist continuum (see Shanker [forthcoming]). Locke’s argu-
ment that ‘in all the visible corporeal world we see no chasms or gaps. All
quite down from us the descent is by easy steps, and a continued series that
in each remove differ very little one from the other’ (Locke 1690: III vi §12);
Whytt’s claim that ‘in all the works of nature, there is a beautiful gradation,
and a kind of link, as it were, betwixt each species of animals, the lowest of
the immediately superior class, different little from the highest in the next
succeeding order’ (Fearing 1930: 75); Herrick’s premise that there is ‘an un-
broken graded series from the lowest to the highest animal species’ (Ibid: 179);
and George’s insistence that there is a cognitive continuum, ‘with simple neg-
ative adaptation (habituation, or accommodation, and tropisms, which are
orientating responses and are known to be mediated by fairly simple physico-
chemical means) at one end, and maze-learning, puzzle-box learning. . . and
ape-learning. . . in stages of increasing complexity, leading to human learning
at the other end’ (George 1962: 180), all spring from the same source as that
which led to Aristotle’s maxim in De Generatione Animalium that ‘Nature
orders generation in regular gradation’ (Aristotle 1938: 186).
It is precisely this psychological issue, however, from which the question of
whether machines can think inadvertently serves to deflect attention. Turing
repeatedly insists that his sole concern is with ‘the meaning of the words
“machine” and “think”’, and that it is his faith in ‘semantic progress’ which
leads him to express his belief ‘that at the end of the century the use of
words and general educated opinion will have altered so much that one will
be able to speak of machines thinking without expecting to be contradicted’
11 Turing and the Origins of AI 299
(Turing 1950: 133, 142). Most of the papers he wrote towards the end of his life
begin with a defence of ‘machine intelligence’. His 1947 ‘Lecture to the London
Mathematical Society’, ‘Intelligent Machinery’ (1948), ‘Computing Machinery
and Intelligence’ (1950), and his 1951 BBC lecture ‘Can Digital Computers
Think?’ all begin by making the same point: Turing’s repeated claim that he
was ‘not interested in developing a powerful brain. . . just a mediocre brain’
(Hodges 1983: 251), and that ‘if a machine is expected to be infallible, it
cannot also be intelligent’ (Turing 1947: 105).
The concept of machine had already undergone radical changes. At the
beginning of the nineteenth century it had been confined to the static mo-
tions dictated by Newtonian mechanics, but by the 1870s it had evolved into
the homeostatic systems conceived by Claude Bernard. These developments
were essential to the transition in the mechanist/vitalist debate from the Life
Sciences to the Human Sciences: from the question whether the body could
be explained in mechanical terms to the question whether the mind could be
so explained.
The problem faced by both physiological and psychological mechanists at
the turn of the century was the same: it stemmed from the widespread doubt
that machines would ever approximate the self-regulating adaptative and se-
lective behaviour which characterizes physiological and mental phenomena.
The key word here is ‘ever’, which signifies that the issue was regarded as
empirical. The obvious solution would be, in G. H. Lewes’s words, to ‘think
through the essentials of such a mechanism’. But ‘An automaton that will
learn by experience, and adapt itself to conditions not calculated for in its
construction, has yet to be made; till it is made, we must deny that organisms
are machines’ (Lewes 1877: 436).
This is precisely the problem which was continuing to preoccupy and frus-
trate mechanists fifty years on; and indeed, might have remained beyond the
reach of their ambitions had Turing not completed the mathematical trans-
formation of the concept of machine. What Turing had proved in ‘On Com-
putable Numbers’ is that an ‘effective function’ is an algorithm that can be
so encoded (e.g., in binary terms) as to be machine-executable. But for the
advocates of strong AI, Turing had proved far more than this: what he had re-
ally accomplished was to transform machines into a species of ‘rule-following
beasts’ (as Hofstadter describes computers). And the manner in which he
achieved this feat was by postulating a category of meaningless (sub-)rules
which could guide the operations of a machine (and/or the brain), thereby
providing the rudiments for a new understanding of ‘machine’ and thence the
creation of artificial intelligence (see Shanker 1987).
For almost three decades, philosophical discussions of Turing’s contribu-
tions to the origins of AI centred on his preoccupation with strong AI, the
nature of consciousness, and in particular, with the significance of the Turing
Test. It is not difficult to account for the overwhelming response which these
issues elicited. Turing’s argument spoke directly to the anxieties of a society
that had just lived through a terrifying war, only to find itself in a world that
300 Stuart Shanker
had not just been transformed, but was continuing to change at a rate never
before experienced in ways that few had envisaged. Turing’s was the bold and,
to many, the reassuring voice of the new vanguard; but the message was as old
as the Renaissance: technological advances cannot be halted, man must adapt
to the inexorable march of progress. And so there followed a flurry of articles
and books in which AI-theorists rhapsodized and humanists anguished, all of
them mesmerized by the debate over man’s position on the Scala Naturae.
Archetypal issues clearly die hard.
With all of the rhetoric about the computational possibilities being opened
up, or the singular phenomenological characteristics of human experience, it
was easy to overlook the fact that, in order to defend his philosophical thesis—
viz., his proof that, if not quite yet, at some point in the future machines
will indeed be capable of thought—Turing was led deeper and deeper into
the development of an appropriate psychological theory: viz., that thinkers
do indeed compute. By the time he came to write ‘Computing Machinery
and Intelligence’ he was explaining how his real goal was that of ‘trying to
imitate an adult human mind’ (Turing 1950: 155). The result was a sublimely
simple mechanist theory whose appeal lay in its claimed ability ‘to resolve
complex psychical phenomena into more elementary process’—a sentiment
which, significantly, was expressed fourteen years before the publication of
‘On Computable Numbers’ (Rignano 1923).
Turing’s psychological theory represents a marriage of Denkpsychologie
and behaviourism. His basic idea is that thinking is an effective procedure (no
doubt because the brain is a digital computer): i.e., the mind proceeds, via
an unbroken chain of mechanical steps, from α to ω, even though the subject
himself may only be aware of α, δ, ξ, and ω. By mapping the subject’s thought-
processes onto a program designed to solve the same problem, we can thus fill
in the intervening—subconscious—steps. This is the thesis underlying Turing’s
observation in ‘Can digital computers think?’ that ‘The whole thinking process
is still rather mysterious to us, but I believe that the attempt to make a
thinking machine will help us greatly in finding out how we think ourselves’
(Hodges 1983: 442).
To be sure, there is nothing particularly novel about this picture of the
unconscious mind. As a matter of fact, Hadamard’s The Psychology of Math-
ematical Invention is about little else; Hadamard cites story after story to
establish how all the important creative work in mathematical discovery is
unconscious (see Hadamard 1947). But Hadamard’s book also vividly illus-
trates why Turing’s psychological thesis was to have such a profound and
widespread impact. For on traditional theories of the cognitive unconscious,
the problem of ‘insight’ was simply shifted down a level, where any amount
of extraordinary cognitive abilities that escaped the conscious mind could
be attributed to the unconscious. But Turing’s argument claims to take no
such metaphysical step; for the processes occurring ‘beneath the threshold of
consciousness’ are all said to be Turing-machine computable. But then, by
placing the emphasis on the question of whether machines can think, Turing
only manages to blur the crucial demarcation lines between materialism and
mechanism: between breaking the machine down into its component parts,
and using the machine as a psychological paradigm for understanding how
the mind works.
Newell and Simon were perhaps the first to realize the significance of this
point. Indeed, it was partly for that reason that they were initially opposed
to the name ‘artificial intelligence’.3 For they were not interested in the philo-
sophical question of whether or not machines can be said to think; rather,
they wanted to place the emphasis firmly on psychological explanation. As
they explained in ‘GPS: A program which simulates human thought’:
We may then conceive of an intelligent program that manipulates

symbols in the same way that our subject does—by taking as inputs
the symbolic logic expressions, and producing as outputs a sequence
of rule applications that coincides with the subject’s. If we observed
this program in operation, it would be considering various rules and
evaluating various expressions, the same sorts of things we see ex-
pressed in the protocol of the subject. If the fit of such a program
were close enough to the overt behaviour of our human subject—i.e.,
to the protocol—then it would constitute a good theory of the sub-
ject’s problem-solving (Newell & Simon 1961: 283).
It is easy to see how this could have been read as an expression of strong
AI: if the computer can solve the same problems as man—and what’s more,
do so in exactly the same steps—then it would have satisfied the demands
for the attribution of ‘intelligence’. But the psychological significance of the
argument is contained in the last line: in the claim that such a program would
constitute a theory of problem-solving. The reasoning here is straightforward:
given that thinking is an effective procedure, then, if a program simulates
a subject’s ‘overt behaviour’—both what the agent did and the ‘fragments’
of his thinking-process which he observed and reported—that program will
explain the ‘hidden processes’ that brought about that subject’s behaviour.
The essential details of Newell and Simon’s argument had already been
sketched by Turing. This is the reason why we find Turing spending so much
time during this period on the development of chess programs. Chess served
as the ideal medium for the computational explanation/simulation of those
processes which a subject(’s mind) employs when solving problems, where this
was seen as a matter of grasping a problem, tentatively experimenting with
various methods of solution, selecting a favoured route, and then verifying the
success of the chosen means. What made chess so suitable for this purpose
was the fact that the decision-making procedures leading up to a move are
highly amenable to hierarchical recursive analysis: a player thinks through a
problem by shuttling back and forth between sub- and primary-goals as he
3
In fact, they refused to use the term for several years, preferring instead to
describe their work as ‘complex information processing’.
302 Stuart Shanker
first formulates a plan, calculates the permutations, deduces whether any of

these will implement the strategy he has in mind, and if none of these succeeds
in realizing the overall goal, formulates a new plan and repeats the process.4
Most important of all was the fact that chess afforded the perfect environment
in which to map out the goal-directed pattern of the mechanical procedures
that constitute problem solving in a fixed domain: given the prior assumption
that these are moves within a calculus. The resulting models may of course be
far removed from anything one might encounter in ordinary chess play; but
the fact that we may not be aware of such processes is merely a confirmation
of the limits of introspection.
Significantly, Turing was not the first to seize on chess as a paradigm for
studying thought processes. Interest in the psychology of chess at the turn of
the century was largely confined to the study of prodigies. The subjects were
primarily masters, and the methodology strictly introspectionist (see Binet
1894, Cleveland 1907). The shift in attitudes towards the psychology of chess
had begun prior to and entirely independent of Turing: largely as a result of
Selz’s work on the ‘laws of cognitive activity’. Proceeding on the basis of the
Würzburg School’s approach to the psychology of thinking, Selz undertook to
confirm the existence of directed associations (i.e., of the manner in which an
Aufgabe influences a subject’s response to a stimulus). Where Selz’s particular
importance for the foundations of AI lies is in his insistence that
The individual analysis of task-conditioned thought processes always
shows an uninterrupted chain of both general and specific partial op-
erations which at times cumulatively (A + B + C) and at times in a
stepwise fashion (B after failure of A) impel the solution of the task.
These operations are continued until a solution is found or up to a
momentary or lasting renunciation of the solution (Simon 1982: 153).
Although Selz’s writings were relatively unknown amongst English
speaking psychologists, the work of one of his followers, Adriaan De Groot,
had an immediate impact on the evolution of GPS.5 De Groot sought to im-
plement Selz’s ideas in an exhaustive investigation of how a broad spectrum
of chess players set about solving board problems. His primary result (pre-
supposition?) was that, as Selz had outlined, such problem-solving processes
4
This theme was clearly expressed by Shannon in ‘A Chess-playing Machine’: The
thinking process is considered by some psychologists to be essentially characterized
by the following steps: various possible solutions of a problem are tried out mentally
or symbolically without actually being carried out physically; the best solution is
selected by a mental evaluation of the results of these trials; and the solution found in
this way is then acted upon. It will be seen that this is almost an exact description of
how a chess-playing computer operates, provided we substitute ‘within the machine’
for ‘mentally’ (Shannon 1956: 2132–3).
5
Although Thought and Choice in Chess was not translated into English until
1965, Simon recounts how, in 1954, he had taught himself Dutch solely in order to
read it (see Simon 1982: 149).
must be based on a linear chain of operations: a point which, as De Groot

noted in the Epilogue to the later English translation of Thought and Choice
in Chess, rendered his findings highly compatible with the requirements of
chess programming.
The picture of thinking which guides De Groot is the same as what we find
in Mechanical Intelligence: it postulates that we cannot—where the ‘cannot’ is
thought to be psychological—hope to capture the full range of our thoughts in
the net of language, either because so much of the thinking process is sublim-
inal, too rapid or too far removed for our powers of introspection, or simply
of a nature that outstrips the present possibilities of linguistic expression.
With training it might be possible to ameliorate some of these deficiencies,
but no amount of laboratory experience can enable a subject to discern the
‘elementary information processes’ out of which human problem solving is
compounded. Computer models thus provide the cognitive psychologist, not
just with a valuable, but in fact, with an essential adjunct to thinking-aloud
experiments. For without them we could never hope to overcome the inher-
ent limitations of introspection. Moreover, it is difficult to see how we could
otherwise hope to explain such phenomena as ‘moments of illumination’ or
the mind’s ability to solve problems of enormous computational complexity
in what, from the vantage point of current technology, seems like an astonish-
ingly small amount of time.
It is important to remember that De Groot laid the foundation for this
argument long before he became familiar with computer models of chess think-
ing. At the outset of his argument he approvingly cites Selz’s dictum that the
psychologist’s goal must be to deliver “a complete (literally: ‘gapless’) de-
scription of the causal connections that govern the total course of intellectual
and/or motor processes” in problem solving (De Groot 1965: 13). De Groot’s
major task was then to explain the phenomenon of pauses in a player’s re-
ports followed by a new approach; for ‘It is often during these very pauses
that the most important problem transformations appear: the subject takes a
“fresh look” at the entire problem’ (184). Given that a ‘subject’s thinking is
considered one continuous activity that can be described as a linear chain of
operations’ (54), ‘transitional phases have to be assumed in order to under-
stand the progress of the thought process even though the written protocol
gives no indication at all’ (113). Thus, we can hypothesize from the program
what must have been going on during the pauses in the player’s mind. For
there can be no such lacunae in the computer model: a program with such
‘gaps’ would literally grind to a halt. And the same must hold true for thinking
itself, given that it too is an effective procedure.
Reading through Mechanical Intelligence, one begins to appreciate just
how propitious Turing’s timing was. If Turing’s major accomplishment in ‘On
Computable Numbers’ was to expose the epistemological premises built into
formalism, his main achievement in the 1940s was to recognize the extent to
which this outlook both harmonised with and extended contemporary psycho-
logical thought. Turing sought to synthesize these diverse mathematical and
304 Stuart Shanker
psychological elements so as to forge a union between ‘embodied rules’ and

‘learning programs’. Through their joint service in the Mechanist Thesis each
would validate the other, and the frameworks from whence each derived. What
is all too often overlooked by AI theorists, however, is that by providing the
computational means for overcoming the impasse in which mechanism found
itself before the war, Turing was committed to the very framework—as defined
by its set of assumptions—which had created it. As important as Turing’s ver-
sion of Church’s Thesis was for the foundations of AI, no less significant was
the psychological thesis which provided the means for the transformation of
Turing’s ‘slave machines’ into ‘intelligent automatons’. It is to the latter that
we must look, therefore, in order to understand, not simply the genesis, but
more importantly, the presuppositions of AI. For it suggests that the gulf be-
tween pre- and post-computational mechanism may not be nearly so great as
has commonly been assumed.
11.2 Turing’s Behaviourist Ambitions
The main reason why it is so tempting to speak of Turing’s ‘computational

revolution’ is because of the contrast between behaviourist and AI attitudes
towards the role of ‘mental states’ in the explanation of actions. But before we
accept the radical divergence commonly postulated between ‘pre-’ and ‘post-
computational’ mechanism, we should consider the extent to which Turing
saw himself as working within the framework of behaviourism: as taking the
behaviourist account of problem solving a step further by treating ‘mental
states’ as ‘machine-state configurations’, thereby allowing for a reductionist
explanation of such higher-level activities as chess playing and theorem prov-
ing. This enables us to see that the route leading from Huxley’s ‘sentient
automatons’ through Jevons and Marquand’s ‘reasoning machines’ to Hull’s
and Turing’s ‘learning systems’ displays far more continuity than is commonly
acknowledged.
In order to understand the significance of Turing’s contribution to the
evolution of behaviourism, it is important to be aware that the thought-
experiment machines portrayed in ‘On Computable Numbers’ are not credited
with cognitive abilities as such; on the contrary, they are explicitly referred
to as ‘devoid of intelligence’. The routines that they execute are described as
‘brute force’: a reminder, not just of the repetitive strategy they use to solve
computational problems, but also, that they belong to the intellectual level
of the brutes—with all the Cartesian overtones which this carries. In Turing’s
words, these machines ‘should be treated as entirely without intelligence’; but,
he continues, ‘There are indications. . . that it is possible to make the machine
display intelligence at the risk of its making occasional serious mistakes’ (Tur-
ing 1946: 41). Just as we can say of a student exposed to ‘teachers [who] have
been intentionally trying to modify’ his behaviour that, ‘at the end of the pe-
riod a large number of standard routines will have been superimposed on the
original pattern of his brain’, so too, ‘by applying appropriate interference,

mimicking education, we should hope to modify the machine until it could
be relied on to produce definite reactions to certain commands’ (Turing 1959:
14).
The key to accomplishing this feat lies in the introduction of ‘learning pro-
grams’: self-modifying algorithms that revise their rules in order to improve
the range and sophistication of the tasks they can execute, thereby satisfy-
ing Lewes’s demand for an automaton capable of adapting to conditions not
calculated for in its construction. The obvious problem which this argument
raises, however, is whether, or in what sense, such programs can be described
as ‘learning’; and should this be deemed inappropriate, how Turing could have
assumed that ‘self-modifying’ means the same thing as ‘learning’. But before
this issue can be explored, there lies the prior question of why Turing should
have seized on this particular notion in order to implement his mechanist
ideas.
There are several reasons why learning assumed such importance in mech-
anist thought vis-à-vis both of the ‘traditions’ outlined in §1. In mythopaeic
terms, the automaton only springs to life once it displays the ability to recog-
nize and master its environment (at which point humanist anxieties invariably
surface in the form of the creator’s loss of control over this now autonomous
creature). In both physiological and psychological terms, questions about the
nature of learning dominated the mechanist/vitalist debates during the nine-
teenth century. And in terms of the history of AI, the first and in some ways
most potent objection raised against the Mechanist Thesis was voiced nearly
a century before the invention of computers.
In her Notes on Menabrea’s ‘Sketch’, Ada Lovelace cautioned that Bab-
bage’s ‘Analytical Engine has no pretensions whatever to originate anything.
It can do whatever we know how to order it to perform. It can follow analy-
sis; but it has no power of anticipating any analytical relations or truths. Its
province is to assist us in making available what we are already acquainted
with’ (Lovelace 1842: 284).6 As he makes clear in ‘Computing Machinery and
Intelligence’, the crux of Turing’s version of the Mechanist Thesis turns on
the very premise which Lovelace denies in this passage. ‘Who can be certain,’
Turing asks, ‘that “original work” that he has done was not simply the growth
of the seed planted in him by teaching, or the effect of following well known
general principles’ (Turing 1950: 150). The important point is that, granted
that the operations of a machine can be guided by rules (however simple these
might be), it should be possible to develop programs of sufficient complexity
to warrant the attribution of intelligence. It was this argument which was to
have so dramatic an effect upon mechanist thought. For Turing was to insist
that the essence of a ‘learning program’ is its ability to simulate the creative
aspect of human learning (see Turing 1947: 103-4).
6
Peirce was to develop this theme in several important papers on ‘logical ma-
chines’. I am indebted to Kenneth Ketner for drawing this to my attention.
306 Stuart Shanker
To serve as a defence of machine intelligence, this argument must assume

that learning ‘denotes changes in the system that are adaptive in the sense
that they enable the system to do the same task or tasks drawn from the
same population more efficiently and more effectively the next time’ (Simon
1983: 28). On first reading, this statement looks like little more than a strained
attempt to tailor the concept of learning so as to mesh with that of mechanical
rules. For if all that learning amounted to were the adaptation of something
to its environment we should be forced to conclude, not just that machines,
but indeed, the simplest of organisms is capable of some primitive form of
learning. But far from seeing this as an objection, the mechanist will respond:
Exactly, that is the whole point of the theory!
One cannot simply assume, therefore, that Turing only succeeded in sub-
verting the concept of learning in his zeal to reduce it to a level commensurate
with the minimal ‘cognitive abilities’ possessed by his machines. For such a
charge would fail to do justice to the manner in which AI evolved from the
union of mathematical and mechanist thought, and the extent to which the
latter had come to dominate learning theory. Moreover, it would ignore the
conceptual evolution of machine which underpins this outcome, and the bear-
ing which this had, not just on Turing’s psychological programme, but as a
result of his influence, on automata theory and thence AI. But most serious
of all, it would obscure the extent to which behaviourist presuppositions were
absorbed into the fabric of AI.
This behaviourist orientation is particularly evident in ‘Intelligent Machin-
ery’, the report which Turing completed for the National Physical Laboratory
in the summer of 1948. The purpose of this paper was to defend the claim
that self-modifying algorithms can legitimately be described as ‘learning pro-
grams’. The opening premise recalls Pavlov and Lashley’s theory that what
we call ‘learning’ is the result of new neural pathways brought about by con-
ditioning (see Pavlov 1927: 4ff). According to Turing, ‘the cortex of the infant
is an unorganized machine, which can be organized by suitable interfering
training’ (Turing 1948: 120). By enabling the system to modify its own rules,
Turing thought he had demonstrated how Turing Machines could in principle
simulate the formation of neural reflex arcs that take place during condition-
ing. The ensuing argument then expands on this notion of conditioning in
terms of the ‘Spread of Effect’ experiments inspired by Thorndike.
In Turing’s eyes, the most important element in this behaviourist theory
is that learning consists in neural stimulus-response connections. He takes
it as given that ‘The training of the human child depends largely on a sys-
tem of rewards and punishments.. . . Pleasure interference has a tendency to
fix the character, i.e., towards preventing it changing, whereas pain stimuli
tend to disrupt the character, causing features which had become fixed to
change, or to become again subject to random variation’. Accordingly, ‘It is
intended that pain stimuli occur when the machine’s behaviour is wrong, plea-
sure stimuli when it is particularly right. With appropriate stimuli on these
lines, judiciously operated by the “teacher”, one may hope that the “char-
acter” will converge towards the one desired, i.e., that wrong behaviour will
tend to become rare’ (Ibid: 121).
The concepts of extinction and positive reinforcement on which Turing
placed so much emphasis in his ‘learning’-based version of the Mechanist
Thesis were thus directly culled from behaviourist writings: it was by em-
ploying ‘analogues’ of pleasure and pain stimuli that he hoped ‘to give the
desired modification’ to a machine’s ‘character’ (Ibid: 124). As he put it in
‘Intelligent Machinery, A Heretical Theory’:
Without some. . . idea, corresponding to the ‘pleasure principle’ of the
psychologists, it is very difficult to see how to proceed. Certainly it
would be most natural to introduce some such thing into the machine.
I suggest that there should be two keys which can be manipulated by
the schoolmaster, and which can represent the ideas of pleasure and
pain. At later stages in education the machine would recognize certain
other conditions as desirable owing to their having been constantly
associated in the past with pleasure, and likewise certain others as
undesirable (Turing 1959: 132).
The mechanist metaphor would now appear to be twice removed from
the established meaning of ‘learning’. Whereas behaviourists had taken the
liberty of depicting habituation as a lower form of learning, Turing went a
step further and added the premise that machines display ‘behaviour’ which
as such can be ‘conditioned’ by ‘analogues’ of ‘pleasure and pain stimuli’ in
what can reasonably be described as ‘training’. Yet Turing was not alone
in this move; at much the same time Hull was arguing that ‘an automaton
might be constructed on the analogy of the nervous system which could learn
and through experience acquire a considerable degree of intelligence by just
coming in contact with an environment’ (Hull 1962: 820).
Far from being a coincidence, the affinity between Turing’s and Hull’s
thinking is a consequence of their shared outlook towards the nature of
problem-solving. This thinking is exemplified–and was to some extent in-
spired–by Thorndike’s experiments on the ‘learning curve’. Thorndike de-
signed a puzzle box to measure the number of times a cat placed inside
would randomly pull on chains and levers to escape. He found that when
practice days were plotted against the amount of time required to free itself,
a ‘learning curve’ emerged which fell rapidly at first and then gradually until
it approached a horizontal line which signified the point at which the cat had
‘mastered the task’. According to Thorndike, his results showed how animal
learning at its most basic level breaks down into a series of brute repetitions
which gradually ‘stamp’ the correct response into the animal’s behaviour by
creating ‘neuro-causal connections’. But repetition alone does not suffice for
the reinforcement of these connections; without the concomitant effects pro-
duced by punishment and reward new connections would not be stamped in.
One’s immediate reaction to Thorndike’s experiment might be that this
manifestly constitutes an example of anything but learning. But to appreciate
308 Stuart Shanker
the full force of Thorndike’s argument, you have to imagine that you were
shown a cat, that had already been conditioned, quickly freeing itself from
a puzzle box. Let us suppose that one’s natural inclination here would be to
say that the cat had clearly learnt how to free itself. The whole point of the
experiment is to show us, in an artificial condition, exactly what processes
had led up to this outcome. To speak of ‘insight’ here is completely vacuous;
and so too, Thorndike wants us to conclude, must it be in all other cases.
That is, we only speak of ‘insight’ when we are unfamiliar with the processes
that have led up to the results we have witnessed.
Herein lies the thrust of the behaviourist version of the continuum pic-
ture outlined in §1. The crucial point is the idea that learning consists in
the formation of stimulus-response connections which themselves require no
intelligence. A plant that turns its leaves towards the sun can be said to have
learned how to maximize its photosynthesis; a dog that is conditioned to sali-
vate at the sound of a bell can be said to have learned that it is about to be
fed. The ‘higher’ forms of learning—e.g., learning how to speak, how to count,
how to solve logical problems—are distinguished from these lower forms by
the complexity of the stimulus-response connections forged in the organism’s
brain. But the mechanical nature of the atomic associations which form the
basis for all levels of learning remain identical. This provides the rationale for
describing what had hitherto been regarded as disparate phenomena—as re-
flexive as opposed to cognitive phenomena—as constituting a learning contin-
uum ranging from simple negative adaptation, habituation, accommodation,
and tropisms, through animal and infant learning, to the highest reaches of
education and scholarship.
Turing’s Thesis was thus tailor-made for behaviourism. For as we saw in
§1, the epistemological significance of his computational version of Church’s
Thesis was said to consist in the demonstration that algorithms are complex
systems of meaningless sub-rules each of which can as such be applied purely
mechanically. The essence of Turing’s version of strong AI is that machine
intelligence is a function of the complexity of the program which the com-
puter follows rather than the individual steps of the algorithm. The difference
between ‘slave’ and ‘learning’ programs lies in the shift from fixed to self-
modifying algorithms. In the former, the Turing Machine repeatedly performs
the same elementary steps; in the latter it alters its program, using heuris-
tics which enable it to augment its knowledge base and/or store of rules, and
thence the range and sophistication of the tasks it can execute.
It is this argument which, according to the standard mechanist interpreta-
tion of Turing’s Thesis, enables us ‘to face the fact that a “human computer”
does need intelligence—to follow rules formulated in a language he must un-
derstand’ (Webb 1980: 220). In order to provide a ‘non-question begging’
analysis of computation, ‘the smallest, or most fundamental, or least sophis-
ticated parts must not be supposed to perform tasks or follow procedures
requiring intelligence’ (Dennett 1978: 83). Thus, ‘Turing’s analysis [of compu-
tation] succeeded just because it circumvented the problem of how a computer
can understand ordinary language’ (Webb 1980: 225). Without ‘meanings’ to

deal with, ‘these atomic tasks presuppose no intelligence’, from which ‘it fol-
lows that a non-circular psychology of computation is possible’ (Ibid: 220).
But not just a psychology of computation; rather, a psychology of thinking
simpliciter. For provided the lowest level of the ‘learning continuum’ can be
simulated, there is no a priori reason why machines should not be capable
of ascending this cognitive hierarchy. And this was exactly the theme which
Turing exploited in the 1940s in his defence of the Mechanist Thesis. In so
doing he saw himself as providing the crucial definition of ‘mechanical’ on
which the behaviourist theory of learning rests, and as a result, opening the
door to a future populated with thinking machines.
11.3 The Continuum Picture
One of the major philosophical questions that emerges as one reads through
Mechanical Intelligence is whether, by forging a union between recursive func-
tion theory and psychology, and within psychology, between Denkpsycholo-
gie and behaviourism, Turing was surmounting or subsuming the conceptual
problems that afflicted the latter, and to what extent this can be said to have
impinged on AI. The mechanist picture of a learning continuum which Turing
embraces is putatively one in which the higher forms of learning are built
up out of simpler components. But in actual fact, the theory proceeds in the
opposite direction: it is only by presupposing that the network of learning
concepts can be applied in an ever-diminishing state to the descending lev-
els of the ‘continuum’ that the converse reductionist analysis of these same
concepts can then be instituted.
The key to this move lies in treating the criteria—what an agent says
or does—that govern the application of learning as evidence of some hidden
transformation in an organism’s CNS. That is, being able to multiply sums
correctly is a criterion for saying ‘S has learnt the multiplication tables’.
But the mechanist continuum presupposes that being able to multiply sums
correctly is only evidence of the formation of the synaptic connections that
constitute learning how to multiply. Hence, an organism undergoing cellular
division also provides us with evidence that it has learnt how to multiply
(albeit at a much lower stage on the multiplication continuum).
The criteria governing the application of learning can thus be hived off
from declining applications of ‘learning’ until the atomic level is reached, where
learning is said to be purely a function of assimilation and accommodation to
stimuli. In other words, the theory postulates that it makes sense to speak of
‘learning’ in the absence of what we would normally regard as rule-following
behaviour: i.e., in the absence of the very possibility of explaining, teaching,
correcting, or justifying the use of a concept. Given that a plant, or a dog, no
less than a child, can respond to discriminably different stimuli, a child’s abil-
ity to learn how to speak, to count, to apply colour or shape or psychological
310 Stuart Shanker
concepts, must simply be the end result of a more complicated cerebral nexus
of exactly the same sort as guides the plant’s or the dog’s behaviour. (And
note that it does indeed make sense on this picture to speak of the plant’s ‘be-
haviour’; to resist this neologism would be to court ‘semantic conservatism’.)
The prior question of whether, or in what sense, we can speak of a plant or a
dog as learning how to distinguish between discriminably different stimuli—as
opposed to reacting to these stimuli—is an issue which the continuum picture
dismisses with the simple expedient of placing problematic uses of the concept
in question in inverted commas.
‘On Computable Numbers’ is filled with such inverted commas. Turing
tells us that a Turing Machine is ‘so to speak, “directly aware”’ of the scanned
symbol. ‘By altering its m-configuration the machine can effectively remem-
ber [the inverted commas are missing here—SS] some of the symbols which
it has “seen” (scanned) previously’ (Turing 1936: 117). Here Turing conforms
to the precedent established by Jacques Loeb, who demonstrated that when
Porthesia Chrysorrhoea are exposed to a source of light coming from the op-
posite direction to a supply of food, they invariably move towards the former
(and perish as a result). In Loeb’s words, ‘Heliotropic animals are. . . in real-
ity photometric machines’: not that far removed, as it happens, from Turing
Machines (Loeb 1912: 41). Loeb designed his experiments to undermine the
vitalist thesis that all creatures are governed by an unanalyzable instinct for
self-preservation. His conclusion was that, ‘In this instance the light is the
“will” of the animal which determines the direction of its movement, just as
it is gravity in the case of a falling stone or the movement of a planet” (Ibid:
40–1). Since it is possible to explain, ‘on a purely physicochemical basis’, a
group of ‘animal reactions. . . which the metaphysician would classify under
the term of animal “will”’, the answer to no less than the ‘riddle of life’—viz.,
the nature of free will—must lie in the fact that ‘We eat, drink, and repro-
duce not because mankind has reached an agreement that this is desirable,
but because, machine-like, we are compelled to do so’ (Ibid: 35, 33).
Loeb treats tropisms as quantitatively, not qualitatively, different from
the mastery of a concept: they are simply a more primitive form of ‘equiv-
alence response’. For on the continuum picture, concepts are ‘the mediating
linkage between the input side (stimuli) and the output side (response). In
operating as a system of ordering, a concept may be viewed as a categorical
schema, an intervening medium, or program through which impinging stimuli
are coded, passed, or evaluated on their way to response evocation’ (Harvey,
Hunt, Schroder 1961: 1). To be sure, on the folk theoretical level of psycho-
logical discourse, to say that ‘S possesses the concept φ’ may be to say that
S has mastered the rules for the use of ‘φ’; but this is seen as nothing more
than the name of a problem, and psychology’s task is to explain in satisfactory
casual terms what this ‘normativity’ consists in.
The reason why the mechanist has no qualms about, e.g., describing cater-
pillars as ‘learning’ where the light was coming from is because of his funda-
mental assumption that when we describe an organism as ‘acquiring knowl-
edge’, ‘we suppose that the organism had some specific experience which
caused or was in some way related to the change in its knowledge state’
(Bower & Hilgard 1981: 13). And this is where Turing’s claim that ‘To each
state of mind of the [human] computer corresponds an “m-configuration” of
the machine’ (Turing 1936: 137) stepped in. With the benefit of this reduction
of mental to machine states, and the step-by-step interaction of these ‘internal
configurations’ with external input, AI was in a position to reintroduce the
various cognitive concepts, hierarchically arranged, at each successive stage
on the phylogenetic continuum.
Turing could thus defend his speaking of a chess program as ‘learning’ from
its past ‘mistakes’ on the grounds that this use of ‘learning’ is no different from
the sense in which Loeb, Thorndike, and Pavlov employed the term. What we
have to remember here is the reductionist animus of these early behaviourists.
Thorndike and Pavlov were not guilty of inadvertently failing to distinguish
between learning and conditioning. The whole point of their theories is that
what we refer to in ordinary language as ‘animal learning’ is in fact merely
a species of conditioning: i.e., that an animal’s behaviour can be explained
without appealing to any ‘mentalist’ concepts. And whatever is true of their
behaviour applies, in virtue of the continuum picture, to human learning.
It is precisely this reductionism—as it applies to the explanation of human
behaviour—which Bruner is reacting to in his repudiation of AI (see Bruner
1990: 4ff). But if one is to challenge the eliminative results of the continuum
picture, then it is the picture itself that must be scrutinized (see Shanker 1992,
1993a). For it is not AI per se that is the problem here, it is the epistemological
framework which underpins the succession of mechanist models described in
§1: a framework which sees adaptation and accommodation as simpler versions
of the same neural phenomena that constitute human learning.
What if we should approach this issue from the opposite starting-point?
What if we should take as the paradigm for applying the concept of learning
the mastery of the rules governing the use of a concept? The application of
‘learning’ in more primitive contexts would then be limited by the extent to
which these resemble standard normative practices: i.e., that doing such-and-
such in appropriate circumstances constitute criteria for what we call ‘learning
how to φ’. In some cases, these criteria can be relaxed to a point that allows
us to speak of ‘primitive’ applications of ‘learning how to φ’. For example, the
very concept of language is such that it makes sense to speak of ‘primitive
linguistic practices’ (see Shanker 1994).
The ‘continuum’ which results from this perspective is grammatical, not
mechanical. The question of whether the behaviour of any particular organ-
ism can be described as a primitive form of ‘learning’ depends on whether
its actions can be described as benefitting from training or experience, as re-
sponding to correction and instruction, as acquiring simple and complex skills.
Has Kanzi learned the meaning of ‘bubbles’: can he use and respond to the use
of words to initiate or participate in the appropriate interindividual routines
(see Savage-Rumbaugh et al. 1993)? Did Pavlov’s dogs learn that food was
312 Stuart Shanker
imminent when a bell sounded? By extinguishing a conditioned stimulus did

Pavlov teach a dog something else, or did he demonstrate how misleading it is
in such circumstances to describe the dog as having learned something? Is the
ability of lichen to survive in the most inhospitable of environments a skill?
By correcting a program have we trained it to φ? Is this really comparable,
as Turing so readily assumes, to the use of punishment and rewards in the
training of a human child?
In all of these cases, the issue is decided by clarifying the logical grammar
of learning. Typically, it is a subject’s ability to respond appropriately to the
use of ‘φ’, to master the rules for the use of ‘φ’, and to explain the meaning
of ‘φ’ (even if only by ostension) that licenses our describing him as having
‘learned how to φ’. These rules can clearly be stretched to a point that allows
us to speak of primitive instances, but not so far as to regard any form of
causal regularity (e.g., a tropism, or a thermostat) as displaying evidence of
having learned x. It is precisely because it makes no sense to speak of a dog
that salivates at the sound of a bell as responding ‘correctly’ or ‘incorrectly’
to the signal that it makes no sense to try to reduce learning to the same
terms. A dog which, bell or no bell, incessantly salivates, has no more made
a mistake than has an electric door that fails to open when its photo-electric
beam is crossed.
The central issue underlying all these questions is whether the beha-
viourists established what learning consists in at the primitive level, or rather
demonstrated that the primitive behavioral responses that they studied should
be explained in mechanical as opposed to normative terms. Indeed, were they
really seeking to explain the concept of learning, or were they simply attacking
its indiscriminate use by comparative psychologists at the turn of the century?
By conditioning an animal to φ, and then extinguishing that conditioned re-
sponse, they showed how inappropriate it would be to describe such behaviour
in normative terms. But that hardly entails that there are no circumstances
in which it makes sense to describe an animal as learning something: that the
concept of animal learning—indeed, learning simpliciter—reduces to the con-
cept of conditioning. If anything, the upshot of their experiments is that they
sharpened the criteria for speaking of ‘conditioning’. But that only serves to
reinforce the categorial distinction between learning and responding to a stim-
ulus by reminding us that the difference between a tropism and, e.g., learning
how to count is one of kind, not degree, insofar as it only makes sense to speak
of explaining the latter in normative terms—viz., of mastering the rules for
the use of the concept—but a tropism or a conditioned response can only be
explained in terms of internal mechanisms.
That is certainly not to deny that reflexes shade into skills, but instead, to
see this observation as drawing attention to the grammatical continua which
frame any psychological investigations into the possible existence of a phylo-
genetic/ontogenetic continuum: scalae linguisticae which range from primitive
to rarefied uses of psychological concept words. Adult human beings serve as
the paradigm subjects for our theoretical discussions about the use of psycho-
logical concepts: any question about the psychological capacities of animals

or infants demands that we compare their behaviour with the relevant adult
human actions which we regard as licensing the use of the concepts in ques-
tion. These grammatical continua are counterpoised against behaviour and
demand ever more complex actions to license the attribution of ever more
complex cognitive skills and abilities.
It is on the basis of the grammatical continuum governing the application
of learning, therefore, that we question whether it makes any sense to describe
Turing’s machines as ‘learning how to play chess’: i.e., the question is whether
there are any justificational criteria—whether there are sufficient grounds in
the machine’s behaviour—for the attribution of ‘learning’ to get a foothold.
That does not mean that we should expect there to be a hard and fast line
between ‘reacting’ and ‘learning’ in all contexts: between the use of causal
and the use of normative terms. When studying a child’s development, for
example, it is often difficult to judge where exactly reacting ends and learning
begins. But to speak of ‘learning’ nonetheless presupposes that it makes sense
to describe the child as mastering the rules for the use of ‘p’, as opposed to
responding in a consistent manner to the sound of ‘p’. That is, if we are to
make the transition from describing a child’s (or an animal’s) behaviour in
causal to describing it in normative terms, his (or its) actions and reactions
must warrant our speaking of him (or it) as e.g., ‘beginning to use or to
respond to the use of “p” correctly’.
This argument illustrates the full gravity of the charge being levelled here
that the continuum picture does not genuinely attempt to ‘build up’ from
bottom to top but rather, proceeds in the opposite direction. By presuppos-
ing that learning can be treated as the same sort of process as assimilation
or accommodation—that habituation constitutes a species of cognition—the
mechanist has already assumed that e.g., when caterpillars are attracted to
a light this signifies a change in their ‘knowledge state’. But is knowledge a
state, much less one that is caused by external stimuli and whose inception
and duration can be measured? Can knowledge be equated with a change
in behaviour? After all, one can behave in a certain way without knowing
what one is doing, or conversely, conceal what one knows. And is it really the
case that ‘we infer someone’s knowledge from inputs to him and outputs from
him, and we infer learning caused by an experience because of before-to-after
changes in his inferred knowledge’ (Bower & Hilgard 1981: 14)? Do we infer
that a child who is able to answer all the adding and subtracting questions
that we put to him knows how to count? Isn’t this just what we call ‘knowing
how to count’, and when a child can perform these sums we see that he has
learnt how to count? Aren’t these all criterial as opposed to evidential issues:
grammatical as opposed to empirical questions?
Like understanding, to which it is so intimately connected, the concept
of learning embraces a wide spectrum of activities that are loosely based on
the development of skills, not ‘cortical connections’ (assuming that there even
is such a thing). To learn how to speak is different from learning a second
314 Stuart Shanker
language; to learn how to tie one’s shoes is different from learning how to play
a Bach fugue. But this grammatical continuum, which progresses from simple
to compound skills, from practice and drill to experience and understanding,
does not ‘descend’ still further, to the point where we can treat the statements
‘The thermostat clicked on’, ‘The leaves of the plant turned towards the sun’,
‘The pigeon pecked the yellow key in order to get a food pellet’, ‘Kanzi pressed
the “drink” lexigram in order to get a drink’, and ‘S has learned how to play
the Toccata and Fugue in D minor’ as all similar in kind: as distinguished
only by their mounting internal complexity.
Just as our descriptions of reflexive movements shade into our descriptions
of automatic behaviour (in the way e.g., that red—or rather, crimson—can be
said to shade into purple), so too causal descriptions of training (of reacting,
associating, repeating or memorizing words) shade into normative descriptions
of learning and understanding: of following, appealing to, explaining, teaching
the rules for the use of words. Different kinds of training lead into different
kinds of normative practice (e.g., how we are taught to count, to describe
colours, or objects, or pains, or actions). The different kinds of practice result
in different kinds of concepts, and to seek to reduce this heterogeneity to a
common paradigm—e.g., the so-called ‘functional’ definition of concepts—is
to embrace the presuppositions which ultimately result in eliminative materi-
alism or mechanist reductionism.
Behaviourism and AI are converse expressions of this point. Both proceed
from the assumption that, since there is a continuum leading from reflexes to
reactions to concept acquisition, the only way for psychology to explain the
mechanics of learning is by having a uniform grammar of description: in the
case of behaviourism, by reducing all explanations of behaviour to the terms
that apply to reflexes, and in the case of AI, by reading in the attributes
which apply to higher-level cognitive abilities and skills into the lowest levels
of reflex actions (e.g., the brain is said to make hypotheses, the nervous system
to make inferences, the thermostat to possess knowledge and beliefs). This is
why it is so often argued these days that AI is just a form of neobehaviourism;
for the fact is that, although the AI scientist may approach learning theory
from the opposite point of view from the behaviourist, he does so because he
shares the same framework as the behaviourist. Thus, on both theories we end
up with caterpillars and even thermostats that think (or perhaps one should
say ‘think’).
This brings us back to the different problems involved in the philosophical
question ‘Can machines think?’ and the psychological question ‘Do thinkers
compute?’ To talk about the ‘mechanics of learning’ is to refer to the impor-
tance of drill, repetition, a systematic training, of allowing scope for creativity
and inculcating intrinsic motivation. We are no more interested in neurophys-
iology when concerned with the mechanics of learning arithmetic than we are
interested in kinesiology when we speak of mastering the mechanics of golf.
It might be possible to design a program that scores the same success rates
on a series of standardized tests as a young child, but could a program be
designed ‘to simulate’ all those things that a child does when learning how
to count: e.g., to find the same problems difficult or easy, to make the same
sorts of excuses, to have its attention wander? Merely designing a program
that matches and even predicts the mistakes a child will make hardly satisfies
our criteria for describing that program as having learned how to count. And
it in no way entails that the operations of that program must shed light on
how a child learns how to count.
The problem here has nothing to do with semantic conservatism. It lies
rather in the fact that, far from explaining, the continuum picture only serves
to undermine the normative foundation on which the concept of learning rests:
the criteria which govern our application of the concept. This is what Turing’s
version of the Mechanist Thesis is all about. Turing’s Thesis only works as a
psychological programme—a ‘model of how the mind works’—on the basis of
the reductionism inherent in the continuum picture. But if the latter should
be a product of unwarranted epistemological assumptions, what are we to say
of the models which it yields? The problems that arise here have nothing to do
with any technical ‘shortcomings’ of computationalism. They are concerned
rather with the question of whether, because of its grounding in the continuum
picture, AI distorts the nature of the cognitive phenomena which it seeks
to explain. There is no need to delve into social or cultural psychology to
see the force of this point. The core of Turing’s argument—his ‘analysis’ of
computation—amply demonstrates the manner in which the framework drives
the theory, rather than the other way round.
11.4 Wittgenstein versus Turing on the Nature of

Computation
On the argument being sketched here, the union which Turing forged between
recursive function theory and behaviourism was grounded in an archetypal
epistemological picture. Furthermore, it was this epistemological picture which
served to establish AI as the paradigm for the burgeoning field of Cognitive
Science. For the cognitive revolution was driven by the premise that ‘Good
correspondence between a formal model and a process—between theory and
observables, for that matter—presupposes that the model will, by appropriate
manipulation, yield descriptions (or predictions) of how behaviour will occur
and will even suggest forms of behaviour to look for that have not yet been
observed—that are merely possible’ (Bruner 1959: 368). The big appeal of AI
lay in the hope that ‘perhaps the new science of programming will help free us
from our tendency to force nature to imitate the models we have constructed
for her’ (Bruner 1960: 23).
Turing was responding, in effect, to a challenge that had been laid down by
Kant, and which had preoccupied psychologists throughout the latter half of
the nineteenth century. Kant had insisted that psychology could ‘become noth-
ing more than a systematic art’. It could ‘never [be] a science proper’, since it is
316 Stuart Shanker
‘merely empirical’ (Kant 1970: 7). What he meant by this last remark is that,
because psychology is based on evidence yielded by the ‘inner sense’, it could
not (pace taxonomic botany) become a true science—i.e., a science on the
paradigm of mechanics—since the latter requires the mathematization of its
subject matter. What Turing was doing in ‘On Computable Numbers’—even
if he did not become fully aware of the fact until about five years later—
was attempting to lay the foundation for just such a science by showing
how, through the use of recursive functions, mental processes could indeed be
mathematized. More is involved here, however, than simply establishing the
bona fides of psychology. More to the point, Turing was responding—using
the tools which Hilbert had developed (see Detlefsen 1993)—to the epistemo-
logical problems that result from Kant’s picture of the mind confronted with
and forced to make sense of reality.
Kant saw his basic task as that of explaining the rule-governed manner in
which the mind imposes order on the flux of sensations that it receives (see
the opening pages of the Logic). The power of Turing’s Thesis lay in the fact
that he was able to analyse these rules in such a way as to dispell the air
of metaphysical speculation surrounding Kant’s argument. For Turing’s ‘em-
bodied rules’ seemed to open up the prospect of a mechanist explanation of
the connection between perception (input) and behaviour (output), between
forming and systematising concepts, mastering and being guided by rules,
acquiring and exercising an ability, hearing a sound or seeing a mark and
grasping its meaning, forming an intention to φ and φing. But these are clas-
sic epistemological, not psychological problems; and the thought-experiment
whereby Turing sought their solution was no more a piece of psychology than
the Transcendental Aesthetic.
Turing discusses the nature of Turing (‘computing’) Machines at two dif-
ferent places in ‘On Computable Numbers’: §§1–2 and §9. In the first instance
he defines the terms of his thought-experiment; in the second he takes up his
promise to defend these definitions. The crucial part of his argument is that
The behaviour of the computer at any moment is determined by the
symbols which he is observing, and his ‘state of mind’ at that mo-
ment.. . . Let us imagine the operations performed by the computer
to be split up into ‘simple operations’ which are so elementary that
it is not easy to imagine them further divided. Every such operation
consists of some change of the physical system consisting of the com-
puter and his tape. We know the state of the system if we know the
sequence of symbols on the tape, which of these are observed by the
computer. . . and the state of mind of the computer.. . . The simple op-
erations must therefore include:
(a) Changes of the symbol on one of the observed squares.
(b) Changes of one of the squares observed to another square within L
squares of one of the previously observed squares.. . .
The operation actually performed is determined. . . by the state of
mind of the computer and the observed symbols. In particular, they

determine the state of mind of the computer after the operation is
carried out.
We may now construct a machine to do the work of this computer. To
each state of mind of the computer corresponds an ‘m-configuration’
of the machine. . . (Turing 1936: 136–7).
Davis remarks of this argument that ‘What Turing did around 1936 was to
give a cogent and complete logical analysis of the notion of “computation”.
. . . Thus it was that although people have been computing for centuries, it
has only been since 1936 that we have possessed a satisfactory answer to
the question: “What is a computation?”’ (Davis 1978: 241). Ignoring for the
moment the validity of this claim, it is important to see how accurately Davis
has represented the tenor of Turing’s argument. It was never intended to
be read as a psychological thesis. It is rather a reductive conceptual analysis
whose epistemological significance is said to lie in the demonstration that
algorithms can be defined as complex systems of meaningless sub-rules, each
of which can as such be applied purely mechanically, from which Turing’s
psychological thesis is said to follow. Davis goes on to explain how
Turing based his precise definition of computation on an analysis of
what a human being actually does when he computes. Such a person
is following a set of rules which must be carried out in a completely
mechanical manner. Ingenuity may well be involved in setting up these
rules so that a computation may be carried out efficiently, but once
the rules are laid down, they must be carried out in a mercilessly exact
way.
Even before one gets to the problems involved in this use of ‘mechanical’
(see Shanker 1987), it is important to consider the accuracy of this first line.
Far from basing his ‘precise definition of computation on an analysis of what
a human being actually does when he computes’, Turing based his ‘analy-
sis’ on a Kantian epistemological picture of what the human being’s mind is
doing when he computes. Computation, according to Turing, is a cognitive
process, large parts of which are hidden from introspection; we can only infer
the nature of these pre-conscious operations from an agent’s observed behav-
iour and self-observations. (Kant presents a similar conception of inference in
the Logic.) Given that the phenomena to be explained in psychology are so
strongly analogous to those with which physics deals, is it any wonder that
both disciplines should be seeking for good correspondence between a formal
model and a process—between theory and observables?
It is precisely because his thought-experiment represents an attempt at
conceptual analysis that Turing’s Thesis has been the source of so much philo-
sophical discussion. And perhaps the most pertinent such investigation is that
conducted by Wittgenstein. Interestingly, Turing’s thought-experiment shares
some affinities with Wittgenstein’s use of artificial language-games to clar-
ify a concept: particularly Wittgenstein’s discussion of ‘reading-machines’ at
318 Stuart Shanker
§§156ff in Philosophical Investigations, which bears an uncanny resemblance

to Turing’s argument.7 Moreover, Turing’s reductionist thesis exemplifies, not
just the target which Wittgenstein is attacking, but the larger point that he is
making at §§122–33 of the Investigations about the dangers of philosophical
theorizing.
The hardest thing to get clear about in Wittgenstein’s discussion of read-
ing is its purpose. Wittgenstein introduces this theme in the midst of his
investigation into the nature of understanding. If the discussion of reading is
to help clarify the nature of understanding it obviously cannot presuppose un-
derstanding. Thus Wittgenstein dwells on reading at its most mechanical: viz.,
‘reading out loud what is written or printed’. But this is not an early version
of the call for a ‘presuppositionless psychology’ (see Dennett 1978). Wittgen-
stein was not looking for the ‘atomic units’ out of which understanding is
‘composed’. Rather, he uses a form of primitive language-game in order to
clarify some of the problems involved in reductionist analyses of psychological
concepts.
The parallel between Wittgenstein’s ‘reading-machines’ and Turing’s com-
puting machines is striking. Reading (counting at PI §§143ff) can serve as
an example of a rule-governed procedure at its most mechanical. The case
Wittgenstein wants to consider is where the agent ‘function[s] as a mere
reading-machine: I mean, read[s] aloud and correctly without attending to
what he is reading.’ Here is a possible key to Wittgenstein’s remark at §1096 of
Remarks on the Philosophy of Psychology (volume I): ‘Turing’s “Machines”.
These machines are humans who calculate.’ That is, what Turing is doing in
his thought-experiment is imagining people performing the most routine of
calculating tasks in order to analyse the concept of calculation. But whereas
Turing’s goal was indeed ‘to break calculation down into its elementary psy-
chic units’, Wittgenstein was looking to clarify the criteria which license us in
speaking of possessing a cognitive ability at its most primitive level, and the
bearing which this has on reductionism.
The target in Wittgenstein’s discussion is the continuum picture. Witt-
genstein distinguishes between three broad uses of the term ‘read’:
(1) Causal, as this applies to a machine reading signs;
(2) Primitive, as this applies to someone who does not understand
a word of what he is reading;
(3) Pardigmatic uses of ‘read’ (i.e., understanding what one is reading).
Wittgenstein argues that there are many situations in which it is difficult to
distinguish between (2) and (3) (e.g., a child learning how to read, a politician
reading from a cue card). And now the interesting question is: does (or could) a
7
The overlap appears to be coincidental, although that in no way diminishes
its significance. Wittgenstein had already formulated his argument by 1934 (see
Wittgenstein 1960: 117ff), and it seems unlikely that Turing had read this before he
composed ‘On Computable Numbers’.
similar ambiguity exist between (1) and (2)? Is the machine-use just a further
extension of the primitive use—as the continuum picture stipulates—and if
so, is this because the internal mechanisms guiding the machine’s operations
are a simpler version of the internal mechanisms guiding the organism? Could
we therefore learn about the ‘processes of reading’ involved in (2) and (3) by
studying the machine’s processes?
Wittgenstein’s response to this problem is to clarify the criteria which
govern the application of ‘reading’ in primitive contexts. He looks at the case
in which we would say that ‘S read p’ even though he had no idea as to the
meaning of p. Even in a limiting case like this, we still demand some criterion
to distinguish between ‘S read p’ and ‘It only seemed as if S read p’ (e.g., ‘The
parrot read the sign’ and ‘It only seemed as if the parrot read the sign’). That
is, no matter how primitive the context, we insist on being able to distinguish
between φing and seeming to φ. The question then is, on what basis do we
draw this distinction. Do (or could) neural considerations play any role here?
Could our judgment that someone is reading be overturned by neurological
evidence? And could the neurophysiologist tell us what the subject is doing
when he is reading?
Wittgenstein makes the distinction between reading and seeming to read
(pretending to read, memorizing the words, repeating sounds) as hazy as
possible. We are at that indeterminate point in a pupil’s training where no
one—not the pupil, not his teacher—can say exactly when he started to read.
The argument is intended to take up the point made in §149. The question
Wittgenstein posed there is: what is the relation between what we call ‘the
possession of an ability’ (‘the acquisition of a concept’) and what we call
‘the exercise of that ability’ (‘application of that concept’). If we treat this
relation as causal—i.e., possession causes exercise—then we shall be drawn
into treating understanding (reading, calculating) as a mental state or process,
and thence, into the construction of ‘hypothetical mind-models’ to explain how
that hidden state or process causes the agent’s observed behaviour:
[A] state of the mind in this sense is the state of a hypothetical mech-
anism, a mind model meant to explain the conscious mental phenom-
ena. (Such things as unconscious or subconscious mental states are
features of the mind model.) In this way also we can hardly help con-
ceiving of memory as of a kind of storehouse. Note also how sure
people are that to the ability to add or to multiply. . . there must cor-
respond a peculiar state of the person’s brain, although on the other
hand they know next to nothing about such psycho-physiological cor-
respondences. We regard these phenomena as manifestations of this
mechanism, and their possibility is the particular construction of the
mechanism itself (Wittgenstein 1960: 117–18).
The conclusion which Wittgenstein draws in §149 is that the pair of con-
cepts conscious/unconscious has no bearing on the relation between posses-
sion/exercise of an ability (acquisition/application of a concept). Wittgenstein
320 Stuart Shanker
returns to this theme in §156. We look for the difference between possessing
and not possessing the ability to read in ‘a special conscious mental activity’,
and because we cannot find any unique phenomenon to distinguish between
reading and seeming to read, we conclude that the difference must be uncon-
sious: where this is construed as either a mental or a cerebral activity. (‘If
there is no difference in what they happen to be conscious of there must be
one in the unconscious workings of their minds, or, again, in the brain’ (PI:
§156).) This, according to Wittgenstein, is a paradigmatic example of the
‘metaphysical “must”’. For if we construe the relation between ‘knowing how
to read’ and ‘reading’ as causal —i.e., if we construe the grammatical relation
between ‘possessing the ability to φ’ and ‘φing’ (viz., that doing x, y, z consti-
tute the criteria for what we call ‘φing’) as causal (viz., the ability to φ consists
in a state-configuration which causes a system to x, y, z)—then we shall be
compelled to identify the ‘two different mechanisms’ which must distinguish
reading from seeming to read.
In the Brown Book Wittgenstein remarks how the failure to observe this
logical distinction proceeds from a picture of the agent as being ‘guided by the
signs’. From this it seems to follow that we can only understand the nature of
the mental activity in which reading consists ‘if we could look into the actual
mechanism connecting seeing the signs with acting according to them. For we
have a definite picture of what in a mechanism we should call certain parts
being guided by others. In fact, the mechanism which immediately suggests
itself. . . is a mechanism of the type of a pianola’ (Wittgenstein 1960: 118).
The problem with this picture has nothing to do with machines lacking con-
sciousness, or with an illicit use of ‘reads’. For, regardless of whether or not
it makes any sense to speak of machines as being conscious, there is indeed a
sense in which we can say that, e.g., a pianola is ‘guided by’ the notes that
it ‘reads’. But then, is the pianola ‘guided by’ the notes that it ‘reads’ in the
same way that a child is guided by the rules he has been taught, (i.e., the way
the child goes through these rules in his mind, checks to see if this instance
applies, automatically repeats certain mnemonics, etc.)?8
8
In §162 of the Investigations Wittgenstein advances a definition of ‘reading’
as ‘deriving a reproduction from the original’. The argument is about whether the
pupil’s behaviour warrants saying that he derived his action from a rule we have
taught him: e.g., if he can be seen to check the rule in the process of reading a
text, if he cites the rule to justify what he has read. But the ‘definition’ proposed
in the first line of §162 is after something much stronger: viz., you can only be
said to be ‘reading’ if this intermediary ‘inferential’ process occurs, and if you are
not conscious of this process then it must have occurred unconsciously. The thesis
under attack here is that of ‘secondarily automatic actions’. Since learning how to
read demands that we be conscious of such intermediary processes, they must have
become automated in the adult reader. Hence, we can deploy a protocol of the
child’s behaviour as the paradigm for the ‘unconscious processes’ that must occur
in a skilled reader. Wittgenstein often touches on this theme in his discussions of
‘calculating in the head’: e.g., we use the long-hand calculations that would be done
The continuum picture presupposes that these two uses of ‘being guided’
are indeed the same: but only because it presupposes that the sense in which
it applies to the child can be reduced to the sense in which it applies to the
machine. On the mechanist argument which Turing embraces, to say that the
child learning how to read is guided by the rules he has been taught is to say
that these rules have been ‘embodied’ in his brain. That is, the child is guided
in exactly the same way as the machine is guided by its program. Only the
differing complexity of their internal state-configurations accounts for their
divergent reading abilities.
Nowhere does Wittgenstein suggest that the kind of use which Turing
makes of inverted commas should be deemed illicit. For who made the philoso-
pher the custodian of good grammar? Can we not understand what Turing
is saying? The point is not that it makes no sense to speak of a Turing Ma-
chine as ‘reading’ or ‘calculating’. It is that this machine use of ‘reading’ or
‘calculating’ is categorially different from the normative sense in which we use
it for children and adults: even in its most primitive applications. Reading is
indeed a family-resemblance concept: a family which ranges from the infant
responding to flash cards to a philosophy class reading the Investigations or
a navigator reading the stars. But when we speak, e.g., of a scanner ‘reading’
a bar code, this is not a still more primitive extension of an infant reacting
to signs.
The mechanist insists that the only difference between the scanner ‘read-
ing’ a bar code and someone reading something mechanically lies in the opera-
tions of the internal mechanisms guiding their respective actions. Accordingly,
it should be possible to design a system that would perfectly simulate—and
thus explain—the operations of the mechanism guiding the human reading-
machine. Wittgenstein responds: ‘By calling certain creatures “reading ma-
chines” we meant only that they react in a particular way to seeing printed
signs. No connection between seeing and reacting, no internal mechanism en-
ters here’ (Wittgenstein 1960: 121). Neurophysiological evidence has no bear-
ing on the application of ‘reading mechanically’: can neither corroborate nor
undermine its use. What misleads us into thinking otherwise is ‘the idea of
a mechanism that works in special media and so can explain special move-
ments.. . . But what is of interest to us in reading can’t be essentially something
internal’ (Wittgenstein 1974: 99). That is, ‘The distinction between “inner”
and “outer” does not interest us’ when clarifying the concept of reading in
paradigmatic or in primitive contexts. The criteria for describing an agent
as ‘reading’ or as ‘reading mechanically’ lie in his behaviour and the situ-
ation in which this occurs: i.e., we call acting thus-and-so in such-and-such
circumstances ‘reading’ or ‘reading mechanically’.
In §157 Wittgenstein returns to the theme that the description of behav-
iour in causal terms shades into the description of behaviour in normative
on paper as the paradigm for what must have gone on in an agent’s mind when the
answer to a problem suddenly occurs to him.
322 Stuart Shanker
terms.9 Causal descriptions of behaviour—of conditioned responses, associa-

tions, repeating sounds—merge into normative descriptions of behaviour: of
reading or calculating (of being able to cite the rules one has followed in read-
ing or calculating). But despite the graduated nature of applying psychological
concepts in primitive contexts, the transition from ‘reacting’ to ‘reading’ in-
volves a fundamental categorial shift: the terms that apply to conditioned
responses do not carry over into the description of rule-following behaviour,
even though it may at times be difficult to identify where ‘reacting’ ends off
and ‘reading’ begins. The crucial point here is that this grammatical contin-
uum does not license the shift to speaking of a mechanist continuum, such
that the sole difference between machine uses, primitive uses, and paradig-
matic uses, is one of degree, not of kind. We do indeed speak of an agent
‘reading like a machine’, but only when we want to signal his failure to attend
to what he was reading, or to describe the manner in which he was reading
(in a monotone, without pauses or inflections).
Wittgenstein does not present us with an empirical generalization here;
rather, with the grammatical observation that ‘it makes no sense’ to ask a
question like ‘What was the first word S read?’. This distinguishes the two
‘mechanical’ uses of ‘reading’ described above: the primitive (2) and the ma-
chine use (1). For the question ‘What was the first symbol the scanner read?’
is the easiest thing in the world to answer. Here we have another useful re-
minder of the different uses of ‘reading’, and more importantly, of the fact
that the relation between possession and exercise is not causal. If it were,
then the question would make perfectly good sense in both normative and
causal contexts.10
§158 ties this argument in to the attack on causal theories of meaning
and understanding initiated at §6 (cf. Wittgenstein 1974: 190, 117–18). The
mechanist misconstrues the grammatical proposition that we cannot speak of
‘a first word in S’s new state’ (i.e., ‘it makes no sense’) as an empirical claim;
maybe, he suggests, this is only a temporary difficulty. But the mechanist is
only drawn into this position because he is already captivated by the idea
that the connection forged in training (between S’s possessing the ability to
φ and his φing) is causal rather than grammatical. And if the changes in S’s
behaviour are brought about by changes in his internal mechanism, then it
must be possible to say or predict what he has read/learnt if only we could
read the brain’s secrets: i.e., the changes occurring in ‘learning’.11
9
But note that not all causal uses imply a lack of understanding. Cf. ‘He reacted
with alarm when he saw the sign of the four’.
10
Note also the subtle distinction drawn in the last line of §157 between ‘change
in his behaviour’ and ‘state-transformations’. If there were no difference between
these two uses then ‘change in behaviour’ would be taken as a sign of an underlying
change in the hypothetical internal mechanism.
11
Note the picture of learning here. Cf. the argument presented at p. 118 of
Wittgenstein 1960 with Turing’s conception of ‘learning programs’.
Wittgenstein’s response to this argument is grounded in his remarks on

psychophysical parallelism (see Shanker, 1993b). His concern here is with the
source of the ‘must’ underpinning the mechanist argument: viz., the miscon-
ception of the relation between possession and exercise of an ability (concept).
Neurophysiological knowledge would not explain but rather presupposes inde-
pendent criteria for using ‘reading’ (‘calculating’). That is, we could only look
for the neural events that take place during the exercise of an ability—i.e., we
can only speak of mapping neural events onto behavioural acts—if we have
independent criteria for saying ‘Now he is φing’. The mistake here is that
of supposing that neurophysiological models of ‘internal mechanisms’ (based
on whatever paradigm: be it that of dynamics, physiological, computational,
connectionist) are essential to the ‘analysis’ of a psychological concept.
The standard mechanist response to this argument is to query: Can we
not at least assume that it is possible to map what an agent does when he
is reading something onto his neural processes? So much is involved in this
seemingly innocent question. Every time I read the Investigations I understand
something different: is the same thing happening in my brain every time I read
the book, or something different? And what about everyone else’s brain? Is it
conceivable that different agents might experience different neural processes
while reading the Investigations? If you answer yes, then does that mean that
a concept must be defined according to the brain that is using it? And if you
say no—as Turing does—then what is the source of the ‘must’ driving your
conceptual analysis? But most important of all, even if everyone did experience
the same neural processes when reading, say, §571 of the Investigations, would
this constitute an analysis of what reading that passage consists in, or just a
correlation?
Suppose a Turing Machine (or an ‘android’) were indeed able to ‘pass’
the Turing Test: this would only be the case on the basis of its satisfying the
criteria that govern the application of ‘reading’ or ‘calculating’—as Turing
himself, and countless mechanists following in his footsteps, have endlessly
insisted. The machine’s internal operations would be completely irrelevant to
this issue: could neither undermine nor corroborate such a judgment. But in
that case, why should we suppose that these same operations could serve as a
paradigm of what reading or calculating consist in? That is, how does Turing
pass from his mechanist to his psychological thesis?
The answer lies in the use which he makes of the continuum picture: i.e.,
in the presupposition that learning consists in the formation of synaptic con-
nections. For Turing could then assume that the machine’s ‘behaviour’ only
satisfies our criteria for saying that it is ‘calculating’ because its internal oper-
ations are isomorphic with those guiding the human computer when he passes
the Turing Test. Thus, given that the human computer’s behaviour is the end
result of mechanical processes, and that any machine can be simulated by a
Turing Machine, it follows that mechanist and psychological theses are inter-
nally related: that machines can be said to think precisely because thinkers
compute. But then, these very ‘processes’ said to be guiding the human sub-
324 Stuart Shanker
ject are themselves the product of the assumption driving this argument: viz.,
that the criteria governing the use of ‘read’ and ‘calculate’ can be trans-
formed into evidence of (embodied) effective procedures (which as such must
be isomorphic with those guiding the machine).
The picture operating here is that of reading or calculating as consisting
in the mental rule-governed transformation of input data. The connections
forged in teaching are these ‘embodied’ rules. In §163 Wittgenstein argues
that this notion of ‘rule-governed’ is empty. These are ‘hidden’ rules: rules
that are inaccessible to both observer and agent. Wittgenstein’s point is not
simply that any number of possible rules could be formulated to satisfy the
required transformation: more importantly, it is that this indeterminacy can
be stretched to a point where it is impossible to distinguish between rule-
governed and random behaviour. His intention is not to force us into the
sceptical conclusion that ‘reading’ or ‘calculating’ are not rule-governed pro-
cedures; rather, it is to emphasize that these rules, qua rules, must be public.
Hence the question raised by the passage is: what misleads us into postulat-
ing these ‘hidden rules’. And the answer is: the epistemological framework
underpinning Turing’s Thesis. The premise that there is a gap between input
and action in the exercise of an ability which must be bridged by a series of
internal operations.
§164 brings us to the heart of this issue. The passage ties in to the family-
resemblance argument at §§65ff. Here the mechanist is accused of looking
for the ‘hidden essence’ of ‘deriving’ or ‘reading’. But the ‘essence’ of these
concepts is that there is no ‘hidden essence’. Hence the mechanist is guilty of
trying to define what can only be explained. (Bear in mind that the point of
these discussions is to clarify the concept of understanding, which is a fortiori
a family-resemblance concept.) Here is the source of Davis’ claim that Turing
succeeded in presenting ‘a cogent and complete logical analysis of the notion
of “computation”’. I.e., the idea that Turing succeeded in defining the essence
of computation.
Instead of asking ‘What does computation consist in?’, we might begin
our investigation of the nature of computation by considering ‘What justifies
our use of “calculating” in such-and-such contexts?’. Waismann has the latter
question in mind when he asks in Principles of Linguistic Philosophy: ‘What,
then, is the difference between a causal connection and a transition in a cal-
culus? What is the difference between a calculation formed by a machine, and
that made by a person? Or else between a musical box playing a tune and a
person playing it?’ The answer lies in a statement of the categorial distinction
between reasons and causes vis-à-vis the use of ‘calculation’ or ‘reading’:
the playing or calculation of a person can be justified by rules which
he gives us when asked; not so the achievements of a machine, where
the question ‘Why do these keys spring out?’ can only be answered
by describing the mechanism, that is by describing a causal nexus. On
the other hand, if we ask the calculator how he comes to his results, he
will explain to us what kind of calculation he is doing and then adduce

certain laws of arithmetic. He will not reply by describing the mode
of action of a hidden machine, say, a machine in his brain (Waismann
1965: 122).
The way that Waismann has phrased this argument invites the mechanist
response that a machine could be programmed to perform exactly the same
linguistic acts; but as we have just seen, such an objection is beside the point.
For what matters here are the criteria for distinguishing between justification
and causal explanation. That is, the important distinction here is between
applying the rules which logically determine whether we can describe a subject
as having read or calculated x, and explaining the causes (if there are any) of
a subject or a machine saying or writing ‘x’.
The idea that Turing succeeded in giving the first cogent and complete log-
ical analysis of computation turns, not just on his success in resolving Church’s
Thesis, but at an even more fundamental level, on the persisting strength of
Kant’s epistemological picture of the hidden rule-governed processes whereby
the mind makes sense of reality. It is precisely this epistemological picture
which Wittgenstein’s argument is designed to subvert. This is why he insists
that it is ‘necessary to look and see how we carry out [calculations] in the
practice of language; what kind of procedure in the language-game [calculat-
ing] is’ (Wittgenstein 1956: I §17). That is, to paraphrase what Waismann
says about concepts at pp. 227–8 of Principles of Linguistic Philosophy, we
need to recognize that ‘what we call “calculating” comes into existence only
by its incorporation in language; that it is recognizable not by one feature
but by a number, which, as it were, constitute its facets’. Hence, to get clear
about what calculating ‘consists in’ we need to clarify the rules governing the
use of ‘calculate’. This is the reason why Wittgenstein remarks, vis-à-vis the
clarification of psychological concepts, that ‘It is very noteworthy that what
goes on in thinking practically never interests us’ (Wittgenstein 1967: §88).
And that is exactly the area where Turing’s ‘analysis’ has had its greatest
influence.
Not only does Wittgenstein’s argument lead us to query from yet another
angle the manner in which Turing sought to analyse the concept of calcu-
lation (viz., as recursive ‘mental process’), but it also brings into question
the whole basis of Turing’s interpretation of the logical relation in which algo-
rithms stand to inferring, reasoning, calculating, and thinking. To be sure, this
barely touches the surface of the issues inspiring the information-processing
approach to thinking: in particular, the basic epistemological picture of the
mind sundered from and forced to make sense of reality. But it does reflect
some of the philosophical problems involved in interpreting the relation in
which programs stand to speaking-aloud protocols.
It also sheds light on the import and relevance of the puzzling argument in
Philosophical Grammar that begins with the question ‘if thinking consists only
in writing or speaking, why shouldn’t a machine do it?’ and concludes with
326 Stuart Shanker
the warning that ‘It is a travesty of the truth to say “Thinking is an activity
of our mind, as writing is an activity of the hand”’ (PG: 105–6). As the
foregoing makes clear, this passage bears directly on Turing’s initial ‘analysis’
of calculation and his subsequent attempt to base an ‘analysis’ of thinking on
this flawed foundation. Moreover, the very fact that Wittgenstein broached
this topic five–six years before the publication of ‘On Computable Numbers’
provides an important insight into the conceptual environment which shaped
Turing’s attitude towards the Mechanist Thesis; for the themes which most
concerned Wittgenstein in the early 1930s also served as the spring-board for
Turing’s ‘post-computational’ entry into the philosophy of psychology.
Here is the reason why Turing’s writings continue to attract such deep
philosophical interest. Turing brought to the fore the consequences of the
epistemological framework that has virtually governed psychology from its
inception. But that framework is itself the product of an archetypal picture
to which Kant, no less than Aristotle, was responding—of the relation be-
tween thought and reality. Turing’s mechanist thesis can either be seen as
the crowning achievement of that framework, or as a reductio, forcing us to
reassess these epistemological premises. Similarly, Turing’s psychological the-
sis can either be seen as one more step in mechanism’s technological advance
(one that has already been displaced), or as forcing us to reassess the conse-
quences of seeking to establish psychology on this epistemological foundation:
of trying to resolve epistemological problems psychologically. Is it any wonder
that Mechanical Intelligence makes for such stimulating and timely reading?
References
Aristotle [1938]: Selections, W. D. Ross (trans. and ed.). New York: Scribner.
Binet, Alfred [1894]: Psychologie des grands calculateurs et joueurs d’échecs.
Paris: Deuxième Partie.
Bower, Gordon H. and Ernest R. Hilgard [1981]: Theories of learning. (5th
ed.). Englewood Cliffs: Prentice-Hall.
Bruner, J. S. [1959]: ‘Inhelder and Piaget’s The growth of logical thinking’, General
Psychology 50, 363–70.
[1960]: ‘Individual and collective problems in the study of thinking’, Annals
of the New York Academy of Science 91, 22–37.
[1990]: Acts of meaning. Cambridge, Mass.: Harvard University Press.
Carpenter, B. E., and R. W. Doran, eds. A. M. Turing’s ACE report of 1946
and other papers. Cambridge, Mass.: MIT Press, 1986.
Cleveland, Alfred A. [1907]: ‘The psychology of chess and of learning to play
it’, The American Journal of Psychology XVIII.
Davis, Martin [1978]: ‘What is a computation?’ in Steen, Lynn Arthur (ed.), Math-
ematics today. New York: Springer Verlag.
De Groot, Adriann D. [1965]: Thought and choice in chess. The Hague: Mouton
Publishers.
Dennett, Daniel [1978]: ‘Why the law of effect will not go away’, in Brainstorms,
Sussex: The Harvester Press.
Detlefsen, Michael [1993]: ‘Hilbert’s formalism’, Revue internationale de philo-

sophie 47, no. 4.
Fearing, Frederick [1930]: Reflex action: A study in the history of physiological
psychology. New York: Hafner.
George, F. H. [1962]: Cognition. London, Methuen.
Hadamard, Jacques [1945]: The psychology of invention in the mathematical field.
New York: Dover.
Harvey, O. J., D. E. Hunt, and H. M. Schroder [1961]: Conceptual systems
and peronality organization. Wiley: New York.
Hodges, Andrew [1983]: Alan Turing. London: Burnett Books.
Hofstadter, Douglas [1980]: Gödel, Escher, Bach: An eternal golden braid. Lon-
don: Penguin Books.
Hull, Clark [1962]: ‘Psychology of the Scientist, VI’, Perceptual and Motor Skills
15.
Kant, Immanuel [1970]: Logic, Robert S. Hartman & W. Schwarz (trans). Indi-
anapolis: Bobbs-Merill.
Lewes, G. H. [1877]: The physical basis of mind. Boston.
Locke, J. [1690]: An essay concerning human understanding, (John Yolton, ed.).
London: Dent, 1961.
Loeb, Jacques [1912]: ‘The significance of tropisms for psychology’ in Fleming,
Donald (ed.), The mechanistic conception of life. Cambridge, Mass.: The Belknap
Press of Harvard University Press, 1964.
Lovelace, Ada Augusta [1842]: ‘Sketch of the analytical engine invented by
Charles Babbage by L. F. Menabrea’, in Babbage, Charles, On the principles and
development of the calculator. New York: Dover Publications, 1961.
Newell, Allen and Herbert A. Simon [1961]: ‘The simulation of human
thought’, in Current trends in psychological theory. Pittsburgh: University of Pitts-
burgh Press.
Pavlov, I. P. [1927]: Conditioned reflexes (B. V. Anrep, trans. and ed.). New York:
Dover.
Rignano, E. [1923]: Psychology of reasoning (W. A. Holl, trans.). London: Kegan
Paul, Trench, Trubner.
Savage-Rumbaugh, E. Sue, Jeannine Murphy, Rose A. Sevcik, Karen E.
Brakke, Shelly L. William, Duane M. Rumbaugh [1993]: ‘Language compre-
hension in ape and child’, Monographs of the Society for Research in Child Devel-
opment 58, Nos. 3–4.
Shanker, S. G. [1987]: ‘Wittgenstein versus Turing on the nature of Church’s
Thesis’, Notre Dame Journal of Formal Logic 28, 615–649.
[1992]: ‘In Search of Bruner’, Language & Communication 12, no. 1, 53–74.
[1993a]: ‘Locating Bruner’, Language & Communication 13, no. 4, 239–
263.
[1993b]: ‘Wittgenstein versus Russell on the analysis of mind’, in Irvine,
Andrew and Gary Wedeking (eds.), Bertrand Russell and the rise of analytic phi-
losophy. Toronto: University of Toronto Press.
[1994]: ‘Ape language in a new light’, Language & Communication 14,
no. 1, 59–85.
[forthcoming]: ‘The mechanist continuum’, in Shanker, S. G. (ed.), Philoso-
phy of the English speaking world in the twentieth century 1: Logic, mathematics and
328 Stuart Shanker
science. Routledge History of Philosophy. (G. H. R. Parkinson and S. G. Shanker,

eds.). London: Routledge.
Shannon, Claude E. [1956]: ‘A chess-playing machine’, in Newman, James R. (ed.),
The world of mathematics. New York: Simon and Schuster.
Simon, Herbert A. [1982]: ‘Otto Selz and information-processing psychology’, in
Frijda, Nico H. and Adriaan D. De Groot (eds.) Otto Selz: His contribution to
psychology. The Hague: Mouton Publishers.
[1983]: ‘Why should machines learn?’, in Michalski, R. S., et. al. (eds.),
Machine learning. Berlin: Springer-Verlag.
Turing, Alan [1936–37]: ‘On computable numbers, with an application to the
Entscheidungsproblem’, Proceedings of the London Mathematical Society (2) 42,
230–265.
[1946]: ‘Proposals for development in the mathematics division of an au-
tomatic computing engine (ACE)’, in Carpenter 1986. Reprinted in A. M. Turing
1992, pp. 1–86.
[1947]: ‘Lecture to the London Mathematical Society on 20 February, 1947’,
in Carpenter 1986. Reprinted in A. M. Turing 1992, pp. 87–106.
[1969]: ‘Intelligent machinery’, in B. Meltzer and D. Michie (eds.), Machine
intelligence 5, Edinburgh University Press: Edinburgh. Reprinted in A. M. Turing
1992, pp.107–128.
[1950]: ‘Computing machinery and intelligence’, Mind 59, 433–460. Re-
printed in A. M. Turing 1992, pp. 133–160.
[1951]: ‘Can digital computers think?’
[1959]: ‘Intelligent machinery: A heretical view’, in Sarah Turing 1959.
Turing, A. M. [1992]: Collected works of A. M. Turing. Vol III, Machine Intelligence
(D. C. Ince, ed.). Amsterdam: North-Holland.
Turing, Sarah [1959]: Alan Matheson Turing, Cambridge: Heffers.
Webb, Judson [1980]: Mechanism, mentalism, and metamathematics. Dordrecht:
D. Reidel Publishing Co.
Waismann, F. [1968]: The principles of linguistic philosophy (R. Harré, ed.). Lon-
don: Macmillan.
Wittgenstein, L. [1953]: Philosophical investigations (3rd ed.) (G. E. M. Ans-
combe, trans.). Oxford: Blackwell.
[1956]: Remarks on the foundations of mathematics (G. H. von Wright,
R. Rhees, and G. E. M. Anscombe, eds., G. E. M. Anscombe, trans.). Oxford:
Blackwell.
[1960]: The blue and brown books. Oxford: Blackwell.
[1967]: Zettel (G. E. M. Anscombe and G. H. von Wright, eds., G. E. M.
Anscombe, trans.). Oxford: Blackwell.
[1974]: Philosophical grammar (Rush Rhees, ed., Anthony Kenny, trans.).
Oxford: Blackwell.
12
Mathematics and Gender: Some
Cross-Cultural Observations∗
Ann Hibner Koblitz
Professor of Women’s Studies, Arizona State University
According to folklore, the eminent mathematician Hermann Weyl occasionally

declaimed: “There have been only two women in the history of mathematics.
One of them wasn’t a mathematician [Sofia Kovalevskaia], while the other
wasn’t a woman [Emmy Noether].” Most people in the mathematical com-
munity today would not venture so far as to endorse this pronouncement.
Yet an astonishing number of mathematicians, mathematics educators, and
historians of mathematics subscribe to the belief that there have been only a
handful of women in the history of mathematics and that women’s numbers in
mathematics-related fields today are uniformly low everywhere in the world.
Moreover, many people would agree with some form of the statement that
the female nature and mathematical thought are incompatible. This paper
challenges these beliefs and argues that the historical legacy and current sta-
tus of women in mathematics are complicated and often contradictory. In the
interests of brevity, I have organized what follows as a series of statements.
STATEMENT 1:
The historical legacy of women in mathematics has been mixed.
There have been women mathematicians since classical times. During at least
the second half of the 19th century as well as the whole of the 20th cen-
tury, many women have been active participants in and contributors to the
mathematical community (Grinstein & Campbell, 1987). Seven women are
generally cited in superficial overviews (Coolidge, 1951; Mozans, 1974; Osen,
1974): (1) Hypatia (370?–415? AD), who according to legend lived in Alexan-
dria, did work on conic sections, and was martyred by Christians; (2) Emilie
du Châtelet (1706–1749), courtier and philosophe, who translated Newton
into French and extensively commented on his work; (3) Maria Gaetana Ag-
nesi (1718–1799), of “witch” of Agnesi fame, who occasionally lectured at the
University of Bologna and turned down an appointment to the Academy of
∗
First published in Gila Hanna, ed., Towards Gender Equity in Mathematics
Education, Dordrecht: Kluwer, 1996, pp. 93–109.
330 Ann Hibner Koblitz
Sciences to devote herself to a religious life; (4) Sophie Germain (1776–1831),

who proved an important case of Fermat’s Last Theorem and whose work on
elasticity won her the Grand Prix of the French Academy of Sciences in 1816;
(5) Mary Fairfax Somerville (1780–1872), an English polymath and popular-
izer whose On the Connexion of Mathematics to the Physical Sciences went
through numerous editions in the 19th century; (6) Sofia Kovalevskaia (1850–
1891), the first woman to receive a doctorate in mathematics (in the modem
sense of the term) and winner of the Prix Bordin of the French Academy of
Sciences; and (7) Emmy Noether (1882–1935), one of the founders of modem
algebra.
These women are the most famous. But it would not be difficult to compose
a list of a hundred or so prominent women mathematicians from many time
periods and diverse cultures.
Women mathematicians have often been the ones to break down educa-
tional barriers and open up professional opportunities for all women. Some
examples:
• The first woman in modern times to be fully integrated into professional,

academic life at the university level in Europe was a mathematician. Sofia
Kovalevskaia joined the faculty of Stockholm University in 1884 and be-
came an ordinary (full) professor there in 1889. Kovalevskaia was also the
first woman to be elected corresponding member of the Russian Imperial
Academy of Sciences; the rules were changed to permit her membership
(Koblitz, 1993).
• The first PhD granted to a woman in any field by Columbia University
(and the first doctorate in mathematics given to any woman by a US
university) was awarded to a mathematician, Winifred Haring Edgerton,
in 1886 (Green & LaDuke, 1990).
• In 1920, Emmy Noether became the first woman in any field to obtain full
qualifications to teach in German universities (Junginger, 1993).
• The first Nigerian woman to obtain a doctorate in any field was a math-
ematician; Grace Alele Williams received her PhD from the University of
Chicago in 1963 (American Association for the Advancement of Science,
1993).
• The first woman to be awarded a full professorship in a scientific or techni-
cal field in Vietnam (where full professor is very rare as a title) was a math-
ematician, Hoang Xuan Sinh. Sinh, a former student of A. Grothendieck
and L. Schwarz, has several times been the only female professor involved
in the International Mathematics Olympiads (in her capacity as a coach
of the Vietnamese team).1
1
Much of my information on women mathematicians of Asia, Africa, and Latin
America does not come from published sources, which are generally conspicuous
by their absence. Rather, the data emerge from extensive personal interviews with
the women themselves, their male colleagues, officials in women’s organizations and
ministries, and so on.
12 Mathematics and Gender 331
Moreover, there are a couple of instances in which the attitude of male

mathematicians toward their female colleagues has been particularly impres-
sive:
• In 1896, a survey of the German professoriate was taken on the question of

whether women should be admitted to universities with the same rights as
men. Mathematicians were unanimously in favor, physicists only slightly
less so. Historians, however, were almost all opposed to the entry of women
(Koblitz, 1993).
• Carl Friedrich Gauss, whom many mathematicians consider the greatest
number theorist who ever lived, attempted to arrange for Sophie Germain
to obtain a doctorate from Göttingen University. Gauss had tremendous
admiration for what Germain had achieved in mathematics. He wrote:
But when a person of the sex which, according to our customs and
prejudices, must encounter infinitely more difficulties than men to
familiarize herself with these thorny researches, succeeds neverthe-
less in surmounting these obstacles and penetrating the most ob-
scure parts of them, then without doubt she must have the noblest
courage, quite extraordinary talents, and a superior genius. (Quoted
in Edwards, 1977, p. 61)
On the other hand, one can chronicle enough cases of blatantly discrim-
inatory conduct of male mathematicians towards their female colleagues to
remove any doubt that male mathematicians are as much a product of their
culture as any other occupational group. Some examples:
• Sophie Germain was a mathematical correspondent of Gauss, Lagrange,
and others, and, as mentioned above, her work on the elasticity of metals
won the Grand Prix of the French Academy of Sciences. Nevertheless, her
name was not on the original list of prize winners on the Eiffel Tower, even
though her work contributed to making the tower itself possible (Mozans,
1974, p. 156).
• Christine Ladd-Franklin, a student of J. J. Sylvester at Johns Hopkins
University, completed all work for the PhD, up to and including her soon-
to-be published dissertation, in 1882. The university did not award her a
degree at the time, and in fact did not do so until 44 years later, in 1926.
Ladd-Franklin’s work earned her one of the very few stars given to women
in early editions of American Men [sic] of Science (Green & LaDuke, 1990,
p. 123).
• D. E. Smith and J. Ginsburg note in their A History of Mathematics in
America Before 1900 that Mathematische Annalen published 15 articles by
U.S. mathematicians in the period 1893 to 1897. They list 14 authors, all
male. The only one they omit is the one woman – Mary Frances Winston,
a student of Felix Klein at Göttingen (Green & LaDuke, 1990, p.117).
• Emmy Noether was an editor of Mathematische Annalen in the late teens
and twenties, as the whole of the European mathematical community knew
well. She was not listed on the masthead of the journal (in contrast to
Sofia Kovalevskaia, who 30 years earlier was listed as editor of Acta Math-
ematica). Moreover, she was sometimes disrespectfully referred to by her
colleagues as “Der Noether.”
• In the course of several years of giving talks in mathematics departments
on Sofia Kovalevskaia, I have sometimes encountered mathematicians who
assume demeaning stereotypes about her. They say, for example, that she
must have been Weierstrass’s mistress, or that the French mathematicians
gave her the Prix Bordin out of gallantry, or (my personal favorite) “no, no,
no, dear, you’ve got it all wrong. Kovalevskaia was an amateur mathemati-
cian; her husband was responsible for the Cauchy-Kovalevskaia Theorem.”
(That particular piece of nastiness has several layers. There was a male
mathematician named Kovalevskii, slightly younger than Kovalevskaia and
not in her field; the Cauchy-Kovalevskaia Theorem is sometimes called
Cauchy-Kovalevskii, due to differing transliterations of her name; her hus-
band was also a well-known scientist – a paleontologist!)
Sometimes, people have a tendency to get rather smug and complacent

when reading about examples of historical discrimination against women or
about the situation in countries far away. They congratulate themselves that
they are not as stupid as the Johns Hopkins professors were when they refused
Christine Ladd-Franklin her degree, or as whoever it was who kept Emmy
Noether off the masthead of Mathematische Annalen. It is well to remember,
though, that just because certain formal prohibitions and rules barring women
have been abolished in academia today, that does not mean that discrimina-
tion against women in mathematical fields no longer exists. It behooves us to
be especially alert to contemporary methods of discrimination as well as their
historical analogs. Although discriminatory practices are in no sense intrinsic
to mathematics, nor necessarily worse in mathematics than in other fields,
nevertheless there are many ways in which women mathematicians can be the
victims of unfairness.
It seems obvious, for example, that a person who can cavalierly and igno-
rantly propagate sexist myths about historical women mathematicians (such
as Hermann Weyl’s insults or the nonsense about Kovalevskaia cited pre-
viously) will not be above concocting analogously scurrilous stories about
present or prospective female colleagues. And that, of course, could be quite
harmful. I shall return to this point in Statement 8.
STATEMENT 2:
There is immense variability in the position of women in mathematics both
historically and cross-culturally. One cannot necessarily predict what the
status of women will be even across ostensibly similar cultural and economic
settings.
The percentage of women receiving PhDs at U.S. universities, for example,

has varied by over 300 percent; I use “varied” as opposed to “increased” be-
cause the percentage has by no means increased steadily. Women begin at a

small non-zero percentage of PhDs awarded,2 and constituted 14.3 percent of
mathematics doctorates in the period 1900 to 1940. This compares satisfacto-
rily to the 14.1 percent of doctorates received by women overall in the same
period. In the 1940s through the 1960s the percentage of women doctorates in
mathematics declined severely, as it did in all fields, so that during the period
1950 to 1969 the figure was always under 6 percent. Not until the 1980s did
the percentage of US women receiving doctoral degrees climb past the earlier
figure of 14.3 percent, and in the late 1980s and early 1990s it is between 18
and 28 percent, depending on whether one counts allied fields like statistics
and computer science. No longer is this close to the percentage of women
PhD recipients overall, however. That has peaked at approximately 36 per-
cent (Almanac, 1991; Green & LaDuke, 1990; Keith, 1991; National Research
Council, 1979). One can see similar up and down trends in other countries. For
example, pre-revolutionary Russia to about 1900 had more women receiving
doctorates in mathematics than in the next 20 years. The numbers start to
rise again in the experimental 1920s (Koblitz, 1988a; Lapidus, 1978).
Although one might assume that the percentages of female mathematicians
in countries with similar cultures or with similar economic indices might be
comparable, in fact there are many obvious (and not so obvious) counter-
examples. Among the obvious are Great Britain and northern Europe in gen-
eral as opposed to the US and southern and eastern Europe. The US, France,
Italy, Portugal, Turkey, and Spain have significantly higher proportions of
female mathematicians than England, western Germany, Sweden, and so on
(Burton, 1990; Lovegrove & Segal, 1991; Ruivo, 1987; Stolte-Heiskanen, 1991).
Nor can we assume that level of material advancement necessarily correlates
positively with percentage of women in mathematics. Japan’s and Singapore’s
numbers are relatively low, while Mexico’s, China’s, Brazil’s, Cuba’s, and
Costa Rica’s are higher (Azevêdo et al., 1989; Burton, 1990; Faruqui, Hassan,
& Sandri, 1991; Graf & Gomez, 1990; Koblitz, 1988b; Ruivo, 1987).
Analogously, neighboring countries with similar cultures can also have dif-
ferent percentages of women in mathematical areas. Costa Rica’s national
university has 12.5 percent women on its mathematics faculty, including a
woman chair. Nicaragua’s national university, by contrast, has virtually no fe-
male professors in mathematics or the physical sciences (though the statistics
department at the Leon branch of the university is chaired by a woman).
In so-called developing countries women appear to be leaders in their de-
partments no less often – and frequently more often – than in the US and
Canada. In preparing this paper, I came up with a list of women who were
or had recently been heads of mathematics, statistics, or computer science
departments in India (Madras Christian College, Lucknow, Mangalore), the
2
Graduate education in the US and Canada did not really get started until the
late 1870s, so there were some women involved in the enterprise almost from the
beginning.
Philippines (University of the Philippines and De La Salle University), the

University of Costa Rica, University of Nicaragua-León, Mexico, Ivory Coast,
Peru, Hong Kong. This does not necessarily mean that the situation for women
in mathematics is better in these countries, merely that certain stereotypes
need to be treated with caution. One cannot assume that academic commu-
nities in Asia, Africa, and Latin America will be more backward on women’s
issues than those of North America and Europe. For example, the Third World
Academy of Sciences membership is 15 percent female, as compared to ap-
proximately 5 percent for the academies of science of the US and the former
USSR (Salam, 1988).
At the student level, too, the picture can be varied and can run counter
to some of our stereotypes and the images put forward in the popular press.
Women’s participation in post-secondary level programs in mathematics and
computer science in some countries (Liberia, Cuba, Indonesia, South Korea,
Kuwait, Saudi Arabia, Turkey, Albania, and Italy) approaches or even exceeds
their percentage in all programs at that level (UNESCO, 1993, Table 3-14).
Other countries (the US, Canada, and most countries of northern and western
Europe) have far lower percentages of women students in mathematics and
computer science (Chipman, Brush, & Wilson, 1985; Stolte-Heiskanen, 1991).
STATEMENT 3:
Periods of social pressure and reaction can have special consequences for
women in “non-traditional” fields like mathematics.
These effects can be either positive or negative (or mixed). In fact, sometimes
“good things” can happen for unusual, or even for the wrong reasons. Take,
for example, the situation in Mexico. Like many countries all over the world,
Mexico is experiencing economic difficulties. The national universities are in
fiscal crisis, and (disproportionately male) professors are leaving in droves for
the private universities and the business sector. At a roundtable discussion of
women mathematicians and scientists in Mexico City in 1991,3 the question
was raised of the impact of the crisis on university women, whether it might
not be a blessing in disguise that the men are abandoning the university to
women. True, the percentage of women on the mathematics faculty is rising.
But prestige and salary are falling, and many women have the feeling that the
best students are not going to be attracted to mathematics anymore.
This is a complex situation, and one that has analogs in many other coun-
tries and fields. If a specialty becomes “saturated” (that is, too many profes-
sionals to make good salaries) or unattractive for some other reason, opportu-
nities for women can increase.4 Also, a field can become stratified. That is, the
teaching of mathematics (and university teaching in general) can become more
3
The discussion was part of a week of activities commemorating the centenary
of Sofia Kovalevskaia’s death.
4
Veterinary medicine in the US, pharmacy in El Salvador, medicine in the USSR
and the Philippines are examples of this phenomenon.
or less the province of women, while the academies of sciences and research
institutes can remain largely or entirely male. One sees this in many countries
that overall have decent percentages of women in mathematics, such as Mex-
ico, India, China, Nigeria, and Costa Rica. The most prestigious positions and
places become or remain largely male preserves.
STATEMENT 4:
One must be very careful about making generalizations from historical and
cross-cultural comparisons.
Mindless use of one or another indicator to make a sweeping generalization
about women’s status in mathematics can give misleading results. It is not
possible to use the same indicators to determine the situation in every coun-
try. The significant statistic might be the percentage of women teaching at
the university level. But it might also be the proportion of women at research
institutes and academies of sciences (and at what level), or the percentage
of women who publish (or who publish in foreign as opposed to domestic
journals), or the proportion of women who go abroad for conferences, post-
graduate study, and so on, or the percentage of women awarded grants by
national and international funding agencies. Indices can have different mean-
ings in different countries, and the prestige of various positions and honors
can vary considerably. This is not to say that it is unimportant that women
constitute a large percentage of professors on a mathematics faculty in a cer-
tain country. But this measure might not be the unique indicator of women’s
success or status in the mathematical world.
STATEMENT 5:
Despite the problems and societal obstacles, women can fare reasonably well
in mathematics for a number of reasons.
There are, after all, some recognized and relatively objective standards in
the mathematical community. Women might have to be better than their
male counterparts to be judged equal, but the standard is not impossible to
achieve. Sofia Kovalevskaia, for example, offered three works to Göttingen in
fulfillment of her degree requirements. She and her adviser Karl Weierstrass
reasoned that as the first woman applying for the doctorate, her case would
have to be especially strong. Three were sufficient, however.
Moreover, there is perhaps not so great an “Old Boy Network” in mathe-
matics as in other fields and more of a tradition of tolerance and eccentricity.
There is also a certain pride in traditions of internationalism – in not wanting
to allow political, ethnic, or ideological considerations to interfere with one’s
mathematical judgments. These factors have worked in women’s favor in a
large variety of contexts.
Women can also fare well in mathematics because of a relatively small
number of men who have served as excellent mentors, in a couple of cases
despite otherwise not terribly progressive politics or overall commitment to
women’s rights. Some random examples:
• Karl Weierstrass performed valuable mentoring functions not only for Sofia
Kovalevskaia but also for her friend Iulia Lermontova. It was through
Weierstrass’s intervention with university officials at Göttingen that Ler-
montova was able to become the first woman in the world to receive her
doctorate in chemistry (Koblitz, 1993).
• Felix Klein supported Mary Winston, Grace Chisholm, and Margaret
Maitby in their admission to graduate studies in mathematics at Göttingen
University in 1893. He later said that the women were fully comparable to
his male students.
• Of the seven PhDs in mathematics awarded to women by Johns Hopkins
University before 1940, five were students of one professor (Morley). Eleven
of the 13 women who completed PhDs at Catholic University were students
of Landry. The University of Chicago granted 46 of the 229 doctorates
given to women in the US through 1939; of these, 30 were students of
either Leonard Eugene Dickson (18 of his 67 students were women) or
Gilbert Ames Bliss (12 of his 52 students were women) (Green & LaDuke,
1990).
• Lee Lorch, now emeritus at York University, performed mentoring func-
tions for many women – including several of the few black women who
later received PhDs in mathematics – during the time he taught at Fisk
University.
STATEMENT 6:
Cross-cultural disparities in female educational and employment patterns in
mathematics and computer science raise serious questions about theories
that claim innate gender differences in mathematical ability.
We need to ask ourselves whether there are other explanations: faulty test
design, socio-cultural factors, and so on. Theories must be examined critically
and tested historically and cross-culturally. Several generalizations have be-
come quite popular in late-20th century US society, yet they rest on dubious
foundation and would fail under historical and cross-cultural scrutiny.
For at least three decades the received wisdom – and the line being pushed
by “objective scientists” like Camilla Benbow and Julian Stanley – was that
girls are better at verbal tests and boys at mathematical ones (Benbow &
Stanley, 1980). Now, though, that picture is breaking down in several im-
portant ways. The so-called gender gap on US standardized tests narrowed
considerably during the 1980s, to the point that specialists at the Educa-
tional Testing Services (ETS) in Princeton are now saying the differences are
not statistically significant. Also, several recent studies have shown that even
in mixed groups where males had performed noticeably better than females
on mathematics Scholastic Aptitude Tests (SATs), on other tests, including
ETS’s Mathematics Achievement Test itself, there were no significant gender
differences. (For a reasonably current review of the literature see Kenschaft,
199lb.)
More importantly, the supposed gender differences are constant neither

across ethnic groups within the US and Canada nor across cultures. A 1987
study noted that even within the mathematics SAT, the gender gap varies con-
siderably. It is largest for Hispanics and smallest for Afro-Americans (Ruskai,
1991). Moreover, Gila Hanna has pointed out in her comparative studies that
differences between countries are much larger than those between boys and
girls, and that the gender gap is larger in countries with low scores (like the
US) than in those with high scores (like Hungary and Japan). On the geome-
try test, for example, US males scored 39.7 while US females scored 37.9 – a
mere 1.8 points. Meanwhile, the advantage over the US of Hungary and Japan
was relatively massive, since all subgroups in both countries scored between
55 and 60 points (Hanna, 1989).
Some countries (Thailand and South Korea, for example) do not ex-
hibit any statistically significant differences between male and female perfor-
mance on mathematics achievement tests (Hanna, 1989; Hanna, Kündiger, &
Larouche, 1990; Kwon, unpublished paper). As Hanna and her collaborators
note, the one clear conclusion that emerges from all the statistics is that one
must doubt the biological explanations of male/female difference in mathe-
matical ability, since it is “very unlikely” to vary between countries (Hanna,
Kündiger, & Larouche, 1990, p. 96).
Other studies have also pointed to the culture-bound nature of many of
our notions of gender difference. For example, spatial ability tests given to
Native American children in Alaska and to central African children show either
no difference or one favoring the females (Fausto-Sterling, 1985; Kenschaft,
1991a; Koblitz, 1987; Ruskai, 1991). Needless to say, these tests are not the
ones reported with fanfare in the New York Times.
Benbow’s and Stanley’s studies have immense cultural problems. Gender
differences are not consistent across ethnic groups, and even girls who did
worse than boys on the Benbow and Stanley test outperform the boys in
school. Moreover, before they test, Stanley’s research centre apparently sends
parents a pamphlet noting that boys do better than girls on the test. Such
a message to parents reveals the researchers’ blindness to issues of bias-free
experimental design (Jackson, 1990; Ruskai, 1991).
Also dubious are generalizations about a fundamental physiological dif-
ference which affects the way men and women reason about mathematics
or their interests in mathematics. The latest manifestation of this type of
pseudo-scientific study is the brain lateralization “research.” This has a long,
inglorious history dating back to early 19th-century phrenology (Alper, 1985;
Bleier, 1984; Dumdell, 1991; Fausto-Sterling, 1985).
STATEMENT 7:
There are fundamental problems with anything that smacks of essentialism;
this includes so-called feminist gender-and-science theory.
“Essentialism” encompasses any theory that attributes gender differences to
biological, genetic, psychosocial, or other immutable factors. One problem
with such theories is that, as indicated above, women’s position in mathe-

matics and the natural sciences is constant neither across cultures nor across
time periods. What in one country or time might be considered unfeminine
might in another country or time not be so. Specifically with regard to the
mathematical sciences, there are several countries, including the Philippines,
Turkey, Kuwait, and Mexico, for example, in which women constitute a rather
high percentage of mathematics-related professions, especially when compared
with the numbers in northern Europe, Canada, or the United States (Faruqui,
Hassan, & Sandri, 1991; Lovegrove & Segal, 1991; Stolte-Heiskanen, 1991;
UNESCO, 1993; United Nations Statistical Division, 1992). Yet most feminist
theorizing about gender and mathematics assumes that women’s participation
in these fields is uniformly low.
The theorists also assume that Victorian-era bourgeois stereotypes con-
cerning femininity and gender polarities are uniform across all cultures,
classes, and historical periods. This is far from being the case, even within
western Europe. In Italy, for instance, the stereotype is different from that in
the US or Sweden or the UK. Women are purported to be “natural” theoreti-
cians (hence the relatively large number of Italian women mathematicians and
computer scientists), while men are supposed to be more practical by nature
(and thus become engineers rather than theoretical scientists).
Moreover, the position of women in mathematics and the sciences can
change quite rapidly for the better (or worse). The changes are far too rapid
to be explainable by biological theories of difference or by psychosocial theories
such as Nancy Chodorow’s (1978).5 Gender and science theorists have the un-
fortunate tendency to make generalizations despite clear historical and cross-
cultural counterexamples. For example, it is simply not true that women’s
status in the sciences has remained unchanged since the Scientific Revolution,
though that is a claim often made by theorists like Evelyn Fox Keller (1985),
Sandra Harding (1986, 1991), and their imitators.
In like manner, I am disturbed by certain aspects of work by Belenky and
her collaborators (1986), Gilligan (1982), and others on women’s purportedly
different ways of knowing. Belenky and her collaborators, for example, assert
that the gender differences they describe are independent of culture, ethnicity,
and class. This is highly questionable. A colleague of mine at Hartwick College,
Katherine O’Donnell, has conducted years of field research with poor migrant
farm women in New York state. At the beginning she had the expectation of
5
In general terms, object-relations theory says that a girl infant never has to
disidentify herself from her mother. Therefore, she never sees the world as alienated
from herself in the same way a boy infant does. A boy baby, on the other hand,
realizes very early that he is not the same gender as his mother. He has to distance
himself from her and thus begins to objectify the world. Because object-relations
theory attributes gender differences in intellectual outlook to an immutable mech-
anism of early childhood, it is almost indistinguishable from a genetic or biological
theory. Moreover, there is no way the theory can account for differences between
individuals or for change over time or across cultures.
finding support for the writings of Belenky and Gilligan (O’Donnell, unpub-
lished paper a) Now, however, she is firmly convinced that, on the contrary,
the interactions of gender, culture, race, and class are far too complex to be
encompassed by a simplistic gender-polarity theory (O’Donnell, unpublished
paper b).
To bring this discussion closer to mathematics, let us take the work of
Sherry Turkle (1984) on children and computers. Under the influence of
object-relations theory and gender-and-science theory, Turkle has created the
concepts of “soft mastery” and “hard mastery” to categorize her analysis of
the use of computers by school-age boys and girls. For Turkle, “hard mastery
is the imposition of will over the machine through the implementation of a
plan . . . the hard masters tend to see the world as something to be brought
under control.” Soft mastery, on the other hand, is more interactive – “the
soft masters are more likely to see the world as something they need to ac-
commodate to, something beyond their direct control.” Though Turkle gives
examples of soft masters of both sexes, she says that girls tend to be soft
masters, while hard masters are “overwhelmingly male.”
Large parts of the theory sound quite plausible at first. Certainly it is true
that in most societies males and females are socialized differently, with female
socialization tending more towards valuing qualities like accommodation and
interaction, and male socialization tending more towards control. Upon re-
flection, however, some people have raised questions about the validity of the
theory and voiced concerns about its social implications.
It has been pointed out (by Beth Ruskai [1990], for example) that the
theory contains certain assumptions about the nature of computer science
that are rather far off the mark. What Turkle dichotomizes as “hard” and
“soft” mastery might be better characterized as two inextricably interwoven
parts of the creative scientific process. The attempt to label people as being
either hard or soft, therefore, misses a crucial point of what it is to do science.
Turkle’s theory makes no allowances for historical and cross-cultural vari-
ation. It ignores the evidence that women’s participation in the sciences in-
cluding computer science – has varied widely from one decade to the next and
even between neighboring countries (Koblitz, 1991; Lovegrove & Segal, 1991).
Moreover, stereotypes regarding women’s innate capacities and their relation
to mathematics and computer science vary from culture to culture.
There is also reason to be uncomfortable with the way Turkle’s theory
dovetails with current western European and North American stereotypes
about women’s intrinsic nature. In practice, these stereotypes can contribute
to discrimination against women in the workplace, and to the segregation of
women in so-called pink-collar ghettoes like data processing.
Sherry Turkle is fairly skillful in inserting caveats in her generalizations and
in reminding us that the picture is complex. (Harding, Belenky, and Gilligan
are not nearly so careful.) Unfortunately, however, the caveats and reserva-
tions rarely make it into popular accounts of the work. What gets picked up
in the media is the rather simplistic idea that girls cannot be attracted into
computer-related fields unless the machine can be portrayed as artistic, rela-

tional, and “soft.” We do not work in a vacuum. If the media can distort a
theory, they will.6
In the US, one of the most common manifestations of sexism and racism
in the classroom is a refusal to intellectually challenge girls and members of
minority groups. They are condescended to and patronized and do not receive
adequate exposure to the more rigorous, thought-provoking, and elegant as-
pects of mathematics. Their understanding thus rarely attains the level of the
systematic and the structural; they seldom arrive at the stage where they can
see much point in doing mathematics. Unfortunately, this phenomenon can
be exacerbated by overzealous followers of the ideas of Turkle, Gilligan, and
Belenky.
STATEMENT 8:
Any discrimination against younger faculty automatically falls
disproportionately on women because of the demographics of the profession.
The injustice is increased because of certain usually unconscious but quite
pervasive attitudes about women’s “natural” roles.
In the US, for example, women are often assigned heavier teaching loads than
men and more courses at the introductory level. This is a triple blow: more
work is devoted to teaching as opposed to research; teaching evaluations are
automatically worse because of the nature of the course (introductory courses
virtually always receive lower evaluations than upper division courses); and
any deviation from stereotypically feminine behavior (such as attempting to
enforce high academic standards) is met with displeasure (and low ratings) by
students. Numerous studies indicate differential treatment of women faculty
on evaluations. For example, students expect to be “nurtured” by women and
punish them for deviations from the ideal “feminine” standard (N. Koblitz,
1990).
Women faculty are caught in a bind. Either they devote tremendous time
to teaching, in which case their research suffers, or they devote as much time
to their research as their male colleagues do, remaining aloof from students,
in which case they are penalized more heavily than men on evaluations.
Moreover, the evaluatory process for granting tenure is in essence a black-
balling system.7 Even if 80 percent of the department have not a sexist bone
6
We had a distressing illustration of exactly this point in Swedish newspaper
coverage of the 1993 ICMI Study Conference “Gender and Mathematics Education”
in Höör, Sweden. The headline of the 9 October 1993 Dagens Nyheter story was
“Mathematicians disagree whether biology makes a difference,” and ICMI Presi-
dent Miguel de Guzmán was misquoted as saying that gender differences manifest
themselves from day one in the classroom!
7
I realize that the concept of tenure (the right to relative job security, granted
after a probationary period of some years) is not institutionalized in all countries.
But I believe that the following discussion is applicable to most other kinds of hiring
and promotion processes as well.
in their bodies, the opposition of a relatively few curmudgeons can sink a

woman’s chances in a variety of ways. It is rather like the old anti-Semitic
blackballing system for admission to US country clubs – the system persisted
for so long because all that was needed was one person in opposition.
There is an analogous blackballing system in place for hiring and tenure
in academic departments today. A couple of people can skew the process and,
in fact, wreck it. Consider the following situations:
• Say a woman has children and resumes mathematical activity after a cou-
ple of years hiatus. How does one interpret this? One could talk about the
fact of her return to active research as indicating resolve and high mathe-
matical ability and dedication. But one typically hears the diehard sexists
referring instead to the “unfortunate gap in her publication record.”
• Women are often held up to far higher standards than men on the pretext
of not lowering standards. For a woman’s appointment, there cannot be
the least shadow of a doubt, while men are often given the benefit of the
doubt.8 We all know at least one senior professor who thinks nothing of
spreading stories to the effect that it would be lowering the department’s
standards to hire a woman, even when the woman being considered is a
far better researcher than he is himself.
In the US a certain amount of sound and fury signifying little or nothing

has sprung up around the issue of Affirmative Action. Departments sometimes
have women come on campus for interviews, under pressure from the admin-
istration, only to have the appointment sabotaged by a couple of diehard
sexists working assiduously to undermine the process. There is a lot of rumor-
mongering that goes on about Affirmative Action. Traditionalists routinely
exaggerate the success of the program, telling female graduate students things
like, “oh, you’ll have no problem getting a job, you’re in fashion these days,”
or jokingly advising their male students to wear a skirt to the job interview.
This kind of childish and disingenuous behavior in itself creates a bad at-
mosphere for women. It conveys the impression that colleagues do not have
confidence in the women they have hired and implies that any women in the
department are there on sufferance.
Conclusion
My goal here clearly has not been the presentation of a definitive account
or a polished treatise. It was my intention to throw out food for thought,
8
This kind of discrimination includes behavior such as the infamous cases of anal-
ogous curricula vitae being sent to chairs with male and female names attached – the
chairs routinely recommended the women for lower positions than the men! Bernice
Sandler has documented many examples of this sort of (unconscious) prejudice in
her “Chilly Climate” series (Sandler, n.d.).
to illustrate some of the curiosities and ironies inherent in women’s position

in mathematics historically and across cultures. The interactions of gender
and culture are never simple or straightforward. The history of women in
mathematics has not been some Whiggish triumphal passage from darkness
into light, but neither has it been a chronicle exclusively of discrimination and
marginalization. The history and present status of women in mathematics
are complicated, and often the picture has elements of contradiction. One
conclusion seems obvious, however. The complex and multifaceted interactions
of gender and mathematics can be understood only if one takes into account
historical and cross-cultural perspectives.
References
Almanac. (1991). The chronicle of higher education, XXXVIII, No. 1.

Alper, J. S. (1985). Sex differences in brain asymmetry: A critical analysis.
Feminist Studies, 11(1), 7-37.
American Association for the Advancement of Science [AAAS]. (1993). Science
in Africa; Women leading from strength. Washington, DC: Author.
Azevêdo, E. S., et al. (1989). A mulher cientista no Brasil. Dados atuais sobre
sua presença e contribuiçao. Ciência e cultura, 41(3), 275-283.
Belenky, M., Clinchy, B., Goldberger, N., & Tarule, J. (1986). Women’s ways
of knowing: The development of self, voice and mind. New York: Basic Books.
Benbow, C. P., & Stanley, J. C, (1980). Sex differences in mathematical ability:

Fact or artifact? Science, 210, 4475, 1262-1264.
Bleier, R. (1984). Science and gender. New York: Pergamon.
Burton, L. (Ed.). (1990). Gender and mathematics: An international perspec-
tive. London: Cassell.
Chipman, S. F., Brush, L. R., & Wilson, D. M. (Eds.). (1985). Women and
mathematics: Balancing the equation. Hillsdale, NJ: Lawrence Erlbaum Asso-
ciates.
Chodorow, N. (1978). The reproduction of mothering: Psychoanalysis and the
sociology of gender. Berkeley; University of California Press.
Coolidge, J. L. (1951). Six female mathematicians. Scripta Mathematica, 17,
20-31.
Durndell, A. (1991). Paradox and practice: Gender in computing and engi-
neering in Eastern Europe. In G. Lovegrove & B. Segal (Eds.), Women into
computing. London: Springer-Verlag.
Edwards, H. M. (1977). Fermat’s last theorem: A genetic introduction to al-

gebraic number theory. New York: Springer-Verlag.
Faruqui, A. M., Hassan, M.H.A., & Sandri, G. (Eds.). (1991) The role of
women in the development of science and technology in the third world. Sin-
gapore: World Scientific Publishing.
Fausto-Sterling, A. (1985). Myths of gender: Biological theories about women
and men. New York: Basic Books.
Gilligan, C. (1982). In a different voice: Psychological theory and women’s
development. Cambridge, MA: Harvard University Press.
Graf, H. B., & Gomez, H. G. (1990). Acerca de las cientfficas en Ia UNAM.
Unpublished paper delivered at the National Autonomous University of Mex-
ico. (I am indebted to the authors for giving me a copy of this paper.)
Green, J., & LaDuke, 1. (1990). Contributors to American mathematics. In
G. Kass-Simon & P. Farnes (Eds.), Women of science: Righting the record.
Bloomington: Indiana University Press.
Grinstein, L. S., & Campbell, P. J. (1987). Women of mathematics: A biobib-
liographical sourcebook. New York: Greenwood Press.
Hanna, G. (1989). Mathematics achievement of girls and boys in grade eight:
Results from twenty countries. Educational Studies in Mathematics, 20, 225-
232.
Hanna, G., Kündiger, E., & Larouche, C. (1990). Mathematical achievement of
grade 12 girls in fifteen countries. In L. Burton (Ed.), Gender and mathematics
(pp. 87-97). London: Cassell.
Harding, S. (1986). The science question in feminism. Ithaca, NY: Cornell
University Press.
Harding, S. (1991). Whose science? Whose knowledge? Thinking from women’s
lives. Ithaca, NY: Cornell University Press.
Jackson, M. V. (1990). SATs ratify white male privilege. Association for
Women in Mathematics Newsletter, 20(5), 9-10.
Junginger, G. (1993). A woman [sic] career in veterinary medicine. Lecture
given at the XIXth International Congress of History of Science, August 1993.
Keith, S. Z. (1991). A statistical overview of American women doctorates,

1988–1989. in P. C. Kenschaft (Ed.), Winning women into mathematics (pp.
59-60). Mathematics Association of America.
Keller, E. F. (1985). Reflections on gender and science. New Haven: Yale
University Press.
Kenschaft, P. C. (1991a). Fifty-five cultural reasons why too few women win
at mathematics. In P. C. Kenschaft (Ed.), Winning women into mathematics
(pp. 11-18). Mathematics Association of America.
Kenschaft, P. C. (Ed.). (1991b). Winning women into mathematics. Mathe-
matics Association of America.
Koblitz, A. H. (1987). A historian looks at gender and science. International
Journal of Science Education, 9(3), 399-407.
Koblitz, A. H. (1988a). Science, women, and the Russian intelligentsia. Isis,
79, 208-226.
Koblitz, A. H. (Ed.). (1988b). La mujer en la ciencia, la tecnologı́a y la medi-
cina. Seattle: Kovalevskaia Fund.
Koblitz, A. H. (1991). Women in computer science: The ‘soft mastery’ con-
troversy. Kovalevskaia fund newsletter, Vl(2), 5-6.
Koblitz, A. H. (1993). A convergence of lives. Sofia Kovalevskaia: Scientist,
writer, revolutionary (2nd. ed). New Brunswick, NJ: Rutgers University Press.
Koblitz, N. (1990). Are student ratings unfair to women? Association for

Women in Mathematics Newsletter, 20(5), 17-19.
Kwon, O. (unpublished paper). U.S.-Korea cross-national studies on college
entrance exams.
Lapidus, G. W. (1978), Women in Soviet society. Berkeley: University of Cal-
ifornia Press.
Lovegrove, G., & Segal, B. (Eds.). (1991) Women into computing. London:
Springer-Verlag.
Mozans, H. J. (1974). Woman in science. Cambridge, MA: MIT Press. [Orig-
inal publication date, 1913.]
National Research Council, Commission on Human Resources. (1979). Climb-
ing the academic ladder: Doctoral women scientists in academe. Washington,
D.C.: National Academy of Sciences.
O’Donnell, K. (unpublished paper a). A class act. Phoebe.
O’Donnell, K. (unpublished paper b). Got nowhere to go and no way to get
there: Migrant women’s struggle for dignity.
Osen, L. M. (1974). Women in mathematics. Cambridge, MA: MIT Press.
Ruivo, B. (1987). The intellectual labour market in developed and developing
countries: Women’s representation in scientific research. International Journal
of Science Education, 9(3), 385-391.
Ruskai, M. B. (1990). Why women are discouraged from studying science. The
Scientist, 4(5), 17, 19.
Ruskai, M. B. (1991). Are there innate cognitive gender differences? Some
comments on the evidence in response to a letter from M. Levin. American
Journal of Physics, 59(1), 11-14.
Salam, A. (1988). Speech given at the Third World Academy of Sciences,
Trieste, October 3.
Sandler, B. R. (n.d.) The campus climate revisited: Chilly for women faculty,
administrators, and graduate students. Project on the status and education of
women. Association of American Colleges.
Stolte-Heiskanen, V. (Ed.) (1991). Women in science: Token women or gender
equality? Oxford: Berg.
Turkle, S. (1984). The second self: Computers and the human spirit. New
York: Simon and Schuster.
UNESCO. (1993). 1992 Statistical yearbook. New York: United Nations.
United Nations Statistical Office. (1992). Statistical yearbook 1990-1991. New
York: United Nations.
Index
Abel, Niels, 194 Aristarchus, 26

absolute algebra, 212 Aristotle, 24, 33, 207, 208, 298, 326
abstract algebra, 212 De Generatione Animalium, 298
abstract algebras, 12 Metaphysics, 33
Acta eruditorum, 90 Aristoxeneans, 39
Acta Mathematica, 229, 233, 332 arithmetic books, 150–152
Adams, Daniel, 148 Arnold, Matthew, 59
The Scholar’s Arithmetic, 148 Arnold, Vladimir Igorovic, 268
Adams, George, 52 artificial intelligence, 297–326
affine transformations, 285 Artin, Emil, 250, 261
Affirmative Action, 341 astrology, 23, 35
Agnesi, Maria Gaetana, 329 Astronomical Journal, 193
Aleksandrov, Pavel Sergeevich, 268 astronomy, 23–43, 52–55, 185, 187
Problemy Gil’berta, 268 atomism, 25
algebraic forms, 253 Autolycus, 26
algebraic geometry, 196, 200, 251 automorphic functions, 200, 289
algebraic number theory, 253 axiom of choice, 278
American Association for the Advance- Bédard Diacre, Thomas, 155
ment of Science, 189 Babbage, Charles, 66, 72, 73, 209, 298,
American Journal of Mathematics, 197, 305
199 Bache, Alexander Dallas, 188
American Mathematical Society, backstaff, 48–52
195–197, 199, 243 Bacon, Francis, 230
analysis, 117–131, 196, 253 Baillargé, Charles, 167
analysis, real-variable, 16 Nouveau traité de géométrie et
Analytical Society, 209 de trigonométrie rectiligne et
André, Frédéric, 169 sphérique, 167
Angus, R. W., 173 Tableau stéréométrique, 168
Annals of Mathematics, 199 Bain, Alexander, 207, 214–216
anti-psychologism, 209 Logic, 215
Arbogast, Antoine, 126 Baire functions, 123
Archambault, Urgel-Eugène, 169 Balète, Emile, 169
Archimedes, 26, 65, 67 Barrow, Isaac
348 Index
Lectiones geometricae, 117 Borel functions, 123

Bashmakova, I. G., 18–19 Boucharlat, Jean Louis, 166
Beatty, Samuel, 171 Bougainville, Louis-Antoine de, 145
Beaupré, Victor-Elzéar, 170, 175, 176 Traité du calcul intégral, 145
Bédard Diacre, Thomas, 153 boundary value problems, 289
behaviourism, 300, 309, 314 Bourbon, Jean, 143
Belenky, M., 338–340 Boutet, Martin, 143
Benbow, Camilla, 336, 337 Bouthillier, Jean-Antoine, 150
Beneke, Friedrich Eduard, 207 Traité d’arithmétique, 151
Bentham, George, 206 Bowditch, Nathaniel, 188
An Outline of a New System of Logic, Brachistochrone Problem, 273, 274
206 Bradley, James, 51
Berkeley, Bishop George, 58, 61, 62, 70 Brahe, Tycho, 50
Bernard, Claude, 299 Brauer, Richard, 177
Bernoulli, Daniel, 61, 85, 90, 93, 95, 123 Bressoud, David, 16
Bernoulli, Jacob, 85, 273 Briand, Mgr. Jean-Olivier, 152
Bernoulli, James, 118 Briggs, Henry, 47
Bernoulli, Johann, 59, 63, 72, 85, 87, British Association for the Advancement
90, 98, 120, 128, 273 of Science, 210
Bernoulli, Johann II, 87 Brouwer, Luitzen Egbert, 279
Bernoulli, Nicolas, 90, 93 Browder, Felix Earl, 267
Bibaud, Michel, 151 Brucker, Margarete, 85
L’arithmétique en quatre parties, 151 Bruner, J. S., 311
bilinear forms, 13
Brunotto, Lorenzo, 176
Binet, Alfred, 302
Buchanan, Daniel, 173
binomial series, 166
Bulletin of the American Mathematical
Birkhoff, George D., 196
Society, 199
Biron, Ernst Johann Reichsgraf von, 97
Burckhardt, Johannes, 85
Bironovshchina, 95
Burke, Edmund, 54, 153, 155
Black, Joseph, 61
Busemann, Herbert, 261
Blackborrow, Peter, 53
Butterfield, Herbert, 74
The Longitude Not Found, 53
Bliss, Gilbert Ames, 336
Blumenthal, Otto, 247 Cajori, Florian, 58, 187
Blumentrost, Laurentius, 97 The Teaching and History of Math-
Bôcher, Maxime, 194 ematics in the United States,
Boltzmann, Ludwig, 279 187
Bolza, Oskar, 195 calculus, 57–75, 154, 166, 170, 209, 221
Bolzano-Weierstrass theorem, 225 calculus of variations, 72, 128, 196, 200,
Bombieri, Enrico, 268 258, 266, 283, 289
Bond, Henry, 45, 52 Cambridge Philosophical Society, 207
The Longitude Found, 53 Canada Educational Journal, 161
Bonnécamps, Father Joseph-Pierre de, Canadian Educational Monthly, 157
144 Canadian Journal, 159
Boole, George, 203, 204, 206, 207, 210, Canadian Journal of Mathematics, 141
213, 215, 216 Canadian Society for History and
Laws of Thought, 207, 215 Philosophy of Mathematics, VII,
Mathematical Analysis of Logic, 203, 3–6
206, 210 Cantor, Georg, 12, 221–239
Index 349
Beiträge zur Begründung der Condorcet, Marquis Marie-Jean-

transfiniten Mengenlehre, 236 Antoine-Nicolas, 114
Grundlagen einer allgemeinen conic sections, 153, 154
Mannigfaltigkeitslehre, 228, 230, Continuum Hypothesis, 228, 231, 235,
232, 233 278
Cantor, Moritz, 74, 259 Cook, James, 108
Carathéodory, Constantin, 128, 243 Cote, Roger, 112
Vorlesungen über reelle Funktionen, Courant, Richard, 194, 261
128 Cox, John, 172
Carnap, Rudolf, 279 Crelle’s Journal, 225, 227, 228
Carroll, Lewis, see Dodgson, Charles L. Croft, H. T., 267
Cartan, Elie, 243 Unsolved Problems in Geometry, 267
cartography, 23 cross-staff, 44, 48
Casault, Abbé J. L., 154
Catherine the Great, 82, 94, 107, 112 d’Alembert, Jean, 62, 66, 67, 69, 70, 73,
Cauchy, Augustin-Louis, 13, 16–18, 58, 123, 131, 166, 282
63, 66, 70, 72, 129 Encyclopédie, 66, 70
Cours d’Analyse, 129 Traité de dynamique, 62
Cauchy-Riemann equations, 17 Davie, George Elder, 73
Cavalleri, P. Antoine, 61 Davies, Charles, 167
Cayley, Arthur, 251 Davis, Ellery William, 191, 199
celestial mechanics, 188 Davis, Martin, 317, 324
Châtelet, Emilie du, 329 de Beaune’s problem, 273
Champlain, 143 De Groot, Adriann, 302
Chandler, George, 172 Thought and Choice in Chess, 303
Chandrasekhar, Subramanyan, 67 De Morgan, Augustus, 206, 207, 213,
Ellipsoidal Figures of Equilibrium, 67 215
Chasles, Michel, 70 Dedekind, Richard, 126, 194, 225, 232,
Chauveaux, Charles, 153, 155 233
Chern, Shiing-Shen, 271 Dehn, Max, 261
Cherriman, J. B., 159, 162 Delisle, Jean-Nicolas, 145
Chessin, Alexander, 199 DeLury, A. T., 171
Chodorow, Nancy, 338 Demers, Abbé Jérôme, 154, 155
Church’s Thesis, 297, 304, 308, 325 Denkpsychologie, 300, 309
Church, Alonzo, 279 Desaguliers, 60
Clairaut, Alexis-Claude, 62, 67, 75 Desaulniers, Abbés Isaac and François,
La Figure de la Terre, 68 156
Cleveland, Alfred A., 302 Descartes, René, 10, 71, 75, 117
Coats, R. H., 173 Deshayes, Jean, 144
cognitive psychology, 298 Despiau, M., 151
cognitive science, 315 Choix d’amusements physiques
Cohen, Paul, 278 mathématiques, 151
Colborne, Sir John, 149 Deutsche Mathematiker-Vereinigung,
Cole, Benjamin, 52 234
Collège de Québec, 145–147 Dewey, Thomas, 157
colour theory, 24 The Psychology of Number, 157
complex variables, 16 Dickson, Leonard Eugene, 196, 336
Comte d’Arcy, Patrick, 105 Dieudonné, Jean, 250, 267
Comte, Auguste, 267, 276 differential equations, 13
350 Index
partial, 72 Introductio in analysin infinitorum,

differential geometry, 200 68, 110, 119, 124, 126
Dijksterhuis, E. J., 10 Methodus inveniendi, 106
Diophantine equation, 288 Opera omnia, 129
Dirichlet’s principle, 268, 289 Theoria motum lunae, 112
Dirichlet, Johann Peter Gustav, 14, Euler, Paul, 85, 87
124, 125, 252 Euler-Maclaurin summation formula,
Dodgson, Charles L., 204 63, 68–69
Symbolic Logic, 204
The Game of Logic, 204 fallibilism, 15
du Bois-Reymond, Emil, 276 Fermat’s Last Theorem, 273, 330
Dupuis, Nathan Fellowes, 159 Fermat, Pierre de, 98
Geometry of the Point, Line and Ferrier, Alan, 173
Circle in the Plane, 161 Fessenden, Arthur, 147
Junior Algebra, 160 Tables, 147
Durfee, William, 191 Fields, J. C., 141, 162, 171, 173
Dwight, H. B., 173 finite basis theorem, 249
finite linear groups, 196
École polytechnique de Montréal,
Fisher, George Egbert, 192
169–170
fluxions, 57–75
École polytechnique, 188
formal algebra, 212
École supérieure de chimie, 175
foundations of geometry, 253
Edgerton, Winifred Haring, 330
foundations of mathematics, 254
Edinburgh Review, 206
Four-Color Problem, 286
Educational Times, 193
Fourier series, 14, 16, 124
ellipsoids, 67
Fourier, Jean Baptiste, 124
elliptic integrals, 63, 69
Frères des Écoles chrétiennes, 163
elliptic partial differential equations,
Nouveau traité d’arithmétique, 163
268
Franquelin, Jean-Batiste Louis, 144
Elton, John, 51
Frederic II, 97
Ely, George, 191
Frederick II, 105
Epicurus, 25
epicyclic model, 28–32 Frederick the Great, 82, 101
Erasmus, 87 Frederick William I, 101
Erdös, Paul, 271 Frege, Gottlob, 213, 214, 216
Euclid, 8–10, 16, 19, 26, 158 Frigon, Augustin, 176
Elements, 8–11, 16, 18 function, 117–131
Optics, 36 function spaces, 120
Euclidean geometry, 248 functional analysis, 264
Eudoxus, 33 Fundamental Theorem of Algebra, 282
Euler, Christoph, 112 Funk, Paul, 261
Euler, Johann Albrecht, 97, 98, 112 Fuss, Nicolas (Nikolaus), 81, 112, 114
Euler, Johann Heinrich, 97
Euler, Karl Johann, 97 Gödel, Kurt, 15
Euler, Leonhard, 13, 59, 61–63, 67–69, Gage, William, 157
72, 73, 81–131, 207, 282 Galilei, Galileo, 47, 118
Dioptrica, 112 Galois theory, 194
Institutiones calculi differentialis, 67, Galois, Evariste, 194
68, 121, 124, 126 Gardiner, 60
Index 351
Gauss, Carl Friedrich, 67, 131, 194, 251, Unsolved Problems in Number
282, 331 Theory, 267
Disquisitiones arithmeticae, 131
Gauthier, Abel, 175 Hadamard, Jacques, 300
Theory of Group Representation by Hadley, John, 50
Matrices, 175 Hadmard, Jacques
Gauthier, Mgr Georges, 174 The Psychology of Mathematical
Gellibrand, Henry, 45 Invention, 300
general relativity theory, 194 Halley, Edmund, 46, 51
Gentzen, Gerhard, 261 Hamel, Abbé Théophile-Étienne, 164,
geography, 98 168, 169
geometric algebra, 19 Hamel, Georg, 261
George, F. H., 298 Hamilton, William, 73, 206, 207, 215
Germain, Sophie, 330, 331 Hankel, Hermann, 84, 213
Gesellschaft Deutscher Naturforscher Hanna, Gila, 337
und Ärzte, 234 Harding, Sandra, 338, 339
Gilbert, William, 44, 46 Hardy, Godfrey Harold, 282
De Magnete, 44 Harkness, J., 171
Gilligan, C., 338–340 harmonics, 23, 24, 26, 27
Ginsburg, Jekuthiel, 187 Harrison, John, 54
A History of Mathematics in America Haskell, Mellon W., 199
before 1900, 187 Hassler, Ferdinand, 188
Gödel, Kurt, 278 Hathaway, Arthur, 192
Goldbach’s conjecture, 273 Treatise on Projective Geometry, 193
Goldbach, Christian, 62, 98 Heath, T. L., 9
Goldstine, Hermann, 74 Hecke, Erich, 261
Golovin, Mikhail Evseyevich, 112 Hedrick, Earle Raymond, 199
Gordan, Paul, 249, 251 Hegel, George Wilhelm Friedrich, 204,
Gosselin, 164 207
Göttinger Nachrichten, 250, 257, 259 Heine, Eduard Heinrich, 223, 226
Gouinlock, G. and J., 150 Heisenberg, Werner, 267
A complete system of practical Henry (Heinrich) Prince of Prussia, 101
arithmetic, 150 Herglotz, Gustav, 250
Gray, Jeremy, 271 Hermite, Charles, 264
The Hilbert Challenge, 273 Herschel, John Frederick William, 66,
Graßmann, Hermann Günther, 213 73, 114, 209
Graßmann, Robert, 213 Hevelius, Johannes, 50
Green’s theorem, 17 Hilbert Syzygy Theorem, 286
Gregory, Duncan F., 73, 210 Hilbert’s basis theorem, 250, 285
Gregory, James, 66 Hilbert, David, 12, 212, 228, 232, 233,
Gromov’s Hilbert tree, 287 243–288, 316
Grothendieck, Alexander, 330 Grundlagen der Geometrie, 243
group theory, 194, 196, 200, 247 Über die invarianten Eigenschaften
Gsell, Katharina, 97 specieller binärer Formen, 247
Gsell, Salome Abigail, 114 Zahlbericht, 243
Gunter, Edmund, 47 Hill, George William, 188
De sectore et radio, 47 Hind, John, 166
Guttman, Vally, 228 Hipparchus, 26
Guy, R. K., 267 Historia Mathematica, VII–VIII, 3–5
352 Index
Hofstadter, Douglas, 299 King, L. V., 173

Holmes, Abbé John, 154 Kirkland, Thomas, 157, 159
homographies, 161 Kleene, Stephen Cole, 279
Hooke, Robert, 46, 53 Klein, Felix, 128, 183, 189, 193–195,
Houdet, Abbé, 155 199, 250, 251, 331, 336
Hull, Clark, 304, 307 Kline, Morris, 74
Hume, David, 61, 207 Klopstick, Friedrich Gottlieb, 58
Hurwitz, Adolf, 247, 256–259 Koenig, Samuel, 105
Hutton, James, 61 Kovalevskaia, Sofia, 329, 330, 332, 335,
hydrography, 143, 146 336
Hypatia, 329 Koyré, Alexandre, 185
hyperelliptic and elliptic integrals and Krafft, Wolfgang Ludwig, 112
functions, 196 Kronecker’s theorem on Abelian fields,
289
Infeld, Leopold, 177 Kronecker, Leopold, 221, 225–229, 233,
integral equations, 264 235, 237, 252, 286
International Congress of Mathemati- Kuhn, Thomas, 185
cians, 173, 243, 251, 255, 266, Kulibin, Ivan Petrovich, 93
273 Kummer, Eduard, 252
invariant theory, 190, 247–251, 285
invariant theory of differential forms, L’Abbé, Maurice, 175
196 l’Hôpital, Marquis de
inversion in the circle, 161 Analyse des infiniment petits, 144
involutions, 161 L’Huilier, Simon, 66, 70, 73
isoperimetric problems, 274 Léveillé, Arthur, 175, 176
Ivanovna, Anna, 97 Lacroix, Silvestre-François, 60, 63, 66,
67, 70, 72, 166
Jacobi, Carl Gustav, 69 Ladd-Franklin, Christine, 331, 332
Jerosch, Käthe, 250 Ladreyt, Casimir, 152
Jevons, William Stanley, 207, 211, Nouvelle arithmétique, 152
214–216, 298, 304 Lagrange, Joseph Louis, 13, 59, 62, 63,
Principles of Science, 214 65, 67, 69, 70, 72–74, 100, 108,
Pure Logic, 214 128, 166, 331
Johnson, Alexander, 172 Analytical Mechanics, 62, 72
Jolliet, Louis, 144 Théorie des fonctions analytiques, 65,
Jones, George William, 192 74, 128
Treatise on Projective Geometry, 193 Lahaille, M. J.-B., 153
Treatise on Trigonometry, 193 Lambert, Johann Heinrich, 131, 205,
207
Kant, Immanuel, 204, 207, 216, 315, Landen, John, 70, 166
317, 325, 326 Langevin, Jean, 166, 168
Kritik der reinen Vernunft, 207 Traité élémentaire de calcul
Logic, 316, 317 différentiel et de calcul intégral,
Kantor, Jean-Michel, 271 166
Kelland, Phillip, 73 Laplace, Pierre Simon, 13, 59, 67, 81
Keller, Evelyn Fox, 338 Laplacian analysis, 209
Kelvin, Lord, 68 LaRue, Abbé Alexandre, 175
Kepler’s sphere problem, 273 latitude, 43
Kepler, Johannes, 46 Laurent series, 122
Index 353
Laurin, Joseph, 152 Treatise of Fluxions, 57–75

Traité d’arithmétique, 152 MacLean, N. B., 173
Laval, Mgr François de Montmorency, MacMurchy, Alexander, 159
145 MacMurchy, Archibald, 157
Lavirotte, Louis-Anne, 61 magnetic compass, 44
Lebesgue, Henri, 123 magnetic variation, 45–46
Lebovitz, Norman, 67 Mainguy, Abbé, 168
Legendre, Adrien-Marie, 67, 69, 167 Maitby, Margaret, 336
Leibniz, Gottfried Wilhelm, 58, 71, 75, Maitland, Lord, 149
81, 94, 103, 117, 128, 166, 205, 273 Manseau, Conrad, 170
Lemoine, Émile, 284 Mascheroni’s constructions, 273
La Géométrographie ou l’Art des Maschke, Heinrich, 195
Constructions Géometriques, 284 mathematical physics, 253
Lermontova, Iulia, 336 Mathematische Annalen, 331
Lewes, G. H., 299, 305 matrix theory, 13
Lexell, Anders Johan, 112, 114 Mauldin, D., 267
l’Hôpital, Guillaume François Antoine, The Scottish Book, 267
Marquis de, 144, 273 Maupertuis, Pierre-Louis Moreau de,
Liard, Louis, 215 103
Les logiciens anglais contemporaines, May, Kenneth O., VII–IX, 3–5, 187
215 McCulloch, Thomas, 158
Lichtenberg, Georg Christoph, 81 McLellan, James Alexander, 157
Lie theory, 251 The Psychology of Number, 157
Lie, Sophus, 194 McMahon, James, 192
Lindemann, Ferdinand, 226, 247 mechanics, 212
Lindsay, Thomas, 206 Mechanist Thesis, 297, 304, 305, 307,
linear differential equations, 289 309, 315, 326
linear groups, 248 Mehmke, Rudolf, 259
Listing, Johann Benedikt, 101 Mémoires de Trévoux, 144
Lobachevsky, Nikolai Ivanovich, 124 Mercator, Gerard, 47
Locke, John, 84, 298 Mersenne, Father, 255
Loeb, Jacques, 310, 311 metamathematics, 280
logarithms, 47, 119, 128 Mill, John Stuart, 204, 207, 214
logic, 203–217 A System of Logic, 214
longitude, 44, 52–55 Miller, Adam, 157
Lorch, Lee, 336 Minkowski, Hermann, 247, 255–259
Lorrain, Paul, 176 Mitchell, Oscar, 191
Lotze, Hermann Rudolf, 214 Mittag-Leffler, Gösta, 229
Loudon, James, 159, 162, 171 Monro, Alexander, 61
Lovelace, Ada, 305 Moore, Eliakim Hastings, 183, 189,
Lovell, John, 156 195–196, 199
Moore, Robert L., 196
MacColl, Hugh, 215 Mower, Nahum, 147
Macfarlane, Alexander, 216 Mumer, Francois, 175
Mackenzie, M. A., 171 Murray, F. H., 173
Maclaurin series, 66 music, 39
Maclaurin, Colin, 57–75
An Account of Sir Isaac Newton’s Napier, John, 47
Philosophical Discoveries, 61 National Academy of Sciences, 189
354 Index
naval theory, 144 Philosophical Transactions, 50, 51

navigation, 43–55, 84, 90, 98, 143 Piatkiewicz,
Stanislaw, 216
Neil, Samuel, 213 Plateau’s problem, 289
New York Mathematical Society, 184, Playfair, John, 72, 167
197, 198 Plotinus, 38
Newcomb, Simon, 188 Ploucquet, Gottfried, 207
Newell, Allen, 301 Poincaré conjecture, 270
Newton, Isaac, 12, 13, 51, 58, 63, 66, Poincaré, Henri, 221, 244, 255, 261, 266
81, 117, 118, 131, 166 Poisson, Siméon, 67
De Analysi, 63, 64 Poivert, Jules, 176
Principia Mathematica, 12, 64, 70 Polya
Newton-Cotes formula, 69 Aufgaben und Lehrsätze aus der
Noether, Emmy, 177, 250, 329–332 Analysis, 267, 271
non-Euclidian geometries, 12 How to Solve It?, 271
non-standard analysis, 121 Let Us Teach It, 271
Normandin, Abbé, 154 Pólya, George, 267, 271
North, Lord, 54 Poncelet, Jean-Victor, 70
number theory, 225 Pönitz, Karl, 222
Popper, Karl, 15, 267
O’Donnell, Katherine, 338 Posidonius, 26
octant, 48–52 post-computational mechanism, 298
Ogilvy, C. S., 267 Pouliot, Adrien, 175
Tomorrow’s Math. Unsolved Problems Premier livre des elements de géometrie
for the Amateur, 267 d’Euclide, 166
Ohm, Martin, 213 projective geometry, 200
Oliver, James Edward, 192 proof theory, 280, 282
Treatise on Trigonometry, 193 psychologism, 209
optics, 23, 24, 26, 36 psychophysical parallelism, 323
Osgood, William Fogg, 194 Ptolemy, 23–42
Almagest, 27–39
Parapatetic physics, 26, 37 Geography, 32, 37
partial differential equations, 268 Harmonics, 39–41
partitions, 190 Optics, 37
Pascal’s cycloid problem, 273 Planetary Hypotheses, 32–35, 38
Pavlov, I. P., 306, 311 Tetrabiblos, 35
Peacock, George, 66, 73, 209 Puiseux series, 122
Treatise on Algebra, 210 Pythagoras’s Theorem, 8–10
Peano, Giuseppe, 213, 216, 259
Pearson, Karl, 69 quadratic forms, 13, 288
Peirce, Benjamin, 188, 192 quadratic reciprocity, law of, 100
Peirce, Charles S., 190, 213 Quebec Herald, 147
Pelletier, Abbé Alexis, 164
Peter II, 95 Rabus, Georg Leonard, 208
Peter the Great, 82, 94 Racine, Michel, 156
Pézénas, R. P., 59 Ratio Studiorum, 155
Phillips, William, 149 Raymond, Abbé J. S., 165
A new and concise system of real and complex function theory, 200
arithmetic, 149 recursive function theory, 298, 309
Philosophia Mathematica, VII–VIII, 3 recursive functions, 316
Index 355
Riehl, Alois, 215 Simson, Robert, 73, 167

Riemann hypothesis, 264, 270, 288 Euclid, Elements, 167
Riemann’s zeta function, 271 Sinh, Hoang Xuan, 330
Riemann, Bernhard, 58, 194, 252 Sisson, Jonathan, 51
Riemannian function theory, 194 Smale, Steve, 268
rigid-body mechanics, 194 Smirnova, G., 18
Rosebrugh, T. R., 173 Smith, Adam, 61
Rosenkilde, Carl, 67 Smith, Barnard, 157
Rossner, Laurence, 67 Smith, David Eugene, 187
Rota, Gian-Carlo, 251 A History of Mathematics in America
Rouse Ball, W. W., 73 before 1900, 187
Rudolff, Christoff, 85 Smith, J. Hamblin, 157
Coss, 85 Snow, Charles Percy, 84
Ruskai, Beth, 339 Société de Mathématiques et
Russell’s paradox, 232 d’astronomie du Canada, 176
Russell, Bertrand, 232, 279 Société mathématique de Montréal, 176
Russian Imperial Academy of Sciences, Société mathématique de Québec, 175
330 Somerville, Mary Fairfax, 330
Ryerson, Egerton, 150 Souza, P. de, 267
Berkeley Problems in Mathematics,
Saint-Pierre, Jacques, 175 267
Salle, Jean-Baptiste de la, 163 Spiess, Otto, 129
Sangster, John Herbert, 156, 159 Stanley, Julian, 336, 337
Sarton, George, 185 Steiner, Jakob, 252
Sauri, Abbé, 146, 151, 154, 167 Stevin, Simon, 44
Institutions mathématiques, 146, 167 Haven-Finding Art, 44
Schoolmaster, The, 161 Stifel, Michael, 85
Schröder, Ernst, 211, 214, 216 Stirling, James, 62, 67, 68
Der Operationskreis des Logikkalkuls, Stoics, 25, 38
211 Stone, Ormond, 199
Vorlesungen über die Algebra der Story, William, 190
Logik, 211 Strachan, John, 147, 149, 158
Schwarz, H. A., 227 Concise introduction to practical
Schwarz, Laurent, 330 arithmetic, 147
Scott, William, 157, 159 Strato, 26
Serrin, James, 268 Stringham, W. Irving, 199
set theory, 12, 221–239 Studley, Duane, 192
abstract, 221 Sullivan, C. T., 173
transfinite, 222 surveying, 143
Shakespeare, William, 230 Sylvester, James Joseph, 183, 189–192,
Shanks, D., 267 197, 199, 251, 331
Solved and Unsolved Problems in symbolic logic, 15
Number Theory, 267 Synge, J. L., 171, 173, 177
Sierpinski, W., 267 syzygy, 280, 285
A Selection of Problems in the Theory Szegö, Gabor, 267, 271
of Numbers, 267 Aufgaben und Lehrsätze aus der
Simon, Herbert A., 301 Analysis, 267, 271
Simplicius, 38
Simpson, Thomas, 70 Takagi, Teiji, 261
356 Index
Tarski, Alfred, 15 Vandaloukis, I. M., 18

Taylor series, 66 Veblen, Oswald, 196
Taylor, Brook, 66, 85 vector theory, 13
tensors, 248 Venn, John, 205, 214, 216
Theon of Smyrna, 27 Symbolic Logic, 205
The Mathematics Useful for Reading Viète, François, 10
Plato, 27 vibrating string, 123
Theophrastus, 26 Vivanti, Giulio, 231
Théorie élémentaire des nombres Viviani’s Florentine enigma, 273
d’après Buler, Legendre, Gauss et Voltaire, Francois-Marie, 84, 105, 117
Cauchy, 168 von Neumann, John, 271
Thorndike, 306, 307, 311
Tietze, H., 267 Waismann, F., 324
Gelöste und ungelöste mathematische Principles of Linguistic Philosophy,
Probleme, 267 324
Toeplitz, Otto, 16 Wait, Lucien Augustus, 192
topology, 101 Treatise on Trigonometry, 193
topology of real algebraic curves and Walkingame, Francis, 148
surfaces, 289 The tutor’s assistant, 148
Tory, Henry Marshall, 171 Ward, Seth, 60
Transactions of the American Mathe- Young Mathematician’s Guide, 60
matical Society, 197, 199 Waring, Edward, 70
transfinite cardinal numbers, 230–231 Webber, W. J., 171
Tremblay, Althéod, 175, 177 Weierstrass, Karl, 58, 63, 122, 128, 194,
Trendlenburg, Adolf, 208 252, 332, 335, 336
trigonometric series, 223, 226, 228 Weil, André, 267
trigonometry, 47, 124, 144, 153, 165, Weinstein, Alexander, 177
167, 171 Wendling, André-V., 176
Weyl, Hermann, 117, 244, 252, 275, 329,
Truesdell, Clifford, 187
332
Turing Machine, 306, 308, 310, 313,
Whately, Richard, 206
316, 321, 323
Elements of Logic, 206
Turing Test, 299, 323
White, Basil, 176
Turing’s Thesis, 308, 315, 317, 324
White, Henry, 199
Turing, Alan, 279, 297–326
Wilkins, T. R., 173
Mechanical Intelligence, 297, 303,
Williams, Grace Alele, 330
309, 326
Williams, Major, 147
Turkle, Sherry, 339, 340
Windelband, Wilhelm, 208
Turnbull, H. W., 73
Winston, Mary Frances, 331, 336
Wittgenstein, Ludwig, 315–326
Ueberweg, Friedrich, 206 Brown Book, 320
System der Logik und Geschichte der Philosophical Grammar, 325
logischen Lehren, 206 Philosophical Investigations, 318, 321,
Ulam, S., 267 323
A Collection of Mathematical Remarks on the Philosophy of
Problems, 267 Psychology, 318
Ulrici, Herman, 215 Wolfenden, H. H., 173
Unguru, Sabatei, 10, 19 Wolff, Christian, 94, 97
universal algebra, 212 Woodhouse, Robert, 70
Index 357
Wren, Christopher, 45 Yandell, Benjamin, 273

Wright, Edward, 44, 47, 52 The Honors Class, 273
Certain Errors in Navigation, 44, 47, Young, Grace Chisholm, 237, 336
52
Xenarchus, 38
Difficulties Addressing the Fifth Zermelo, Ernst, 279
Element, 38 Zermelo-Fraenkel axioms, 278

2005 Book MathematicsAndTheHistorianSCra PDF

Uploaded by

Copyright:

Available Formats

2005 Book MathematicsAndTheHistorianSCra PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

2005 Book MathematicsAndTheHistorianSCra PDF

Uploaded by

Copyright:

Available Formats

Editors-in-Chief

Mathematics and the

Mathematics Subject Classification (2000): 01-06, 01A85

Library of Congress Control Number: 2005923503

ISBN-10: 0-387-25284-3 Printed on acid-free paper.

© 2005 Springer Science+Business Media, Inc.

Printed in the United States of America. (EB)

Glen Van Brummelen and Michael Kinyon

Annual Meeting Themes & Kenneth O. May Lecturers Since 1990

2003: Maritime Mathematics (Halifax, NS)

2002: In Memory of Kenneth May (Toronto, ON)

2001: French Mathematics (Québec, PQ)

2000: History of Mathematics at the Dawn of a New Millennium

1999: Joint meeting with the British Society for

1998: Late 19th-Century Mathematics (Ottawa, ON)

1997: Science and Mathematics (St. John’s, NF)

1996: Ancient Mathematics (St. Catharines, ON)

1995: Mathematics Circa 1900 (Montreal, PQ)

1994: History of Mathematics in the United States and Canada

1993: Philosophy of Mathematics (Ottawa, ON)

1992: Ethnomathematics (Charlottetown, PEI)

1991: Women in Mathematics (Kingston, ON)

1990: History and Pedagogy (Victoria, BC)

CSHPM/SCHPM Proceedings Editors

Introduction: The Birth and Growth of a Community

2 Ptolemy’s Mathematical Models and their Meaning

3 Mathematics, Instruments and Navigation, 1600-1800

4 Was Newton’s Calculus a Dead End? The Continental

6 Mathematics in Canada before 1945: A Preliminary Survey

7 The Emergence of the American Mathematical Research

10 Hilbert and his Twenty-Four Problems

11 Turing and the Origins of AI

Thomas Archibald Judith V. Grabiner

Jim Bennett Ivor Grattan-Guinness

Joseph W. Dauben Ann Hibner Koblitz

Ph.D. Program in History, Karen Hunger Parshall

The Canadian Society for History and Philosophy of Mathematics, or Société

mathematics. The responses that May received show an enthusiastic reception

sixteen years, it is time to present some of that material to a wider audience.

Middlesex University at Enﬁeld

To the fond memory of John Fauvel (1947–2001)

1.1 Interest and Disagreements

1.2 Pythagoras’s Theorem, Euclid Style

We recognise it as saying of the sides AB, AC, and BC that

1.3 Some Principal Diﬀerences between History and

Feature History Heritage

Motivation(s) to N Important issue; maybe Probably only of

Types of inﬂuence Can be negative as well Likely to draw only

Relationships of N Major issue; diﬀerences Important issue:

Handling unclarities Reconstruct them, and Recognise them, but

Successful Very important: but also Likely to be the

Role of chronology Usually important; can Beyond broad details,

Determinism? Preferably not claimed: May carry a determinist

Foundations Dig down to them, and Lay them down and

Level of importance Can vary over time, Not normally considered;

1.4 Changing Habits

1.5 Some Philosophical Background

histories of other sciences (including technology, engineering, and medicine),

1.6 General Remarks about History in Mathematics