Kant and The Problem of Knowledge - Sanet.st

KANT AND THE PROBLEM
OF KNOWLEDGE
This book examines Kant’s contributions to the theory of knowledge and

studies how his writings can be applied to address contemporary epistemological
issues. The volume delves into the Kantian ideas of transcendental idealism,
space, naturalism, epistemic normativity, communication and systematic unity.
The chapters in the volume study Kant’s theories from a fresh perspective
and offer new arguments for assenting that knowledge cannot account for
itself without acknowledging the fundamental role of the cognitive subject.
In doing so, they suggest that we reconsider Kant’s views as a powerful
alternative to naturalism.
Featuring readings by well-known Kant specialists and emerging scholars
with unorthodox approaches to Kant’s philosophy, the volume fills a significant
gap in the existing scholarship on the philosopher and his works. It will be
of great interest to scholars and researchers of knowledge, philosophy and
epistemology.
Luigi Caranti is Professor of Political Philosophy at the Università di

Catania. He focuses on Kant, human rights, peace studies and distributive
justice. Principal investigator of numerous EU-funded research projects, he is
currently coordinating the Marie Curie Rise project “Kant in South America”.
Among his recent publications are The Kantian Federation (2022), (ed. with
D. Celentano) Paradigms of Justice: Redistribution, Recognition and Beyond
(2021) and Kant’s Political Legacy: Human Rights, Peace, Progress (2017).
Alessandro Pinzani is Professor of Ethics and Political Philosophy at the

Federal University of Santa Catarina, Florianópolis (Brazil), and, since 2006,
a fellow researcher of CNPq (Brazilian Research Council). His publications
include Jürgen Habermas (2007), An den Wurzeln moderner Demokratie
(2009) and Money, Autonomy, and Citizenship (with W. Leão Rego, 2018).
KANT AND THE
PROBLEM OF
KNOWLEDGE
Rethinking the Contemporary World
Edited by Luigi Caranti and

Alessandro Pinzani
First published 2023
by Routledge
4 Park Square, Milton Park, Abingdon, Oxon OX14 4RN
and by Routledge
605 Third Avenue, New York, NY 10158
Routledge is an imprint of the Taylor & Francis Group, an informa business
© 2023 selection and editorial matter, Luigi Caranti and Alessandro
Pinzani; individual chapters, the contributors
The right of Luigi Caranti and Alessandro Pinzani to be identified
as the authors of the editorial material, and of the authors for their
individual chapters, has been asserted in accordance with sections
77 and 78 of the Copyright, Designs and Patents Act 1988.
All rights reserved. No part of this book may be reprinted or
reproduced or utilised in any form or by any electronic, mechanical,
or other means, now known or hereafter invented, including
photocopying and recording, or in any information storage or
retrieval system, without permission in writing from the publishers.
Trademark notice: Product or corporate names may be trademarks
or registered trademarks, and are used only for identification and
explanation without intent to infringe.
British Library Cataloguing-in-Publication Data
A catalogue record for this book is available from the British Library
Library of Congress Cataloging-in-Publication Data
A catalog record for this book has been requested
ISBN: 978-0-367-90316-9 (hbk)

ISBN: 978-0-367-50674-2 (pbk)
ISBN: 978-1-003-05074-2 (ebk)
DOI: 10.4324/9781003050742
Typeset in Sabon
by Apex CoVantage, LLC
CONTENTS
List of contributors vii

Acknowledgments ix
List of sigla xi
Introduction 1
L U I G I CA R A N TI A N D AL E SSAN DRO P IN ZAN I
1 The transcendental deduction and transcendental

idealism: a retrospective 9
H E N RY E . A L LISO N
2 Kant on scientific laws 25

LEA YPI
3 Normativity and reflective casual inference 44

PATR Í C I A K AUARK- L E ITE
4 Consciousness as representation of representation: Kant on

the human and animal capacity for representation and
consciousness 55
S I LV I A A LTM AN N
5 Critique and development of the Kantian theory of space

in Gerold Prauss 80
B E R N D D Ö R F L IN GE R
6 Kant on communication 94
L U CA F O N N E SU
v
CONTENTS
7 Kant’s space of theoretical reason and science:

a perspectival reading 109
L O R E N Z O SPAGN E SI
8 Can physics explain physics? Anthropic principles

and transcendental idealism 136
RO B E RT H AN N A
Index 146
vi
CONTRIBUTORS
Henry E. Allison is Professor Emeritus at the University of California, San

Diego and Boston University. He has published fifteen books, most
recently An Introduction to the Philosophy of Spinoza (Cambridge 2022)
and numerous articles on Kant and other figures in the history of modern
philosophy. He is a former president of the Pacific Division of the Amer-
ican Philosophical Association, winner of the International Kant Prize
(2005) and the De Gruyter Kant Prize Lecturer (2014).
Silvia Altmann is Full Professor at the Federal University of Rio Grande
do Sul, Brazil, and a CNPq researcher fellow and vice-president of the
Brazilian Kant Society. Her PhD dissertation was on Kant’s category of
existence. Her main research interests are about the notions of judgement,
concepts and consciousness in the history of philosophy, with publica-
tions about Descartes, Kant and Wittgenstein.
Bernd Dörflinger completed his studies in philosophy and worked as a
researcher and lecturer at the University of Mainz from 1974 to 2000.
He finished his doctoral thesis (on the Critique of Judgment) in 1986
and his habilitation (on the Critique of Pure Reason) in 1994. He held a
chair as Professor of Modern Philosophy at the University of Trier from
2000 to 2019. From 2004 to 2019, he served as President of the Kant-
Gesellschaft. Since 1999, he has served as Co-Editor of the journal Kant-
Studien.
Luca Fonnesu is Full Professor of Moral Philosophy at the University of
Pavia and President of the Società italiana di studi kantiani (2016–2022).
He is the author of Antropologia e idealismo: La destinazione dell’uomo
nell’etica di Fichte (1993), Storia dell’etica contemporanea: Da Kant
alla filosofia analitica (2006) and Per una moralità concreta: Studi sulla
filosofia classica tedesca (2010). He is coeditor (with Lucia Ziglioli) of
System und Logik bei Hegel (2016).
Robert Hanna is an independent philosopher, Director of the online phi-
losophy mega-project Philosophy Without Borders, Director of The
vii
CONTRIBUTORS
Contemporary Kantian Philosophy Project and also Editor of two jour-

nals: Borderless Philosophy and Contemporary Studies in Kantian Phi-
losophy. Philosophically, he’s a resolute generalist, with a broadly but
also radically Kantian orientation. He’s published over 150 articles and
twelve books and has also written four other books for universal free
sharing online.
Patrícia Kauark-Leite is Associate Professor in the Philosophy Department
at Federal University of Minas Gerais (UFMG-Brazil) and Researcher at
National Council for Scientific and Technological Development (CNPq-
Brazil). She was a visiting scholar at Stanford University (2011–2012),
Martin Luther Universität Halle-Wittenberg (2019–2020) and Universi-
dade de Lisboa (2020). Her book Theorie quantique et philosophie tran-
scendantale: dialogues possibles (Paris: Hermann, 2012) won the Prize
Louis Liard (2012) from Académie des sciences morales et politiques
(France). Her research interests focus on Kant’s theoretical philosophy
and the significance of transcendental idealism for contemporary science.
Lorenzo Spagnesi is a postdoctoral researcher at the University of Trier. He
is Research Fellow in the DFG project Inductive Metaphysics, and he is
part of the Kant-Forschungsstelle Trier. He received his PhD in philoso-
phy from the University of Edinburgh in 2021. His research is focused on
Kant, the philosophy of science and metaphysics.
Lea Ypi is Professor in Political Theory at the London School of Economics
and Political Science and Adjunct Professor in Philosophy at the Austra-
lian National University. She is the author of Global Justice and Avant-
Garde Political Agency, The Meaning of Partisanship (with Jonathan
White) and The Architectonic of Reason. Her latest book, a philosophical
memoir entitled Free: Coming of Age at the End of History, was recently
shortlisted for the Baillie Gifford Prize, Britain’s biggest literary prize for
nonfiction. Her work has been translated into more than twenty lan-
guages.
viii
ACKNOWLEDGMENTS
Does Kant still have something to say to us after more than 200 years from
the publication of his major works (the three Critiques, the Groundwork, the
Metaphysics of Morals, the political writings and those on the philosophy of
history)? Can his thinking still serve as a guide to navigate the turbulence of
a globalized world, of an imprescriptible social reality in which moral val-
ues and ethical life models have lost their stability, while sciences – even the
most exact ones – are painfully aware of the precarity of their foundations,
and the borders between domestic and international issues are increasingly
blurred?
The editors of these volumes are convinced that these questions are to be
answered positively and that Kant remains a source of inspiration for deal-
ing with the latest developments in areas that include ethics, politics and the
theory of knowledge. Guided by this belief, we have invited Kant specialists
from different backgrounds to discuss contemporary epistemological, moral
and political issues from a Kantian perspective. Some have chosen to stay
close to Kant’s texts, and others have used his work merely as a source of
inspiration; all of them, we believe, have shown how his critical philosophy
retains a capacity for interpreting reality in salient ways and for offering
solutions to our problems – be they new ones, which Kant himself could
not foresee, or old ones, which have haunted humankind from its beginning.
Editing these volumes was a journey that began organically some years
ago, first through personal conversations between us, then through com-
mon projects. The decisive pivot occurred during the 8th Multilateral Kant
Congress, held in Catania October 11–13, 2018, which was in turn made
possible by a grant from the European Commission that enabled a four-
year-long structured cooperation between Kant scholars from Europe and
South America (Marie Curie RISE n.777786 “Kant in South America”). The
present book constitutes the main scientific output of that project. Its articu-
lation in three volumes, devoted separately to knowledge, ethics and poli-
tics, reflects the three scientific work packages into which that project was
divided. Many of the authors who participated in this publication presented
ix
AC K N OW L E D G M E N T S
a first draft of their contributions in Catania, while others joined the project
later. Our gratitude goes to all of them. Special thanks to Routledge and to
our editor there, Aakash Chakrabarty, for believing in the validity of our
idea and for being willing to publish the book despite its dimensions.
x
SIGLA
References to Kant’s works follow the sigla reproduced in the following.

The English translation is that of the Cambridge edition unless otherwise
indicated by the authors.
Kant, Immanuel: Gesammelte Schriften Hrsg.: Bd. 1–22 Preussische Akad-
emie der Wissenschaften, Bd. 23 Deutsche Akademie der Wissenschaften zu
Berlin, ab Bd. 24 Akademie der Wissenschaften zu Göttingen. Berlin 1900ff.
AA Akademie-Ausgabe
Anth Anthropologie in pragmatischer Hinsicht (AA 07)
AP Aufsätze, das Philanthropin betreffend (AA 02)
BDG Der einzig mögliche Beweisgrund zu einer Demon-
stration des Daseins Gottes (AA 02)
Br Briefe (AA 10–13)
DfS Die falsche Spitzfindigkeit der vier syllogistischen
Figuren erwiesen (AA 02)
DI Meditationum quarundam de igne succincta delinea-
tio (AA 02)
EaD Das Ende aller Dinge (AA 08)
EACG Entwurf und Ankündigung eines Collegii der phy-
sischen Geographie (AA 02)
EEKU Erste Einleitung in die Kritik der Urteilskraft (AA 20)
Ethica (1763) Ethica Philosophia (AA 27)
FBZE Fortgesetzte Betrachtung der seit einiger Zeit wah-
rgenommenen Erderschütterungen (AA 01)
FEV Die Frage, ob die Erde veralte, physikalisch erwogen
(AA 01)
FM Welches sind die wirklichen Fortschritte, die die
Metaphysik seit Leibnizens und Wolff’s Zeiten in
Deutschland gemacht hat? (AA 20)
FM/Beylagen FM/Lose Blätter FM: Beylagen (AA 20)
FM Lose Blätter (AA 20)
xi
SIGLA
FRT Fragment einer späteren Rationaltheologie (AA 28)

GAJFF Gedanken bei dem frühzeitigen Ableben des Herrn
Johann
Friedrich von Funk (AA 02)
GMS Grundlegung zur Metaphysik der Sitten (AA 04)
GNVE Geschichte und Naturbeschreibung der merkwürdig-
sten Vorfälle des Erdbebens, welches an dem Ende des
1755sten Jahres einen großen Theil der Erde erschüt-
tert hat (AA 01)
GSE Beobachtungen über das Gefühl des Schönen und
Erhabenen (AA 02)
GSK Gedanken von der wahren Schätzung der lebendigen
Kräfte (AA 01)
GUGR Von dem ersten Grunde des Unterschiedes der Gegen-
den im Raume (AA 02)
HN Handschriftlicher Nachlass (AA 14–23)
IaG Idee zu einer allgemeinen Geschichte in weltbürgerli-
cher Absicht (AA 08)
KpV Kritik der praktischen Vernunft (AA 05)
KrV Kritik der reinen Vernunft
KU Kritik der Urteilskraft (AA 05)
Log Logik (AA 09)
MAM Mutmaßlicher Anfang der Menschheitsgeschichte
(AA 08)
MAN Metaphysische Anfangsgründe der Naturwissen-
schaft (AA 04)
MoPh Metaphysicae cum geometria iunctae usus in
philosophia naturali, cuius specimen I. continet mon-
adologiam physicam (AA 01)
MpVT Über das Mißlingen aller philosophischen Versuche
in der Theodicee (AA 08)
MS Die Metaphysik der Sitten (AA 06)
RL Metaphysische Anfangsgründe der Rechtslehre (AA 06)
TL Metaphysische Anfangsgründe der Tugendlehre
(AA 06)
MSI De mundi sensibilis atque intelligibilis forma et prin-
cipiis (AA 02)
NEV Nachricht von der Einrichtung seiner Vorlesungen in
dem Winterhalbenjahre von 1765–1766 (AA 02)
NG Versuch, den Begriff der negativen Größen in die
Weltweisheit einzuführen (AA 02)
NLBR Neuer Lehrbegriff der Bewegung und Ruhe und der
damit verknüpften Folgerungen in den ersten Grün-
den der Naturwissenschaft (AA 02)
xii
SIGLA
NTH Allgemeine Naturgeschichte und Theorie des Him-

mels (AA 01)
OP Opus Postumum (AA 21 und 22)
Päd Pädagogik (AA 09)
-PG Physische Geographie (AA 09)
PhilEnz Philosophische Enzyklopädie (AA 29)
PND Principiorum primorum cognitionis metaphysicae
nova dilucidatio (AA 01)
Prol Prolegomena zu einer jeden künftigen Metaphysik
(AA 04)
Refl Reflexion (AA 14–19)
RezHerder Recensionen von J. G. Herders Ideen zur Philosophie
der Geschichte der Menscheit (AA 08)
RezHufeland Recension von Gottlieb Hufeland’s Versuch über den
Grundsatz des Naturrechts (AA 08)
RezMoscati Recension von Moscatis Schrift: Von dem körperli-
chen wesentlichen Unterschiede zwischen der Structur
der Thiere und Menschen (AA 02)
RezSchulz Recension von Schulz’s Versuch einer Anleitung zur
Sittenlehre für alle Menschen (AA 08)
RezUlrich Kraus’ Recension von Ulrich’s Eleutheriologie (AA
08)
RGV Die Religion innerhalb der Grenzen der bloßen Ver-
nunft (AA 06)
SF Der Streit der Fakultäten (AA 07)
TG Träume eines Geistersehers, erläutert durch die
Träume der Metaphysik (AA 02)
TP Über den Gemeinspruch: Das mag in der Theorie
richtig sein, taugt aber nicht für die Praxis (AA 08)
TW Neue Anmerkungen zur Erläuterung der Theorie der
Winde (AA 01)
UD Untersuchung über die Deutlichkeit der Grundsätze
der natürlichen Theologie und der Moral (AA 02)
ÜE Über eine Entdeckung, nach der alle neue Kritik der
reinen Vernunft durch eine ältere entbehrlich gemacht
werden soll (AA 08)
ÜGTP Über den Gebrauch teleologischer Principien in der
Philosophie (AA 08)
UFE Untersuchung der Frage, ob die Erde in ihrer Umdre-
hung um die Achse, wodurch sie die Abwechselung
des Tages und der Nacht hervorbringt, einige Verän-
derung seit den ersten Zeiten ihres Ursprungs erlitten
habe (AA 01)
VAEaD Vorarbeit zu Das Ende aller Dinge (AA 23)
xiii
SIGLA
VAKpV Vorarbeit zur Kritik der praktischen Vernunft (AA 23)

VAMS Vorarbeit zur Metaphysik der Sitten (AA 23)
VAProl Vorarbeit zu den Prolegomena zu einer jeden künfti-
gen Metaphysik (AA 23)
VARGV Vorarbeit zur Religion innerhalb der Grenzen der
bloßen Vernunft (AA 23)
VARL Vorarbeit zur Rechtslehre (AA 23)
VASF Vorarbeit zum Streit der Fakultäten (AA 23)
VATL Vorarbeit zur Tugendlehre (AA 23)
VATP Vorarbeit zu Über den Gemeinspruch: Das mag in der
Theorie richtig sein, taugt aber nicht für die Praxis
(AA 23)
VAÜGTP Vorarbeit zu Über den Gebrauch teleologischer Prin-
cipien in der Philosophie (AA 23)
VAVT Vorarbeit zu Von einem neuerdings erhobenen vorne-
hmen Ton in der Philosophie (AA 23)
VAZeF Vorarbeiten zu Zum ewigen Frieden (AA 23)
VBO Versuch einiger Betrachtungen über den Optimismus
(AA 02)
VKK Versuch über die Krankheiten des Kopfes (AA 02)
VNAEF Verkündigung des nahen Abschlusses eines Tractats
zum ewigen Frieden in der Philosophie (AA 08)
V-Anth/Busolt Vorlesungen Wintersemester 1788/1789 Busolt (AA
25)
V-Anth/Collins Vorlesungen Wintersemester 1772/1773 Collins (AA
25)
V-Anth/Fried Vorlesungen Wintersemester 1775/1776 Friedländer
(AA 25)
V-Anth/Mensch Vorlesungen Wintersemester 1781/1782 Menschen-
kunde, Petersburg (AA 25)
V-Anth/Mron Vorlesungen Wintersemester 1784/1785 Mrongovius
(AA 25)
V-Anth/Parow Vorlesungen Wintersemester 1772/1773 Parow (AA
25)
V-Anth/Pillau Vorlesungen Wintersemester 1777/1778 Pillau (AA
25)
V-Eth/Baumgarten Baumgarten Ethica Philosophica (AA 27)
V-Lo/Blomberg Logik Blomberg (AA 24)
V-Lo/Busolt Logik Busolt (AA 24)
V-Lo/Dohna Logik Dohna-Wundlacken (AA 24)
V-Lo/Herder Logik Herder (AA 24)
V-Lo/Philippi Logik Philippi (AA 24)
V-Lo/Pölitz Logik Pölitz (AA 24)
V-Lo/Wiener Wiener Logik (AA 24)
xiv
SIGLA
V-Mo/Collins Moralphilosophie Collins (AA 27)

V- Vorlesungen zur Moralphilosophie (hrsg. von Werner
Stark,
Mo/Kaehler(Stark) Berlin – New York, de Gruyter, 2004)
V-Mo/Mron Moral Mrongovius (AA 27)
V-Mo/Mron II Moral Mrongovius II (AA 29)
V-Met/Arnoldt Metaphysik Arnoldt (K 3) (AA 29)
V-Met/Dohna Metaphysik Dohna (AA 28)
V-Met/Heinze Metaphysik L1 (Heinze) (AA 28)
V-Met/Herder Metaphysik Herder (AA 28)
V-Met-K2/Heinze Metaphysik K2 (Heinze, Schlapp) (AA 28)
V-Met-K3/Arnoldt Metaphysik K3 (Arnoldt, Schlapp) (AA 28)
V-Met-K3E/Arnoldt Ergänzungen Kant Metaphysik K3 (Arnoldt) (AA 29)
V-Met-L1/Pölitz Metaphysik L1 (Pölitz) (AA 28)
V-Met-L2/Pölitz Metaphysik L2 (Pölitz, Original) (AA 28)
V-Met/Mron Metaphysik Mrongovius (AA 29)
V-Met-N/Herder Nachträge Metaphysik Herder (AA 28)
V-Met/Schön Metaphysik von Schön, Ontologie (AA 28)
V-Met/Volckmann Metaphysik Volckmann (AA 28)
V-MS/Vigil Die Metaphysik der Sitten Vigilantius (AA 27)
V-NR/Feyerabend Naturrecht Feyerabend (AA 27)
V-PG Vorlesungen über Physische Geographie (AA 26)
V-Phil-Th/Pölitz Philosophische Religionslehre nach Pölitz (AA 28)
V-PP/Herder Praktische Philosophie Herder (AA 27)
V-PP/Powalski Praktische Philosophie Powalski (AA 27)
V-Th/Baumbach Danziger Rationaltheologie nach Baumbach (AA 28)
V-Th/Pölitz Religionslehre Pölitz (AA 28)
V-Th/Volckmann Natürliche Theologie Volckmann nach Baumbach
(AA 28)
VRML Über ein vermeintes Recht, aus Menschenliebe zu
lügen (AA 08)
VT Von einem neuerdings erhobenen vornehmen Ton in
der Philosophie (AA 08)
VUB Von der Unrechtmäßigkeit des Büchernachdrucks
(AA 08)
VUE Von den Ursachen der Erderschütterungen bei Gele-
genheit des Unglücks, welches die westliche Länder
von Europa gegen das Ende des vorigen Jahres betrof-
fen hat (AA 01)
VvRM Von den verschiedenen Racen der Menschen (AA 02)
WA Beantwortung der Frage: Was ist Aufklärung? (AA
08)
WDO Was heißt sich im Denken orientiren? (AA 08)
ZeF Zum ewigen Frieden (AA 08)
xv
INTRODUCTION
Luigi Caranti and Alessandro Pinzani
How does Kant contribute to today’s epistemology? Answers legitimately

vary if one looks at specific critical themes, such as the role of each cognitive
faculty in the constitution of the object of experience; the role of reflective
judgment in the sciences; the unity of science; the role of representation, con-
sciousness, imagination; the account of causality; the role of constitutive and
regulative principles; and so on. But if one were to identify the Kantian theme
that has contributed most to the contemporary epistemological debate, it
would be quite safe to say – and this could hardly be a surprise – that this is
the possibility of synthetic a priori knowledge, with the strictly related theme
of the possibility of a very special subset of synthetic a priori knowledge,
which Kant labeled “transcendental”, that concerns itself “not so much with
objects but rather with our mode of cognition of objects insofar as this is
to be possible a priori” (B25). The debate on synthetic a priori knowledge
occupied, from logical empiricism on, much theoretical philosophy of the
last century, laying the foundation not only of analytic philosophy but also
of the tradition now perceived as essentially opposed to it – phenomenology
(Hanna 2008). Most importantly, the stance philosophers take today on the
viability and nature of this kind of knowledge (and of its transcendental
subset) is going to affect decisively their position in regard to naturalism,
arguably the philosophical orientation still dominant in our times. If one
believes transcendental knowledge is possible, then philosophy has its own
specific domain, and science, even the most advanced, will not be able to
provide all theoretical answers, as methodological naturalism holds. If one
denies the viability of transcendental philosophy, then the only alternative to
naturalism seems to be some form of skepticism. Indeed, Kant’s epistemol-
ogy is still here to remind us that there is an alternative, a third way between
thinking that humans are denied access to objective, universally valid knowl-
edge and the belief that in science one must find the answers to all legitimate
questions, including those of epistemology itself. Even a cursory look at the
recent history of epistemology will suffice to prove that this is the case.
While the effort to harmonize Kant with the evolution of the exact sciences
before Einstein’s relativity had been the central concern of Neo-Kantianism,
DOI: 10.4324/9781003050742-1 1
LUIGI CARANTI AND ALESSANDRO PINZANI
the general epistemological orientation after relativity was largely inspired by

and oriented to a complete rejection of synthetic a priori knowledge or to a
more or less profound reform of it. Logical empiricism was indeed grounded
on the belief that all necessary a priori truth was analytic truth, while all
“extensive”, synthetic knowledge was necessarily a posteriori. The logical
empiricists most respectful of the Kantian heritage, for example, Reichenbach
in his early work The Theory of Relativity and A Priori Knowledge (1920),
attempted to reform rather than reject the critical philosophy by forging a
conception of the “relativized a priori” turning on the idea that constitutive
scientific principles can be both a priori and revisable (de Boer 2011).
In the second half of the century, at least in the Anglo-American world,
the attempt to isolate in Kant a more general philosophical core no longer
dependent on the truth of the mathematics and physics of his time continued,
followed by a large debate on the nature and viability of transcendental argu-
ments (Stroud 1968). In the hugely influential The Bounds of Sense (1966),
Strawson suggested to search for a doctrine of the “necessary conditions of
the possibility of experience” more general than a theory about the neces-
sary presuppositions of Euclidian geometry and Newtonian physics and inde-
pendent of transcendental idealism, considered helplessly dogmatic and, as
such, unworthy of being salvaged. The resulting “metaphysics of experience”
would thus identify necessary features of anything that we can intelligibly
consider experience, such as the reliance of our self-knowledge on some (rela-
tively) permanent objects in the external world, a commitment to the unity of
the spatio-temporal world in which any experience has to find its place and
certain very abstract (yet not merely logical) principles such as permanence
and causality. Strawson’s “descriptive metaphysics”, however, had no room
for genuine synthetic a priori knowledge, let alone for transcendental knowl-
edge as Kant intended it. The necessary features of our experience are not
conceived as elements the knowing subject “lends” to the objects of experi-
ence. Certainly, these features are not to be found in sensations; hence they
are a crucial component of experience no empiricist could account for. And
yet they are not “given” by the knowing subject either, as the transcendental
idealist would say. They are just components of a conceptual scheme without
which experience would not be intelligible for us. The task of philosophy is
to identify these features and articulate the necessary connections between
them (Strawson 1966: 22–3). Moreover, echoing Reichenbach’s notion of
a relativized and historicized a priori (Reichenbach 1920), Strawson thinks
that these conceptual features may vary as science proceeds: “They are not
static schemes, but allow of that indefinite refinement, correction, and exten-
sion that accompany the advance of science and the development of social
forms” (1966: 44). By so doing, Strawson was opening to a form of soft
methodological naturalism (to be distinguished by the hard type championed
by Quine) in which the only thing unchangeable in our knowledge, and as
such not influenced by scientific progress, were logical necessary truths.
2
INTRODUCTION
A more recent representative of soft naturalism is Michael Friedman. He

thinks that the ultimate point of the Critique is to explain how Euclidean
geometry and Newtonian physics are possible (Friedman and Bird 1998: 122).
Thus, “Kant articulated and contextualized the Newtonian framework” and
“the logical empiricists did the same for the Einsteinian” (Friedman and Bird
1998: 129). From this perspective, vindicating a central role for Kant in con-
temporary epistemology is rather hard, because much of Kant’s philosophy
would be inseparable from a scientific worldview by now surpassed. Similarly,
Kant’s anti-naturalistic resources are weakened, if not lost completely. Philoso-
phy is no longer distinguished from science in the sense that the latter is about
objects while the former is about “our mode of cognition of objects” (B25)
but merely in the sense that they operate at two different levels of generality
about the same set of objects. Philosophy articulates and contextualizes the
paradigms or constitutive frameworks that science uses and takes for granted
during its “normal” (in the Kuhnian sense) periods. Hence its task today is
evidently that of articulating and letting emerge the specific framework con-
temporary science employs or, if one considers the status of “permanent revo-
lution” physics offers, the numerous frameworks all competing revolutionary
theories assume. Friedman thinks that this does not amount to a complete
surrender to naturalism because the philosophical articulation is not, as Quine
would have it, a particular application of science to the problem of identifying
its general framework, and its result would not be just one extra piece to be
added to the web of beliefs that holistically constitutes knowledge. It would
still be “transcendental” philosophy because we aim for the identification of
the conceptual necessary conditions of (contemporary) science.
One may obviously have all sorts of doubts about this attempt to reconfig-
ure transcendental philosophy. If its role is that of articulating the necessary
conditions of contemporary science, transcendental philosophy is rendered
fully dependent on what science does and where it proceeds. It is, notwith-
standing all efforts by Friedman, nothing but the science of today’s science’s
main tacit assumptions. The whole project of reforming metaphysics to set it
on a secure path is gone, and the very expression “metaphysics of experience”
has acquired a distinctly un-Kantian undertone. Other scholars are more
optimistic as to the possibility of separating Kant from any specific scientific
worldview. Graham Bird, for example, thinks, correctly in our opinion, that
Kant’s transcendental philosophy defines the general conditions of experi-
ence and only derivatively explains how those conditions can make sense of
a specific science (which of course Kant considered certain, definitive and
immune from future falsification or reform as his transcendental philoso-
phy). The Metaphysical Foundations of Natural Science may be about the
general articulation Friedman thinks is all one can hope from philosophy,
but the Critique of Pure Reason operates at a different, more general level.
Even in the Analytic of principles, where Kant seems to be closest to a con-
cern with the justification of the basic principles of Newtonian mechanics,
3
he makes clear that his focus is rather on more general principles that enable
the scientific (Newtonian) principles: “We treat only of the principles of
pure understanding in their relation to inner sense. . . . It is through these
principles of the pure understanding that the special principles of mathemat-
ics and physical dynamics become possible” (KrV B201–2). Similarly, Kant
establishes a fundamental difference between the metaphysical exposition
of the concept of space (and time) and the transcendental exposition of the
same concept(s). The former is fully independent of Euclidean geometry;
the latter, which as Kant says is about the “explanation of a concept as a
principle from which insight into the possibility of other synthetic a priori
cognitions can be gained” (KrV B40), is about how one can make sense of
the synthetic a priori character of that specific geometry. Perhaps even more
forcefully, in the B Introduction, Kant is quite explicit that the question
“How are synthetic judgments a priori possible?” is more general and more
fundamental than the questions about the a priori status of the sciences. As
Kant puts it, “[i]n the solution of the above problem there is at the same
time contained . . . the solution to the questions: ‘How is pure mathematics
possible? How is pure natural science possible?’” (KrV B20).
Obviously, even if a part of Kant’s theory of knowledge can be separated
from the sciences of his time, this does not mean that the redeemed part is
compatible with the sciences of our time. Certainly, a rather meager con-
solation would be to say that the critical philosophy is merely about the
manifest aspect of reality, that is, about the way we perceive the world in
our everyday experience, and has no ambition to account also for the way
in which science tells us the world is. This solution reduces the critical phi-
losophy to the analysis of the conditions of the possibility not of experience
but of something very close to what Kant calls illusion (Schein) or, at best,
to secondary qualities.
The very project of a critical philosophy, in the editors’ opinion, stands
or falls with the possibility of identifying in the objects of experience, all of
them, including those described and analyzed by the most advanced sciences,
very general features that can only be accounted for by taking into consid-
eration the presence and influence on the final cognitive product of a finite
knowing subject. Kant had his own set of transcendental conditions, and
he may have been wrong about them. Nothing major would happen to the
critical philosophy if we were to discover, through the progress of science or
thanks to some philosophical argument, that these conditions are different,
slightly or significantly, from those Kant identified. But a Kantian should
not expect that these conditions are merely conceptual hidden assumptions
embedded and fully reducible to the scientific picture of contemporary sci-
ence (Friedman’s quasi-naturalistic suggestion). By presenting us with new
objects of experience, scientific progress may show the necessity to revise
what we thought were the true conditions. But the intellectual tension that
guides the search for what Allison called epistemic conditions (2004), that
4
INTRODUCTION
is, conditions of experience in general, not of the most recent science, should
never be abandoned if one wants to remain faithful to the spirit of the criti-
cal philosophy. This would be disproved in its very essence, by scientific
progress or by some other source, if there were compelling reasons to think
that it is open to human cognizers to know the world (cognizers included)
as if they had no influence whatsoever on the cognitive product – as if, to
use the common metaphor, we could experience the world with a view from
nowhere. This would evidently be the triumph of transcendental realism
and of naturalism, which is nothing but the former’s contemporary face. We
go back to the point from which we started. Probably the most significant
contribution Kant offers to the contemporary theory of knowledge is a clear
alternative to naturalism. This contribution does come at a price that many
(Kantians included) find too high to pay: the acceptance of transcendental
idealism, at least of the very minimal quota of it we just sketched.
The chapter collected in this volume speak to many of the problematic
dimensions highlighted above. Henry E. Allison’s chapter starts from
Strawson’s attempt to isolate Kant’s supposedly valuable contribution to
the theory of knowledge from his transcendental idealism. As we saw pre-
viously, Strawson reads the argument in the “Refutation of Idealism” not
merely as an ad hominem argument against Cartesian skepticism but as
the central theme of the Transcendental Deduction. While much could be
said (and something has been said, e.g. Caranti 2007) on whether Kant’s
anti-Cartesian arguments (in the first and second edition of the Critique)
can be construed as independent of transcendental idealism, Allison here
reacts against the Separation Thesis by focusing on its application to
the Transcendental Deduction. Allison argues that Kant’s idealism, prop-
erly understood, is neither disreputable in itself nor dispensable in the
logic of the Transcendental Deduction. Key in this reading is the rejec-
tion of Strawson’s idea that the Transcendental Deduction should not
aim, as Kant thought, to establish a necessary conformity between our
categories and appearances but merely a conditional necessity. Repeated
after Strawson by philosophers such as Harrison and Cassam, whom
Allison briefly discusses, this thought could be expressed in the sense that
appearances must conform to the conditions of the unity of conscious-
ness if they are to constitute experience. In Allison’s analysis, not only is
the stronger conformity thesis, which obviously rests on transcendental
idealism, required by the Transcendental Deduction, but it is defensible
on its own account.
Lea Ypi reinterprets the contemporary debate on scientific laws, in par-
ticular the question of whether there are any laws of nature and, if so,
whether they can be known by us, through the lenses of the Kantian idea
of systematic unity. After identifying four major interpretations of the way
in which scientific laws should be understood (best system account, deriva-
tions from a priori laws account, necessitarian account, reflection account),
5
Ypi suggests that, in the course of his career, Kant moved from a proto-
necessitarian account – the necessity of the laws of nature is grounded in
the “essential nature” of things and is independent of human cognition – to
a reflection-based account – the necessary character of laws is derived from
a priori principles combined with an ideal of unity that orients our efforts
to arrive at lawlike generalizations. Ypi further suggests that what led Kant
to this reconsideration is the realization that he could not reconcile his pre-
critical views on natural purposiveness with the analysis of scientific laws
put forward in the first Critique.
Patrícia Kauark-Leite discusses the role of reflective judgments in Kant’s
theory of epistemic normativity. She argues that the reflective inferences of
the teleological power of judgment are necessary, in addition to the concep-
tual rules of the understanding, in order to constitute our knowledge of the
physical world as an organic and unified system. In particular, and this is
the center of the chapter, she argues that teleological judgment performs an
epistemic role not only for biology and history but also for the domain that
would appear least in need of final causes, that is, physics.
Silvia Altmann reconsiders the relation between understanding, conscious-
ness and self-consciousness in Kant. By looking at some notes taken by Kant
students, she reconstructs a sense, left implicit in the Critique of Pure Rea-
son, in which consciousness can be the representation of a representation
without being self-attribution of a representation. This connects with con-
temporary debates about the difference between the mental life of human
and non-human animals. According to Kant, animals have the capacity to
discriminate objects through representation but no consciousness. Lacking
consciousness (even in the weak sense reconstructed by Altmann), they can-
not develop the capacity of understanding. In human beings, the capacity
to represent our representations permits the development of the capacity of
understanding, which, in turn, allows for a stronger form of consciousness,
that is, the self-attribution of representations.
Bernd Dörflinger’s chapter focuses on a recent important discussion of
Kant’s theory of space as a form of our intuition of outer objects, the one
offered by Gerold Prauss’s Die Einheit von Subjekt und Objekt. Kants Prob-
leme mit den Sachen selbst. Prauss’s central claim is that Kant’s idea of the
subjectivity of space is not developed enough. Kant merely says that space
pertains “to the subjective constitution of our mind, without which” it could
not be “attributed to any thing” (KrV B 38/A 23). According to Prauss, this
stands in need of completion, which is going to be the thesis that space is
produced or generated by the subject, in a radical sense that escaped the
nonetheless revolutionary theory Kant proposed.
Luca Fonnesu moves from Onora O’Neill’s reading of Kant’s conception
of the public use of reason to focus on the problem of communication. By
looking at the section devoted to “Having an opinion, knowing and believ-
ing” in the “Canon” of the “Doctrine of method” of the first Critique and
6
INTRODUCTION
at some passages of the Critique of Judgment, Fonnesu uncovers the episte-

mological presuppositions that govern communication in Kant and explains
why communication is central in Kant’s entire philosophy. Indeed, in differ-
ent contexts (not only the political one), communicability plays the role of the
touchstone or criterion for the distinction between the private and the public.
Lorenzo Spagnesi focuses on the unity of science thesis, namely the idea that
science is both about one single coherent world and that all scientific theories
should harmonize towards a systematic description of reality. Most contem-
porary philosophers of science tend to be skeptical regarding the epistemic
component of the thesis, thus favoring pluralism, which can be minimally
characterized as the idea that the existence of a plurality of at times irrecon-
cilable theories is not a defect of science to be accepted only until scientific
progress will (re)establish unity. Spagnesi focuses on this aspect of the thesis
and raises a number of interesting questions:
Does the acceptance of pluralism inevitably entail the exclusion

of unity from a purely epistemic point of view? Is pluralism suf-
ficient as the only epistemic principle guiding scientific research?
Isn’t it possible to redefine scientific pluralism and monism as non-
conflictual principles?
In the attempt to suggest a positive answer, Spagnesi looks at a perspec-

tival approach to the problem Kant presents in the Critique of Pure Reason
and suggests that Kant’s perspectivism might inspire the current debate.
The last chapter in this collection takes us back to the general orientation
we defended in this introduction. Against naturalism, Robert Hanna argues
that physics is explanatorily incomplete in the sense that it cannot explain
itself as a cognitive enterprise by appealing merely to concepts and methods
of natural science. Only if a moderate Anthropic Principle, construed as
equivalent to a suitably weak version of Kant’s metaphysics of transcenden-
tal idealism, is added to the standard models of the world at the micro and
macro level (particle physics and cosmology) can physics explain itself and
thereby complete itself to that extent. By defending a weak version of tran-
scendental idealism as a necessary component of a complete physics, Hanna
offers a new argument for assenting that knowledge cannot account for
itself without acknowledging the fundamental role of the cognitive subject,
a role irreducible to that of an object among others.
References
Allison, H. E. 2004. Kant’s Transcendental Idealism. New Haven: Yale University
Press.
Caranti, L. 2007. Kant and the Scandal of Philosophy. Toronto: University of Toronto
Press.
7
de Boer, K. 2011. “Kant, Reichenbach, and the Fate of A Priori Principles”, Euro-
pean Journal of Philosophy, 19: 507–31. https://doi.org/10.1111/j.1468-0378.
2010.00397.x
Friedman, M. and Bird, G. 1998. “Kantian Themes in Contemporary Philosophy”,
Proceedings of the Aristotelian Society, Supplementary Volumes, 72: 111–51.
Hanna, R. 2008. “Kant in the Twentieth Century”, in Dermot Moran (ed.) The
Routledge Companion to Twentieth Century Philosophy. London: Routledge.
Reichenbach, H. 1920. Relativitätstheorie und Erkenntnis Apriori. Berlin: Springer.
Trans. M. Reichenbach as The Theory of Relativity and a Priori Knowledge.
Berkeley and Los Angeles: University of California Press, 1965.
Strawson, P. F. 1966. The Bounds of Sense. London: Methuen.
Stroud, B. 1968. “Transcendental Arguments”, The Journal of Philosophy, LXV(9):
241–56.
8
1
THE TRANSCENDENTAL
DEDUCTION AND
TRANSCENDENTAL IDEALISM
A retrospective
Henry E. Allison1
The publication of P. F. Strawson’s The Bounds of Sense in 1966 was a major

event in Anglophone Kant interpretation, because it appeared to make it
possible for philosophers trained in the analytic tradition to have their Kant
without the disreputable idealism that was generally thought to be inseparable
from it. Strawson accomplished this by distinguishing sharply between what
he regarded as the “analytic achievement” (Strawson 1966: 271) of the Cri-
tique and the transcendental idealism with which it is intertwined. Strawson’s
positive focus was on the Transcendental Analytic, and he considered Kant’s
achievement insofar as it is contained in this portion of the Critique “analytic”,
since, in his reading, it attempted to demonstrate that certain general features
are necessary ingredients in any coherent conception of experience that one
can ascribe to oneself. Paramount among these is that the experienced world
must contain at least relatively permanent re-identifiable objects and an overall
regularity in their interactions and changes (Strawson 1966: 24, 28). Strawson
characterized Kant’s idealism in numerous ways, but the most striking is as the
doctrine that “reality is supersensible and that we can have no knowledge of it”
(1966: 38) According to Strawson, the reason Kant appeals to transcendental
idealism in support of his claims about the structure of experience is to account
for their presumed necessity. In this widely shared view, the fundamental idea
underlying Kant’s so-called “Copernican Revolution” is that necessity can be
established only on the basis of the metaphysically idealistic premise that the
order of nature is imposed by the human mind, which Strawson describes as
the theory of the “mind making Nature” (1966: 22). And closely related to this
is his dismissal of Kant’s account of the mental processing through which this
“making” is supposedly achieved, which he refers to as “the imaginary subject
of transcendental psychology” (Strawson 1966: 32).
DOI: 10.4324/9781003050742-2 9
H E N RY E . A L L I S O N
Strawson’s work has spurred ongoing debates regarding the nature of

Kant’s idealism and its relation to the Transcendental Deduction, as well as
the broader issue of the force of transcendental arguments in which, to para-
phrase Robert Stern, it is claimed that the possibility of some supposedly
indisputable fact about us or our mental life, for example, that we have expe-
riences, presupposes some state of affairs, for example, the existence of endur-
ing material objects, that the skeptic has called into question (Stern 2000: 6).
One of the familiar objections to such arguments is that they rest on some
undefined idealistic commitment, which is frequently considered sufficient
to undermine their force (Stern 2000: 49–58). My concern here, however, is
with the bearing of this objection on Kant’s Transcendental Deduction, and
my intent is to defend Kant against it by showing that this idealism, properly
understood, is neither disreputable nor separable from the argument of the
Transcendental Deduction. The discussion is divided into three parts: (1) an
analysis of Strawson-inspired criticisms of the Deduction, which turn on the
propositions that Kant’s mistake was to attempt to demonstrate the uncondi-
tional necessity that appearances conform to the categories, when all that is
either required or possible is a merely hypothetical or conditional necessity;
(2) an explanation of why the stronger claim is required; (3) a consideration
of the connection between the argument of Transcendental Deduction and
transcendental idealism from the perspective of Lucy Allais’s discussion of the
issue in her recent book (2015), in which she both defends this idealism and
argues that certain aspects of the Deduction are separable from it.
1. Kant after Strawson

Until recently, much of the post-Strawson Anglophone Kant literature has
been devoted to the project of separating as much of Kant’s theoretical phi-
losophy as possible from the supposedly excess and disreputable baggage of
transcendental idealism. I shall here consider the views of two philosophers
who have played a significant role in the development of this line of thought
but, as far as I know, do not consider themselves Kantians: Ross Harrison
and Quassim Cassam.2
(A) Harrison: Harrison’s concern is to defend transcendental arguments,
understood in roughly the manner described by Stern, namely as anti-
skeptical arguments in which the negation of a skeptical conclusion is shown
to be a necessary condition of a premise that the skeptic denies, against the
objection raised by Bernard Williams and others that they presuppose some
kind of idealism (Harrison 1982: 211–24). Assuming that a dependence on
idealism in any form would undermine such an argument, because it allows
the skeptic to reply that all it could show is how things must seem to us,
his self-appointed task is to demonstrate that these arguments can succeed
without any dubious idealistic premises But though his primary concern is
with contemporary forms of transcendental argumentation, he does focus a
10
T R A N S C E N D E N TA L D E D U C T I O N A N D I D E A L I S M
significant amount of attention on Kant, whom he considers the philosopher

who both invented this form of argumentation and the exemplar of the folly
of attempting to provide it with an idealistic grounding.
Harrison’s major move, which was anticipated by Strawson and subse-
quently endorsed by many others, is to isolate and call into question the
underlying assumption on the basis of which idealism is supposedly intro-
duced into Kant’s core transcendental argument that any world experienca-
ble by beings like ourselves must be subject to the categories. Characterizing
Kant’s “transcendental” claim in general terms as the proposition that the
world is necessarily judgable, Harrison accuses Kant of a modal fallacy by
conflating the merely hypothetical necessity that the world must be itself
judgable if we are able to judge or experience it, with the quite different
proposition that the world is necessarily judgable tout court (Harrison 1982:
215–16). And, in explanation of this fallacy, he appeals to Kant’s concern to
establish certain synthetic a priori propositions about possible experience,
which he can only do by assuming that the truth conditions of these proposi-
tions are imposed by the mind. As Harrison puts it,
Because appearances [for Kant] depend on us there is not a crucial

distinction between hypothetical necessities in which something is
only true if it is judged and categorical necessities which are true
in themselves. For appearances are not things in themselves; their
existence depends on judgment.
(Harrison 1982: 216)
Harrison’s proposal for saving transcendental arguments from the fatal

infection of idealism consists of lowering the bar by insisting that a success-
ful argument of this type need not establish such necessities. According to
him, “a transcendental argument works without establishing the necessity
of anything apart from hypothetical statements. It therefore does not need
the idealist premise that what exists is necessarily judged” (Harrison 1982:
216). Harrison attempts to illustrate this by considering an argument of the
Kantian type, which shifts attention from the objects of our knowledge to
the “medium by which they are apprehended by us” (1982: 220). On this
basis, he suggests that it is possible to establish indirectly propositions about
the nature of whatever is cognized by beings like us through this medium,
but he insists that the procedure goes awry when, instead of drawing from
this conclusions about how things will necessarily appear to us in virtue
of this medium, one draws conclusions about how they are necessarily in
themselves, for then the objects do not have any existence apart from this
medium, which is full-fledged idealism.
The most interesting feature of Harrison’s account is his response to a
worry that he thinks it might generate, namely the merely contingent nature
of the agreement between the cognitive requirements of the mind and the
11
nature of the world. He suggests that without idealism, there appear to be

two alternative explanations for this contingency: either it “was pure luck or
that it was the product of the conscious design of some external cause” (Har-
rison 1982: 222). But rejecting an appeal to luck as a non-explanation and
citing Kant’s own skepticism regarding the appeal to pre-established har-
monies to dismiss the latter, Harrison suggests a third explanation, namely
natural selection. In this view, the agreement between our forms of judgment
and the world is the result of “normal causal processes”, such that “any indi-
vidual who thinks (or acts) as if there is no causation in the world, or that the
future does not resemble the past, is an unsuccessful mutation, ill-adapted
to survive or breed” (Harrison 1982: 223). In a concluding paragraph, Har-
rison acknowledges that this explanation holds only up to the present time
and that if causal relations cease to hold in the world we will no longer be
able to judge it, but he insists that this is beside the point, since “[the] pres-
ent harmony is all that it is required to explain and for which idealism was
thought to be the only possible explanation” (1982: 223).
(B) Cassam: Although Cassam expresses skepticism regarding the viability
of transcendental arguments under the best of conditions, the bulk of his
account consists of a defense of a version of Strawson’s “objectivity argu-
ment” purified of any idealistic component (Cassam 1987: 355–78). As with
Harrison, Kant is used as the model to illustrate the futility of appealing to
idealism in order to answer a skeptic by means of such an argument. But
inasmuch as Cassam’s account pays far greater attention to Kant’s text, par-
ticularly to the Transcendental Deduction, which he considers “as profound
as it is opaque” (1987: 356), it calls for a separate treatment.
According to Cassam, “Transcendental arguments are concerned with
the specification of conceptually necessary conditions of the possibility of
experience” (1987: 356). After dismissing as a side issue Stroud’s objection
that such arguments unavoidably reduce to some form of verificationism,
he maintains that any argument of this type must contain two components,
which he names a “Conceptual Component” and a “Satisfaction Compo-
nent” (Cassam 1987: 357). The former is a claim about what counts as
experience, while the latter refers to a state of the world, which must hold
if experience, so understood, is to be possible. But recognizing that, so
described, this argument-schema is too weak, since it leaves it open to the
skeptic to challenge the assumed conception of experience, Cassam rede-
scribes the Conceptual Component as the necessity of the possibility of dis-
tinguishing within experience “a component of recognition or judgment,
which is not simply identical with, or wholly absorbed by, the particular
item which is recognized, which forms the topic of the judgment” (1987:
359).3 Correlatively, the Satisfaction Component is a degree of regularity in
experience that is sufficient to make this possible. And it is assumed that the
skeptic must accept the former, because it is either analytic or some other
form of conceptual truth (Cassam 1987: 358–9).
12
Cassam tests his proposal by applying it to the A-version of the Transcen-

dental Deduction. He admits that his reading is “somewhat artificial”, since
it deals with only one aspect of the Deduction (its “objective side”) while set-
ting aside for later consideration another aspect (its “subjective side”), and
his intent is to show how closely his schema corresponds to this aspect of the
Deduction. Eschewing details, the “Conceptual Component” is equated with
the necessary unity of consciousness, and it is claimed that,
The Satisfaction Component then makes a claim about how the

world must be if the unity of consciousness is to be possible, the
upshot being that the truth of something which is explicitly claimed
to be doubtful is necessary if the unity of consciousness requirement
which with luck, the sceptic will already have been persuaded to
accept, is to be met.
(Cassam 1987: 358–9)
Correlatively, the Satisfaction Component is that “appearances must display

such unity and interconnectedness as is possible only if they are appearances
of objects” (Cassam 1987: 361). Accordingly, there is a close fit between
Cassam’s proposal and the objective side of the A-Deduction. But the most
interesting part of Cassam’s account is the portion that he had initially
neglected, which corresponds to the “subjective side” of the Deduction,
where idealism makes its appearance. At this point, Cassam’s account echoes
Harrison’s and Strawson’s and consists of two parts. First, suppressing for
dialectical purposes a worry about the viability of transcendental arguments
as such, he claims that,
It is quite sufficient to say that what a transcendental argument

shows is that the world must be a certain way for experience to
be possible; it is a further unnecessary step to insist that the world
referred to is merely the world of appearances, and that it is in a way
constructed by the understanding.
(Cassam 1987: 364)
Having dismissed any need for transcendental idealism, Cassam attempts

to explain why Kant could have failed to realize this obvious point. His,
answer, like Strawson’s and Harrison’s, is that it resulted from a conflation
of two senses of necessity, namely between a merely conditional and an
unconditional necessity (Cassam 1987: 366). The former, which is assumed
to be warranted by Kant’s objectivity argument, is that appearances must
conform to the conditions of the unity of consciousness, that is, contain
sufficient order and coherence, if they are to constitute experience. The lat-
ter is the unconditional necessity that appearances conform to this unity
tout court. And this conflation is deemed responsible for the introduction
13
of idealism, because it is only on the assumption that the requisite order-

liness of appearances is imposed by the mind that their conformity to the
conditions of the unity of consciousness can be guaranteed.
The final step is to account for Kant’s erroneous assumption that such a
guarantee is needed. Cassam sees this as grounded in what he, like Harrison,
regards as Kant’s misguided worry about contingency, and he attributes to
this worry Kant’s insistence on the necessity for a transcendental synthesis
of the imagination as the objective ground of the affinity or associability of
appearances (Cassam 1987: 366–7). Moreover, he dismisses this worry by
distinguishing between two questions: (1) “What must appearances be like
if they are to provide a basis for the unity of consciousness?” and (2) “[W]hy
are appearances such as to provide for the unity of consciousness?” (Cassam
1987: 370). Since it is precisely the kind of question that a transcendental
argument on his view is designed to answer, it is not surprising that Cassam
finds the first a “genuine philosophical question” (1987: 366–7). The second
question, however, is not treated so kindly, for he questions its legitimacy,
despite the fact that Harrison appears to have provided a naturalistic answer
to it. Instead, in dismissing the second question, Cassam appeals to a Witt-
gensteinian quietism, claiming that,
It is enough . . . that appearances do conform to the conditions of the

unity of apperception; they might not have done and they might not
continue to do so, and whilst this might give rise to a certain sense of
insecurity, this degree of insecurity simply has to be tolerated.
(1987: 370)
2. Response to Cassam
Although I accept Cassam’s diagnosis of the motivation underlying Kant’s
appeal to transcendental idealism, I do not accept his claim that Kant was
incorrect in doing so. To reduce the central issue to a single point, the ques-
tion is whether the possibility that appearances might not conform to the
conditions of the unity of apperception is, as Cassam maintains, simply cause
for a certain degree of insecurity that must be tolerated or, as I have argued
elsewhere, constitutes a specter that must be exorcized if the Transcendental
Deduction is to succeed.4 Kant’s clearest statement of his view is in a passage
contained in the introductory portion of the Deduction contained in both
editions, where he writes:
Appearances could . . . be so constituted that the understanding

would not find them in accord with the conditions of its unity, and
everything would then lie in such confusion that, e.g., in the succes-
sion of appearances nothing would offer itself that would furnish
a rule of synthesis and thus correspond to the concept of cause
14
and effect, so that the concept would therefore be entirely empty,

nugatory, and without significance. Appearances would nonetheless
offer objects to our intuition, for intuition by no means requires the
functions of thinking.
(A90–91/B122–23)
It is crucial to recognize that Kant regarded this as a genuine worry that

must be dealt with, rather than, as is sometimes thought, one that can be
quickly set aside as a remnant of a discarded dogmatism. This is because it is
the direct consequence of what I have termed Kant’s discursivity thesis, that
is, the thesis that human cognition consists of two distinct elements (sensible
intuitions and concepts), which must work together to constitute cognition.
Since experience is empirical cognition for Kant, it follows from this thesis
that the agreement between what is presented in sensible intuition (appear-
ances) and the concepts through which these appearances are unified in
thought is a necessary condition of possible experience. And it also follows
that the requisite agreement is an immanent one between two species of rep-
resentations rather than between a unified set of representations in the mind
and a world composed of things in themselves, where agreement would
seem to be either purely a matter of luck, a pre-established harmony, or, as
Harrison suggests, the result of natural selection. But though this may be
said to reframe or “immanentize” the question of the conditions of empiri-
cal cognition, it does not resolve the problem posed by Kant’s critics, for it
does not explain why anything more than a de facto agreement is required,
which would involve a merely conditional necessity. Accordingly, the task is
to understand why Kant insisted that a more than conditional necessity is
called for and how it is to be understood.
Let us begin with the passage in which Kant first introduces this notion of
necessity in the A-Deduction in connection with his account of the relation
between a representation and its object in cognition. He there writes:
We find . . . that our thought of the relation of all cognition to

its object carries something of necessity with it, since . . . the lat-
ter is regarded as that which is opposed to our cognitions being
determined at pleasure or arbitrarily rather than being determined
a priori, since insofar as they are to relate to an object our cogni-
tions must necessarily agree with each other in relation to it, i.e.,
they must have that unity that constitutes the concept of an object.
(A104)
The main question posed by this passage concerns the sense of “necessity” to
which Kant appeals. Since he is describing the relation between all cognition
and its object, it is one that pertains to empirical as well as a priori cognition,
which is the source of the puzzle. Indeed, it has sometimes been suggested that,
15
despite his clear insistence on the contingent nature of empirical propositions,

Kant effectively rendered them necessarily true.5 At first glance, Kant’s charac-
terization of the object as that which “is opposed to” in the sense of preventing
our cognitions from being determined in an arbitrary manner might suggest
a causal necessity, such as is affirmed in causal theories of knowledge, but it
seems evident that this is not what Kant had in mind. Moreover, since this
account is intended to apply to empirical cognition, the claim that the relation
between the representations is “determined a priori” can only mean by an a
priori rule, which Kant will proceed to equate with a category.
For these reasons, I maintain that the necessity that Kant claims to be
involved in empirical cognition is normative rather than either causal or
logical.6 I further maintain that it is grounded in the nature of concepts as
rules, which are inherently normative. For example, Kant says that “in the
case of the perception of something outside us the concept of body makes
necessary the representation of extension, and with it that of impenetrability,
of shape, etc.” (A106). Since Kant maintained that the proposition “bodies
are extended” is analytic, and their possession of a certain shape (as well
as other properties such as divisibility) is entailed by their extendedness, it
might be thought that the necessity to which he refers is merely logical. That
this is not what Kant meant is clear, however, from the inclusion of impen-
etrability, which does not follow analytically from the concept of body, as
well as the “etc.”, which suggests other properties such as weight (or heavi-
ness) and attraction, which he regards as necessarily possessed by bodies in
experience without being contained in the concept. Moreover, this rule has
a normative necessity in the two-fold sense that it gives a kind of epistemic
right to expect, for example, that if something is extended and impenetrable,
it will exercise an attractive force, and to require the acquiescence of others
who possess the concept of body.
Like all claims of a right or entitlement, this one requires a warrant, which
in the A-Deduction Kant locates in the affinity of appearances, that is, the
uniformity of nature. As he put it at one point,
If cinnabar were now red, now black, now light, now heavy, . . . then
my empirical imagination would never even get the opportunity to
think of heavy cinnabar on the occasion of the representation of
the color red; or if a certain word were attributed now to this thing,
now to that, or if one and the same thing were sometimes called this,
sometimes that, without the governance of a certain rule to which
appearances are already subjected in themselves, then no empirical
synthesis of reproduction could take place.
(A100–101)
Although Kant is referring to the conditions of the possibility of reproduc-

tion by the imagination rather than cognition, it is relevant to the latter,
16
because reproducibility is a necessary (though not a sufficient) condition of

forming a concept corresponding to the word “cinnabar”. And from this,
Kant eventually concludes that the uniformity of nature, under the descrip-
tion the “affinity of appearances”, is a necessary condition of the possibility
of experience. Save for the inclusion of the word “rule”, this line of reason-
ing echoes Hume’s and seems relatively unproblematic. Problems begin at
the next step, however, where Kant goes beyond Hume by claiming that,
[I]f this unity of association did not also have an objective ground,
so that it would be impossible for appearances to be apprehended
by the imagination otherwise than under the condition of a possible
synthetic unity of this apprehension, then it would be entirely contin-
gent whether appearances fit into a connection of human cognition.
(A121)
It is here that the objections of Harrison, Cassam, and others come into
play. Thus, it is granted that without what Kant calls an “objective ground”,
the uniformity of nature, that is, the conformity of appearances to a pos-
sible synthetic unity (of apperception), would be merely contingent and that,
lacking this uniformity, experience would not be possible. But it is insisted
that, though it is necessary for cognition that there be uniformity, there is
no need to assume that nature is necessarily uniform. Once again, Kant is
charged with conflating a merely conditional or de dicto with an uncondi-
tional or de re necessity.7
I shall address this issue by taking the liberty of putting words into Kant’s
mouth through a hypothetical response to a question posed by Hume, which
I believe gets to the heart of the matter. In beginning his skeptical reflections
on the operations of the understanding in the Enquiry, Hume notes that, “It
may . . . be a subject worthy of curiosity to enquire what is the nature of that
evidence, which assures us of any real existence and matter of fact beyond
the present testimony of our senses, or the records of our memory” (1999:
108). For Hume, the question was merely rhetorical, because he denied that
there was any such “evidence”. Kant agreed with him on this point, but
he also realized that it pointed to a genuine problem, namely the need for
a ground or warrant to justify projections beyond what is being presently
perceived or recollected. Moreover, the reason for this is not, as Cassam sug-
gests, that this is a quasi-psychological matter providing a sense of security
about the future course of things; it is rather a normative issue regarding the
objective validity of one’s present “experience”. Expressed in the terms used
in the Prolegomena, it is a matter of whether one’s present “cognition”, for
example, that the sun warms the stone, is to be considered a judgment of
experience possessing objective validity or merely a judgment of perception.8
The reason for this lies in the nature of concepts (including empirical con-
cepts) as rules. I have claimed that, as rules, concepts possess a normative
17
force, which can be understood as involving a right to expect both that any
object falling under the concept will have certain properties and that other
cognizers who possess the concept would concur with one’s judgment. Oth-
erwise expressed, concepts possess both an objective and a subjective uni-
versality, which amounts to a two-fold projectibility: to future experiences
and to other cognizers. Consider the empirical concept of gold, understood
as a yellowish, malleable metal, soluble in aqua regia. When I predicate these
properties of some object = x, I am not saying merely that it seems to be
gold to me or that it is gold at the present time but, for all I know, may not
remain such in the future. Rather, I am assuming that it was gold yesterday,
even though I did not then perceive it, that it will be gold tomorrow, and that
all others possessing the concept of gold ought to judge likewise. Of course,
I may be mistaken, as is evidenced by the phenomenon of so-called “fool’s
gold”. But that is beside the point, for the issue concerns the grounds of the
projectibility of legitimate and correctly applied concepts. Moreover, this
explains why Kant would reject the claim that a contingent conformity of
the world to the requirements of the understanding is not only the best that
we can hope for but all that is required to account for the possibility of expe-
rience. Simply put, since the use of a concept in a judgment presupposes its
projectibility and this presupposes the uniformity of nature, the latter must
be assumed to be more than contingent, though less than logically necessary,
since, as Hume pointed out, its denial does not yield a contradiction.
3. Response to Allais
Having argued against the view that Kant was in error in endeavoring to
argue for more than the conditional necessity that appearances conform to
the requirements of the understanding, if experience is to be possible, and
that this led to the introduction of transcendental idealism to account for
a stronger form of necessity, I shall now consider the relation between the
actual argument of the Transcendental Deduction and transcendental ideal-
ism. But rather than dealing with the matter directly by spelling out my own
view of this idealism, I shall approach it more obliquely through a consider-
ation of the approach to the topic by Lucy Allais in her recent book, which
contains both a critique of my interpretation of transcendental idealism and
a somewhat different account of its relation to the Deduction.
Allais provides what she terms a “moderate metaphysical interpretation”
of transcendental idealism. She claims that what makes her interpretation
metaphysical is that it maintains that the things of which we which have
knowledge (phenomena), “have a way they are in themselves that is not
cognizable by us, and that the appearances of these are genuinely mind-
dependent, while not existing merely in the mind” (Allais 2015: 9). What
makes this interpretation “moderate” is that it is only in one of the ways in
which things are that they are mind dependent. Allais contrasts this with two
18
“extreme” metaphysical interpretations, which classify Kant as either a nou-

menalist or a phenomenalist, and a “deflationary” interpretation. By “nou-
menalism”, she understands the familiar two-world view, which holds that
Kant affirms the existence of a domain of intelligible, non-spatio-temporal
entities that is ontologically distinct from the mind-dependent phenomenal
world with which alone we are acquainted (Allais 2015: 9). By “phenom-
enalism”, she understands the view that “empirically real objects are men-
tal entities or constructions out of mental entities, mental states, or mental
activities” (Allais 2015: 9). And by “deflationary” interpretations, she has in
mind mainly, though not exclusively, my view, which she evidently labels as
such because it is not metaphysical in the sense that it understands the dis-
tinction between appearances and things in themselves as between two ways
in which things can be considered rather than as between two ways in which
they are, namely as mind-dependent entities with respect to their “manifest”,
that is, spatiotemporal properties and as mind-independent with respect to
their intrinsic properties, about which we have no knowledge.
Despite this difference, our interpretations share significant common
ground, which makes a consideration of Allais’s view particularly useful.
Specifically, I am in basic agreement with her on three fundamental points
beyond her rejection of noumenalism and phenomenalism. The first is her
analysis of Kant’s core argument for idealism in the Aesthetic, which, follow-
ing the brief account in the Prolegomena (4: 282), she takes as an inference
from the concept of an a priori intuition to its mind-dependence, combined
with the claim that the representations of space and time have this status,
which is established in the Metaphysical and Transcendental Expositions
in the Critique (Allais 2015: 194–6). The second is her view that the target
of the Deduction is not, as is frequently assumed, a Cartesian skepticism
regarding the external world or the objectivity of experience but an empiri-
cism that assumes that we have experience in the sense of objectively valid
cognitions about the empirical world but denies that this requires any a
priori component. The third, which follows from this, is that the central
concern of the Deduction is to demonstrate that empirical concepts, which
we are assumed to possess, presuppose pure ones, which are in doubt.
My main disagreement concerns her view of the relationship of the Deduc-
tion to transcendental idealism. Even here, however, the disagreement is not
complete, since we both affirm a connection between them. Rather, it is a
matter of degree, for whereas I maintain that the Deduction is inseparable
from this idealism, she argues for a somewhat looser connection, which
supposedly leaves room for a significant realist dimension. Allais provides
a clear statement of her position in a footnote in her introductory chapter,
where she contrasts it with the interpretive stances of both those like Straw-
son and his followers, who “think that everything that is of value in the
Critique can be entirely separated off from Kant’s transcendental idealism”,
and those, presumably including myself, who “think that seeing any aspect
19
of Kant’s arguments as not dependent on, leading to, otherwise intimately

embedded in his idealism is to fail to take transcendental idealism seriously
and to miss the coherence of Kant’s work”. In opposition to both views, she
writes,
I think that we should take transcendental idealism seriously, and it

is clearly the key to Kant’s arguments and positions in the Critique,
but it is compatible with this that Kant may have had some argu-
ments and epistemological insights which are separable from his
idealism.
(Allais 2015: 11)
Accordingly, the question becomes exactly what arguments and epistemo-

logical insights Allais has in mind when she denies that any aspect of them
is connected to Kant’s idealism. As far as I can see, she is thinking primar-
ily of two things. The first is that the central argument of the Deduction
is epistemological rather than metaphysical, by which she evidently means
that it is concerned with what is required to have a determinate object of
thought rather than for such an object to exist (Allais 2015: 285). Although
this is certainly true, it is not clear to me why this precludes any connection
with transcendental idealism, unless she has in mind merely logical consid-
erations, or this idealism is understood in crude impositional terms, wherein
objects are generated by imposing unity on chaotic sensations, which is obvi-
ously not how she understands it. To be sure, Kant claims to have shown
that, as conditions of the possibility of experience, the categories are also
conditions of the possibility of objects of experience (A111), but it is clear
that this refers to the cognition of these objects, not to their very existence.
Allais’s second point that supposedly mitigates the connection between
transcendental idealism and the Deduction is the claim that the latter deals
with a problem that would be recognized by a realist, which she describes in
one place as “what it takes for us to be in a position to think of anything as
an object”, to which she adds that “giving an account of the requirements
of referential thought can also be taken to be a problem by a realist” (Allais
2015: 285). By the first part of this claim, I assume that Allais has in mind
that an object must in some manner be given in order to be thought as an
object, which certainly is the case for Kant, if “thought” is taken as equiva-
lent to “cognized” rather than simply imagined or postulated, and I have no
doubt that most realists would accept this. But the realist (certainly not the
transcendental realist) cannot be expected to endorse Kant’s account of the
manner of its givenness, since this involves an appeal to a priori forms of
sensibility. As for her second point, everything turns on what is understood
by “the requirements of referential thought”. If it is taken, say, as a problem
in the philosophy of language, it could well be acknowledged by a realist,
but it would have little bearing on the Transcendental Deduction. And if, as I
20
suspect, it is to be understood as a problem in normative epistemology, then

the realist would certainly be interested in the problem broadly construed
but not in the manner in which it is posed by Kant, which concerns the rela-
tion between two cognitive faculties, rather than between the unification of
representations in the mind and an an sich reality.
Given this, together with her thoughtful defense of transcendental ideal-
ism, one might wonder why Allais is so concerned with loosening its con-
nection to the Deduction rather than simply pointing out against critics
like Strawson, Harrison, and Cassam that it is not in any way connected
with transcendental idealism as they interpret it. Judging from her overall
account, it seems to be largely a consequence of her antipathy to so-called
conceptualist readings of Kant, which for better or worse assume the insepa-
rability of the Deduction from this idealism, because they view his idealism
as fundamentally conceptual in nature, that is, as consisting of the role of
the categories in determining what counts as an object. Like many recent
commentators, Allais attaches great importance to the distinction between
conceptualist and non-conceptualist readings of Kant. Following Allais’s
account, the conceptualist denies that intuitions present objects in the sense
of distinct particulars to consciousness apart from the synthesizing, con-
structive activity of the understanding, whereas the non-conceptualist “takes
seriously and is informed by [Kant’s] claim that intuitions are singular and
immediate representations that give us objects, and that this is something
that concepts could never do” (Allais 2015: 150).
Although this is correct, it relocates rather than lessens the dependence
of the argument of the Deduction on transcendental idealism. Consider the
proof-structure of the B-Deduction. As I have argued at length, its first part
(§§15–20) considers the categories a priori rules for discursive thought as
such, assuming that something is sensibly given but abstracting from the
manner in which it is given, whereas it is only in the second part (§§21–27)
that Kant achieves the goal of the Deduction, which is to establish the valid-
ity of the categories for objects qua given under our a priori forms of sen-
sibility (Allison 2015: 327–432). In this interpretation, the argument of the
first part is purely conceptualist, which is why it is of itself incapable of fully
establishing the categories. And though she does not refer specifically to the
argument of the B-Deduction in this context, I suspect that this is at least
part of what Allais had in mind when, evidently referring to the views of
Strawson and Cassam, she remarks that,
A serious objection to making the argument of the Deduction

depend on idealism with respect to the categories [which is argu-
ably what a conceptualist reading does] is that this would establish
the conclusion of the argument too easily, and render the result
unsatisfactory.
(2015: 287)
21
It is unsatisfactory because, as an analytic argument, concerned solely with

the conditions of the discursive thought of an object, a purely conceptual form of
the Deduction of the categories, which is what the first part of the B-Deduction
is, can consider only the concept of an object, not actual objects as they are
given under our forms of sensibility. Moreover, this points to two significant
problems for an overall conceptualist reading of the B-Deduction. (1) Any ide-
alism that it is thought to involve would be solely conceptual in nature, based
on the constitutive function of the categories and having nothing to do with
sensibility. (2) It renders the second part of the Deduction either redundant or
utterly mysterious, which is perhaps why it is often either completely ignored
or treated as an afterthought in the literature. Accordingly, a serious consid-
eration of the B-Deduction as a whole leads to a non-conceptualist reading.9
This is not to suggest that it is only the B-Deduction that rests on a non-
conceptualist premise, as if Kant were a conceptualist in 1781 and a non-
conceptualist in 1787. In fact, we have already seen that in a passage in
the introductory section contained in both versions of the Deduction, Kant
raised the possibility, which I have characterized as a specter, that “Appear-
ances could . . . be so constituted that the understanding would not find
them in accord with the conditions of its unity”, as a question with which
the Deduction must deal. This assumes that appearances, considered objects
qua given in intuition, have a constitution that is independent of any that
might be imposed on them by the understanding, which is presumably the
defining feature of non-conceptualism as understood by Allais. But though
the non-conceptual element is present in the A-Deduction, its role is more
perspicaciously presented in the B-version, specifically in its second part.
More importantly, this entails the inseparability of the Deduction as a
whole from transcendental idealism. If the latter be doubted, consider what
a non-conceptual position would be like without this idealism. Since non-
conceptualism claims that things are given as objects in intuition apart from
the spontaneous activity of the understanding, it would mean that they are
given as they are in themselves. But in that case, what is the role of the under-
standing? It could consist either of logical operations or clarification, that is,
rendering what is initially (as given) obscure and indistinct clear and distinct,
which is precisely the Leibnizian view. In other words, as is demonstrated
by the development of Kant’s thought during the “Silent Decade”, without
the ideality or mind-dependence of the sensibly given, not only would the
project of the Transcendental Deduction be inconceivable, but the discursiv-
ity thesis, as Kant understood it, would be rejected (Allison 2015: 85–142).
Although it does not deal with the question of the success of the Tran-
scendental Deduction, if the preceding analysis is correct, I believe that it
establishes three points that are essential to its interpretation. (1) Since the
Deduction is concerned with the a priori conditions of empirical cognitions
and, as was argued, this involves a normative necessity, the conformity of
appearances to these conditions cannot be considered as a merely contingent
22
matter. (2) Since it contains an essential non-conceptual component, the

argument of the Deduction cannot be understood in purely conceptualist
terms. (3) Despite the endeavor of many critics to separate them, in view
of its non-conceptual component, the argument of the Deduction is insepa-
rable from transcendental idealism, which, together with the fact that its
target is the empiricist rather than the external world skeptic, distinguishes
it decisively from most present-day transcendental arguments. And to under-
score this point, I shall end this chapter as I ended my commentary on the
Deduction by shamelessly mimicking Kierkegaard, who in his response to
Hegelian efforts to present a reasonable Christianity lacking in paradox and
absurdity, chose as the motto for his treatment of the topic in Philosophi-
cal Fragments the line from Shakespeare’s Twelfth Night: “Better well hung
than ill wed” (Kierkegaard 1962: 2).
Notes
1 The chapter was initially published in 2016 by the European Journal of Philoso-
phy (24 (4): 920–33). I wish to thank the editors for their kind permission to
reprint it.
2 Among professed Kantians, the best-known proponent of this line of interpreta-
tion is Paul Guyer. But since I have discussed his views on several occasions, I shall
focus here on the accounts of Harrison and Cassam, whose views are more closely
related to Strawson’s. For my most recent response to Guyer on this issue, see
Allison (2015: 444–53).
3 The citation is from Strawson (1966: 100).
4 I initially introduced the specter analogy in Allison (2004) and developed it further
in Allison (2015).
5 This was raised as a worry by Schulz in his favorable review of Ulrich’s critique
of Kant. I discuss it in Allison (2015: 309 and 441). In recent times, it has been
expressed by Guyer, who attributes to Kant the “shocking” view that “empirical
judgments are forms of necessary truth” (1987: 114).
6 This is a central thesis for which I argue at length and in several contexts in Allison
(2015).
7 This line of objection has been frequently pressed by Guyer, for example, in Guyer
(1987: 54–61, 122–4, 132, 140, 142, 144, 363–9, 379–83).
8 I analyze this distinction in Allison (2015: 362–3).
9 This is not to suggest that Kant’s position on this issue in the second part of the
B-Deduction is completely transparent. Two contentious aspects of the text are the
much-discussed note to B160 and account of perception as subject to the catego-
ries. For my discussion of these topics, see Allison (2015: 406–30).
References
Allais, L. 2015. Manifest Reality, Kant’s Idealism & His Realism. Oxford: Oxford
University Press.
Allison, H. 2004. Kant’s Transcendental Idealism. New Haven: Yale University Press.
Allison, H. 2015. Kant’s Transcendental Deduction. Oxford: Oxford University
Press.
23
Cassam, Q. 1987. “Transcendental Arguments, Transcendental Synthesis and Tran-

scendental Idealism”, The Philosophical Quarterly, 17: 355–78.
Guyer, P. 1987. Kant and the Claims of Knowledge. Cambridge: Cambridge Univer-
sity Press.
Harrison, R. 1982. “Transcendental Arguments and Idealism”, in G. Vessey (ed.)
Idealism Past and Present. Cambridge: Cambridge University Press: 211–24.
Hume, D. 1999. An Enquiry Concerning Human Understanding, edited by L. Tom
Beauchamp. Oxford: Oxford University Press.
Kierkegaard, S. 1962. Philosophical Fragments, translated by David F. Swenson and
revised by V. Howard Hong. Princeton, NJ: Princeton University Press.
Stern, R. 2000. Transcendental Arguments and Scepticism. Oxford: Oxford Univer-
sity Press.
Strawson, P. F. 1966. The Bounds of Sense: An Essay on Kant’s Critique of Pure
Reason. London: Methuen.
24
2
KANT ON SCIENTIFIC LAWS
Lea Ypi
This chapter explores Kant’s analysis of systematic unity in relation to the

question of whether there are any laws of nature and, if so, whether they can
be known by us. The chapter begins by noting how Kant’s theory of system-
atic unity has inspired different accounts of the laws of nature and explores
the relation between a necessitarian account of laws and a reflection-based
one. Focusing in particular on the Critique of Pure Reason, I suggest that
Kant’s analysis of systematic unity relies on a principle of purposiveness
which is not based on the analogy with reason’s practical causality but on
an assumption of natural purposiveness indebted to Kant’s earlier theory of
germs and dispositions. The difficulty of reconciling his pre-critical views on
natural purposiveness with the analysis of scientific laws put forward in the
first Critique is, I suggest, what leads Kant to abandon the proto-necessitarian
theory in favor of an alternative, reflection-based, account.
1. Kant’s contribution to the debate on the laws of nature

In recent years, there has been a surge of interest in Kant’s arguments on the
systematic unity of knowledge to vindicate an account of scientific explana-
tion centered on the unifying power of laws (for a recent state of the art, see
Massimi and Breitenbach 2017). At the heart of this enterprise is an attempt
to reflect on the importance of systematicity for understanding the char-
acter of laws as regularities reduced to a limited set of patterns balancing
explanatory strength and simplicity. While traditionally associated to a rigid
defense of a priori statements in tension with the fallibilistic commitments of
modern scientific theories, more charitable readings have paved the way to a
different interpretation of Kant’s reflections on scientific method (see Gava
and Stern 2015) Kant’s argument about systematic unity has been invoked
in different contemporary accounts of the laws of nature.
These can be divided into four groups of theories. First there is an account
of laws of nature understood as “best systems” for the classification of regu-
lar empirical occurrences where laws are seen as regularities in a system of
empirical generalization subject to a requirement of specificity of intervening
DOI: 10.4324/9781003050742-3 25
LEA YPI
connections (Kitcher 1986). In this “best systems” version, a scientific theory

is understood as a “a projected order of nature”, whereby methodological
directives such as the directive to obtain “systematic unity in our knowl-
edge” act as heuristic devices enabling us to evaluate scientific proposals
even before these are tested (Kitcher 1992: 213). Then there is an account of
laws of nature understood as “derivations” from a priori laws. In this second
family of interpretations, only those particular empirical generalizations that
can be connected to a priori laws have necessary status and therefore count
as genuine laws. (Buchdahl 1992; Friedman 1992). Third, Kant’s theory is
often invoked in the context of a “necessitarian” account of laws whereby
the necessity of the laws of nature is grounded in the “essential nature” of
things and independently of human cognition (Kreines 2009, 2017; Messina
2017). Finally, there is a “reflection” account, according to which the neces-
sary character of laws is derived from a priori principles combined with an
ideal of unity that orients our efforts of comparing and contrasting particu-
lars so as to arrive at lawlike generalizations (Breitenbach 2018).
Common to all these interpretations is an attempt to capture adequate
relations of dependency among natural phenomena while vindicating the
distinctiveness and necessary character of laws vis-à-vis contingent regu-
larities (for a view that objects to this framework, see Watkins 2014). The
difficulty, for all of them, is how to account for the relation between two
interconnected claims that we find throughout Kant’s contribution to the
question of unification of the laws of nature. The first is an epistemic claim
about the heuristic role of systematic unity for the purpose of constructing
stringent generalizations with causal explanatory power and that satisfy cri-
teria of necessity. The second claim is a more controversial, ontological claim
about the character of the regularities that laws are supposed to capture
and whether such regularities can be seen as inherent to nature. For, as one
author puts it, “how could we be justified in adopting explanatory patterns
from considerations of unification, if unity of causes, laws and powers was
not intrinsic to nature” (Kitcher 1994: 262). Without a further ontological
commitment, the logical demand would seem driven by “an obsessive phan-
tasy that might in no way match the order of natural phenomena” (Kitcher
1994: 262).
While some authors consider the ontological claim unnecessary to the
epistemic one (Guyer 2005), others take seriously its implications for a Kan-
tian contribution to our understanding of the necessity of laws (Kreines
2009). For how can one account for the necessity of the laws of nature while
being agnostic or even denying that nature really does exhibit the neces-
sary regularities that our explanations capture? “An explanation”, we are
told, “must provide information about an underlying condition on which an
explanandum really depends” (Kreines 2009: 531). Yet the problem, from a
widely shared Kantian perspective, is that while we can postulate the neces-
sity of the regularities of nature, we cannot know that the regularities we
26
K A N T O N S C I E N T I F I C L AW S
postulate are necessarily laws of nature, regardless of our postulating them

as such. The identification of patterns of regularity depends on our own
methodological directives, not on the way things are. We are therefore left
with a dilemma: either the laws of nature are not laws in a strict sense after
all (i.e. we don’t know if they capture necessary relations in nature), or we
grant that there are laws of nature but that our a priori principles are (at
best) only contingently related to them.
One way out of this dilemma is to develop a “reflection” account of the
laws of nature, grounded on a principle of purposiveness orienting our sci-
entific enquiry (Breitenbach 2018; Ginsborg 2017). Like the necessitation
account, the reflection account finds laws of nature by presupposing an
ideal of systematic unity based on natural purposiveness and which orients
empirical investigation. Like derivation accounts, that enquiry is also rooted
in a priori laws. And like best system theories, it is open to the rectification
of previous errors through new empirical observations: where lawlike gen-
eralizations sustain further scrutiny, they can be called laws. Knowledge of
the laws of nature is, in the reflection account, neither entirely inaccessible
(as with necessitation theories) nor entirely given to us (as with derivation
accounts) nor knowledge of laws with reduced modal force (as with best
system theories). It is progressively acquired with the help of the assumption
of the purposive unity of nature and evidence of our own observations.
One problem with the reflection account, however, is Kant’s insistence
that the principle of purposiveness at the basis of the systematic unity of
nature, while being necessary from a logical perspective (i.e. to orient our
empirical investigations), is also contingent from an ontological point of
view (Ginsborg 2017). In other words, while we can postulate the systematic
unity of nature according to purposes, we cannot be sure that nature really
is so arranged. But if we cannot assume that, how can the reflection account
vindicate the necessity of laws, as opposed to their contingent generality?
What really distinguishes this account from best-system theories? And how
can it address the objections about weak modality that best-system theories
usually attract?
Kant has two kinds of answers to the problem of what gives necessity to
the laws of nature. They are related but in tension with each other. Both try
to explain why the principle of purposiveness can help us discover necessary
rather than merely contingent generalities in nature. Where they differ is in
how they think about necessity. The more familiar answer, which is at the
heart of reflection accounts, directs us to Kant’s remarks on the empirical
laws of nature in the Critique of Judgement, where the transcendental status
of the principle of purposiveness is clarified in connection to the constitu-
tion of organic beings and the analogy with human intentional action. It is
reflection on the way reason sets and pursues practical purposes, we are told,
that legitimizes the use of purposive principles in our effort to systematize
the laws of nature.
27
LEA YPI
The other, less familiar, answer is the one I want to explore in this chapter.
It stems from an interpretation of Kant’s remarks on systematicity at the
very end of the first Critique, in the section devoted to the Architectonic of
Pure Reason. Here, too, Kant tries to explain why systematic unity is neces-
sary, how reason achieves it and in what relationship it stands to nature.
Here, too, the analogy with organic beings is vital. Here, too, the principle
of purposiveness plays a key role in explaining what confers systematicity
to laws and why some of the regularities we establish following empirical
observation are also necessary. Yet, as I try to show, Kant’s answers in the
first Critique vindicate a proto-necessitarian rather than a reflection account
of laws of nature.
Kant’s version of necessitarianism is different from contemporary neces-
sitarian accounts, however. As we shall shortly see, Kant’s reflections on
systematic unity imply that there is inherent purposiveness in nature, that
there are laws of nature that reflect necessary relations between kinds but
also (and this is where his theory differs from contemporary necessitarian
accounts) that they can in principle be accessed by us. The reason they can
be accessed by us is that the constitution of reason mirrors the constitution of
nature: the same principle of purposiveness guides the relation between the
whole and the parts in both. While in the reflection account, the systematic
unity of nature is a projection of the systematic unity of reason based on
the analogy with reason’s disposition to pose practical ends, in the proto-
necessitarian account that Kant holds prior to the third Critique, the sys-
tematic unity of reason and the systematic unity of nature are connected
through the use of the ideas of reason. This, as I try to show in what fol-
lows, is difficult to square with other core aspects of Kant’s critical work. It
also explains Kant’s later abandonment of necessitarianism in favor of the
reflection account.
2. The metaphor of reason as an organism

The distinction between aggregative and systemic knowledge is at the heart
of all rigorous scientific enquiry and crucial to an analysis of laws as uni-
versal regularities as opposed to merely contingent generalizations. Such a
distinction is central to the section of the Critique of Pure Reason devoted
to the Architectonic of Pure Reason, one of the lesser-known parts of Kant’s
wider discussion of philosophical method and the only one to explicitly
tackle the problem of the unity of the system (Ypi 2011a, 2011b; Manches-
ter 2008). An aggregate of knowledge, we are told here, is like a rhapsody
where accidental knowledge is amassed without a proper account of the
relation between the parts and of their distinctive function within the whole.
A system, on the other hand, is an arrangement where cognitions fit together
in an ordered way and where the parts and the whole mutually support each
other acting as a condition of reciprocal development (A 832, B 860).
28
Kant’s suggestion in the Architectonic of Pure Reason is that, in the case

of a system, the necessary relation between the whole and the parts can be
understood with reference to the purposive nature of our cognitive activities.
Reason, he claims, contains essential purposes, purposes which structure
its unifying efforts compatibly with an idea of the whole that stands in an
organic relation to the development of the parts. Such an idea of the whole
acts both as the basis for the development of the different parts of the system
while at the same time representing its only “supreme and internal end”
(A 833, B 861).
Kant’s analysis of systematic unification has two components. On the one
hand, systematization consists of the effort to structure the relation between
different parts of a system in such a way that new empirical findings can be
coherently incorporated in a body of knowledge already available. On the
other hand, Kant argues, that systematic unity must be presupposed if we
strive toward the unification of disparate principles. The latest part is often
interpreted as evidence of the dynamic character of methodological enquiry
(even of Kant’s affinities with a fallibilistic method of scientific enquiry)
(Gava 2015). However, Kant’s further argument on how exactly to conceive
this unifying effort and the particular remarks on purposiveness that we
find in the Architectonic are perplexing. We are told, first, that the idea of
a unitary system must be presupposed given the purposive nature of our
cognitive activities and that this presupposition grounds our efforts to find
a coherent relation between disparate principles and cognitions (A 832, B
860). This is a logical requirement of unification and a rather straightfor-
ward one at that. But we are also told, much more problematically, that
the idea of the whole of the system underpins the purposive structure of
reason and can be explained in analogy with the constitution of an organic
body (A 833, B 861). As Kant puts it, pure speculative reason contains “a
truly articulated structure of members in which each thing is an organ”
(B XXXVII–XXXVIII).
This means, Kant suggests, that “everything is for the sake of each mem-
ber, and each individual member is for the sake of all, so that the least frailty
whether it be a mistake (an error) or a lack, must inevitably betray itself in
its use” (B XXXVII–XXXVIII).
The metaphor of reason as an organic being is crucial to understand the
kind of methodological principle on the basis of which the Critique seeks
to bring the multiplicity of cognitions in one systematic whole. Systematic
unity is essential, as we can see, to reason’s ability to be internally self cor-
recting and for the adequate use of its more specific principles. As Kant also
explains in the Preface to the Prolegomena, since pure reason is an isolated
domain,
there is nothing outside of it that could correct our judgment within

it, the validity and use of each part depends on the relation in which
29
LEA YPI
it stands to the others within reason itself, and, as with the structure
of an organized body, the purpose of any member can be derived
only from the complete concept of the whole.
(Prol 4: 263)
Scientific knowledge for Kant requires incorporating the “form of a whole,

in so far as the conception determines a priori not only the limits of its con-
tent, but the place which each of its parts is to occupy”. The very definition
of a system as “the unity of various cognitions under one idea” (A832/B860)
entails a rational redescription of the idea of the whole as both anticipat-
ing a priori the possibility of arranging the parts as parts of one system and
analyzing the system as the concrete unity to which the parts tend. As Kant
emphasizes, the scientific “idea” of a system contains “the end and the form
of the whole which is in accordance with that end” (A832/B860). That dis-
tinction, in turn, relies on an “idea of the whole” with regard to which it is
possible to assume that the parts of the system belong together in dynamic
relation to each other.
The unity of reason is therefore grounded on the idea of a dynamic whole
which can contain a multiplicity of cognitions in their constant develop-
ment, a development whose possibility is in fact also presupposed. This way
of articulating the idea of the whole is reminiscent of the Leibnizian concep-
tion of monads. For Leibniz, the monad was conceived as a “primitive force”
or “originary activity” which also contained the conditions of possibility for
its own future development. By expanding and revisiting the Aristotelian
concept of entelecheia, Leibniz sought to show how this primitive force was
at the basis of his account of the development of organic bodies and their
distinction from inanimate objects.1
Interestingly, like Leibniz, Kant mentions the similarity with the activity
of organic beings to explain the characteristics of the idea of the whole on
which the architectonic-systematic structure of reason is grounded.2 The
whole, he argues, is in this case
articulated (articulatio) and not heaped together (coacervatio); it

can, to be sure, grow internally (per intus susceptionem) not exter-
nally (per appositionem), like an animal body, whose growth does
not add a limb but rather makes each limb stronger and fitter for its
end without any alteration of proportion.
(A 833, B 861)
The analogy with an animal body and the explicit reference to the devel-
opment of organic beings are very important to understand the principle
of unity on which we rely to distinguish between systematic and aggrega-
tive knowledge. Most Kant scholars refer to Kant’s Critique of Judgment
and to the principle of purposiveness in nature to explain how the analysis
30
of biological nature can help us understand the necessity of scientific laws

more generally (Breitenbach 2009, 2018). However, in the third Critique, the
purposive analysis of organic beings is oriented by the causality according
to purposes that reason displays in the practical domain. In the Critique of
Pure Reason, Kant makes no reference to an autonomous domain of practi-
cal reason and to its distinctive, practical causality in accordance to ends.
The contrast between the two is instructive. In the Critique of the Power
of Judgment, Kant argues, everything is reciprocally means and end (KU 5,
376). It is difficult to explain the development of organic beings by refer-
ence to mechanic causes, guaranteed by the application of the concepts of
the understanding to objects of experience. What one needs to presuppose
instead is a particular type of causality, “the idea of an effect as the very
condition of possibility of that effect” (KU 5, 367). Thus, organized beings
ought to be analyzed as “ends of nature” (Naturzwecke) (KU 5, 372–4).
However, Kant insists that the expression is also inadequate, since the source
of that type of cause is the capacity for reflective judgment.
In the third critique, Kant explains that the hypothesis of final causes can-
not be formulated by invoking the principles of the understanding (KU 5,
234). The concept of an end must, on the contrary, be explained in relation
to practical reason and the determination of the will in conformity with the
moral law.3 The causal relation established by the principles of the under-
standing, Kant argues, can only help us articulate the link between efficient
causes (nexus effectivus) which relates causes to effects but not the other
way around. Attempts to reflect on the causality of organic beings rely on
a nexus established by analogy with a concept of reason (ends), which, as
Kant says
if considered as a series, would carry with it descending as well as

ascending dependency, in which the thing which is on the one hand
designated as an effect nevertheless deserves, in ascent, the name of
a cause of the same thing of which it is the effect.
(KU 5, 372)
Such a nexus, he clarifies further, is established when: (1) the parts are only
possible due to their relation to the whole and (2) the parts are combined
into the whole by being reciprocally both cause and effect of their form.
Only then, Kant argues, can we say that the idea of the whole determines the
form and combination of all the parts as “a ground for the cognition of the
systematic unity of the form and the combination of all of the manifold that
is contained in the given material for someone who judges it” (KU 5, 372).
The principle that therefore helps to explain the development of organic
beings is the principle of the “internal purposiveness of nature” (innern
Zweckmäßigkeit der Natur) (KU 5, 376–8; 247–9), a principle that belongs
to the capacity for reflective judgment, that is transcendental and regulative,
31
LEA YPI
that differs from all other regularities established with the help of the under-
standing and that is more similar to the concepts of reason. This principle in
turn relies on an “idea”, the idea of the whole of nature as a regulated system
of ends (die Idee der gesammten Natur aus eines System nach der Regel der
Zwecke) (KU 5, 379). Yet this idea, Kant explains in the third Critique, does
not expand knowledge beyond the limits of experience, and we cannot infer
from it anything on the existence of a potential “architect” of such system.
The principle of purposiveness serves as a rule to the faculty of judgment in
reflecting on the multiplicity of the laws of nature where it can no longer be
assisted by the principles of the understanding.
Now, if we return to the Critique of Pure Reason in the light of Kant’s
remarks on purposive causes in the third Critique, we find that to analyze
the laws of nature from a systematic perspective also requires the idea of
purposive unity. We find, too, an explanation of purposive unity with refer-
ence to the analogy between a system and an organism where everything
is reciprocally a means and an end. But in the third Critique, we can only
ascribe purposes to organic beings if we rely on a principle of internal purpo-
siveness, which in turn is grounded on an idea of nature as a system of ends,
which is in turn based on the necessity of practical reason. Does this analogy
work to also explain the nature of the systematic principle at the basis of the
architectonic unity of reason in the first Critique? Should we also postulate
an internal conformity to ends within the system of pure reason?
On the one hand, this seems plausible and indeed compatible with Kant’s
suggestions that the rational scientific concept of a system contains the “end
and the form of the whole that is congruent with it”.4 On the other hand,
the way in which purposiveness is discussed in the first Critique is peculiar,
since here practical reason has no domain in which its principles are legisla-
tive and no necessitation of its own. In the first Critique, Kant cannot rely on
the practical causality of reason to explain why projecting to nature a con-
formity to purposes is necessary. Here, the role that the principle of reflective
judgment plays in the third Critique is played by the “ideas” of reason (Ver-
nunftbegriffe) which connect the concept of the whole required to conceive
of a unitary system with the concept of a purpose (Zweck), which explains
the assumption of a purposeful internal coordination of the parts within it.5
The unity of an end, Kant argues, “to which all parts are related and in the
idea of which they also relate to each other, allows the absence of any part
to be noticed in our knowledge of the rest” (A 832; B 860).
In the third Critique, then, the principle of purposiveness helps explain
how we can postulate the systematic unity of the laws of nature guided by
a postulate of purposive unity which helps construct ever more appropriate
sequences of generalizations. When we ask the question of where the “neces-
sity” of these generalizations come from, we can only go back to the analogy
with the constitution of our faculties, in particular the nature of practical
reason whose capacity to pose practical ends guides our postulate of the
32
systematic unity of the laws of nature. This vindicates a reflection account

of the laws of nature. But in the Critique of Pure Reason, Kant makes no
mention of the distinctive causality of reason in the practical domain. Yet
the analogy with organic beings persists and plays a major role in defending
the necessity of postulating systematic unity in scientific enquiry. In the next
section, I explain the different way in which purposiveness is articulated in
the first Critique, why it supports a proto-necessitarian account of laws and
why the systematic unity of reason mirrors the systematic unity of nature
rather than the other way around (as is the case with Kant’s later work).
3. Germs and dispositions

Nobody, Kant argues in the first Critique, “attempts to establish a science
without grounding it on an idea” (A 834, B 862). Although this argument
might at first appear like a dry formalistic and by now outdated remark in
favor of the methodological unity of sciences, its relevance becomes obvious
once we turn to its implications for the kind of enquiry that the Critique
of Pure Reason seeks to advance: systematic enquiry based on laws rather
than aggregative enquiry based on contingent generalizations. For Kant, phi-
losophy is not a discipline that operates in isolation from other sciences. It
is a rigorous science that ought to integrate the findings of all others and
complement them with a reflection on the proper use of their tools, their
respective methods of enquiry and their relation to our cognitive faculties
and practical attitudes. But philosophy can aspire at all this only because it
relies on reason alone, without appeal to external authorities and refining
its own methodological criteria in dialogue with other sciences. It is impos-
sible to prove the validity of its tools more geometrico, by presupposing
certain premises and definitions and by equating its method to that of phys-
ics or mathematics (as Descartes or Spinoza maintained). The legitimacy
and appropriate use of every principle depends on its organic relation with
others and on the ability of all of them to help understand the conditions of
possibility of experience in its complexity. Therefore, philosophy, like every
other science, needs to be conceived with reference not to “the description
given to it by its founder” but “rather in accordance with the idea, grounded
in reason itself, of the natural unity of the parts that have been brought
together” (A 834, B 862).
This presupposition of the unity of parts rests on a purposive relation of
the whole and the parts which has both a logical/heuristic and an ontologi-
cal component. Kant argues that the unity of scientific knowledge and the
purposive character of the idea of the whole that lies at its basis provides the
cornerstone for the “scientificity of knowledge in general”. The assumption
of systematic unity is an important logical presupposition required for build-
ing stringent generalizations. Yet to have genuine necessity, it must, for Kant,
also reflect the way things are, the way nature is intrinsically organized. To
33
LEA YPI
explain where this necessity comes from, we need to turn to the purposive
structure of reason and explain in what relation it stands to the purpo-
sive structure of nature.
To explain how reason and nature support each other, Kant invokes an
idea of purposiveness which helps us bring together reason’s multiple cogni-
tions and which confers a character of necessity to its systematic unifying
attempts. On the one hand, such an idea is introduced in the Architectonic
of Pure Reason as the principle of possibility of systematic experience, the
foundation of our logical assumptions for the unity of lawlike empirical gen-
eralizations. On the other hand, Kant also emphasizes that such an idea is
already given to us in the (schematic) representation of the whole of knowl-
edge before its complete execution. The relationship is here a functional
one: the emerging idea of the sum total of cognitions already presupposes
and determines the order of the parts in the present.6 This does not mean
that the correspondence is flawless; indeed, the full “definition of science”,
Kant argues, only rarely corresponds fully to its idea from the beginning. Yet
such an idea of the whole is already contained in reason “like a seed, all of
whose parts still lie very involuted and are hardly recognizable even under
microscopic observation” (A 834, B 862).
The clarification of the nature of ideas provided here appeals to an anal-
ogy between organism and system that is crucial to understanding why the
generalizations of reason reflect necessary relations in nature. Just like a sys-
tem is in previous pages of the Architectonic compared to a living being, so
the idea of unity that lies at its basis is compared to a “germ” in accordance
with which the organism develops. Systems, Kant clarifies here, appear to be
formed “like maggots” by a “generatio equivoca from the mere confluence
of aggregated concepts, garbled at first but complete in time, although they
all had their schema, as the original germ (ursprünglichen Keim) in the mere
self-development of reason” (A 834, B 862).
One might argue that Kant’s insistence on the capacity of reason for self-
correction suggests here a best-system or reflection theory of the develop-
ment of scientific knowledge. But such an interpretation would be difficult
to square with the use of the term “ursprünglichen Keim” (original germ)
to clarify the conditions of development of the idea of the system and to
explain why Kant insists that such an idea is always already present in us.
To explain the reference to germs and their origins requires going back to
the reflections on the nature of organisms that surround the Critique of Pure
Reason and on which Kant’s conception of systematic unity relies.
Kant’s reflections on these issues must be situated in the context of a
familiar 18th-century debate about the unity of the species and the modality
of transmission of character traits from one generation of organisms to the
next. To better understand their relevance to the idea of natural purposive-
ness, one should focus on the trilogy of essays in which the natural history
of human beings is explicitly discussed: Of the Different Races of Human
34
Beings (1775), Determination of the Concept of a Human Race (1785) and

On the Use of Teleological Principles in Philosophy (1788). In all these
essays, reflecting on the reasons for the origin of different races, Kant offers
a very similar explanation to the one quoted subsequently:
the grounds of a determinate unfolding which are lying in the nature

of an organic body (plant or animal) are called germs (Keime), if
this unfolding concerns particular parts; if however it concerns only
the size or the relation of parts to one another, then I call them
natural predispositions. . . . This care of Nature to equip her crea-
ture through hidden inner provisions for all kinds of future circum-
stances, so that it may preserve itself and be suited to the difference
of the climate or the soil, is admirable. . . . Chance or the universal
mechanical laws could not produce such agreements. Therefore, we
must consider such occasional unfoldings as preformed.
(VvRM 2, 89–90)
Kant’s remarks on germs and predispositions are here situated in the con-
text of a longstanding 18th-century scientific debate between defenders of
epigenetic theories of natural development and preformist accounts (see, for
different discussions, Huneman 2007). The former, revived in the 18th cen-
tury in the defense of “mechanistic” epigenesis, offered by G. L. L. Buffon’s
Histoire naturelle générale et particulière that appeared in 1749, explained
organic development by referring to the action of a moule intérieure, a kind
of vital force understood in analogy with Newtonian microforces, which
organized the interaction of the various molécules organiques of which
living matter was composed. Here, the generation of new organisms was
illustrated with reference to a capacity inherent in matter to transform itself
and generate new organic forms. Preformist accounts, on the other hand,
developed in a climate of skepticism about epigenetic theories. They gained
particular prominence in Germany through the work of Albrecht von Haller,
later also reinforced by the analysis and microscopic observations of the
Swiss naturalist Charles Bonnet. Both Haller and Bonnet refined existing
accounts of preformation, drawing on Malebranche and Leibniz by referring
to the existence of preformed germs, which were thought to be present in all
natural beings and which contained the seeds for their future development.
Germs therefore pre-existed the fully formed organism, not in the sense that
all the properties of a fully formed organism could be interpreted as already
developed in the germs but as seeds which required an ordering cause to
facilitate their growth. Preformist theorists like Haller and Bonnet could
thus respond to both biological and theological disputes about the relation
of God to living matter, reconciling the natural development of organic parts
with the defense of a purposeful intervention in the way natural forces with
innate teleological direction could organize and develop.
35
LEA YPI
As many interpreters have pointed out, Kant’s general position towards

epigenesis is complex (Zammito 2007). Although in the first Critique, he
went as far as calling the entire system “an epigenesis of pure reason” (A 3,
B 167), his commitment to a version of preformation is consistently clear in
his early writings, particularly in his essays on race. What makes this debate
particularly interesting for our purposes is Kant’s assumption of purposive-
ness as intrinsic to nature, not a projection of the practical causality of rea-
son. In the writings on race, Kant notes that germs and dispositions should
be understood as purposive conditions for the development of natural beings
that specify their capacity to adapt and survive in particular environmental
and atmospheric conditions. They represent innate structures, independent
of mechanical causes, which precede the empirical development of organisms
yet contain the seeds for their future growth and allow them to adapt in dif-
ferent environments (VvRM 2, 89–90). Faced with the usual difficulties of
invoking mechanical laws to analyze the unity of a species and explain how
the characteristics of a particular organism could be preserved and transmit-
ted to the next generation, a version of “preformist” theories is endorsed to
account for the evolution of human traits that are already contained in it as
germs and predispositions. Yet, while predispositions refer to certain condi-
tions of development with regard to the size and relation of parts (organs) in
a living being, germs are conditions for the development of new features. This
then allows Kant to explain the unity and the diversity of the human species
as well as its capacity to adapt to different external circumstances.
The implications of this theory are crucial for understanding Kant’s analysis
of the historical development of the human species and the role of reason in
this development (see also Ypi 2014 for more discussion of this point). Kant
explains how germs account for the different characteristics inherited by every
race within the same human species but also emphasizes how the influence
of a particular environment, the character of the soil or certain atmospheric
conditions establish differences in human traits. But the larger point that mat-
ters for our purposes is that, as Kant puts it, “the purposive character” of an
organism is the general reason (allgemeine Grund) from which we can infer
“a preparation that is originally placed in the nature of a creature with this
intent” and from which we conclude to innate germs (VvRM 2, 102–3).
The Critique of Pure Reason does not contain an explicit analysis of the
principle of purposiveness. What Kant has to say here is not different from his
views in the earlier anthropological writings on the concept of human race.
Kant’s biological theory of germs is essential to explain the status of the idea of
purposiveness in the first Critique: it is only with reference to natural purpo-
siveness that we can explain why the logical postulate of the systematic unity
of the laws of nature is grounded on an ontological assumption of necessary
relations between natural phenomena, as necessitarian theories emphasize.
But the kind of proto-necessitarian theory that Kant embraces in the first Cri-
tique is problematic. Kant’s theory of germs and predispositions seeks to solve
36
the difficult problem of the relationship between logical and ontological pur-
posiveness by introducing the idea of natural conditions for development able
to account for both immutability and change in the course of human history.
On the one hand, the postulate of purposive unity is logically necessary to
understand the development of scientific theories as the progressive reveal-
ing through philosophical reflection of specific patterns linked to each other
in a non-accidental way: this is also what the reflection account of the laws
of nature maintains. But Kant adds more; he adds that the unity of purpose
that reason displays is a reflection of a unity of purposes inherent to nature.
This is where his theory is much closer to necessitarian account of the laws.
But why assume that the contingent development of reason and the pur-
posive arrangement of nature are connected in this way? Even if we say
that the purposiveness of reason mirrors the purposiveness of nature, what
guarantee is there that the regularities reason identifies and the patterns that
nature exhibits overlap? What is the relation between the logical and the
ontological assumption of systematic unity of empirical knowledge?
4. The germ in the system

Kant tries to address this problem by assigning to germs a dual purposive
function, a function crucial to understand the passage from the laws of nature
to the patterns established by reason. Ideas of reason are to philosophy what
germs are to anthropology. The history of reason is nothing else than a series
of discovery of patterns of explanation that are necessarily connected to each
other if we ascribe nature an immanent end that provides unity to our under-
standing of natural phenomena. This immanent end is given in the purposive
character of the ideas of reason, which schematize the purposive order of
nature. But it is important to insist that there is a crucial difference in how the
reflection account and the necessitarian account of laws explain this develop-
ment. From the third Critique onwards, Kant’s analysis of the laws of nature
resembles the reflection account and is based on the discovery that order in
nature is not immanent but projected with reference to human reason and the
way it operates in the practical domain. In the first Critique, Kant’s endorse-
ment of the theory of germs and the analogy with ideas as subject to both
natural and historical development contributes to a different theory of the
laws of nature, one that is much closer to necessitarian theories.
Prior to the Critique of Judgment, and the identification of a specific prin-
ciple of judgment that grounds the purposive analysis of nature on reason’s
capacity to set practical ends, Kant’s answer to the problem of the laws of
nature relies on nature’s inherent purposiveness. The solution to the ques-
tion of the conditions of possibility of systematic knowledge is contained in
the theory of germs and dispositions and in the analogy with the purposive
development of an innate idea of the whole which contains at the same time
the condition of possibility of the development of the parts.
37
LEA YPI
Both in the Architectonic of Pure Reason and in his essays on history, Kant
emphasizes the difference between an aggregative and a systematic account
of knowledge and highlights the need for an organic connection of the whole
and the parts. In both cases, Kant also stresses how systematic unity can be
defended in accordance with an idea which is already contained almost like
a germ within human reason. Systems, Kant writes in the Architectonic,
often appear formed “like maggots” by a “generatio equivoca” which all had
their schema, “as the original germ (ursprünglichen Keim) in the mere self-
development of reason” (A 835, B 863). The presence of a such a germ is not
only observable if we turn our attention to particular systems and a given
body of scientific work but to the development of philosophical enquiry as
a complete discipline. It is not just that each body of work is articulated in
accordance with an idea, but all are rather in turn
purposively united with each other as members of a whole in a system

of human cognition, and allow an architectonic to all human knowl-
edge, which at the present time, since so much material has already
been collected or can be taken from the ruins of collapsed older edi-
fices, would not merely be possible but would not even be very difficult.
(A 835, B 863)
These passages confirm Kant’s interest in a conception of philosophy

articulated compatibly with the demands of the Architectonic of Pure Rea-
son and perceived as a culmination of the demand for systematicity and
classification in all the sciences. The requirement of systematicity cannot be
considered fully satisfied already from the outset but develops in accordance
with particular dispositions and follows the historical evolution of reason.
Reason does not always contain what it needs to achieve the unity of the
laws of nature but acquires it following a process of reflection and self-
explication of its own purposes in an autonomous endeavor to shape the his-
tory of the different sciences. Philosophy as the discipline that embodies the
highest form of systematic unity is here the “mere idea of a possible science”,
nowhere given in concreto, but which one seeks to approach in vari-

ous ways until the only footpath, much overgrown by sensibility, is
discovered and the hitherto unsuccessful ectype, so far as it has been
granted to humans, is made equal to the archetype.
(A 838, B 866)
Notice, however, that this mere idea of a possible science is not a chimera.
The efforts of those involved with particular sciences to find their place within
an ideal of complete knowledge are not risky attempts to achieve the kind of
metaphysical insight prohibited by the critique of reason and required by its
practical use. Whoever commits to doing philosophy, Kant argues also in the
38
Wiener Logik, erects his building “on the ruins of another” (V-Lo/Wiener 24,
799), yet no work has ever reached a stage in which it can be considered stable
in all of its parts. We do not know in advance what the common ground of all
the laws of nature and of philosophy as a system of sciences looks like. The
particular connection between different parts of the system must be identified
dynamically and cannot be considered given already at the outset.
The history of scientific discovery consists of a series of efforts on the basis
of which one can eventually come to see a path whose visibility is initially
hindered to human knowledge. Reason begins by following this path in a
way that is disoriented, without being prescribed any predefined or conclusive
purposes. It proceeds chaotically at the start. But its construction finds more
stability when the architectonic key to the systematic unity of all cognitions is
discovered in the form of a principle of conformity to purposes which helps
us explain the necessary connections between phenomena. In virtue of this
principle, a researcher can organize scientific cognitions in a systematic way
and see them as interconnected with each other while also allowing for plural-
ity (Breitenbach and Choi 2017). Kant explains this point by arguing that it is
too bad that it is first possible for us to glimpse the idea in a clearer
light and to outline a whole architectonically, in accordance with
the ends of reason, only after we have long collected relevant cogni-
tions haphazardly like building materials and through them techni-
cally with only a hint from an idea lying hidden within us.
(A 834; B 862)
Approaching the existent body of scientific knowledge from a historical

perspective that enables us to observe how reason learns from its past mis-
takes is essential. But a purely historical outlook, we learn in the Critique of
Pure Reason, is not enough. The process of identifying both reason’s present
limitations and its orientation to future purposes is assisted by a purposive
arrangement of nature which contains the necessary relations between the
different phenomena humans observe. That same purposive arrangement is
at the core of the analysis of germs and dispositions with reference to which
Kant explains the function of the ideas of reason. The strong, ontological
assumption of a purposive order in nature on which Kant’s theory is grounded
prior to the third Critique brings his account of systematic unity much closer
to necessitarian accounts. Once we discover the key to Kant’s architectonic,
lawlike regularities are both inherent in nature and accessible to us.
Conclusion
Kant’s contribution to contemporary debates on the systematic unity of
the laws of nature is less straightforward than contemporary adaptations
of Kant’s thought make it sound. This is because it is difficult to isolate
39
LEA YPI
Kant’s concern with the unity of empirical laws of nature from his more
general concern with the unity of reason and the justification of the method
of philosophy as a systematic discipline. For Kant, every particular body
of knowledge occupies a specific place in the history of philosophy. In the
course of the development of the human species, a body of knowledge that
was once considered coherent and plausible but became subsequently irrel-
evant is similar to an archeological repository from which one can learn
about the past so as to better plan for the future. Kant thinks of his own
critique as both a philosophical synthesis of the work of his predecessors
but also as a tool for the discovery of a clearer and more comprehensive
systematization of reason’s cognitions, starting from an appropriate account
of the relation of its different purposes and the justification of their con-
nection to the unity of reason. This, he claims, can be achieved, because of
the emergence “in a clearer light” of an idea that has guided reason at first
haphazardly and almost in a hidden way and now finally discovered as the
basis for its architectonic system.
Yet, as we have already emphasized, when taken seriously, this account of
the architectonic unity of knowledge also runs into peculiar difficulties. The
historicization of the idea of systematic unity at the basis of the entire philo-
sophical system requires overcoming a conception of the history of science
as mere cognitio ex datis (A 836, B 864) and opens up to a philosophical
account of it on the basis of an idea for a purposeful history of knowledge.
But how can the postulate of coherence and classification under higher and
higher principles remain a subjective postulate without any claim to objec-
tivity in the phenomenal world? Is such an idea a mere illusion? And if not,
why not? Where does its necessity come from?
Kant’s answer here, as we have seen, is to invoke the concept of purpo-
siveness to explain how the idea of systematic unity is both subjective and
necessary. On the one hand, the aspiration of systematic unity is a merely
subjective ambition, an intellectual concern. Yet Kant also invites us to think
of such an idea in analogy with an original “germ” always present in reason
but that can only grow and develop further in appropriate circumstances of
enquiry, sensitive to distinctive methodological constraints. This implies that
the possibility of the harmonization of particular systems into a whole that
is purposefully oriented is in some ways always contained in the nature of
reason, even if the full development of such an idea depends on contingent
scientific discoveries. Just as, in the case of organic beings, environmental
conditions are crucial to the full development of innate germs and disposi-
tions, in the case of the idea of reason, historical circumstances play an
analogous role. The evolution of the system is therefore a dynamic process
through which reason disciplines and limits itself, seeking order and creating
unity as it discards unnecessary elements found along the way. Each stage
in the development of this process is a step towards the process of reaching
further maturity.
40
But we should be cautious in reading Kant’s methodological stance here

a pragmatic-fallibilist one. The different elements that contribute to the
unity of the system are not merely connected to each other but also to the idea
of the whole that is at the basis of their development, an idea that Kant
insists uncovers a unity that is already there. It is as if this idea of the whole
always preceded the coming together of the unity of the parts; the whole is
thought of as in some ways a determining cause of the parts. This account
is much closer to Leibniz’s proto-necessitarian account: the laws of nature
are already there and already valid; human reason merely uncovers them,
notwithstanding the mistakes made along the way. The principle of general
purposefulness and order of nature that is at the basis of such an idea looks
like a critical adaptation of the concept of unitary laws central to the old
metaphysical tradition, whereby conformity to ends is an inherent charac-
teristic of nature. Although Kant’s explanation of the regulative role of ideas
in the Appendix to the Transcendental Dialectic is supposed to have shown
the implausibility of such interpretations, the return to the concept of germs
and their status within Kant’s philosophy of biology, as well as his remarks
about the analogy between reason and organic beings, suggest that by the
end of the first Critique, the possibility of jumping from the logical use of
ideas to an ontological view of nature as containing a purposeful arrange-
ment of empirical laws cannot be discarded. In the Critique of Pure Reason,
Kant proceeds much less carefully in reflecting on the link between the logi-
cal principle of systematic unity of experience and the ontological analysis of
nature as inherently purposefully oriented. This latter assumption is fully dis-
carded only in later works, especially in the Critique of Judgment. But to see
how and why the demand for unity grounded in the heuristic role of ideas
progressively makes way to the concept of reflexive judgment as satisfying
the same demand for systematic unity, we need to show how reason oper-
ates autonomously from the constraints of nature in the practical domain.
It is only when Kant eventually succeeds in finding an autonomous domain
where practical reason is legislative, and where the demand for unity in sci-
ence is justified in the light of the demand for unity in moral action, that one
is able to understand how the principle of conformity to purposes is both
necessary and contingent. And it is only at that point that Kant abandons
the proto-necessitarianism of his initial position for a reflection-based stance
influenced by his theory of reason in practice.
Notes
1 On the relation between Leibniz’s conception of monad and Kant’s (and Hegel’s)
account of purposiveness, see Chiereghin (1990).
2 For two relevant discussions of the relation between reason and the activity of
organic beings, see Mensch (2013) and Ferrarin (2015).
3 Indeed, in the first version of the introduction to the third Critique, Kant uses the
term Zweckmäßigkeit to explain both the specific principle of purposiveness at
41
LEA YPI
the basis of reflective judgment and the concept of practical conformity to ends
explaining how the latter is a kind of “Zweckmäßigkeit, die zugleich Gesetz ist” a
form of purposiveness that is at the same time also law (KU 5, 245). The fact that
Kant uses the same term for both suggests that it is hard to distinguish the concept
of conformity to ends from the practical domain. Indeed, it is only in the second
introduction to the third Critique that Kant distinguishes the term Zweckmäßig-
keit, reserved to the capacity for reflective judgment from the Gesetzmäßigkeit of
the understanding and the Endzweck of reason. See on the use of these terms in
the first and second introductions: Menegoni (1988).
4 “Der scientifische Vernunftbegriff enthält also den Zweck und die Form des Gan-
zen, das mit demselben congruirt”.
5 See on the importance of the idea of the whole for the Kantian system and for
analysis of the issue also in the Architectonic of pure reason: Driesch (1924).
6 See for more on this functional explanation of Kantian teleology: Breitenbach
(2009).
References
Breitenbach, A. 2009. “Teleology in Biology: A Kantian Perspective”, Kant Year-
book, 1(1): 31–56.
Breitenbach, A. 2018. “Laws and Ideal Unity”, in W. R. Ott and L. Patton (eds.) Laws
of Nature. Oxford: Oxford University Press.
Breitenbach, A. and Choi, Y. 2017. “Pluralism and the Unity of Science”, The Monist,
100(3): 391–405.
Buchdahl, G. 1992. Kant and the Dynamics of Reason: Essays on the Structure of
Kant’s Philosophy. Oxford: Blackwell.
Chiereghin, F. 1990. “Finalità e idea della vita. La recezione hegeliana della filosofia
di Kant”, Verifiche, 19: 127–9.
Driesch, D. 1924. “Kant und das Ganze”, Kant Studien, 29: 365–76.
Ferrarin, A. 2015. The Powers of Pure Reason: Kant and the Idea of Cosmic Philoso-
phy. Chicago: University of Chicago Press.
Friedman, M. 1992. Kant and the Exact Sciences. Cambridge, MA: Harvard Uni-
versity Press.
Gava, G. 2015. “The Fallibilism of Kant’s Architectonic”, in G. Gava and R. Stern
(eds.) Pragmatism, Kant and Transcendental Philosophy. London: Routledge:
46–66.
Gava, G. and Stern, R. (eds.). 2015. Pragmatism, Kant and Transcendental Philoso-
phy. London: Routledge.
Ginsborg, H. 2017. “Why Must We Presuppose the Normativity of Nature”, in A.
Breitenbach and M. Massimi (eds.) Kant and the Laws of Nature. Cambridge:
Cambridge University Press: 71–88.
Guyer, P. 2005. Kant’s System of Nature and Freedom: Selected Essays. Oxford:
Oxford University Press.
Huneman, P. (ed.). 2007. Understanding Purpose: Kant and the Philosophy of Biol-
ogy. Rochester: University of Rochester Press.
Kitcher, P. 1986. “Projecting the Order of Nature”, in R. E. Butts (ed.) Kant’s Phi-
losophy of Physical Science: Die Metaphysische Anfangsgruende der Naturwis-
senschaft, 1786–1986. Dordrecht: Reidel Publishing: 201–38.
42
Kitcher, P. 1992. “Kant’s Philosophy of Science”, in A. Wood (ed.) Self and Nature in
Kant’s Philosophy. Ithaca: Cornell University Press: 185–215.
Kitcher, P. 1994. “The Unity of Science and the Unity of Nature”, in P. Parrini (ed.)
Kant and Contemporary Epistemology. Dordrecht: Kluwer Academic: 253–272.
Kreines, J. 2009. “Kant on the Laws of Nature: Laws, Necessitation, and the Limita-
tion of Our Knowledge”, European Journal of Philosophy, 17(4): 527–558.
Kreines, J. 2017. “Kant on the Laws of Nature: Restrictive Inflationism and Its Philo-
sophical Advantages”, The Monist, 100(3): 326–341.
Manchester, P. 2008. “Kant’s Conception of Architectonic in Its Philosophical Con-
text”, Kant-Studien, 99(2): 133–51.
Massimi, M. and Breitenbach, A. 2017. Kant and the Laws of Nature. Cambridge:
Cambridge University Press.
Menegoni, F. 1988. “Finalità e scopo finale nelle introduzioni alla kantiana ‘Critica
del Giudizio’”, Verifiche, 17: 327–51.
Mensch, J. 2013. Kant’s Organicism: Epigenesis and the Development of Critical
Philosophy. Chicago: University of Chicago Press.
Messina, J. 2017.“Kant’s Necessitation Account of Laws and the Nature of Natures”,
in M. Massimi and A. Breitenbach (eds.) Kant and the Laws of Nature. Cam-
bridge: Cambridge University Press: 131–49.
Watkins, E. 2014. “What Is, for Kant, a Law of Nature?”, Kant Studien, 105(4):
471–90.
Ypi, L. 2011a. “The Problem of Systematic Unity in Kant’s Two Definitions of Phi-
losophy”, in S. Bacin, A. Ferrarin, C. L. Rocca, and M. Ruffing (eds.) Proceedings
of the 11th Kant Congress. Berlin: De Gryter: 773–86.
Ypi, L. 2011b. “Teleology and System in Kant’s Architectonic of Pure Reason”, in
H. Williams, S. Baiasu, and S. Philstrom (eds.) Politics and Metaphysics in Kant.
Cardiff: University of Wales Press: 134–51.
Ypi, L. 2014. “Commerce and Colonialism in Kant’s Philosophy of History”, in K.
Flikschuh and L. Ypi (eds.) Kant and Colonialism: Historical and Contemporary
Perspectives. Oxford: Oxford University Press: 99–126.
Zammito, J. H. 2007. “Kant’s Persistent Ambivalence Towards Epigenesis, 1764–1790”,
in P. Huneman (ed.) Understanding Purpose: Kant and the Philosophy of Biology.
Rochester: University of Rochester Press: 51–74.
43
3
NORMATIVITY AND
REFLECTIVE CASUAL
INFERENCE
Patrícia Kauark-Leite1
The aim of this work is to discuss the role of reflective judgments in Kant’s
theory of epistemic normativity. Konstantin Pollok (2017), in his book Kant’s
Theory of Normativity, seems to restrict epistemic normativity merely to the
domain of the conceptual rules of the faculty of understanding. In contrast
to him, I would like to argue that, beyond the conceptual rules of the faculty
of understanding, the normative function of reflective inferences of the teleo-
logical power of judgment is also crucial not only for biology or history but
also for physics, in order to constitute our knowledge of the physical world
as an organic and unified system. This paper is divided into four parts. First,
I will argue in what sense theoretical judgments about physical nature can
be considered normative even though nature is the realm of what actually is,
and not the realm of what ought to be. Second, I will discuss the normativity
of the two kinds of causality: real and ideal. Third, I’ll examine the norma-
tive function of the concept of the purposiveness of nature when it is applied
to two different domains: that of organic beings and that of physical science.
Finally, I conclude by stressing the specific normative role of reflective causal
inference in physical theories.
1. The problem of the normativity of theoretical

judgments
Robert Brandom, in his seminal book Making It Explicit (1994), in the
chapter entitled “Toward a Normative Pragmatics”, recognizes that Kant’s
great fundamental insight is that human beings are essentially normative
creatures. According to him, “one of [Kant’s] cardinal innovations is his
introduction of the idea that conceptually structured activity is distinguished
by its normative character” (Brandom 1994: 8). In this sense, not only the
practical actions but also the theoretical actions are acts performed by ratio-
nal agents, who are, according to Brandom, responsible for them. What
44 DOI: 10.4324/9781003050742-4
N O R M AT I V I T Y A N D C A S U A L I N F E R E N C E
Brandom’s analysis of Kant’s fundamental insight shows is that the same

normative character is presented both in practical acts of doing and in the-
oretical acts of judging. In this sense, the Kantian discursive function of
the understanding and of the determining power of judgment is grasped by
Brandom as a linguistic practical activity, which exhibits a specific form of
normativity.
However, the fundamental problem with this kind of approach is how
to identify this normative character in theoretical judgments. Kant himself
acknowledges that moral judgments are sharply distinct from judgments
about physical nature. The former are about what morally one ought to do,
while the latter are about what is or exists. According to Kant:
In nature the understanding can cognize only what exists, or has

been, or will be. It is impossible that something in it ought to be
other than what, in all these time-relations, it in fact is; indeed, the
ought, if one has merely the course of nature before one’s eyes, has
no significance whatever. We cannot ask at all what ought to hap-
pen in nature, any more than we can ask what properties a circle
ought to have; but we must rather ask what happens in nature, or
what properties the circle has.
(KrV A547/B575)
If we agree with Kant’s claims that “in nature the understanding can cognize
only what exists” and that “it is impossible that something in it ought to
be other than what . . . it in fact is”, then what is the point in saying that
theoretical judgments about nature are also normative? In view of that, what
I want to do first is to clarify the distinctively normative features of our
theoretical inferences about the physical world, as acts performed by our
determining power of judgment. The challenge is to explain how it is pos-
sible to reconcile the normative character of our theoretical acts of judging
with the fact that the real world cannot be other than it is.
The solution to this challenge, however, is not so very difficult. It rests,
as we know, in the heart of the Copernican revolution proposed by Kant.
It is sufficient that we consider that the external world is distinct from our
knowledge of it; that is to say, that nature and the knowledge of nature are
two absolutely distinct things. So, we can recognize without contradiction
that, on the one hand, nature is non-normative and mind independent but
also that, on the other hand, our knowledge of nature is necessarily norma-
tive and mind dependent. In the case of our acts of theoretical judging, the
faculty of understanding subsumes the natural events of the objective world
under a priori conceptual rules. The determining power of judgment “is the
faculty of subsuming under rules, i.e., of determining whether something
stands under a given rule (casus datae legis) or not” (KrV A132/B171). The
lawlikeness that is laid down by the determining power of judgment is the
45
PAT R Í C I A K AU A R K - L E I T E
distinctive mark of the normative character of our objective judgments of

the natural world. This led Brandom to argue that:
Kant understands concepts as having the form of rules, which is to

say that they specify how something ought (according to the rule)
to be done. The understanding, the conceptual faculty, is the faculty
of grasping rules – of appreciating the distinction between correct
and incorrect application they determine.
(1994: 8)
In a passage in Jäsche Logic, Kant is absolutely explicit about this distinc-

tive normative character, also emphasized by Brandom, regarding all logical
principles, whether they belong to general logic or transcendental logic, that
is, whether they are analytic or synthetic. Kant says:
Some logicians, to be sure, do presuppose psychological principles

in logic. But to bring such principles into logic is just as absurd as
to derive morals from life. If we were to take principles from psy-
chology, i.e., from observations concerning our understanding, we
would merely see how thinking does take place and how it is under
various subjective obstacles and conditions; this would lead then to
cognition of merely contingent laws. In logic, however, the question
is not about contingent but about necessary rules; not how we do
think, but how we ought to think. The rules of logic must thus be
derived not from the contingent but from the necessary use of the
understanding, which one finds in oneself apart from all psychol-
ogy. In logic we do not want to know how the understanding is
and does think and how it has previously proceeded in thought,
but rather how it ought to proceed in thought. Logic is to teach
us the correct use of the understanding, i.e., that in which it agrees
with itself.
(Log 9: 14)
As we see in this passage, what Kant intends is not to show what the under-
standing is but instead to show how the understanding ought to proceed in
order to obtain a successful and objective meaningful knowledge of phe-
nomena. In this sense, whether analytic or synthetic a priori, the logical
principles of the understanding are not descriptive but instead prescriptive
(or normative) rules of thought. In the specific case of the determining power
of judgment, its normative character is made explicit by the way the thinking
agent connects the perceptual data according to its a priori conceptual rules
for producing epistemic judgments of experience.
46
2. The normativity of causal inferences

By recognizing that the conceptual rules of the understanding are inher-
ently normative, one important issue that comes up is whether the Kantian
epistemic normativity is restricted to the rules of the determining power of
judgment that yields the objective judgments of experience. Konstantin Pol-
lok (2017), in Kant’s Theory of Normativity, seems to hold, particularly in
the seventh chapter, entitled “The Normativity of Judgments of Experience”,
that the pure forms of the understanding alone are sufficient conditions for
taking into account Kant’s theory of epistemic normativity. In this sense,
only the determining judgments, that is, the judgments constituted by the
subsumption of perceptual data under a determinate concept, count as nor-
mative judgments for the possibility of objective experience. As spelled out
in the first Critique, all the determining judgments of empirical science fall
under the synthetic a priori judgments that form the System of Principles of
Pure Understanding “in accordance with which everything (that can even
come before us as an object) necessarily stands under rules, since, without
such rules, appearances could never amount to cognition of an object cor-
responding to them” (KrV A159/B198).
However, in opposition to Pollock’s approach, I would like to emphasize
that, in addition to determining judgments, reflective judgments, introduced
by the third Critique, also play a special role in Kant’s theory of epistemic
normativity.
From here on, I intend to provide a clarification of the differences
between the two kinds of normative functions – constitutive and regulative –
performed by the two powers of judgment, the determining one and the
reflecting one. To do this, I will restrict my analysis to the two types of causal
inference proper to each of these faculties: empirical or phenomenal causal-
ity and transcendental or noumenal causality.2 My thesis is that not only the
former but also the latter is necessary for the normative conceptualization
of human experience.
In the second part of his Critique of the Power of Judgment, the “Cri-
tique of the Teleological Power of Judgment”, in section 65, Kant makes
an explicit distinction between two different kinds of causal inference: on
the one hand, a causal inference that proceeds according to a principle of
mere natural mechanism and, on the other, a causal inference that proceeds
according to a principle of natural conformity to ends. The first has its struc-
ture defined by the faculty of understanding in the form of the Second Anal-
ogy of Experience, presented in the Analytic of Principles of the first Critique
(KrV B 232–56), and is called efficient causality (nexus effectivus), or the
connectivity of real causes and their effects. It is properly characterized as a
deterministic, mechanical causality. The second has its transcendental form
provided by the faculty of reflective judgment and is called final causality
47
(nexus finalis), or connectivity of ideal causes and their effects. This second
meaning of “causality” is also called teleological connectivity in the Tran-
scendental Dialectic of the first Critique (KrV A687/B715), but it is only in
the second half of the Critique of the Power of Judgment that Kant develops
it in more detail as an irreducibly different second kind of causality. Accord-
ing to the third Critique, “there cannot be more than these two kinds of
causality” (KU 05: 373). Although each one of these kinds of causal infer-
ence has distinctive characteristics and, correspondingly, different domains
of applicability, both perform fully normative functions.
The real and efficient causal inference, in particular, provides a rule for
connecting sense data such that the empirical series of effects and causes
generated by this rule is always taken as a “descending” series; that is, the
effect descends from its cause, and the opposite is not possible. As Kant
shows in the Second Analogy of Experience, this kind of descending suc-
cession is necessarily temporal; that is, the efficient causal connection is
intimately bound up with time’s arrow. So, the cause-effect relation has a
one-way direction that is oriented in time. According to this normative rule,
any cause in the real world of appearances always precedes its effect in time,
and it is impossible for any effect to be either the cause of itself or the cause
of its own cause. For instance, this rule ensures that it is impossible for you
to travel back in time, as in the grandfather paradox, meet your grandfather
before he produces any children (that is, your father/mother) and kill him.
You causally descend from your grandfather, and it is impossible that you
were the cause of your grandfather’s death.
In what sense can we say that this kind of efficient causality is normative and
not merely descriptive? As Kant points out in the proof of the Second Analogy,
the causal connection between the two perceptions in time “is not the work of
mere sense and intuition, but is here rather the product of a synthetic faculty
of the imagination, which determines inner sense with regard to temporal rela-
tions” (KrV B233). This means that we cannot derive causal relations from
mere observations, that is, merely seeing or perceiving that and how percep-
tions succeed one another. But any contingent series that we can derive from
our experience presupposes a necessary and a priori rule that prescribes how
two perceptions ought to be connected in an efficient causal temporal way.
Nevertheless, as distinct from this mechanical causal inference, in the ideal
and final causal inference, the “one-way direction” in time between cause
and effect is broken and there is no requirement for the cause to be prior
to the effect. According to a reflective causal inference, the effect can be the
cause of its own cause. Suppose, for instance, that you are the only child
of a generation descending from John’s son or daughter. From the point of
view of real and efficient causation, you are the effect of John’s son/daughter,
who in turn is the effect of John in a series of cause and effect that is always
descending and not interchangeable. Nevertheless, from the point of view of
ideal and final causation, we can say without contradiction that John is the
48
cause of your being a grandchild and also, conversely, that you are the cause
of John’s being a grandfather.3 In this latter case of ideal causal inference,
performed not by the determining power of judgment but by the reflecting
power of judgment, the relation between cause and effect is clearly not unidi-
rectionally temporal. It is important to emphasize that this connection is ideal
and not real and therefore a regulative and not constitutive concept. This is
because such a connection does not refer to the efficient cause in which the
understanding thinks about the empirical relationship of production of one
thing from another, or an effect from its cause, but to the way in which reason
judges based on its own concept, which conceives the parts as possible only in
relation to the whole. In Kant’s words, it requires “that its parts be combined
into a whole by being reciprocally the cause and effect of their form” (KU
5: 373). This special kind of either timeless or non-unidirectionally temporal
causality, namely that of ends, in contradistinction to that which makes up
the system of all synthetic judgments of pure understanding, does not need
inner sense and its a priori form, unidirectional time, as the medium for the
synthesis between the a priori concepts (KrV A155/B194). This kind of infer-
ence is a teleological judgment in which the effect is the underling condition
of the possibility of the causality of its cause (KU 5: 367). As Kant puts it:
In contrast, however, a causal nexus can also be conceived in accor-

dance with a concept of reason (of ends), which, if considered as a
series, would carry with it descending as well as ascending depen-
dency, in which the thing which is on the one hand designated as
an effect nevertheless deserves, in ascent, the name of a cause of the
same thing of which it is the effect.
(KU 5: 372)
In view of this, we can think of this special transcendental kind of final cau-
sality as a type of relational normativity in which each of the two basic terms
that make up the normative judgment – subject and predicate – is the condi-
tion of possibility of the other. So, in contradistinction to the rule of efficient
causality, in which there is no possibility of an exchange between the subject
and its predicate, or the effect and its cause, in the ideal or final causal rule,
subject and predicate, or the effect and its cause, are interchangeable. Here is
the distinctive normative mark of this kind of causal inference. The possibil-
ity of representing an object as a natural end implies applying the following
definition given by Kant: “I would say provisionally that a thing exists as a
natural end if it is cause and effect of itself” (KU 5: 370).
What is also especially interesting about Kant’s definition is that, for him,
a real and efficient causal inference is not in opposition to an ideal and final
causal inference, but, on the contrary, one can be “conceived in accordance
with” another. So, empirical causality and the noumenal causality are not in
conflict with each other but are complementary, or at the very least compatible.
49
The distinction between real causal inference and ideal causal inference is
grounded on the role of the former as the constitutive principle of experience
and the latter as a merely regulative principle. Only the understanding can
provide constitutive principles for the derivation of effects from their causes.
The regulative principles refer to the transcendental ideas of reason and can
never directly refer to what is given in human sensible intuition but instead
refer to the empirical use of understanding, as Kant had already pointed
out in the Appendix to Transcendental Dialectic of the first Critique, where
he first describes and defends the regulative use of the ideas of reason (KrV
A643–644/B671–672).
In light of the distinction between mechanical (real, efficient) causal infer-
ence and teleological (ideal, final) causal inference, as presented in the Cri-
tique of the Power of Judgment, what I would like to argue now is that the
ideal and final causal inference, presented in the third Critique and per-
formed by the reflecting power of judgment, plays an important and spe-
cifically normative role in the context of a scientific explanation of nature,
which, in turn, cannot be ignored or played by mechanical causal inference
as it is presented in the Second Analogy of Experience. My leading ques-
tion here is: what counts from a Kantian normative perspective as a causal
scientific explanation?
3. The normative role of the purposiveness of nature in

organized beings and physical science
As we know, according to Kant, the concept of the purposiveness of nature
in organized beings derives from the regulative activity of reason that is
characteristic of reflective judgments insofar as they are applied to cer-
tain empirically conditioned objects (EEKU 20: 216). As a regulative ideal,
the connectivity of final causes and their effects provides us an explicative
ground for everything that happens. This principle works as a rule of rea-
son that prescribes to nature an objective purposiveness and also an overall
systematicity that are jointly necessary for constituting our knowledge of
nature as an organic and unified system.
Thus, Kant reserves a privileged space within his critical system for the
ideal and final causal inference that is not filled by the Second Analogy of
Experience. This in turn is shown by the famous passage from section 75 of
the third Critique, where Kant says:
[I]t is quite certain that we can never adequately come to know the
organized beings and their internal possibility in accordance with
merely mechanical principles of nature, let alone explain them; and
indeed this is so certain that we can boldly say that it would be
absurd for humans even to make such an attempt or to hope that
50
there may yet arise a Newton who could make comprehensible even
the generation of a blade of grass according to natural laws that no
intention has ordered; rather, we must absolutely deny this insight
to human beings.
(KU 5: 400)
By virtue of ideal and final causality, reason through its reflecting faculty of
judgment can provide for everything in nature, including organisms, animals
and human beings, a purposive explanation, beyond real, efficient, deter-
ministic, mechanical causal explanation, to account for the organic and
systematic character of nature (Kauark-Leite 2008). Kant also reserves a
relevant role for teleology in the field of reflection on history. Because of the
crucial role of teleology in the domain of the phenomena of biological life,
many interpreters have been led to consider teleological causality exclusive
to organic and human beings; hence there would be no place for it in physi-
cal science in general and in Newtonian mechanics in particular. However,
I find this assumption very problematic from the standpoint both of the
Critique of the Power of Judgment and the Appendix to the Transcendental
Dialectic of the Critique of Pure Reason. As I have argued in previous works
(Kauark-Leite 2012, 2018), the concept of ideal and final causality has a
wider scope not restricted to biological or historical beings that performs a
specifically normative function in physical science. Nevertheless, the kind of
normativity that this concept of teleological causality has in the biological
domain is quite different from the kind of normativity it has in Newtonian
physics. Let me try to explain why.
First, it is crucial to recognize that the wider scope of ideal and final cau-
sality is not equivalent to claiming that teleological explanation has constitu-
tive epistemic force. On the contrary, Kant’s explicit view is that teleological
normativity has at most regulative epistemic force. Second, at the biological
level, the concept of teleological causality is required to reflect on the object
of experience, that is, on living beings. However, in physical science, the con-
cept of teleological causality is necessary only at a metatheoretical level, not
reflecting the physical object itself as given in empirical intuition but instead
on the products of the theoretical action of the understanding, namely sci-
entific theories. It is the teleological power of judgment that completes the
process started by the faculty of understanding by taking into account sci-
entific theories not as aggregates of empirical laws but instead as a coherent
system of such laws.
What transcends the scope of the Transcendental Analytic in the first Cri-
tique is the account of the “internal possibility” of nature itself. At the level
of the faculty of the understanding, nature has its objectivity established by
the a priori principles of the determining power of judgment that jointly
constitute nature as series of causal events. Nonetheless, the determining
51
power of judgment doesn’t provide all the elements for establishing the
ground of natural-scientific activity in its task of explaining nature as a
whole. We must continue in a progressive way through a synthesis that
can only be reached by the activity of reason and of the reflecting power
of judgment.
At this higher level of synthesis, or at a synthetic a priori meta-level, the
situation is different from the one established by the constitutive action of
the understanding. At the synthetic a priori meta-level, reason is employed
for the creation not of empirical laws but of scientific theories. The system-
atic unity that theories appear to have is not given but projected. Reason
does not create any objects but postulates theoretical unities. This procedure
is more unilateral than in the case of the understanding, where the catego-
ries and the manifold of the intuition are placed in a relation of reciprocity
between sensibility and understanding: on the one hand, sensible intuition
validates the employment of categories, and, on the other, the categories
make experience intelligibly possible. In the case of reason, its autonomy is
preserved at any price.
This results in a tension within Kantian philosophy. On the one hand,
reason must act autonomously. On the other, it must restrict its action to
empirically conditioned objects. Although the task of reason is regulative, it
plays a fundamental explanatory role. It is reason that offers a set of prin-
ciples or maxims that, once introduced in the cognitive dynamics of human
reasoning, produce a theoretical unit, from which stems the notion of the
objective purposiveness of nature. Only by means of the activity played by
these two superior faculties – reason and teleological power of judgment – is
it possible to arrive at an organic system of empirical laws.
Conclusion
Finally, in conclusion, I would like to emphasize two points. First, I want to
point out that the principles or maxims of reason together with the principle
of the teleological judging of nature in general as a system of ends are not,
strictly speaking, either true or false. They are not descriptive but instead
prescriptive. They are principles that human reason projects onto nature
according to its own desires and interests. If we lose this critical perspective
and ascribe to them a real ontological value, we fall into insoluble antino-
mies. Nevertheless, they do ground true counterfactual conditionals about
how nature would be under teleological principles, hence as if it had teleo-
logical properties.
Second, another crucial point is that the scientific explanation of nature
appeals both to the method of induction and also to a heuristic method of
hypothesis proposal. The inductive method is guided by the mechanical real,
efficient causal inference and makes it possible to constitute physical nature
as an aggregate of empirical laws. On the other hand, without the heuristic
52
method that provides a creative anticipation of nature as if it were teleological,

an organic and systemic construction of scientific theories would not be pos-
sible. In the absence of the heuristic method, there would be no interconnected
use of the understanding and no sufficient conditions for empirical truth. The
maxims of reason together with the judgment of purposiveness provide us a
systematic unity of nature as objectively valid and necessary (KrV A651/B679).
I conclude, therefore, that the concept of an ideal and final causality, as a
transcendental principle in its theoretical use, as applied to the realm of pos-
sible experience, cannot have its explanatory function filled by the real and
efficient, determining causal inference, even in the case of Newtonian phys-
ics. The ideas of reason, although they are not directly and constitutively
applied to any object of experience, have a regulative use that is entirely
legitimate and absolutely indispensable for completing the task of the under-
standing. While the understanding aims at unifying the multiple of experi-
ence by means of its concepts, reason aims at the unification of all empirical
laws by means of its ideas, seeking the maximum possible expansion of our
representation of the world of actual and possible experience (KrV A644/
B672f.). In this sense, reason is more demanding than the understanding. It
demands that the sufficiency of the scientific explanation must be sought in
the idea of the complete unity of the concepts of the understanding, even if
only in a hypothetical or counterfactual way. But these uses of the concept
of ideal and final causality entail necessary hypotheticals or counterfactuals
that make manifest the explanatory inadequacy of deterministic, real and
efficient mechanical causal explanation.
Notes
1 This project has received funding from the European Union’s Horizon 2020
research and innovation program under the Marie Skłodowska-Curie grant agree-
ment No 777786, and from the Fundação de Amparo à Pesquisa do Estado de
Minas Gerais (FAPEMIG), grant agreemente Nº APQ-000179-18, as part of the
activities of Kant in South America (KANTINSA) Project, and also from the Coor-
dination of Superior Level Staff Improvement (CAPES). I would like to deeply
thank Luigi Caranti for his critical comments that greatly improved the chapter.
I am also grateful to Damián Bravo Zamora, Robert Hanna and Luigi Caranti
for inviting me to discuss some of these ideas in conferences organized in Mexico
City and Catania. Finally, I would like to thank Alessandro Pinzani and Luigi
Caranti for the invitation to publish this work in this collective volume dedicated
to Kantian thought.
2 See Rescher (1983: 17–30); Buchdahl (1992). Rescher names this principle nou-
menal causality and Buchdahl names it transcendental causality, in order to dis-
tinguish it from the principle of causality in the “Second Analogy”.
3 This is an example of a final cause analogous to what Kant offers us in § 65 of
the Critique of the Teleological Power of Judgment: “the house is certainly the
cause of the sums that are taken in as rent, while conversely the representation of
this possible income was the cause of the construction of the house. Such a causal
connection is called that of final causes (nexus finalis)” (KU, AA 05: 372).
53
References
Brandom, R. 1994. Making It Explicit: Reasoning, Representing and Discursive
Commitment. Cambridge, MA: Harvard University Press.
Buchdahl, G. 1992. Kant and the Dynamics of Reason. Oxford: Malden.
Kauark-Leite, P. 2008. “Causalité empirique et causalité transcendantale: vers une
approche plus holiste de la théorie de la science de Kant”, in Valério Rohden,
Ricardo Terra, Guido A. Almeida, and Margit Ruffing (eds.) Recht und Frieden
in der Philosophie Kants. Akten des X. Internationalen Kant-Kongresses. Bd 2.
Berlin: De Gruyter: 481–92.
Kauark-Leite, P. 2012. “A Propósito das Distinções Kantianas entre Física e Biologia”,
in U. R. A. Marques (ed.) Kant e a Biologia. São Paulo: Barcarolla: 109–26.
Kauark-Leite, P. 2018. “Kant and Scientific Explanation Beyond Mechanical Causa-
tion”, in V. L. Waibel, M. Ruffing, and D. Wagner (eds.) Natur und Freiheit. Akten
des XII. Internationalen Kant-Kongresses. Bd. 2. Berlin: De Gruyter: 1261–8.
Pollok, K. 2017. Kant’s Theory of Normativity: Exploring the Space of Reason.
Cambridge: Cambridge University Press.
Rescher, N. 1983. “Kant on Noumenal Causality”, in N. Rescher (ed.) Kant’s Theory
of Knowledge and Reality: A Group of Essays. Washington, DC: University Press
of America: 17–30.
54
4
CONSCIOUSNESS AS
REPRESENTATION OF
REPRESENTATION
Kant on the human and animal capacity for
representation and consciousness
Silvia Altmann1
Kantian theses about our cognitive capacities, especially about the differ-
ence and relation between sensibility and understanding, have always been a
focus of interest not only in Kantian exegesis but in the theory of knowledge
in general. Today this interest is once again revived, seeking in Kant elements
to better understand the specificity of the human capacity of representa-
tion and consciousness. Kant clearly attributed to non-human animals the
capacity for representation but denied them the capacity for understanding
or thinking, while it is not clear whether and in what sense he attributes
consciousness to animals.2 For Kant, it is through our faculty of under-
standing (in particular, because of pure concepts of object in general, i.e.,
the categories) that humans are capable of the self-attribution of representa-
tions Hence, if by the term “consciousness”, we mean the self-attribution
of representations, animals would certainly have no consciousness for him.
In this chapter, I will suggest that Kantian story about human and animal
consciousness is more articulated. By looking at numerous statements in the
students’ notes,3 I will suggest that (1) Kant seems to make room for a sense
of consciousness weaker than the self-attribution of representations, (2) this
form of consciousness is construed as a necessary yet insufficient condition
for the faculty of understanding and (3) Kant seems to deny animals even
this weaker form of consciousness. As a consequence, there is a Kantian
sense in which not only do animals have no consciousness because they do
not have understanding, but it is also true that they do not have understand-
ing because they do not have consciousness (in the weaker sense).4
Given this working hypothesis, the chapter is articulated as follows. The
first half outlines a sense of consciousness as representation of representation,
DOI: 10.4324/9781003050742-5 55
S I LV I A A L T M A N N
although not necessarily as self-attribution of representation. The first sec-

tion presents the sense in which Kant thinks we should attribute representa-
tions to animals. Section 2 indicates how every conscious apprehension can
be said to involve a representation of representation. Section 3 examines the
link between consciousness, understood as representation of representation,
and the capacity for self-consciousness. The second half of the chapter seeks
to explain how consciousness is the condition of the understanding, even if
the understanding is the condition for the self-attribution of representations.
Section 4 focuses on the relation between consciousness and cognition, while
Section 5 deals with consciousness and perception. On the basis of the pre-
vious analysis, the final section explains how the understanding depends
on consciousness, while the capacity of self-attribution of representations
depends on the understanding.
1. Representation, consciousness and apperception
1.1. Representation and nonhuman animals

Let us first see why, according to Kant, we should attribute to animals some-
thing at least similar to what in ourselves we call representation. It seems
that this is inevitable insofar as they are living beings, animated, with a soul,
with functions irreducible to matter:
To live, properly speaking, means to have a faculty for performing

actions in conformity with one’s representations. We call an animal
alive because it has a faculty to alter its own state as a consequence
of its own representations.
(V-Met/Volckmann 28: 449)
Soul [anima] is the animating principle of an animal. Matter cannot

live for itself.
(V-Met K2/Heinze 28: 753)
Life is the faculty of a being to act in accordance with laws of the

faculty of desire. The faculty of desire is a being’s faculty to be by
means of its representations the cause of the reality of the objects of
these representations.
(KpV 5: 12)
Let us say that my cat, hearing the noise of its food bowl, comes to the kitchen
to eat. We should say that in some sense, the noise of the plate was the cause
of its movement, but it does not seem reasonable to say that this occurred by
something mechanical such as when a car pulls a trailer or even by something
more sophisticated, such as when a voice command makes a robot move. It
56
R E P R E S E N TAT I O N O F R E P R E S E N TAT I O N
seems that we must talk about something in the “animating principle” or soul
of the cat which “takes the place” of the food bowl and, to that extent, causes
its movement. And it seems that this something which “takes the place” of the
bowl should be called “representation” or “representing”: something in the
soul or animating principle of the cat “takes the place” of the plate of food
and, with this, the cat can, as a consequence of this representation, alter its
state and be the cause of the reality of the object of representation.5
On the other hand, Kant is explicit that nonhuman animals do not have
cognition, because that requires understanding. In a passage from the hand-
written version of the Anthropology, Kant wrote:
The irrational animal [perhaps] has something similar to what we

call representations . . . but no cognition of things; for this requires
understanding.
(Anth 9: 141, crossed out in H.)
We should then examine why Kant denies understanding to animals and

therefore denies them cognition. Animals do not cognize because the under-
standing is required for cognition. Given that cognition involves concepts
and the understanding is the faculty of thinking objects through concepts,
obviously without understanding, nonhuman animals are not capable of
cognitions. The interesting point, however, is to examine what seems to be
Kant’s reason for denying them understanding: their lack of consciousness.
We perceive in ourselves a specific feature of the understanding and

of reason, namely consciousness.
(V-Met/Volckmann 28: 448–9)
It is crucial to remark that this is not to say animals have no consciousness

(in the Kantian sense) because they have no understanding. On the contrary,
it seems that, according to Kant, they have no understanding because they
have no consciousness:
Due to the lack of consciousness, even animals are not capable of

any concept.
(V-Log/Dohna 24: 702, emphasis added)
Consciousness is the principle of the possibility of the understand-

ing, but not of sensibility.
(V-Met/Mron II 29: 878 emphasis added)
Now, what does consciousness consist of?
57
1.2. Consciousness
According to the Critique of Pure Reason, we have the characterization of
consciousness as the capacity to differentiate objects:
Clarity is not, as the logicians say, the consciousness of a representa-

tion; for a certain degree of consciousness, which, however, is not
sufficient for memory, must be met with even in some obscure rep-
resentations, because without any consciousness we would make no
differentiation6 in the combination of obscure representations; yet
we are capable of doing this with the marks of some concepts (such
as those of right and equity, or those of a musician who, when impro-
vising, hits many notes at the same time). Rather a representation is
clear if the consciousness in it is sufficient for a consciousness of the
difference between it and others. To be sure, if this consciousness suf-
fices for a differentiation,7 but not for the consciousness of the dif-
ference, then the representation must still be called obscure. So there
are infinitely many degrees of consciousness down to its vanishing.
(B414–5 note, modified translation)8
Kant takes up the example of the musician improvising a melody as evi-

dence that we have unconscious representations in the Anthropology. We do
not find the distinction between obscure representations with and without
consciousness in the Anthropology (nor in the Lectures on logic),9 and the
example of the musician is offered simply as evidence of obscure or uncon-
scious representations. Still, what in the Anthropology is offered as evidence
for obscure representations is clearly what in the Critique is evidence for
obscure representations with some degree of consciousness. And the evidence
is this: we know that the musician has some representation of the notes (rep-
resentation that is simply obscure, in the formulation of the Anthropology,
or obscure with some consciousness, in the case of the Critique) because “a
single stroke of the finger not in accordance with the harmony would imme-
diately be heard as a discordant sound” (Anth 07: 136).
If being conscious were for Kant simply the capacity to differentiate
objects, it seems it would be impossible for him to deny consciousness to
animals: it seems obvious that animals would also notice things equivalent
to a dissonant note, like, say, a different configuration of a visual field to
which they are accustomed to. In pre-critical texts, Kant is explicit about
the capacity of non-human animals to differentiate objects, and in the first
introduction to the Critique of the Power of Judgment, Kant even writes that
there is a sense in which we can say animals reflect:
To reflect (to consider) . . . is to compare and to hold together

given representations either with others or with one’s faculty of
58
cognition, in relation to a concept thereby made possible. . . .

Reflecting . . . goes on even in animals, although only instinc-
tively, namely not in relation to a concept which is thereby to be
attained but rather in relation to some inclination which is thereby
to be determined.
(UUEK 20: 211)
It seems unavoidable to say that, in order to explain the discriminatory

capacity of animals, it is necessary to say that in some sense they compare
and hold together representations, which, by association, influence their
behavior. And saying that animals reflect is very close to saying they are
conscious of these representations.
However, we must be careful and examine whether, for Kant, any capac-
ity for differentiation implies consciousness. Leland (2018) explains how, in
pre-critical texts, Kant refused Meyer’s position, who attributed conscious-
ness to animals because of their discriminatory abilities. In the text The False
Subtlety of the Four Figures of Syllogism of 1762, Kant writes:
It is one thing to differentiate things from each other, and quite

another thing to cognize10 the difference between them. The latter
is only possible by means of judgments and cannot occur in the
case of animals, who are not endowed with reason. The following
division may be of great use. Differentiating logically means cog-
nizing11 that a thing A is not B; it is always a negative judgment.
Physically differentiating means being driven to different actions
by different representations. The dog differentiates the roast from
the loaf, and it does so because the way in which it is affected by the
loaf (for different things cause different sensations); and the sensa-
tions caused by the roast are a ground of desire in the dog which
differs from the desire caused by the loaf, according to natural
connection which exists between its drives and its representations.
This consideration may induce us to think more carefully about
the essential difference between animals endowed with reason and
those not so endowed.
(DfS 02: 59–60, translation modified)
This passage cannot by itself suffice to say that Kant would deny any kind
of consciousness to animals, especially because what is explicitly denied
here is the capacity for cognition, judgment and reasoning. One could
therefore contemplate the possibility that Kant’s position on conscious-
ness becomes more sophisticated with the development of the critical
philosophy. In particular, since the critical Kant sees sensibility as a fac-
ulty with its own a priori forms, one may expect the introduction of a
concept of consciousness independent of the understanding. However, as
59
Leland (2018) observes, Kant never attributes consciousness to animals

in later texts. Moreover, in Kant’s and students’ notes we repeatedly find
the denial of attribution of consciousness to animals, even in the critical
period. All this suggests that the capacity of consciousness, for Kant, might
be more than the capacity to discriminate objects by having and associat-
ing representations.
In students’ notes, we seem to find confirmation of this conjecture. Con-
sciousness is characterized as representation of representation:
Consciousness is really a representation that another representation

is in me.
(Log 9: 33)
consciousness (representation of our representation).

(V-Log/Dohna 24: 701)
Given this characterization of consciousness as representation of representa-

tion, it is tempting to say that Kant is reserving the term “consciousness” for
the self-attribution of representations, when the object of my representation
is my own representative states. However, as again Leland (2018) observes,
it is hard to say that all consciousness for Kant is self-attribution of repre-
sentations and, in this sense, a second-order representation. In particular, in
the example of the musician improvising, it seems impossible to say that,
because he is differentiating the notes when improvising, the musician is rep-
resenting some of his representations. Therefore, it does not seem adequate
to say that the specificity of the consciousness that Kant attributes to human
beings is the self-attribution of representations.
I would like to explore here the possibility of a notion of consciousness
proper to human animals that is, in some sense, a representation of repre-
sentation but that is not the self-attribution of states (although closely linked
to this capacity for self-attribution). For this, we should first consider, even
if only in general terms and with many suppositions,12 the relation between
consciousness and apperception. There are many notes from Kant’s stu-
dents in which the expressions “apperception”, “consciousness” and “self-
consciousness” appear as equivalent:
One kind of representation can accompany all our representations,

this is the representation of our self. The representation of our self
is called consciousness, apperception.
(V-Met-L2/Pölitz 28: 584)
Representations may be combined also with apperception – the

consciousness of the representation.
(V-Log/Dohna 24: 701)
60
Consciousness is also called apperception, which accompanies the

represented object.
(V-Met-K3E/Arnoldt 29: 970)
The first and major representation is that of the I or the conscious-

ness of my self, apperception (as Prof. Kant calls it in his Critique).
(V-Met/Mron 29: 878)
Given this evidence, it seems to me difficult to dissociate consciousness from

apperception, although obviously the two expressions are not synonymous.
We have to understand why, despite the use of different terms (“conscious-
ness”, “self-consciousness” and “apperception”), it would be possible, in
different contexts, to use them as equivalents.
Self-consciousness and apperception is consciousness of oneself, and Kant
distinguishes two senses in which we become conscious of ourselves: pure
apperception and empirical apperception.
I call it the pure apperception, in order to distinguish it from the

empirical one, or also the original apperception, since it is that
self-consciousness which, because it produces the representation I
think, which must be able to accompany all others and which in all
consciousness is one and the same, cannot be accompanied by any
further representation.
(B132)
Discursive consciousness (pure apperception of one’s mental activity) is

simple. The “I” of reflection contains no manifold in itself and is always
one and the same in every judgment, because it is merely the formal
element of consciousness. On the other hand, inner experience contains
the material of consciousness and a manifold of empirical inner intu-
ition, the I of apprehension (consequently an empirical apperception).
(Anth 7: 141–2)
One must therefore distinguish pure apperception (of the under-

standing) from empirical apperception (of sensibility). The latter,
when the subject attends to himself, is also at the same time affected
and so calls out sensations in himself, that is, brings representations
to consciousness.
(Anth 7: 142, crossed out in H.)
In pure apperception, we are merely conscious of ourselves as an activity,

as a “simple self”, a consciousness that, precisely because of this simplicity,
does not furnish us any self-knowledge nor involve the self-attribution of
any specific representation. For self-knowledge, besides this consciousness
61
of oneself common to all representations, we also need the empirical con-

sciousness of the affection by our representations – the material of inner
empirical intuition. Finally, it is worth remembering that although we can
simply become conscious of ourselves as an activity (as a simple “I” of the “I
think”), this is a consideration by abstraction, since without some empirical
representation, no thinking could occur.13
It is therefore trivial to say that apperception or consciousness of oneself
involves consciousness. However, the converse does not seem to hold: it does
not seem possible to say that every consciousness involves apperception in
the sense of the actual self-attribution of states or even in the sense of the
explicit recognition of my existence as a thinking being. Thus, to say that
every consciousness is representation of representation is not to say that it
is always representation of representation in the sense that, when I have a
representation with consciousness, the object of my representation is my
representation or myself. If I am consciously representing a table, the object
of my representation is the table, not myself or my mental state. If that is so,
in what sense would it still be possible, if at all, to maintain that conscious-
ness is representation of representation?
The first step in arguing for that possibility is to remember how weak the
Kantian sense of representation is: according to Kant, we may be represent-
ing without having the slightest idea that we are representing:
The field of sensuous intuitions and sensations of which we are not

conscious, even though we can undoubtedly conclude that we have
them; that is, obscure14 representations in the human being (in thus
also in animals), is immense.
(Anth 7: 135)
The saying “clothes make the man” holds to a certain extent even for
intelligent people. To be sure, the Russian proverb says: “One receives
the guest according to his clothes and sees him to the door accord-
ing to his understanding.” But understanding still cannot prevent the
impression that a well-dressed person makes of obscure representa-
tions of a certain importance. Rather, at best it can only have the
resolution afterwards to correct the pleasing, preliminary judgment.
(Anth 7: 137)
e.g., when we observe it with the naked eye, we are not conscious
to ourselves that the Milky Way consists of sheer small stars, but
through a telescope we see that. Now we infer that since we have
seen the entire Milky Way, we must also have seen all the individual
stars. For were that not so, we would have seen nothing. But what
we have seen we must have represented to ourselves.
(V-Met/Mron 29: 879, emphasis added)
62
The conditions to say that “we represent something to ourselves” are very
undemanding: it is enough that something in the “animating principle” of
the subject “takes the place” of the object and affects the subject’s behavior.
The host favorably predisposed by the visitor’s beautiful garments could even
eternally deny having ever represented these garments, because perhaps he has
not even apprehended his representation of the garments. But, according to
Kant, if the garments affected his behavior, it is because some representation of
the garments interacted with other representations and favorably influenced
the host’s disposition. In the example of the Milky Way, although we were
not aware of the individual stars, we were always representing the individual
stars;15 otherwise, we would never have represented the whole of the Milky
Way. That is, even if we had not developed telescopes and had never learned
the individual stars that compose the Milky Way, if we saw the whole, we
always represented (in this weak sense of representing) the individual stars.16
It is important to note that the case of the individual stars is completely
different from the case of the musician improvising. The musician would have
noticed a dissonant chord; he just was not paying attention to the notes. On
the contrary, in the case of the Milky Way, we would not notice with the naked
eye a change in some individual stars: our representations of the individual
stars are really absolutely unconscious (which is stronger than to say that I
have not paid attention or cannot recover them by memory). And, as in the
case of the guest’s beautiful garments, it seems the same can be said of many
things that might have affected our behavior (and been represented in that
weak sense) without us ever acknowledging it.17 So to say that by consciously
apprehending a table, I also have a representation of representation does not
necessarily mean that the object of my representation is one of my representa-
tions. I may be representing representations in the same way that, when I see
the Milky Way from a distance, I am representing the individual stars.
It seems therefore possible, from a Kantian point of view, to speak of repre-
sentation of representation without this being the self-attribution of represen-
tations.18 Still, if this explains the possibility of saying that, in the conscious
representation of a table, I have representation of representation without say-
ing that the object of representation is the representation (as opposed to the
table), that still does not explain why we should say that, when consciously
representing the table, we have representation of representation. To under-
stand how we can do justice to the statements in student notes according to
which conscience is representation of representation, we must remember that
in order to be conscious of representing the table, we must apprehend it, and
for this to happen, there must be an appearance of the inner sense:
all representations, whether or not they have outer things as their

object, nevertheless as determinations of the mind themselves belong
to the inner state.
(A34/B51)
63
Wherever our representations may arise, whether through the influ-

ence of external things or as the effect of inner causes, whether they
have originated a priori or empirically as appearances – as modifi-
cations of the mind they nevertheless belong to inner sense.
(A98–9)
Perception (empirical intuition with consciousness) could be called

merely appearance of inner sense.
(Anth 7: 144, marginal note in H.)
When I perceive a table, I have an appearance of the outer sense, but not only
that: there is also an appearance of inner sense, and the inner sense is affected
by myself as the subject of representations: by means of inner sense “the mind
intuits itself, or its inner state” (A22/B37), the object of inner sense is the soul
(B427) and inner sense is how we are affected by the mind (Anth 7: 153).
Once again, this does not mean that, when I consciously apprehend a table,
I have a second-order representation which has as its object my represen-
tation of the table. And, depending on how we understand the notion of
inner sense, to deny that, in consciously apprehending the table, I have as the
object of my representing the representation of the table means denying two
things. First of all, as we have seen, to say that in any conscious apprehension,
we have representation of representation does not mean that the intention-
ally “aimed” object of my representation becomes my representation or my
mental state – this only occurs in cases of self-attribution of representations
(having representation of representation is a necessary but not sufficient con-
dition for such attribution). But not only that. It also does not seem possible19
to say that we initially have representations of outer sense and then (though
unconsciously) this representation (a mental state) affects my inner sense in
a way analogous to that in which a physical object affects outer sense. If we
look at the previous passages on the inner sense, it is presented as the capac-
ity to be affected by ourselves. In the case of the apprehension of objects of
outer sense, I am affected by my act of apprehension of this sensory outer
content.20 It is important to emphasize that this does not mean that, in the
apprehension of outer objects, I am affected only by my act of synthesizing
in the sense of pure perception. It is necessary that the content of outer sense
affect my inner sense.21 And inner sense is affected by us and our states, not
by physical objects. Thus, for the apprehension of the object of outer sense,
besides the affection by the physical object, we must also affect ourselves and
at least in this very weak sense represent ourselves and our representations.22
1.3. Apperception
With this, we naturally move on to the question of whether every conscious-
ness is apperception, in the sense that it depends on the representation of
64
oneself. I think the answer depends on what we mean by “representing one-

self”. If by self-representation, we understand the representation of oneself
as “oneself” or, in the case of empirical apperception, as the self-attribution
of certain states, it is of course not possible to say every consciousness is
apperception. The representation “I think” can accompany all my represen-
tations but does not necessarily accompany them (B131); this representation
“I think” is produced by pure or original apperception (B132), and the rela-
tionship with the identity of the subject happens through the consciousness
of the synthesis of representations (B133). If pure or original apperception
produces the representation “I think”, then consciousness of the synthesis of
representations cannot depend on self-attribution of representations.
On the other hand, however, if consciousness is always representation of
representation, there is a sense in which some representation of oneself is pres-
ent in all conscious representation: in the trivial sense that, if consciousness
is representation of representation, it is a representation of oneself (though,
again, not necessarily as a self). If by pure apperception we understand merely
the representation of oneself as a “representation of a simple unit that accom-
panies all our representations”, without this representation of oneself being
the reflexive consciousness of oneself, we can say that this representation of
oneself is present in every conscious representation of, for example, a table
(even if, once again, this representation of oneself is not the object targeted
by consciously representing a table). And if by empirical apperception, we
understand merely the representation of our state (in the sense of our state
“being something to” our inner sense, not in the sense of taking that state as
the object consciously targeted), then we can say that some particular empiri-
cal apperception accompanies each particular perception. In both cases, it
would be a representation of oneself, but not as oneself; it would be a repre-
sentation of one’s own representations, but not as one’s own representations.
On the other hand, if the term “apperception” should designate the pure
consciousness of oneself (expressed, for example, in “I exist”) or the self-
attribution of states (in the case of empirical apperception), then we would
have a representation as: the representation of ourselves as the simple logical
subject of the act of representing (pure apperception of what is always one
and the same in every representation) or the representation of ourselves as
having a certain empirical representation (empirical apperception).
Thus, if we understand “apperception” in the first sense (not the self-
attribution of existence or of certain mental states but simply the repre-
sentation of oneself), and if every consciousness is a representation of
representation, then we can understand the “substitutability” between
apperception and consciousness: every consciousness is representation of
oneself, even if not as oneself. If we understand apperception as the second
meaning, the indissociability is less direct, since we can talk about conscious
representations without apperception in this second sense (because we can
talk of consciousness without self-attribution). Even in this case, however, it
65
is worth remembering that, for Kant, every representation has a necessary

relation with possible empirical consciousness and this latter, in turn, has a
necessary relation with original apperception:
All representations have a necessary relation to a possible empirical

consciousness: for if they did not have this, and if it were entirely
impossible to become conscious of them, that would be as much as
to say they did not exist at all. All empirical consciousness, however,
has a necessary relation to a transcendental consciousness (preced-
ing all particular experience), namely the conscious of myself, as
original apperception.
(A117 note)
If empirical consciousness were not possible, it would be as if these repre-

sentations did not even exist. And this empirical consciousness depends on
transcendental consciousness. Explaining and justifying this necessary relation
depends on many additional considerations and how we interpret the tran-
scendental deduction that will be assumed.23 I only intend to have provided
evidence that, although limited to the consideration of certain representations
(those that may be something for us),24 there is a reciprocity between “con-
sciousness” and “apperception”, which can be explained if these expressions
are understood, respectively, as representation of representation and self-
representation, where self-representation is not necessarily the self-attribution
of states but the capacity for this self-attribution. And that Kant thinks of
“apperception” as a capacity is explicit:
There are, however, three original sources (capacities or faculties of

the soul), which contain the conditions of the possibility of all expe-
rience, and cannot themselves be derived from any other faculty of
the mind, namely, sense, imagination, and apperception.
(A94/B127)
If inner sense depends on this at least on the capacity of representing our-

selves,25 and if the conscious apprehension of any object depends on the
affection of inner sense, then at least as a capacity, it involves apperception,
and we can therefore understand the terms “consciousness” and “appercep-
tion” can often be used inseparably and interchangeably.
2. Consciousness as a condition of understanding
2.1. Cognition
We can now return to the dependence of cognition on consciousness. If
consciousness is representation of representation, it is easy to understand in
66
which sense cognition depends on consciousness or apperception. In order

to have cognitions, we need to relate something as a representation to some-
thing as an object of this representation. This is explicit, for instance, when
Kant says that “I cannot stop with these intuitions, if they are to become cog-
nitions, but must refer them as representations to something as their object”
(BXVII) or that by the faculty of cognizing an object through concepts, the
object is “thought in relation to that representation [by which the object
is given] (as a mere determination of mind)” (B74): in thinking the object,
something is referred as representation (as a mere determination of mind) to
something as its object.
Now, if cognizing involves referring something as representation to
something as an object, that involves representing a representation,26 and
if consciousness is representing a representation, then having cognition or
understanding depends on consciousness. If I cognize this table, it is not
simply the case that I have something in me that, so to say, “takes the place”
of the table and interacts with my inclinations and other representations: in
cognition, I take some representation as ground of cognition of the object
and in that sense represent the representation (I must represent the represen-
tation if I must refer it to the object as ground of cognition).
Again, to say that “I take some representation as ground of cognition of
the object” obviously does not mean that I am taking my representation
as the object of my attention. After all, it is one thing to be conscious and
another to pay attention: “consciousness with the power of choice is atten-
tiveness” (V-Met/Mron 29: 878). However, if it is merely possible to explain
what I am thinking, it is because, even when I am not doing that “explain-
ing” (even when I am not taking my thoughts as object of my attention in
explaining what I am thinking), I am, in some sense, having a representa-
tion of my representations, because my representations “are not nothing to
me” (if they were, I could never retrospectively make them the object of my
attention). If this is so, we understand why consciousness is a condition of
the possibility of the faculty of cognition. And, given the connection between
consciousness and apperception, apperception is a condition for the faculty
of cognition:
we assume that animals can endure to eternity and that their powers
can steadily grow, and yet not attain to understanding, because first
an essential piece must then be added to their sensibility, through
which alone understanding becomes possible. Namely, apperception.
(V-Met/Mron 29: 906)27
The self underlies consciousness and is what is peculiar to spirit. But

we can consider this self in three ways: I think as intelligence, i.e.,
the subject of thinking is intelligence. I think as subject which has
sensibility, and am a soul. I think as intelligence and soul, and am a
67
human being. A body which is animated only by a soul, and not by

intelligence, is an animal.
(V-Met/Mron 29:878)
We have so far seen that cognition depends on understanding which depends

on consciousness or apperception, because it depends on representing our
representations. Since Kant denies non-human animals’ consciousness, non-
human animals, without being able to represent their representations, could
never relate something as a representation to something as an object, that
is, cognize something.
2.2. Perception and apperception

While the dependence of cognition on apperception seems clear, since to
cognize is to relate something as a representation to something as an object,
things get less simple if we remember that not all human conscious repre-
sentation involved in cognition is itself a cognition. In particular, we have to
keep in mind the distinction between perception and cognition.
In the first Critique, we find that “the first thing that is given to us is
appearance, which, if it is combined with consciousness, is called percep-
tion” (A119–120); that the synthesis of apprehension is the “composition
of the manifold in an empirical intuition, through which perception, i.e.,
empirical consciousness of it (as appearance), becomes possible” (B160);
and that “experience is an empirical cognition, i.e., a cognition that deter-
mines an object through perception” (B218).
We have to distinguish the conscious apprehension of the matter of an
empirical intuition: the fact that we are affected in a certain way (a percep-
tion), which is not necessarily cognition but a necessary step in order to
explain our cognitions as to their content, and we have to distinguish this
apprehension from the attribution of this appearance such as it appears to
an object. If we see a round plate from the side, we apprehend an oval shape,
but we do not determine or classify the plate as oval. In order to determine
the object, we take into account conditions of possible experience, and we
determine the object not according to particular conditions of our own per-
ception but according to those connections of perception that can be attrib-
uted to something not solely from a particular subjective point of view but
according to rules of possible experience.28
It is important to note that this distinction between the apprehension of
something in the perception and the conceptual determination of an object
by means of what is apprehended in perception is, at least to a certain extent,
independent of the question of whether the perception itself depends on the
use of some empirical concept. It is undeniable that, as a conclusion of the
transcendental deduction, we have the dependence of perception in rela-
tion to the functions of understanding and, as a consequence, the necessary
68
conformity of perception to the categories. And it might well also be possible

to argue that every perception depends further on some empirical concept,
but I will not address this point here.29 What is important to highlight is
that apprehending the manifold of perception is not “by itself” a cognition.
Now, unlike the refusal to animals of the capacity of thought or cognition,
I do not know if there are passages, from Kant or his students, in which he
denies animals perception. On the contrary, a famous passage by Jäsche sug-
gests that we should attribute perceptions to animals:
In regard to the objective content of our cognition in general, we

may think the following degrees, in accordance with which cogni-
tion can, in this respect, be graded:
The first degree of cognition is: to represent something;
The second: to represent something with consciousness, or
to perceive (percipere);
The third: to be acquainted with something (noscere), or
to represent something in comparison with other things,
both as to sameness as to difference;
The fourth: to be acquainted with something with con-
sciousness, i.e., to cognize it (cognoscere). Animals are
acquainted with objects too, but they do not cognize them.
(Log 9: 64–5)
It might seem natural to conclude that if animals have the third degree, they
have the second, being able to perceive or represent with consciousness.
They would therefore, according to Kant, have representations, perceptions
and acquaintances, only not cognitions, and since perception involves con-
sciousness, non-human animals would, according to Kant, be capable of
consciousness.
It is true that, as Leland (2018) rightly observed, this is not a text by Kant
but by Jäsche, and its authenticity as a Kantian position can be doubted.
However, I do not think that this settles things, if we take into account the
passage from the first introduction to the Critique of the Power of Judgement
already mentioned, according to which it seems that we should attribute to
animals the ability to differentiate objects, although only instinctively (see
KU 20: 210). So, although the mention of animals having the third degree
is not, as Leland pointed out, present in other similar passages in other lec-
tures, it seems hard to deny that Kant attributed to animals something very
close to being acquainted or noticing objects, that is, “representing them
in comparison to other things, both as to sameness and difference” (Log
9 64–5). And this ability to instinctively differentiate and react to objects
seems very close to the obscure consciousness that the note of B414–5 in the
Critique of Pure Reason attributes to humans.
69
However, I believe that, if we take into account the specifically Kantian

sense of perception to which Tolley (2013, 2016, 2018) in particular has
been drawing attention, as well as the purpose of Jäsche’s passage, this
text does not necessarily imply attributing consciousness or perception to
animals.
Although it seems natural to think that the passage in question implies that
animals have the second degree (because they have the third), it is important
not only to point out that this is presented as a description of the degrees of
human cognition30 but also that it is a description of the possible degrees of
intensification of our cognition as to its objective content.31 That is, it is not
necessarily a description of steps in cognitive development. All the passage
implies is that, in our (human) case, when explaining the content of our
cognition, it seems we could use the following description.32 The affection
by an object generates a representation (a, say, “mental state” in the weakest
sense). Since we are beings capable of representing representations, we can
apprehend this. Nonhuman animals cannot do this (or, at least, according to
Kant, we have no reason to suppose they do). But nonhuman animals, even
without perception, can have the equivalent of the third degree: they can dis-
criminate objects thanks to the representations generated by the objects, but
without representing their representations, they would not perceive in the
Kantian sense.33 In order to discriminate objects, animals need something
like representation in the sense of something that, in them, “takes the place”
of the object in order to interact with other representations so that these
representations may be the cause of a change in their state: say, get access
to food. But nothing in this seems to demand that they represent all of this,
represent their representations. We have reasons to assume that, just as we
do, non-human animals associate representations, but that by itself is not
necessarily a reason to assume that this association presupposes the capac-
ity representing their representing.34 So they could perfectly well have some
equivalent of degrees one and three without the second.
Let us now consider the human case and try to understand our “second
degree”. As already anticipated, an important point is to consider the speci-
ficity of the Kantian sense of perception. The expression is now currently
used to refer simply to a representation of something given to sensibility and,
as Tolley (2013, 2016, 2018) pointed out, Kant’s specific understanding of
the term “perception” has frequently been ignored in the recent debates. As
we have seen, perception, for Kant, essentially involves an act of becoming
consciously aware of something, which, as the passage from the Jäsche Logic
says, involves more than merely representing something, although it is dis-
tinct from cognizing an object.
We can now go back to some characterizations of perception that we find
in the Critique of Pure Reason. We saw that “experience is an empirical cog-
nition, i.e. cognition that determines an object through perception” (B218),
and therefore perceiving has to be distinguished from cognizing: the same
70
perception can be used to determine the object differently, as when we assign

or not ovality to a dish perceived as oval depending on something that goes
beyond the perceived (relying in particular on rules of possible experience).
But even if perceiving is not by itself determining the object, it is crucial to
note the specificity of this apprehension: it is “perception, i.e., empirical
consciousness of it (as appearance)” (B160, emphasis added). Now, what
does it mean to have consciousness of something as appearance? It depends
on the subject of consciousness already “mobilizing” what is expressed by
the concepts of object and representation: appearance is always the appear-
ance of something to a subject, and if the subject is to be conscious of that
(consciousness of something as appearance), in some sense, that subject is
using the notions of object and representation. That is not surprising, since
this apprehension is a “step” in the increase of content of cognitions. In the
Anticipations of Perception, Kant is quite explicit: appearances, as objects
of perception, contain
the materials for some object in general . . ., i.e., the real of sensa-
tion as a merely subjective representation by which one can only be
conscious that the subject is affected, and which one relates to an
object in general.
(B207–8, emphasis added)
Thus, when apprehending something as an appearance, we have the ref-

erence to an object in general, even if to an object undetermined, and that
will be determined taking into account conditions for an object in general
other than those of the mere apprehension. As Kant formulates at the begin-
ning of the Analogies of Experience, empirical cognition is “a cognition that
determines an object through perceptions” (A176/B218). The analogies of
experience teach us how to use perceptions for cognition, how to move from
the perceived (say, the ovality of a plate) to the determination of the object
(as oval or round, depending on the connection of perceptions according to
the conditions of possibility of experience). Even so, although perception is
not by itself the conceptual determination of an object,35 it is the conscious-
ness of something as the appearance of an object.
Finally, this also explains the specificity of our “third” degree in relation
to animals. In the quoted passage from the first introduction to the Critique
of the Power of Judgment, we saw that, according to Kant, animals compare
and hold together representations, but they do that only instinctively, “in
relation to some inclination which is thereby to be determined” (UUEK 20:
211). We, on the contrary, compare representations in order to arrive at the
fourth degree, that is, we compare things as to sameness and difference, “in
relation to a concept which is thereby to be attained” (EEKU 20: 211). If this
is so, it would indeed be surprising if Kant attributed to non-human animals
the consciousness of perception: it is, as I have already pointed out and as
71
Jäsche’s own text makes clear, a degree of intensification of the content of

cognition.
2.3. Consciousness and the faculty of understanding

In order to see that, although the capacity to self-attribute representations
depends on the faculty of understanding, we can nevertheless sustain that
this capacity of understanding in turn depends on the capacity of conscious-
ness, I would like, in conclusion, to consider another passage, now from
Kant himself, which is often taken as evidence that he would consider attrib-
uting consciousness to non-human animals. In a letter to Marcus Herz, from
May 26th, 1789, we read the following:
all other concepts of things (which are not thus conditioned [by
form of sensibility and functions of the understanding]) are for us
empty and utterly useless for cognition. But not only that; all sense
data for a possible cognition would never, without those conditions,
represent objects. They would not even reach that unity of con-
sciousness that is necessary for cognition of myself (as object of
inner sense). I would not even be able to know that I have sense
data; consequently for me, as a cognizing being, they would be
absolutely nothing. They could still (if I imagine myself to be an
animal) carry on their play in an orderly fashion, as representations
connected according to empirical laws of association, and thus even
have an influence on my feeling and desire (assuming that I am
even conscious of each individual representation, but not of their
relation to the unity of representation of their object, by means of
the synthetic unity of their apperception), without my cognizing the
slightest thing thereby, not even what my own condition is.
(Br 11: 52)36
The passage seems to indicate the possibility of some kind of conscious-

ness (of individual representations) without relation to apperception. How-
ever, the text does not necessarily imply the possibility of consciousness of
individual representations without relation with the unity of the represen-
tation of its object by means of the synthetic unity of apperception. At the
most, the text implies the possibility of consciousness of individual repre-
sentations without consciousness of that relation. That at the most, for Kant
writes “assuming that I am even conscious of each individual representa-
tion, but not of their relation to the unity of representation of their object”
(emphasis added).37
About the role of that assumption (mere individual consciousness) in
Kant’s considerations here, it is illuminating to take into account the con-
text of that remark in the letter to Marcus Hertz. Kant is opposing his
72
own characterization of the faculty of understanding to that of Salomon

Maimon’s:
Herr Maimon’s theory consists basically in the contention that an

understanding (indeed, the human understanding) not only is a fac-
ulty of thinking . . . but is actually a faculty of intuition, where
thinking is only a way of bringing the manifold of intuition (which
is obscure because of our limitations) into clear consciousness. I, on
the other hand, conceive of the understanding as a special faculty
and ascribe to it the concept of an object in general (a concept that
even the clearest consciousness of our intuition would not at all
disclose. In other words, I ascribe to the understanding the syn-
thetic unity of apperception, through which alone the manifold of
intuition (of whose every feature I may nevertheless be particularly
conscious), in a unified consciousness, is brought to the representa-
tion of an object in general.
(Br 11:50)
Insofar as what is being compared are two conceptions of the faculty of

understanding, Kant’s point may simply be to insist that understanding
is originally the capacity of reference to an object, which could never be
explained by the Maimonian conception of understanding as mere par-
ticular consciousness, since mere particular consciousness would never
yield the possibility of self-attribution of representations. If we only
represented our representations, if we were only conscious of scattered
representations but not conscious of the relation of the representations to
the unity of the representation of their object, the representations “would
not even reach that unity of consciousness that is necessary for cognition
of myself” (Br 11: 52). In other words, understanding, in particular the
application of the categories and the reference to objects in general, is,
according to the lesson of Transcendental Deduction, a condition to say
that we know we have representations, it is a condition for self-attribution
of representations.
On the other hand, this is compatible with saying that consciousness (rep-
resenting representations) is, in turn, a condition of the faculty of under-
standing. If that is so, would it not then be possible for non-human animals
to have consciousness (representation of representations) but, because they
lack understanding, only not be capable of the self-attribution of represen-
tations? I think that Kant’s answer to that question might be: even if it
were possible, what grounds do we have to concede such a capacity to non-
human animals? As we read in Volckmann’s Metaphysics:
if many beings exhibit a large degree of the effects which can arise
in human beings through reason, it still does not at all follow from
73
that that they also would have reason, for, if they are lacking con-
sciousness, then they are also missing understanding and reason.
(V-Met/Volckmann 28: 448–9, emphasis added)
We could perhaps say that each one of us knows to be a conscious being,

because we know that we represent our own representations, and by con-
sidering others as human beings like us, we project to those we recognize
as persons this ability to use the word “I” for themselves as a representing
subject. On the other hand, it seems that, according to Kant, as long as the
behavior of animals can be explained in a way that is compatible with the
absence of consciousness as the capacity to represent its own representations,
it would be arbitrary to attribute consciousness to them. And, if animals have
no consciousness, they would not have understanding or reason.
Concluding remarks
It is important to note what does not follow from what I have tried to sug-
gest about representation and consciousness in the Kantian sense. It does not
follow that it is not possible to attribute to non-human animals (and perhaps
to ourselves) a sense of consciousness other than the Kantian one, and even
a sense that could nevertheless be compatible with Kantian positions. Nor
does anything follow about whether, abstracting from the capacity of apper-
ception and considering only our sensibility, we have representations in the
same sense that non-human animals have and, therefore, whether we can
have sensory representations regardless of our capacity for apperception.
It is true that a student’s note suggests that, due to the accompaniment by
apperception, even the faculty of sensibility should have different names in
our case and in that of non-human animals:
All three of these cognitive faculties [sense, reproductive power

of the soul or imagination and the faculty of anticipation] can be
accompanied by apperception or not. When they are, then they
belong only to human beings, when not – then animals also have
them. We ought, therefore, to have two different names for these,
but this is only for one, namely, the reproductive power of imagina-
tion; for this is called memory when accompanied by apperception.
(V-Met/Mron 29: 884)38
However, besides being a note that is not Kant’s, this remark does not in itself
exclude what Conant (2016) called the “layer-cake” conception of cognitive
faculties: it is compatible with saying that we have a sensory capacity com-
mon to animals, by itself able to represent (and, what is more, perhaps also
to notice or be acquainted with objects, comparing them and distinguish-
ing them without consciousness of doing so), thanks to the unconscious
74
(unconscious in the Kantian sense) power of imagination, shared with ani-

mals – a capacity that, when (and only when) associated with apperception,
yields a particularly human capacity.
For the same reason, stating that Kant attributes representations but denies
consciousness to non-human animals is not by itself decisive to the question of
whether all sensory human representation depends on the capacity of under-
standing. As Thomas Land (2018) has shown, this will depend on whether
we have an “additive” or “transformative” interpretation of the “accompani-
ment” of sensibility by our superior cognitive faculties. To understand the
relation between sensory, imaginative and thinking faculties would depend on
examining not only the relation between perception and understanding and
conceptualization (which would depend on analyzing the relation between
apperception and the unity of apperception) but also the relation between apper-
ception, unity of apperception and intuition, not even mentioned here.
My purpose was to suggest that consciousness in the Kantian sense is
always a representation of representation and therefore a representation of
oneself, even if this “self-representation” is not the same as thinking about
oneself or assigning to oneself representations. Non-rational animals repre-
sent and associate representations and differentiate objects thanks to these
representations. They are therefore affected by their representations, and
these representations contribute to the determination of their behavior. I
have tried to suggest, however, that for Kant, this behavior, which involves
the capacity of representation, does not imply the capacity to represent rep-
resentations. More precisely, the capacity to discriminate objects through
representations does not imply the capacity to represent representations as
representations (that is, the self-attribution of representations) and, as a con-
sequence, does not imply a weak sense of consciousness as representation of
representations (without self-attribution of representations). This weak sense
of consciousness, although it is not the self-attribution of representations or
the consciousness of oneself as a self, is nevertheless a representation of one-
self. Moreover, for Kant, this consciousness does not seem to be a necessary
component of animal behavior, while it is a necessary condition for the devel-
opment of human intellectual capacities. It would, then, be in this sense that
consciousness is a condition for the possibility of understanding, even though
understanding is a condition for the self-attribution of representations.
Notes
1 This work was carried out with the support of CNPq, National Council for
Scientific and Technological Development – Brazil (Proc. no. 310731/2017–8)
and CAPES, Higher Education Improvement Coordination – Brazil. Several
parts of this chapter have been published in Portuguese in Altmann (2018) and
in Altmann (2019). I would like to thank Antoine Grandjean, François Calori,
Giovana Dalmás, Gabriel Hickmann e Fernando Silva e Silva, Inga Römer, Paulo
Licht, Pedro Rego, Raphaël Ehrsam, Renato Fonseca and Virgínia Figueiredo for
75
discussion and suggestions on previous versions of this chapter, as well as two

anonymous referees who reviewed another version of parts of it. I would like in
particular to thank Berndt Dörflinger for, when presenting an earlier version of
this text at the Catania Multilateral Kant Colloquium, insisting on the need for
emphasis to be placed on the notion of consciousness as the specificity of human
cognitive capacities; Werner Stark for looking at the originals and helping me
with a quote from the Lectures (see note 38), and especially Luigi Caranti, for a
careful reading and valuable contributions to an earlier version of this text.
2 Just to mention some of the examples in recent discussions, see Allais (2015),
Hanna (2008), Land (2018), Leland (2018), McLear (2011) and Tolley (2016,
2018).
3 To back up my thesis, I will use several notes from Kant’s students. I will consider
them a faithful description of Kant’s considered position only if recurrent and
compatible with the published texts.
4 It is important to point out that my purpose is limited to trying to better under-
stand Kant’s use of the term “consciousness”, leaving aside the question of
whether he could accept that we share with animals another sense in which we
today might use the term.
5 Of course, we cannot say that the animal “creates” the food (nor do we, by that
matter, in that sense make real the objects of our representation). I take that,
because Kant is here talking about the faculty of desire, the representation of the
object (in this case, food), is represented connected with the feeling of pleasure,
and so the object the action makes real is, say, the “food pleasing the animal”.
6 I have modified the translation of “Unterschied” from “distinction” to “dif-
ferentiation” in order to avoid confusion with “distinction” in the sense of
“Deutlichkeit”.
7 I used “differentiation” for “Unterscheidung” (see previous note).
8 In the case of the Critique of Pure Reason, I will make reference only to the first
and second edition page numbers and, in the case of other Kantian works, to
the volume and page of the Akademie-Ausgabe. The English translations used are
listed in the bibliographic references. I will leave aside here many difficulties in
interpreting this note of the first Critique, especially taking also into account what
we find about clarity and distinction in the Anthropology and the Lectures on
Logic (I tried to deal with some of them in Altmann 2019). For my purposes here,
we must retain the following: some obscure representations involve some degree
of consciousness as long as we are capable of some differentiation. (For an expla-
nation of why we can say that there are infinite degrees of consciousness until
extinction, see Schulting 2012, p. 293 ff. On the importance of different degrees
of consciousness in explaining the content of our cognition, see also Heidemann
2012).
9 I tried to explain Kant’s reason for ignoring this sophistication due to the subject
matters and the purposes of the disciplines anthropology from a pragmatic point
of view and general logic in Altmann (2019).
10 I have modified the translation for “erkennen” from “recognize” to “cognize”.
11 Again, I used “cognize” for “erkennen”.
12 I will assume here a general line of reading of the transcendental deduction of the
categories that can be said conventional (though by no means uncontroversial).
As an example of this line of interpretation, see Allison (2004 and 2015) (but it
is an example, because I believe that what I will assume is, albeit differences in
several aspects of different reconstructions of the argument, common to other
readings of the transcendental deduction).
13 “Only without any empirical representation, which provides the material for
thinking, the act I think would not take place” (B423 note).
76
14 Keeping in mind that the Anthropology treats obscure representations as equiva-

lent to unconscious representations.
15 A similar point also appears in Jäsche’s Logic and in the so-called Answer to
Eberhard: when we see a house from too far away to distinguish its parts, we
still see the parts of the house (otherwise we wouldn’t see the house) (see Log 9:
34 and ÜE 08: 217).
16 On the Kantian notion of unconscious representation and the relation with the
Leibnizian school, see, for example, Heidemann (2012) and Tolley (2016, 2018).
17 As Leland insists, “the representation ‘completely escapes consciousness’ in the
sense that it operates exclusively at the unconscious lever of the lower faculty”
(2018).
18 I will not be concerned here with whether we nevertheless could always become
conscious of any representation.
19 For reasons presented, for example, in Rosefeld (2019) (who on that point takes
up a line of argument from Schmitz 2015).
20 As Rosefeld writes, in a formulation that takes into account details of the tran-
scendental deduction argument: “Affection from the inside, on the other hand,
consists in the causal impact of the activity of the understanding that affects
sensibility by synthesizing the manifold provided by outer sense. This affection
does not result in meta-representations with an entirely new content but rather
adds temporal features to the old content” (2019).
21 As Grandjean points out: “in order for me to become conscious of the content of
outer sense, it is necessary that my inner sense be affected by this content” (2015:
78, my translation). (The author further maintains that it would be a mistake to
reduce all the role of inner sense to this self-affection by the action of apprehen-
sion of the understanding, but, for my purposes, what interests me is only this
aspect of inner sense.)
22 As we read in one of the Lectures: “Every apprehension (Auffassung) of objects
of the senses is an action of the mind by which man affects himself (my transla-
tion) (V-Met-K2/Heize 28: 713).
23 However the argument of the transcendental deduction of categories is recon-
structed, it goes through something completely ignored here, namely the role of the
concept of object in general, without which, according to Kant, self-attribution
of representations would never be possible. I will briefly come back to this point
at the end of this chapter.
24 Again, leaving aside the question of whether it is possible to talk about representa-
tions that are nothing to me, that is, that could not be accompanied by “I think”.
25 See V-Met/Mron 29: 882: “Apperception is the ground of inner sense”.
26 In fact, it involves much more, since cognition depends on the concepts of object
in general, but I would like to stress here a necessary (in no way sufficient) condi-
tion of cognition: to be able to represent representations.
27 It is worth noting that this observation is not incompatible with a note in the
Anthropology, where Kant suggests that an orangutan or chimpanzee could
develop the necessary structure for understanding (see Anth 7: 328 note). For
that, the note makes clear, there would have to be an upheaval in nature, where
orangutans developed new organs – that is, it would not be enough for animals
to live eternally with the capacities that, according to Kant, we must attribute to
them as they are now.
28 In mere perception, we would simply have an “appearing”, or appearance or
phenomenon or indeterminate object (Erscheinung in this sense) as opposed to
an appearance or phenomenon in the sense of a determinate object of cognition
(Erscheinung in this other sense or Phaenomenon). For this distinction and a sug-
gestion about the implications of appearance being in space as to their relation
77
with apperception, see also Fonseca (2017). On the peculiarity of the role of
categories as a condition of perception in contrast to their role in judgments, see
Torres (2008).
29 About the sense in which, without concepts, our intuitive representations of
objects are blind, see Zöller (2013).
30 As already noted by, for instance, McLear (2011), who nevertheless thinks the
passage lends support to attributing consciousness to animals.
31 Passages similar to this one in other logic lectures do not say explicitly that they
are about the objective content of cognition but are nevertheless also about what
is necessary to produce cognition: “degrees of representations of all cognitions”
(V-Log/Blomberg 24: 132), “degrees of distinctness” (V-Log/Wiener 24: 846),
“produce a cognition” (V-Log/Dohna 24: 732).
32 This description is oversimplified (because it does not take into account the role of
the unity of apperception and concepts of object in general), but it is enough for my
point here, that is, to underline the relation between perception and apperception.
33 Besides acknowledging that the passage deals with human cognition, McLear
(2011) also considers the possibility of animals having only degrees one and
three but dismisses this interpretation as artificial, partly by taking the citation
as describing steps in cognitive development and partly supporting other texts
(which, however, do not seem decisive to me either).
34 Hannah Ginsborg (1997, 2006) explores this peculiarity of humans of not only
associating but being conscious of the association.
35 Once again, leaving aside the question of whether it is or is not necessarily associ-
ated with some empirical conceptual determination.
36 I modified the translation of “Erkentniss” to “cognition” instead of “knowledge”
and used “cognizing being” for “erkennendes Wesen”.
37 I believe that Leland (2018) is quite convincing in noting that Kant’s assumption
of the consciousness of scattered perceptions to myself is an additional assump-
tion to the consideration of myself as an animal, and therefore the passage does
not imply the attribution of consciousness to animals; quite the contrary.
38 I have slightly modified the English translation of the last sentence given in the
Cambridge Edition, which was the following: “We ought, therefore, to have two
different names for these, but for this [faculty] there is only one, namely, the
reproductive power of imagination; for this is called memory when accompanied
by apperception”. The German text in the Academy edition is the following:
Alle diese 3 Erkenntniss Vermögen [Sinn, Reproductions Kraft der Seele oder der
Imagination, Facultas praevisionis] können mit Apperception begleitet werden
oder nicht. Sind sie es, so gehören sie bloss dem Menschen zu, wo nicht – so haben
sie auch die Thiere. Wir sollten daher von iedem 2 verschiedene Namen haben, aber
das ist nur von einem, nehmlich der reproductiven Einbildungskraft; denn diese
heißt, mit Apperception begleitet, Gedächtniss”. Werner Stark suggested the fol-
lowing slight change in the transcript from the manuscript: “Wir sollten dafür v.
iedem 2 besondere Namen haben aber das ist nur v. einem nehmlich der reprod.
Einbildkr; denn dieß heißt, mit Apperception begleitet Gedächtniß” (where v. =:
von/vor [i.e. für]).
References
Allais, L. 2015. Manifest Reality: Kant’s Idealism and His Realism. Oxford: Oxford
University Press.
Altmann, S. 2018. “Representações Sensíveis e Apercepção Originária”, Studia Kan-
tiana, 16(6): 27–45.
78
Altmann, S. 2019. “Consciência como Base para a Distinção entre Forma de Cog-
nição em Kant”, in E. M. Rocha, L. Levy, E. da R. Marques, M. A. Gleizer, and
L. C. Pereira (eds.) Caminhos da Razão. Rio de Janeiro: Nau Editora: 203–26.
Conant, J. 2016. “Why Is Kant Not a Kantian”, Philosophical Topics, 44(1): 75–125.
Fonseca, R. D. 2017. “Percepção Objetiva, Espaço Egocêntrico e Não-Conceitualismo
Kantiano: Uma Primeira Aproximação”, Kant E-prints, 12: 104–30.
Ginsborg, H. 1997. “Lawfulness Without a Law: Kant on the Free Play of Imagina-
tion and Understanding”, Philosophical Topics, 25(1): 37–81.
Ginsborg, H. 2006. “Thinking the Particular as Contained Under the Universal”, in
R. Kukla (ed.) Aesthetics and Cognition in Kant’s Critical Philosophy. Cambridge:
Grandjean, A. 2015. “Uma Anterioridade do Mundo: sobre o Sentido Interno como
‘Apercepção Cosmológica’”, Analytica, 19(1): 69–91.
Hanna, R. 2008.“Kantian Non-Conceptualism”, Philosophical Studies, 137(1): 41–64.
Heidemann, D. 2012. “The ‘I Think’ Must Be Able to Accompany All My Represen-
tations”, in P. Giordanetti, R. Pozzo, and M. Sgarbi (eds.) Kant’s Philosophy of the
Unconscious. Berlin: De Gruyter: 37–60.
Land, T. 2018. “Conceptualism and the Objection from Animals”, in V. Waibel,
M. Ruffing, and D. Wagner (eds.) Natur und Freiheit: Akten des XII. Internation-
alen Kant-Kongresses. Berlin: De Gruyter: 1269–76.
Leland, P. R. 2018. “Kant on Consciousness in Animals”, Studi Kantiani, 31: 75–107.
McLear, C. 2011. “Kant on Animal Consciousness”, Philosophers’ Imprint, 11(15):
1–16.
Rosefeld, T. 2019. “Kant on Imagination and the Intuition of Time”, in G. Gentry
and K. Pollock (eds.) The Imagination in German Idealism and Romanticism.
Cambridge: Cambridge University Press: 48–65.
Schmitz, F. 2015. “On Kant’s Conception of Inner Sense: Self-Affection by the Under-
standing”, European Journal of Philosophy, 23(4): 1044–63.
Schulting, D. 2012. “Non-Apperceptive Consciousness”, in P. Giordanetti, R. Pozzo,
and M. Sgarbi (eds.) Kant’s Philosophy of the Unconscious. Berlin: De Gruyter:
271–304.
Tolley, C. 2013. “The Non-Conceptuality of the Content of Intuitions: A New
Approach”, Kantian Review, 18(1): 107–36.
Tolley, C. 2016. “Between ‘Perception’ and Understanding, from Leibniz to Kant”,
Estudos Kantianos, 4(2): 71–98.
Tolley, C. 2018. “The Meaning of Perception in Kant and His Historical Context”,
in V. Waibel, M. Ruffing, and D. Wagner (eds.) Natur und Freiheit: Akten des XII.
Internationalen Kant-Kongresses. Berlin: De Gruyter: 3244–55.
Torres, J. C. B. 2008. “Determinação Categorial e Síntese da Apreensão”, Studia
Kantiana, 6(7): 62–81.
Zöller, G. 2013. “Not Seeing and Seeing Nothing. Kant on the Twin Conditions of
Objective Reference (Não Vendo e Vendo nada. Kant sobre as Duplas Condições
de Referência)”, Kant e-Prints, Series 2, 8(2): 1–21.
79
5
CRITIQUE AND DEVELOPMENT
OF THE KANTIAN THEORY OF
SPACE IN GEROLD PRAUSS
Bernd Dörflinger1
1. Introduction
In 2015, Gerold Prauss published a monumental transcendental-philosophical
theory of the space-time continuum. In style and content, this is not some-
thing that could be classified as secondary literature on Kant. Instead, it is the
work of someone who thinks for himself in an original and innovative fashion,
although he conducts his self-thinking activity through a continuous, detailed
examination of Kant’s texts – eventually going beyond these texts but still more
oriented towards them than someone who merely wants to know what Kant
may have meant in this or that passage. Prauss’s ambition is no less than that of
having completed what was left uncompleted in Kant or having arrived at the
place to which, as Prauss says, Kant’s “transcendental philosophy . . . was only
on its way” (2015: 236). The license to such a project of completion is provided
by Kant himself, who stresses, in light of all “present attemps[.]” in philosophy,
the “reservation of the right of reason” to “investigate and to confirm, or to
reject” these attempts in their “sources” (A 838/B 866), according to what
presents itself as a general principle in them. Downright famous is the passage
where Kant declares that it is not at all unusual with regard to an “author . . .
to find that we understand him even better than he understood himself, since he
may not have determined his concept sufficiently and hence sometimes spoke,
or even thought, contrary to his own intention” (A 314/B 370).
The special general principle which will be our focus in what follows is
supposed to be space, that is, the a priori and hence strictly universal and
necessary formal principle according to which the outer sense intuits that
which is outer in an external world. Prauss calls his transcendental phi-
losophy, which is according to his ambition completed under this aspect,
“fundamental geometry” (2015: 236). The fundamental aspect of this tran-
scendental doctrine of space is supposed to be that it reaches back to the
central place of the origin or the genesis of the formal intuition of space.
80 DOI: 10.4324/9781003050742-6
K A N T I A N T H E O RY O F S P A C E I N G E R O L D P R A U S S
Insofar as Kant’s theory is viewed as standing in need of supplementation

with regard to this point, the criticism is (still in general terms) that Kant did
not advance to this origin but rather began from subsequent, secondary or
derivative aspects of the intuition of space.
In a transcendental philosophy that considers ways of cognizing or rep-
resenting objects prior to all talk about objects (A 11f./B 25), the aforesaid
central place of the origin of the formal intuition of space can be no other
than the subject of cognition or representations. It can thus be expected that
the disagreement between Kant and Prauss will not least concern the ques-
tion of how precisely one ought to understand the subjectivity of space as
a form of intuition. Kant expresses this subjectivity in a manner that is not
fully clear, for instance, when he says that the form of intuition pertains “to
the subjective constitution of our mind, without which” the predicate, here
that of spatiality, could be “attributed to no thing” (A 23/B 38). Prauss – this
much can be anticipated – will oppose the somewhat ambivalent talk about
the constitution pertaining to the subject with something definite, namely
with the thesis that space has its origin in the subject in the sense of being
produced or generated by the subject.
2. Criticisms of the Kantian theory of the space

continuum
What blocks Kant’s recognition of the original disclosure of space by the
subject is, according to Prauss, that Kant ascribes to space from the begin-
ning a structure consisting of parts; this means that spatial extension is
always already being presupposed, and thus the origin of the generation
of spatial extension cannot become thematic. Kant conceives space (and
also time) from the outset as continuous magnitudes, and he determines
his concept of spatial (and temporal) continuity (following, by the way, the
Aristotelian Physics) as follows:
The property of magnitudes on account of which no part of them

is the smallest (no part is simple) is called their continuity. Space
and time are quanta continua, because no part of them can be given
except as enclosed between boundaries (points and instants), thus
only in such a way that this part is again a space or a time. Space
therefore consists only of spaces, time of times. Points and instants
are only boundaries, i.e., mere places of their limitation; but places
always presuppose those intuitions that limit or determine them,
and from mere places . . . neither space nor time can be composed.
(A 169/B 211)
Although in another place Kant qualifies the aforesaid thesis that space con-
sists of spaces and is thereby a composition of partial spaces, but he does
81
BERND DÖRFLINGER
not completely repeal it. He says here: “Properly speaking, one should call
space not a compositum but a totum because its parts are possible only in
the whole, and not the whole through the parts” (A 438/B466). The qualifi-
cation consists of the statement that the entire space cannot be exhausted or
measured through the composition of subspaces. Nevertheless, the remain-
ing presupposition is that the totum, albeit inexhaustible, is supposed to
internally display the structure of the composition from parts. Immediately
after the cited passage concerning the totum, Kant expresses his view that
this composition is still assumed as a necessary condition of spatial exten-
sion. Here he says that there could, “if I remove all composition from it
[space], nothing, not even a point, be left over; for a point is possible only as
the boundary of a space (hence of a composite)” (ibid.).
Kant’s main thesis concerning the space-continuum is according to the
aforesaid: Space consists of subspaces, regardless of the fact that space can-
not be completely measured through these subspaces and that none of these
subspaces can be considered the smallest. Essential parts of Kant’s definition
of this continuum are unextended points as boundary – intersection – or
division points, as well as extension between such points; nothing here indi-
cates that one of these elements is supposed to have priority. Kant seems to
consider them coequal in the sense of mutual conditions, so that points as
unextended boundary points are supposed to have spatial extension as their
condition and, conversely, such extension is supposed to be possible only
under the condition of discrete subspaces that are bounded by being punctu-
ated. It is these thematic relationships that Gerold Prauss discusses and with
regard to which he expounds problems, in order to eventually reject Kant’s
concept of the space continuum, although, as was indicated earlier, this is not
a rejection in every respect – rather, it is an opposition against the idea that
the original intuition of space could be comprehended through the concept of
continuity that has been sketched previously. The questions that are thereby
raised and call for answers are, essentially, the two following ones: (1) Which
concept of continuity is adequate to the original formal intuition of space?
and (2) What leads to that secondary representation of space which attaches
to the original formal intuition of space and with regard to which the Kantian
concept of continuity need not be rejected but can be retained as valid?
With regard to answering the first question, one must consider the essen-
tial aspects of what Gerold Prauss expounds in detailed fashion concern-
ing the previously mentioned relationship between the punctuated nature
of spatial boundaries and spatial extension. In the given context of a prin-
cipled clarification of this relationship, one distinction, which otherwise is
quite important, can be neglected: namely whether this punctuated nature,
as the punctuated nature of the boundary point of a one-dimensional line (as
the punctuated nature of the line itself), bounds the two-dimensional surfaces
or whether this punctuated nature is eventually conceived as the punctuated
nature of the surface which functions as boundary in three-dimensional space
82
(Prauss 2015: 108).2 The one-dimensional line, which is a boundary between

surfaces in two-dimensional space, and the two-dimensional surface, which
bounds as surface area in three-dimensional space, both retain an essentially
punctuated character, namely the character of being non-extended, without
which they could not be boundaries in the strict sense. At any rate, accord-
ing to Prauss, there is no coequal relationship between punctuated character
and extension in the sense of mutual conditions; rather, there is a unilateral
conditioning relationship according to which spatial extension is the condi-
tion of boundaries and consequently of discrete subspaces that exist in spatial
extension but where boundaries and subspaces are not conversely conditions
of spatial extension. This is the result of his thought experiment (Prauss 2015:
44–6) where one tries to remove, on the one hand, the boundaries and, on the
other hand, the extension from the intuition of space. In the case of removing
the boundaries, this intuition is retained as the representation of undivided
extension, whereas in the case of removing the extension, this intuition col-
lapses entirely. In order to spotlight the primordial intuition of spatial exten-
sion, one must focus on what lies in between or external to the unextended
boundaries and is, for its part, undivided and indeterminate as condition of
division and determination. Insofar as there is always something that lies
in between and external to two boundaries, the original spatial extension
can also be called infinite under these two aspects. With this, the essential
moments of the Praussian concept of continuity (as relating to the original
intuition of space) are brought together. According to this concept, the origi-
nal space continuum is the continuum of the undivided “unlimited-infinite
space”, which Prauss at the same time calls the “epitome of the infinite” and
which he designates as the “precondition of the discrete-determinate” (2015:
43). The latter has merely the secondary status of something conditioned.
The concept of continuity sketched previously is obviously opposed to the
Kantian-Aristotelian one, according to which space is from the beginning,
that is, already at the level of the fundamental intuition of space, supposed to
be characterized through its part-structure. Prauss’s concept of continuity is
specifically opposed to Kant’s already cited assertion (which is supposed to
concern what space always already is) that space consists of spaces (Prauss
2015: 47; A 169/B 211). If the original intuition of space concerns the inde-
terminately undivided space, then this space cannot consist of that for which
boundaries that have a punctuated character are conditions, namely pre-
cisely for subspaces. To repeat: subspaces, including their boundaries, which
have a punctuated nature, are not wholly rejected, and we are still await-
ing the explanation of their secondary and derivative character. The cru-
cial point is, rather, that there is something being presupposed in subspaces,
namely precisely the intuiting of spatial extension that is by itself in no way
interrupted and that by itself does not encounter any boundaries.
If, according to this presupposition, the representation of undivided spatial
extension has, as the prior and fundamental representation, an independent
83
BERND DÖRFLINGER
character, whereas the representation of subspaces has a strictly dependent

character, then it follows that that which lies at the basis of all counting,
measuring and calculating in relation to space – namely discrete subspaces –
presupposes something which eludes all such counting, measuring and cal-
culating. Expressed in Kantian terminology: the application of the categories
of quantity presupposes something which, although fundamental for this
application, also eludes the reach of this application. In Kant, this distinction
seems to be blurred insofar as he (this is what Prauss criticizes) conceives
space from the beginning as a quantity, namely (as we saw) as a quantum
continuum or elsewhere as an extensive magnitude.
With regard to something that has a magnitude, it must be possible to
determine a unity as a measure. Moreover, one must pose and be able to
answer the question of how many of these units, one being added to the
other, in the sum yield its magnitude. The application of this procedure for
determining magnitudes obviously presupposes discrete elements and rela-
tions among these elements. However, if the original formal intuition of
space is continuous in a Praussian sense, namely if it is in itself without
boundaries and in itself unobstructed, that is, not interrupted through any-
thing concrete, then it follows that the previously mentioned conditions for
conceiving the intuited space as a magnitude are not satisfied. If, according
to the modified concept of continuity, the original intuition of space (as dis-
closed from the side of the subject) is indeterminate and thus undivided, then
other marks which in Kant are obviously supposed to characterize space
from the beginning, and hence ineluctably and not merely in a derivative
way, do not apply to primordial space either. This concerns Kant’s concep-
tion of space as the order in which multiple things are represented as “out-
side and next to one another” (B 38) and as “simultaneous” (A 33/B 50).
Elements which relate to other elements in terms of externality, adjacency
and simultaneity cannot be distinguished in the original merely formal intu-
ition of space insofar as this intuition qua indeterminate exhibits no discrete
unities. Space as intuited in this manner offers nothing for initiating the
procedure of determining magnitudes, namely no determinate unities which
can be synthesized and combined into a multiplicity. The originally intuited
space is therefore not extensive magnitude or quantum continuum but, for
Prauss, pure quality (2015: 84 and elsewhere).
3. The spontaneous subject as the origin of the undivided

space continuum
In order to approach an understanding of that of which the qualitative (not
quantitative) unity of the intuition of space consists and of why this intuition
can claim the privilege of primordiality (over space viewed as a quantity), let us
presuppose as an example a concrete perceptual situation, namely an empirical
intuition of space. Thereupon, we can consider, on the one hand, what in this
84
perception is the formal, that which originates in the subject and that which
is a priori conditioning, and, on the other hand, what in this perception is the
material, that which is not produced by the subject, hence that which is the a
posteriori conditioned. Let the exemplary spatial intuition be that of a 100-
meter sprinter at the start, who has in view the finishing line and a tribune.
By the way, through the construction of this example, that is, through
talking about 100 meters, it is conceded from the beginning that the intuited
space can be considered a magnitude; the question is only whether quantity
belongs to its primordial constitution. In order to capture this constitution,
one must draw certain distinctions concerning the perception of our exem-
plary sprinter. The material of that which he has in view, that which is not
produced by him but which is instead based on sensibility as receptivity, are
the empirical contents corresponding to the sensations of the outer senses,
thus, for instance, the red of the track or the gray of the tribune’s concrete.
Of these contents, it can indeed be said that they appear external and adja-
cent to each other and simultaneous. In their composition, the empirical con-
tents yield the intuitable properties of the appearing objects that are present
in perception. Of these empirical objects, it can be said that they are discrete
units in relation to one another which occupy subspaces. Furthermore, some
determinate subspace that is occupied by such an object can be designated as
a measure. Equipped with this measure, for instance, the meter unit, one can
conduct a determination of magnitudes according to concepts; that is, one
can conduct the measuring and counting process that concerns subspaces.
This process consists of the addition of unit to unit and their combination
into a determinate multiplicity. In the chosen example, the answer to the
question about the quantity, that is, about the length of the distance that
must be completed until the finishing line, is 100 meters.
However, none of the aspects distinguished thus far is suitable for char-
acterizing the sought-after formal element that originates in the subject and
which is the a priori condition for the possibility under which empirical
spatial objects can appear. We can comprehend this fact through a thought
experiment which reverses the determination of perception that is due to
the empirical contents obtained by means of receptivity. In the course of
this abstraction, what vanishes are not only these empirical contents (such
as the red or gray content) but also the relations of being external, adjacent
or simultaneous that depend on the occurrence of these contents. Moreover,
along with the empirical objects, the discrete subspaces occupied by these
objects have also vanished, namely that which must be presupposed so that
one can regard space as a quantity or, to put it differently, so that one can
even meaningfully raise the question of how large something is.
However, the abstraction from the contents of perception, which at once
removes the previously mentioned spatial relations and the representation
of space as a magnitude, does not remove this perception in every respect.
What remains – going back to the sprinter example – is the outward look
85
BERND DÖRFLINGER
extending from the sprinter as the subject of perception. In light of the fact
that there are multiple outer senses, it may still for now be permitted to keep
talking about “looking”, pars pro toto as it were. To be sure, considering
outer sense a generic concept for the outer senses calls for more general for-
mulations, such as the one that Gerold Prauss sometimes uses, namely that
the subject of perception “opens” (2015: 96) itself through its perceptual
activity (something we will need to return to).
In any case, the outward look through which the subject of perception
produces outwardness as space in the first place can rightfully be called for-
mal, insofar as it does not by itself involve any content. And, insofar as this
outward look is the necessary condition for encountering empirical contents
a posteriori at all, it is the a priori condition of the possibility of appear-
ances. Since this look does not by itself produce any empirical contents
through which alone determinate partial spaces are represented, this look
as well as the original space that it generates are not divided and hence are
indeterminate. For the same reason, since that look and the original space
that lies within it do not by themselves involve anything that bars the way as
a kind of obstacle and thereby introduces finitude (as is eventually the case
at the secondary level where empirical contents occur), they extend into the
open, towards infinity. Representations of spatial magnitude must, however,
be kept distinct from all these characteristics of the Praussian original space,
because it, qua merely formal, undivided-indeterminate and infinite, offers
no occasion for even meaningfully posing the question of quantity, that is,
the question concerning how many units there are.3
The quantitative indeterminacy or, put in positive terms, the qualitative
unity of the space that originates in the subject is retained even once empiri-
cal contents have entered into formal space, that is, once the question con-
cerning quantity becomes meaningful at the subsequent level of appearances
through which subspaces are occupied materialiter. This question remains
meaningless, according to Prauss, with regard to “the perceptual conscious-
ness” (2015: 142). This means that it would be meaningless to assume, with
respect to the aforesaid 100-meter sprinter, that his look from the start to
the finishing line is a look that exhibits the length of 100 meters. The esti-
mation of quantity becomes possible only once appearances enter into the
per se undivided, and thus on its part wholly non-quantitative, look – once
these appearances, as Prauss puts it, emerge “as discretions vis-à-vis discre-
tions” within the continuum that originates in the subject (2015: 143). If the
sprinter abstracts from himself as the subject of perception and talks about
his appearing body, he can rightfully say that he is 100 meters away from the
finishing line. But the perception of his body as an object appearing in space
presupposes precisely this subject of perception, which discloses space in the
first place through its outward-directed perceptual activity.
The subject itself and as such cannot, qua origin and condition of spatial
extension as the conditioned, exhibit traces of this conditioned. Hence, the
86
subject is to be conceived as non-extended and thus in a special sense as hav-

ing a punctuated character (Prauss 2015: 15, 218 and elsewhere). A special
sense of a punctuated character is needed here, because the punctuated nature
of the previously mentioned boundary – or intersection point, which already
presupposes extension (that is, a quantitative extension which is divided into
discrete unities) – cannot be at issue (Prauss 2015: 116). Thus, as Prauss says,
“the relation of intersection among point and extension . . . [is] not the only
and original one” (2015: 15). The punctuated character of the subject of per-
ception precedes all intersection relations. From this it follows that no place
in space can be attributed to the subject if it is considered as such. In order
to consider the subject of the perceiving consciousness as such, it is required
that one not already represent it as connected with the body, which for its
part already presupposes (again, discrete quantitative) spatial extension. Of
course, the body qua spatial appearance is at some place.
Prauss’s view that the subject of perception is without a place is a radi-
calization, compared to Kant’s doctrine, since Prauss goes back to the inner
structure of subjectivity where spatial extension cannot at all be presupposed
in order to reach the point where spatial extension is generated. Certainly, as
we know, in Kant, the subjectivity of the form of intuition is asserted as well.
The “Transcendental Aesthetic” states accordingly: “For in order for certain
sensations to be related to something outside me . . . the representation of
space must already be their ground” (A 24/B 38). If this could be interpreted
as meaning that the active subject contributes the outward relation to the sen-
sation which belongs to inner sense and which is thus non-spatial, so that the
external world arises in the first place, then the subject would be originally
disclosing space in Kant’s doctrine as well. However, Kant’s further elabora-
tion of this relation does not support this interpretation. Kant’s further elabo-
ration is that the sensation is referred “to something in another place in space
from that in which I find myself” (ibid.). The assignment of a place to the
subject which is expressed here already presupposes the subject as something
corporeal and thereby presupposes space. According to this view, space must
be considered something that is discovered as complete, not something that
is generated. We will see later, when we look beyond the only seemingly com-
plete and final doctrine that one finds within the “Transcendental Aesthetic”,
that the notion of the generation of space was not at all foreign to Kant either.
According to Gerold Prauss, space as the form of intuition is in no respect
something that is discovered as wholly complete and therefore static. Rather,
it is the product of a dynamic “extending-oneself of a subject” (Prauss 2015:
96). Prauss, who presents his theory of the continuum as an “insight into the
inner structure of subjectivity” (2015: 141), holds that formal space – where,
again, we can think pars pro toto of the look of a subject of perception – is
produced from within this subject as an undivided and hence indeterminate,
internally continuous extension which proceeds towards the infinite open.
Insofar as this extension is grounded in the subject’s inwardness, that is, in
87
BERND DÖRFLINGER
its special, not already spatial kind of punctuated character, the extension at
issue here is that of the “exten[ding] . . . act[ing] point” (Prauss 2015: 134f.).
If extension is to be comprehended according to its origin, then it must be
(in Prauss’s view) the “extension of something”, “which can only be a non-
extension” (2015: 110). For, although there is another kind of genesis of
extension, namely that of enlargement, the latter obviously proceeds from an
already presupposed extension and must therefore miss the original genesis
of extension.
The fact that for Prauss, the “original-a priori production of that exten-
sion of . . . space” (2015: 148) (and, by the way, of time as well) is explicitly
grounded in “subjectivity as spontaneity” (2015: 96) does not contradict
Kant’s repeated claim that the forms of intuition are forms of receptivity.
According to Prauss, “subjectivity as spontaneity” must “lie at the basis of its
receptivity, because only in that way can it reach from within itself beyond
itself towards other things: to an external world” (2015: 96). According
to this conception, “something like a receiving of this or that content” can
“take place” (Prauss 2015: 96) only subsequently as part of the formal intu-
ition of space that has been produced antecedently and spontaneously by
the subject. Put differently, it is only under the presupposition of the subject
opening itself that that which Kant calls affection can take place, from which
empirical intuition results that can eventually be elevated to experience by
the judging understanding (Prauss 2015: 97).
4. The beginnings of a modified concept of

continuity in Kant
The presentation of Kant criticism that has been provided thus far has essen-
tially been concerned with Kant’s concept of continuity (which goes back
to Aristotle), according to which space is presupposed from the beginning
as consisting of subspaces and, relatedly, is from the beginning presumed to
be a quantity.4 According to Prauss, both of these features obstruct the pos-
sibility of going back towards the origin of the representation of space, that
is, towards the inner structure of a subjectivity that actively produces space.
However, he does not miss and indeed appreciates the fact that one can find
in Kant (in the Critique of Pure Reason but also going beyond that work)
approaches towards further modifications and developments, which point
in precisely this direction. These approaches, as is yet to be shown, indeed
exist and extend to the opus postumum. They include passages where Kant
presents spatial representing at a very fundamental level, as depending on
active synthesis and thereby on spontaneity. For instance, there is the follow-
ing passage: “I cannot represent to myself any line, no matter how small it
may be, without drawing it in thought” (A 162/B 103). Kant calls the syn-
thesis that is operative here the “synthesis of the productive imagination, in
the generation of shapes”, upon which the entire “mathematics of extension
88
(geometry) with its axioms” (A 163/B 204) is supposed to be grounded.

Without synthesis, according to the A-Deduction, we can represent neither
the “line in thought” nor the “time from one noon to the next” (A 102) and
hence neither spatial nor temporal extension; without synthesis, none of
these “thoughts, not even the purest and most fundamental representations
of space and time could ever arise” (ibid.).
Despite all the appreciation of the synthesis that Kant newly introduced,
Prauss’s verdict, which is based on the closer consideration of Kant’s specifi-
cations and explications regarding this synthesis, is that Kant in the end got
stuck halfway in between. Still generally speaking, this is supposed to mean
that, for Kant, the productivity of the subject with respect to the disclosure
of the fully developed spatial representing is only a necessary but not a suf-
ficient condition, or that Kant presupposes something further with regard
to this representing that is external to the productivity of the subject. This is
also clear from the fact that Kant, as Prauss discusses in great detail (2015:
72–82 and elsewhere), comprehends synthesis as a combining activity (A
162/B 202), which apparently presupposes something given that is consti-
tuted as a non-composite and hence disparate plurality. To this plurality, one
would have to add through synthesis only such a unity that yields a composi-
tion of what was previously disparate.
Kant does not provide a wholly unified account of that which is presup-
posed by the act of composition. Sometimes he simply calls it the manifold
(B 160), which in the case of space arguably leads inevitably to the represen-
tation of disparate places or space points. The synthesis would have to be
applied to the latter in order to yield the unity of the composition as a result.
This account accords with the expressions through which Kant repeatedly
characterizes synthesis. Synthesis is here regarded as an act of going through,
taking up, combining or connecting, of gathering and unifying (A 77f./B 102;
B 164). Even without a precise presentation of how one must conceive the
cooperation of all these activities (which is perhaps impossible), it is clear that
that which precedes this synthesis, namely the plurality of disparate space
points, already presupposes the representation of spatial extension. This is
even more obvious from the fact that Kant sometimes suggests that the mate-
rial for the synthesis of composition already involves spatial parts (A 102).
Prauss rightfully points out that such an act of combining is an act of
extending qua enlarging, namely “an extending of extension which already
exists” (2015: 109). This also applies to Kant’s account of the aforesaid
drawing of a line. When Kant says that all of its parts, thus including the
first part, are generated from a point (A 162f./B 203), this point is already
being presupposed as a space point and is therefore not regarded as that
subjectivity which has a punctuated character. For insofar as this subjectiv-
ity is supposed to be the radical origin of spatial representing as such, it
cannot be thought as already being at a place. In the end, all variations of
the synthesis of the composition already presuppose the representation of
89
BERND DÖRFLINGER
spatial extension so that the question concerning the original genesis of this
representation cannot be raised in this context but remains concealed here.
As has already been indicated, Kant went beyond the position that has
been sketched and whose problems have been expounded previously after all,
namely in the opus postumum and in late Reflexionen. There he approaches
the notion of a formal intuition which has as a whole been generated sponta-
neously through subjectivity and the notion of a receptivity which was made
possible through this subjective positing. With respect to both space and time,
he talks about “product[s] of my capacity of representation”, about “forms
of sensible intuition . . . where the subject exercises powers” (OP 22: 70), even
about forms of the “self-determination of the subject” (OP 22: 74). Similarly,
he says: “The capacity of representation proceeds from within through some-
thing that it [in the German: sie; perhaps better: es; Kant is probably thinking
of “facultas”] posits itself . . . the space and the time of intuition” (OP 22: 73).
Concerning the receptivity that is made possible through such spontaneity,
one reads: “The receptivity for the cognition (receptivit.) is grounded on the
faculty to produce it within oneself” (OP 22: 52).
The late Kant also expresses the qualitative, and hence not quantitative,
unity of the one singular infinite space, which he now designates as the “origi-
nal space [.]” (20: 419), insofar as it is the condition for subspaces within
it with regard to which alone the concept of quantity can be meaningfully
applied. According to Kant, “the unity [is] not always that which upon fre-
quent repetition constitutes quantity” (Refl 18: R 5729). He means the quali-
tative unity, which he both distinguishes from and the same time relates to
the quantitative unity as follows: “Unity used in the singular is qualitative,
used in the plural quantitative. Qualitative unity is to be considered like the
ground of the whole, quantitative like a part of the whole” (Refl 18: R 5663).
The passages which have been collected in the concluding part of this chapter
concern the formal intuition of space which is wholly and without presupposi-
tion grounded in spontaneous subjectivity, as well as the original qualitative
unity of this intuition. These passages may be taken to prove that Kant himself
yet prepared the ground for a further development of the doctrine of space
that can be derived from his published works, a doctrine which is only appar-
ently complete and consistent. Since these Kant passages have indeed been
collected partly from remote corners, and since they yield a sketch rather than
a fully developed system, it is all the more pleasing that with Gerold Prauss,
a contemporary philosopher did present such a system and made progress on
the path where Kant, as Prauss says, “was on the way” (2015: 236).
Notes
1 Translated into English by Markus Kohl, University of North Carolina, Chapel
Hill. The translator would like to thank Karl Ameriks and Stephanie Basakis for
their helpful advice and suggestions.
90
2 Thus, according to Prauss (2015), three-dimensional space can no longer be rep-

resented as the boundary or (respectively) the surface of a further space and hence
no longer exhibits anything that corresponds to the character of a point to which
the extension of a further dimension could be connected. However, despite the
fact that Prauss justifies these claims in an innovative manner which is not drawn
from Kant’s texts, his resulting view is in agreement with Kant. His view asks us
to consider three-dimensional Euclidean space, just as Kant did, that space which
adequately corresponds to the pure formal intuition of space that is represented
by the subject of transcendental philosophy (cf. Prauss 2015: 218f.). The status
of Euclidean space and Euclidean geometry has been subject to a still-ongoing
controversial philosophical discussion, which was initiated after Kant’s time
through developments in mathematics (such as the non-Euclidean geometries of
Carl Friedrich Gauß and Bernard Riemann) and physics (in the first place through
Albert Einstein’s general theory of relativity). One central issue among those that
are raised in this debate concerns the rectilinearity of the expansion of space that
is presupposed in Euclidean geometry. This presupposition, it seems, is at least
relativized through non-Euclidean geometries (as an element of just one among
multiple possible geometrical models) and is declared false by the general theory
of relativity, according to which space is curvilinear. In Kant scholarship, there are
(broadly speaking) two camps which have reacted differently to these challenges.
A prominent representative of the first camp is Michael Friedman 1994, who
argues that Kant would have had to revise his thesis that Euclidean geometry (the
only geometry known in his time) is a priori valid in light of the later insights of
mathematics and the physics of space. Michael Wolff can be regarded as a protag-
onist of the other camp (2001). Wolff defends the validity of Euclidean geometry
for the pure intuition that is the focus of Kant’s theory; by contrast, he classifies
the general theory of relativity as an empirical theory, which transforms geometry
“into a kinematics of light signals and systems of reference”; according to Wolff,
one is not forced “to choose between Euclidean and non-Euclidean geometry with
regard to their validity” (2001: 227), since they consider space in wholly different
respects.
3 The explanations offered thus far allow us to place Prauss’s theoretical approach
within the context of a contemporary controversy among Kant scholars regarding
the relationship between understanding and sensibility or, respectively, concept
and intuition. This is the controversy between so-called conceptualists and non-
conceptualists. Among the central claims of the conceptualists is the denial that
intuiting has a self-contained and independent role to play in the generation of
cognition. Instead, they stress the dependence of even intuiting (as such) on the
understanding, that is, on the capacity for concepts. Passages in Kant that seem
to support this view include, for instance, those where he declares that “intuition
without concepts” is “blind” (A 51/B 75) or those where he claims that “even
the manifold in a given intuition” stands “necessarily under concepts” (B 143).
Protagonists of conceptualism include, for instance, Robert Pippin und Hannah
Ginsborg (Pippin 2013: 91–109; Ginsborg 2008: 137, 65–77). There are, how-
ever, passages in Kant with a contrary implication regarding the thematic problem
of the relation between concept and intuition, which force upon us systematic
considerations that go beyond textual interpretation. These contrary passages
support the main thesis of non-conceptualism, namely the claim that intuition
can present objects even independently of concepts. They include, for instance,
the passage where Kant says that “appearances can certainly be given in intuition
without functions of the understanding” (A 90/B122). The main representatives
of non-conceptualism include Robert Hanna und Lucy Allais; the latter puts spe-
cial emphasis on the independence from concepts of the intuition of space (Allais
91
BERND DÖRFLINGER
2009: 383–413; Hanna 2008: 41–64) Since for Prauss, the formal intuition of the
undetermined and infinite space is a precondition for the occurrence of discrete
empirical contents, which in turn is the precondition for the application of the
pure intellectual concept of quantity, it seems that Prauss can clearly be assigned
to the non-conceptualist camp. Dennis Schulting has made a contribution which
discusses the subtleties and difficulties of both sides (2015: 561–80).
4 In contemporary mathematics, the reigning theory of the continuum stands, in
essential respects, in the tradition of the criticized Aristotelian-Kantian concept
of continuity. This is set theory (going back to Georg Cantor), which most con-
temporary mathematicians accept as an adequate theory of the continuum. This
theory considers the continuum a set of points which possesses the cardinality of
the uncountably infinite; the real numbers correspond to the uncountably infi-
nitely many points (Knerr 1989: 267). The mentioned set of points can, accord-
ing to Prauss, be understood in no other way than the set of discrete intersecting
points (Prauss 2015: 307). However, the discreteness of intersecting points rests
on a presupposition and is in that sense secondary; it stands under the presup-
position of the original undivided and undetermined continuum, which exhibits
no correspondences to quantification through numbers (cf. Prauss 2015: 307).
Insofar as set theory begins from the discreteness of intersecting points, it misses
the original continuum in a way that is similar to how Kant’s official defini-
tion (A 169/B 211) misses it, which begins from the givenness of points and
parts (although, as we will see in the concluding remarks, there are in Kant
some relativizing considerations that go beyond his official definition). How-
ever, besides those mathematicians who stand in the still-dominant tradition of
set theory, such as, for example, Bernulf Kanitscheider, for whom “all domains
of objects can in the end be mathematized” (2013: 335), there have been other
voices as well in recent times. These include, for instance, Thomas Bedürftig,
who denies that real numbers which correspond to points can exhaust the con-
tinuum (Bedürftig and Murawski 2012: 201f.) and Rudolf Taschner, who has a
critical attitude towards the “idea of a dissection of a line into infinitely many
points” (2013: 208). For a survey of mathematical theoretical approaches, see
Langwitz (1997).
References
Allais, L. 2009. “Kant, Non Conceptual Content and the Representation of Space”,
Journal of the History of Philosophy, 47(3): 383–413.
Bedürftig, T. and Murawski, R. 2012. Philosophie der Mathematik. 2. Aufl. Berlin:
de Gruyter.
Friedman, M. 1994. Kant and the Exact Sciences. Cambridge, MA: Harvard Uni-
versity Press.
Ginsborg, H. 2008. “Was Kant a Nonconceptualist?”, Philosophical Studies, 137:
65–77.
Hanna, R. 2008. “Kantian Non-Conceptualism”, Philosophical Studies, 137: 41–64.
Kanitscheider, B. 2013. Natur und Zahl. Die Mathematisierbarkeit der Welt. Heidel-
berg: Springer Spektrum.
Knerr, R. 1989. Knaurs Buch der Mathematik. München: Droemer Knaur.
Langwitz, D. 1997. “Mathematische Modelle zum Kontinuum und zur Kontinuität”,
Philosophia Naturalis, 34(2): 265–313.
Pippin, R. 2013. “What Is Conceptual Acitivity?”, in J. K. Schear (ed.) Mind, Reason,
and Being-in-the-World. London and New York: Routledge: 91–109.
92
Prauss, G. 2015. Die Einheit von Subjekt und Objekt. Kants Probleme mit den
Sachen selbst. Freiburg and München: Karl Alber.
Schulting, D. 2015. “Probleme des kantianischen Nonkonzeptualismus im Hinblick
auf die B-Deduktion”, Kant-Studien, 106: 561–80.
Taschner, R. 2013. Die Zahl, die aus der Kälte kam. Wenn Mathematik zum Aben-
teuer wird. München: Goldmann Verlag.
Wolff, M. 2001. “Geometrie und Erfahrung. Kant und das Problem der objektiven
Geltung der Euklidischen Geometrie”, in G. Volker, R. Horstmann, and R. Schum-
acher (eds.) Kant und die Berliner Aufklärung. Akten des IX. Internationalen
Kant-Kongresses. Berlin and New York: De Gruyter: 209–32.
93
6
KANT ON COMMUNICATION
Luca Fonnesu1
1. Communication as a philosophical problem

In 1962, Jürgen Habermas gave Kant an outstanding position not just in
the history of philosophy but in the Strukturwandel der Öffentlichkeit.
With this title, the German philosopher analyzed the genesis and develop-
ment of the “public sphere” or of “public opinion”.2 More than 20 years
later, the 2015 Kant Prize recipient and distinguished Kant scholar Onora
O’Neill published a brilliant and seminal article concerning the peculiar dis-
tinction between public and private use of reason in Kant’s essay What Is
Enlightenment? (O’Neill 1989a). Both contributions were of extraordinary
importance for subsequent Kant-Forschung.
My subject, however, is not Kant’s political philosophy, but his epistemo-
logical presuppositions concerning the question of communication and the
private/public opposition. In Kant’s thought, this opposition plays an essential
role also – or already – outside and before the political domain and has, just in
the dimension of communication, probably its deepest philosophical justifica-
tion. This does not mean that I question or even underestimate the political
nature of Kant’s philosophy, demonstrated brilliantly by Onora O’Neill, who
also refers to the rich political metaphors which run through the whole criti-
cal enterprise (1989b). On the contrary, what I would like to show is the close
connection of significant epistemic notions to their political implications in
Kant’s philosophy. Beyond metaphors, the political nature of Kant’s thought
seems to be the consequence of his analysis of reason. It receives a justification
of an epistemological nature based on some concepts of his philosophy, where
this is, however, not immediately apparent. The notion of “communication”
can be found just at the beginning of O’Neill’s contribution on the public use
of reason. Since her task was to explain the distinction between the public and
the private use of reason, O’Neill’s aim was mainly to give a sound interpreta-
tion of the distinction in Kant’s 1784 essay, an interpretation that builds on
the idea that the two uses of reason are directed toward two different audi-
ences. The extent of the audience determines the public or the private char-
acter of the use of reason and therefore of communication. There is private
94 DOI: 10.4324/9781003050742-7
K A N T O N C O M M U N I C AT I O N
communication and public: the audience of the former is restricted, because

it is the communication of officers, clergy, civil servants and taxpayers, while
the audience of the latter is an enlarged one, the world at large (WA, 8: 38) In
this context, there are at least two forms of communication: the private and the
public, but only the latter is communication proper. Kant’s conception of the
private is not – x a conception of the merely individual or personal (O’Neill
1989a: 33–4). And this is, I think, absolutely correct.
I do not disagree with O’Neill’s interpretation of Was ist Aufklärung?
Her reading of the Kantian text is absolutely right. My point is rather that
Kant, in perhaps his most famous essay, uses the opposition private/public
in a way which is different from – but not necessarily in contrast with – the
most important and most fundamental application of this opposition in his
published and unpublished writings and works.
O’Neill introduces the extremely interesting distinction between expression
and communication. She understands the former to be private – as we have
seen – thanks to the extension of the audience but also thanks to the attitude
of this audience towards the subject that expresses and/or communicates some-
thing. Her aim is to offer a richer concept of toleration than the mere “liberal”,
negative one. O’Neill distinguishes between two different attitudes that an audi-
ence can have. First, the audience can consider the expressions of a person – for
example, words – merely private expression and consequently tolerate them in
a merely negative, “liberal” and indifferent way. Second, it can consider these
expressions a proper communication. Her thesis is that toleration is something
more than mere non-interference,3 and the audience has to look at other human
beings as subjects of proper communication, not of mere expression. On this
basis, O’Neill brilliantly develops Kant’s important suggestion.
I want to move in a different, although not contrasting, direction. I suggest
that one can make use of O’Neill’s distinction between expression and
communication – relating it to the private/public opposition – not from the
point of view of the audience or the receiver but from the point of view of the
subject of the communication. My proposal is to take into account the mental
states of that subject and the possibility to communicate them – what Kant calls
their communicability. This latter question – the investigation of the very pos-
sibility of communication – is therefore approached in the following pages in a
very determinate (and limited) sense, that is, by considering the nature of some
mental states as conditions of communication. While O’Neill does pay atten-
tion to the intersubjective side of communication, considering its audience, I try
to evidentiate its intrasubjective or epistemic nature, that is, those mental states
which are respectively capable of having either a private or a public character.
The latter is grounded upon their communicability. As we will see, private men-
tal states can be (e.g. linguistically) expressed but not communicated.
I maintain that long before Habermas’s theory of communicative action
and – in the analytic tradition – Grice’s classic attempt to understand com-
munication in terms of the intentions of a person making an utterance
95
LUCA FONNESU
(Habermas 1984; Grice 1957: 377–88), Kant pays great attention to com-
munication understood as a specific philosophical problem. His discus-
sion of the question in a remarkable and interesting way concerns different
aspects of his philosophy, connecting them in a kind of unified vision.
Two preliminary questions: The first has at the same time a terminological
and conceptual nature. Kant’s philosophical terminology derives – of course –
from the philosophical tradition and the German philosophical tradition in par-
ticular, which developed a language of its own in the course of the 18th century.
This is not the case for communication. Kant is the first who gave philosophical
relevance to the German word Mitteilung.4 The question of communication is
nevertheless not only a terminological one. The philosophical analysis of com-
munication is – as far as I know – a novelty of Kant’s thought, although the
problem of communication is a prominent one for the whole movement of the
Enlightenment. The very idea of an “enlightenment” has to do with communi-
cation, and Kant is one of its most representative figures5 in this sense as well.
A second preliminary problem that deserves attention concerns the exten-
sion of the communicative community. Who are its members? The question
may appear to be a trivial one, and it may in fact be. Humans are certainly
members of this community, and animals – at least for Kant – are not. But
what about God? God has – so to speak – an ambiguous status in Kant’s
philosophy. On the one hand, from the theoretical point of view, he is epis-
temologically not as accessible as every other object of traditional metaphys-
ics, but from another point of view, the practical one, moving from morality,
we can speak of his properties, and thanks to these properties, we can grasp
the highest good and its realizability.
The point of view of communication excludes the possibility of considering
God a member of this community. Kant does not directly consider the ques-
tion, but the borders of the communicative community seem to be the same
as those of the proper moral community, in the sense in which Kant defines or
determines it at the end of the Metaphysics of Morals as a doctrine of duties:
in ethics, as pure practical philosophy of internal lawgiving only the

moral relations of human beings to human beings are comprehensi-
ble by us. The question of what sort of moral relation holds between
God and the human being goes completely beyond the bounds of
ethics and is altogether incomprehensible to us.
(MS, 6: 491)
Communication has the same character, and if the moral community of

duties does not include God as a member, he cannot be a member of the
communicative community either, also because Kant has a radically negative
position with regard to prayer. The most typical “communicative” ambition
of the Christian believer with regard to God – to pray – is in fact judged in
a clearly negative way by Kant:
96
Praying, conceived as an inner ritual service of God and hence as a

means of grace, is a superstitious delusion (a fetish-making); for it
is only declaring of a wish to a being who has no need of any dec-
laration regarding the inner disposition of the wisher; through wish
nothing is accomplished.
(RGV, 6: 194)
So there is, with God, no communication at all.
2. Communication in the first Critique: conviction,

persuasion, faith6
My starting point is the consideration that, for Kant, there are mental states
that we can in a certain sense communicate and others that we cannot. This
is a very early, important and at the same time original Kantian idea, which
can already be found in the Blomberg Logic, one of the Nachschriften of
Kant’s lectures on logic, which Kant notoriously held on the basis of Georg
Friedrich Meier’s handbook Auszug aus der Vernunftlehre (1752) during
the whole of his academic career.7 This thesis is also present in the Critique
of Pure Reason, in a chapter of the “Doctrine of Method” which has been
neglected up to very recent times.8
The title of the chapter, “On having an opinion, knowing and believing”
(“Vom Meinen, Wissen und Glauben”), refers to three states that are all, in
Kant’s terminology, forms of conviction/Überzeugung. But the grounding
notion of the chapter is a word – and a concept – that most scholars (includ-
ing me) have thought to be a Kantian neologism. Recently, however, Gabriele
Gava has shown that Kant takes the word from Crusius (Gava 2019: 53–75),
an important source also for other aspects of Kant’s thought, as we know. The
notion in question is what Kant calls Fürwahrhalten, holding-to-be-true, which
doesn’t appear earlier in the Critique but plays an important role in Kant’s lec-
tures and notes on logic.9 This notion has to do with the mental attitude of the
subject in any form of (in a broad sense) cognitive act of the mind, something
that contemporary philosophers would call propositional attitude. The holding-
to-be-true is presented by Kant as “an occurrence (Begebenheit) in our under-
standing that may rest on objective grounds, but that also requires subjective
causes in the mind of him who judges” (A 820/B 848). In the Lectures on logic,
Kant proposes a correction of Meier’s definition of certainty, claiming that what
Meier is actually discussing is the holding-to-be-true, not truth. Meier gives the
following definition: “Certainty . . . is the consciousness of truth, or the clear
consciousness of truth” (Meier 1752: § 155), and Kant comments:
Here the talk is not of truth but of holding-to-be-true. This is judg-

ment in relation to the subject. We want in this case to know not
the grounds of the truth, but only those of the holding-to-be-true.
(V-Lo/Dohna, AA 24: 731)
97
LUCA FONNESU
Kant here pays particular attention to the subjective aspect of knowledge.

Let me be clear: here “subjective” does not (or: not only) refer to the tran-
scendental subject but to a particular subject and his mental acts. This trait
is apparent in another characterization of the holding-to-be-true:
Truth is an objective property of cognition, the judgment through

which something is represented as true; the relation to an understand-
ing, and thus to a particular subject, is, subjectively, holding-to-be-true.
(Log, 9: 65–6)
Having an opinion, knowing and believing are forms of conviction, we said.

The question is to understand what conviction is:
If the holding-to-be-true is valid for everyone merely as long as he

has reason, then its ground is objectively sufficient, and in that case
holding-to-be-true is called conviction. If it has its ground only in
the particular constitution of the subject, then it is called persua-
sion./Persuasion is a mere semblance (Schein), since the ground of
the judgment, which lies solely in the subject, is held to be objec-
tive. Hence such a judgment also has only private validity, and this
holding-to-be-true cannot be communicated.
(A 820/B 848)
Communicability is therefore an epistemic criterion. Not, to be clear, because

Kant maintains a conception of truth as consensus gentium but because, since
we are speaking of an individual, personal attitude, the communicability can be
a first and perhaps most important criterion for distinguishing a propositional
attitude – conviction – which depends on some common ground (reason, at
least in this case) from one, that is persuasion, which depends on the particular,
individual constitution/Beschaffenheit of the subject. Here we can make use of
O’Neill’s distinction between expression and communication: mere expression
is always possible for every kind of mental state, including a private one such
as persuasion. However, simple expression does not imply proper communi-
cation, which is in Kant’s eyes both a deeper and more difficult notion and a
deeper and more difficult action that presupposes certain mental “properties”.
In the form of persuasion, the holding-to-be-true consists of an attitude
which exhausts its validity in its own nature as a subjective and private
mental state, with no objectivity. The subjective cause or root of the holding-
to-be-true constitutes its nature, and the judgment expressed by this form
is subjective and private because it belongs to private, personal individual
features of the subject. And not only that. In Kant’s reconstruction, persua-
sion is a Schein, semblance or illusion, because not only does the judgment
have only subjective, private validity, but its ground is held by the subject to
be objective and universally valid (A 821/B 848).
98
Schein is an essential feature of persuasion: the mistaking of a subjective

ground or reason (Grund) for an objective one. In fact, there are other forms
of holding-to-be-true which do not have objective validity (for example,
having an opinion),10 but the main defect of persuasion is its illusionary
character, its taking subjective grounds to be objective. In this case, one
believes in the truth of a proposition because one takes its ground, reason
or justification to be universally valid and objective, while it is grounded in
the subject’s individual nature, psychology, prejudices, inclinations, wishes,
ignorance and so on, that is, in some kind of individual, private element of
the subject’s nature.
With psychological realism, Kant is aware of the difficulty for the subject
himself of distinguishing conviction and persuasion: if one believes some-
thing, it is not at all obvious that he is capable of understanding that this
(in a broad sense) belief is the fruit of individual conditioning factors such
as those mentioned previously or not: Subjectively, that is from our own
personal perspective, “conviction cannot be distinguished from persuasion”
(A 821/B 849). This difficulty of distinguishing conviction and persuasion
means, therefore, as we can read in the Logik-Busolt, that sometimes “some-
one believes he is convinced when he is persuaded” (V-Lo/Busolt, 24: 639).
That’s why Kant suggests some touchstones which – though problematic –
can help to understand if the mental attitude in question is a genuine con-
viction or a persuasion. This explains the essential role of communicability.
Communicability is the first touchstone to avoid persuasion:
The touchstone of whether holding-to-be-true is conviction or mere

persuasion is therefore, externally, the possibility of communicating
it and finding the holding-to-be-true to be valid for the reason of
every human being.
(A 820/B 849)11
Again: Having an opinion, knowing and believing are forms of conviction.

But the question is about believing in the most important sense, that is,
moral faith.12 This is not at all obvious, because Kant stressed already in
the Critique and in many lectures and reflections that moral faith is not
communicable. The question becomes still more complicated because of
the ambiguity of the German word Glauben or Glaube,13 and in the first
Critique, this ambiguity provokes many problems that have recently been
analyzed (for example, by Andrew Chignell 2007: 323–60). Belief is gener-
ally the form of holding to be true that is only subjectively sufficient and
is at the same time held to be objectively insufficient (A 822/B 850). I limit
myself to moral faith, which is for Kant the most important and which is
also the main subject of Kant’s considerations in the unpublished writings. It
is difficult to question the incommunicability of faith. Already in the chapter
of the “Canon”, Kant writes: “Of course, no one will be able to boast that
99
LUCA FONNESU
he knows that there is a God and a future life. . . . All knowing can be com-
municated” (A 828–829/B 856–857), a declaration that seems to imply that,
on the contrary, faith cannot, a thesis which in fact frequently occurs in the
reflections and in the lectures on logic. A few examples:
One cannot communicate to someone else his own faith (Refl.

2498, 16: 394); Faith is a conviction which is not communicable
(for the subjective grounds). Knowing has to be communicated, and
imposes agreement (Refl. 2489, 16: 391); Faith is a private holding-
to-be-true (Privatfuerwahrhalten), sufficiently certain only for me.
(V-Lo/Dohna, 24: 732)
The outcome of Kant’s understanding of faith is the reinterpretation of the

traditional expression moral certainty, a reinterpretation which attributes
this kind of certainty exactly to moral faith. The expression concludes Kant’s
argument for moral faith in the Canon but occurs many times in the unpub-
lished writings: with this expression, Kant explicitly asserts the personal,
private character of moral faith:
No, the conviction is not logical but moral certainty, and, since it
depends on subjective grounds. . . . I must not even say “It is morally
certain that there is a God”, etc., but rather “I am morally certain” etc.
(A 828–829/B 856–857)14
Notwithstanding the private character of faith, Kant is not Kierkegaard.

Moral faith does have reasons, namely moral reasons grounded in a need,
which, accordingly, Kant calls a need for reason – Bedürfnis der Vernunft.
Following Kant’s most clear formulation in the Critique of Practical Reason,
this rational need is articulated through different aspects, which we could
respectively call the existential aspect, the aspect of moral impartiality and
that expressing God’s justice (one can say: theodicy). For the highest good,
happiness is also required, and that not merely in the partial eyes of
a person who makes himself an end but even in the judgment of an
impartial reason, which regards a person in the world generally as
an end in itself. For, to need happiness, to be also worthy of it, and
yet not to participate in it cannot be consistent with the perfect voli-
tion of a rational being that would at the same time have all power,
even if we think of such a being only for the sake of the experiment.
(KpV, 5: 110)
There are therefore discursive arguments for the existence of God, which
Kant even calls “proofs”, though only moral, and proofs, of course, can be
communicated. This is the properly rational – in the sense of practical reason
– side of Vernunftreligion, rational religion. Nevertheless, the proof can be
100
insufficient in Kant’s eyes, because doubt is always not only possible but also
allowed, as Kant says on several occasions. As Bernd Dörflinger has shown
in an excellent article, morality does not conduce inevitably to religion or,
better, to faith, but up to its threshold (Dörflinger 2004). Atheism is for Kant
as dogmatic as traditional metaphysics, which, allegedly, attempted to prove
God’s existence theoretically. But agnosticism is both conceptually and mor-
ally a perfectly comprehensible position.
Kant’s position about having an opinion, knowing and believing seems to
maintain that the only form of holding-to-be-true that is not communicable
is faith, which is declared a conviction, but does not have – paradoxically –
the most important feature of conviction, that is, communicability. Probabi-
listic considerations leading to forms of theoretical belief in a non-religious
sense seem on the contrary to be communicable, but this is only one kind of
Glauben Kant takes into account in the first Critique. Glauben can refer to
both belief and faith (recall the ambiguity of the term I signaled previously).
Intuitively, it seems that the uncertain belief of a physician – this is Kant’s
own example (A 824/B 852) – can be communicated to other physicians.
Claiming the contrary would be too counterintuitive. This aspect of Kant’s
thought will be modified later: in the Critique of Judgment, Glauben will
mean only moral faith. This has probably two motives. First, Kant tries to
avoid ambiguity, all the more so for such an important and delicate ques-
tion. Second, he wants to distinguish his position from that of Jacobi, the
fideist who had attacked rationalism by appealing to exactly this ambiguity
in the term Glauben. Probabilistic considerations – so he writes in the third
Critique – can become knowing, something that faith will never become,
and a certain kind of belief, for example, historical or geographical belief,
is or has been for somebody just knowing. The highest good:
together with the sole conditions of its possibility that are conceiv-
able for us, namely the existence of God and the immortality of the
soul, are matters of faith (res fidei), and are indeed the only ones
among all objects that can be so designated.
(KU, § 91, 5: 469)
3. Communication in the Critique of Judgment:

agreeable, beautiful, sublime
It is well known, and an important subject of philosophical investigation
in the Kant-Forschung, that the question of communication – or, better, of
communicability – plays a grounding role in the aesthetic part of the Critique
of Judgment. What is important to point out here is that the field of discur-
sive thought and argumentation – such as knowing, or anything analogous
to knowing – is not the only one in which communication plays a role. In
the aesthetic field, too, the private/public opposition is of great importance
101
LUCA FONNESU
for Kant, also in this case thanks to the concept of communicability. The
language of – in the sense mentioned – publicity or, better, borrowing Onora
O’Neill’s terminology, publicizability (O’Neill 1989a: 34), is the main tool
for Kant’s distinction between different forms of Wohlgefallen. There is a
private and a public form of taste – that is, the agreeable (angenehm) and
the beautiful (schön) – and the criterion suggested by Kant for drawing
this distinction runs right along the same lines of the distinction between
the private and the public. The judgment on the agreeable “makes merely
private judgments about an object, insofar as the second [the judgment on
the beautiful] makes supposedly generally valid (public) judgments” (KU, 6:
214). The “supposedly” has of course to do with the Kantian idea of a claim
or demand or – as Paul Guyer suggests (1997: 124) – “title” to universal
validity. In this case – thanks to a distinction between subjective and private
that Kant did not apply in the first Critique, Kant identifies a subjective judg-
ment which has a title to universality, that is, which is not private. This is
the central and essential feature of the judgment on the beautiful. The oppo-
sition between subjective and objective is not a complete disjunction any
more, because a third element emerges, which seems to be new: universality
can be not only of an objective but also of a subjective sort. The judgment
on beautiful is not private (La Rocca 2003: 116, footnote 130). A proof that
Kant is aware of the novelty of his terminology – I guess – may be that the
word “private” occurs many times in these pages of the third Critique. The
novelty needs to be stressed.
The question is therefore that there are mental states that we can call
public in the sense of publicizability, exactly because they are communicable.
Among these, we have, of course, discursive thoughts of different kinds,
knowing in a general sense – including also moral knowing. But not only
that: for Kant emphasizes that feelings also or, better, certain kinds of feel-
ings, can be communicated, although on different grounds. The taxonomy
cannot be complete, because there are, for example, feelings such as delight
(Wohlgefallen) in the morality of an action. In this case, the delight,
is by contrast not a pleasure of enjoyment, but of self-activity and

of its appropriateness to the idea of its vocation. This feeling, how-
ever, which is called moral, requires concepts; and does not exhibit
a free, but rather a lawful purposiveness, and therefore also cannot
be universally communicated other than by means of reason, and, if
the pleasure is to be of the same kind in everyone, by means of very
determinate practical concepts of reason.
(KU, § 39, 5: 292)
Kant never thought that he had to give a justification of the communicability

of thoughts – this is in a certain way obvious and not problematic at all – but
he did consider it necessary to justify the communicability of a sensation,
102
as the title of § 39 shows. Here, it is sufficient to point out that there are at
least two different forms of communicability in aesthetic judgments. One
concerns the beautiful, the other the sublime. These feelings and the related
judgments are substantially different in their nature and, even more impor-
tantly, in their justification. The judgment on the beautiful is grounded in the
sensus communis, which, unsurprisingly, O’Neill suggested should be trans-
lated as public sense (1989a: 45). The judgment on the sublime, which is
also communicable, has a different justification, because “the disposition of
the mind to the feeling of the sublime requires its receptivity to ideas” (KU,
§ 29, 5: 265). The communicability of the sublime is therefore grounded in
concepts, moral concepts, and in their development. Lacking these concepts,
what we consider the sublime would assume a completely different nature
and provoke a completely different reaction in the subject:
without the development of moral ideas, that which we, prepared

by culture, call sublime will appear merely repellent to the unrefined
person. He will see in the proofs of the dominion of nature given by
its destructiveness and in the enormous measure of its power, against
which his own vanishes away to nothing, only the distress, danger,
and need that would surround the person who was banished thereto.
(KU, § 29, 5: 265)
Although we can presuppose morality in every man, there is a certain ambi-

guity in Kant’s position, which has to do with culture
But just because the judgment on the sublime in nature requires cul-
ture (more so than that on the beautiful), it is not therefore first gen-
erated by culture and so to speak introduced into society merely as a
matter of convention (konventionsmässig); rather it has its founda-
tion in human nature, and indeed in that which can be required of
everyone and demanded of him along with healthy understanding,
namely in the predisposition to the feeling for (practical) ideas, i.e.,
to that which is moral.
(KU, § 29, 5: 265)
The distinction among different forms of communicability offers a further

element for the analysis of the nature of this concept in Kant’s thought,
which in turn also illuminates the more general notion of communication.
Kant’s analysis of the sublime shows that communicability can have certain
limits, exactly because one cannot always presuppose that some conditions
are fulfilled in others, where culture gives the most relevant example in this
case. A consequence of this is that there are degrees of communicability.
To explain Kant’s general point of view, it is useful to recall once more
O’Neill’s distinction between expression and communication and my
103
LUCA FONNESU
interpretation of it. Expression is not in itself communication, and com-

munication is not just to speak or to say something. Expression can become
communication when the mental attitude and/or its content15 is a commu-
nicable one, that is, according to Kant’s perspective, when I can presuppose
certain conditions in others which permits them to be the adequate audience
of communication (reason, but not only that).
To summarize: Private mental states such as persuasion, or the agreeable
or – of course it is not the same thing – moral faith, cannot be communi-
cated. Public means first of all communicable, and the main philosophi-
cal meaning of the opposition private/public in Kant’s thought is grounded
exactly on communicability.
4. The negative side of communication: private states and

the forms of egoism
But what about private mental states and incommunicability? The field of
the “private” is a very interesting domain in Kant’s thought, which partially
concerns, so to speak, the negative side of communication, or the denial of
its possibility.
Of course, it is not at all necessary or obvious that private mental states
or experiences have to be judged negatively. Agreeable sensations or feelings
are good and, at least in many cases, morally indifferent. On the contrary,
Persuasion – Überredung – is in Kant’s opinion almost always seen in a
negative light.16 Persuasion has to do with prejudices: “Prejudice in sensu
subjectivo is the inclination to persuasion. Persuasion is the holding-to-be-
true based on merely subjective causes which are erroneously held to be
objective” (V-Lo/Pölitz, 24: 547).17 In this kind of epistemic error, we are
often motivated by our personal interest or wish:
in this way the human mind is actually subjected to very many illu-
sions and deceptions; frequently we take something to be certain merely
because it pleases us, and we take something to be uncertain merely
because it displeases or annoys us. This certainty or uncertainty is
not objective, however, but subjective.
(V-Lo/Blomberg, 24: 198)
Of course, faith is not persuasion, but this kind of holding-to-be-true involves

a risk nonetheless. Let me clarify: In this case as well, there is no possibility of
communication beyond the moral appeal to the need of reason. Furthermore,
it is very difficult – as we know – to distinguish between conviction and per-
suasion from our subjective point of view. Faith has to be moral in the deepest
sense – that is, sincere – otherwise it can become something very similar to per-
suasion, as when we believe what we like or wish. In the latter case, to believe
means to fall into a sort of minority in the Kantian sense: “Who is persuaded
104
most easily? Children. For their judgments are merely subjective. They believe
what they wish” (V-Lo/Wiener, 24: 854). And in an analogous way:
Faith – I would gladly believe what I wish for, if only I had a reason
for it. [Later addition: but it is not on that account easy, rather I
seek to persuade (überreden) myself that I can hope for that which
is good from a practically necessary point of view.] – If, however, it
is a duty to wish for something (for there is no duty to believe), then
I am right to believe it if I can.
(Refl 2503, 16: 395)
But faith also leads to another risk, which again has to do with its private
character: its solidity and firmness, which is a sort of consequence of its
incommunicability: in the case of faith, I am not at all inclined to listen to
my opponent’s grounds (V-Lo/Blomberg, 24: 148). Practical faith is often
firmer than every knowing. In the latter one listens to the reasons, but not in
the former (V-Lo/Pölitz, 24: 543).
Finally, faith opens the door to a third risk, that of egoism, a notion that
occurs not only in Kant’s lectures on logic and anthropology (Hinske 1999:
101 f.) but also in the published Anthropology (1798). Egoism is the per-
manent risk of every individual, private “conviction” which cannot be com-
municated and shared with others. The Kantian interpretation of the notion
of “egoism”, which also occurs in Wolff and Meier, is of great interest in this
context. In the Anthropology, Kant offers a taxonomy of egoism where the
most interesting part, from the epistemological point of view, is its logical
form: “The logical egoist considers it unnecessary also to test his judgment
by the understanding of others; as if he had no need at all for this touchstone
(criterium veritatis externum)” (Anth, 7: 128) (where, again, the echo of the
distinction between conviction and persuasion is noteworthy). As we have
seen, in Kant’s view, faith avoids, given its very nature, intersubjective dia-
logue. Therefore, it needs all possible certainty and Gewissenhaftgkeit in order
not to become logical egoism. Egoism is not only logical: Kant’s taxonomy
includes an aesthetic egoism beside logical and ethical egoism, which obtains
when one is satisfied with one’s own taste. The aesthetic egoist “deprives him-
self of progress toward that which is better when he isolates himself with his
own judgment” (Anth, 7: 129). Egoism as a general attitude is the denial of
communication. On the contrary, the space of communication and dialogue is
the space of culture, of science, of knowledge, including moral – that is, ethi-
cal and juridical – knowledge and aesthetical culture, whose communicability,
writes Kant “almost infinitely increases its value” (KU, 5: 297). The leading
value – against egoism – is “pluralism: the way of thinking in which one is not
concerned with oneself as the whole world, but rather regards and conducts
oneself as a mere citizen of the world” (Anth, 7: 129). The normative applica-
tion of this perspective is the maxim of a broad-minded way of thinking in
105
LUCA FONNESU
which the subject sets himself “apart from the subjective private conditions of
the judgment, within which so many others are as if bracketed, and reflects on
his own judgment from a universal standpoint” (KU, 5: 295).
Conclusion
My purpose in this chapter was not to deal with the whole question of
communication in Kant’s thought. My point was to highlight how Kant
developed a theory of communication by distinguishing among subjective
attitudes of different kinds where, in turn, this theory is closely connected
with his idea of human reason. Communicability and communication, as
we have seen, do give a peculiar and interesting explanation of the pub-
lic/private opposition, which, of course, also has further, properly political,
implications. The private domain contains very different elements, which
can be (1) important in a negative and dangerous way – persuasion, preju-
dice, egoism; (2) substantially good and usually morally irrelevant – the
agreeable; and (3) personally of a great importance in a positive sense – faith,
when faith is sincere but, when it is not, even dangerous, precisely because
of the firmness and importance of faith: we do not die for mathematical
propositions, but we die for our faith. All of these mental states are private
because they are uncommunicable.
In fact, communicability has not only its grounding, subjective, epistemic
side. There are other factors that make thoughts communicable or incom-
municable, and it is exactly here that the connection between subjective and
objective, or political, aspects of communicability is to be found. Here, it is
the proper political outcome of ourselves as subjects that is at stake. Com-
municability is conditioned by the opportunity of concrete communication
with other human beings through writing and speech. This is why freedom
of speech, of expression, of the press are key.
Of course, it is said that the freedom to speak or to write could be

taken from us by a superior power, but the freedom to think cannot
be. Yet how much and how correctly would we think if we did not
think as it were in community with others to whom we communi-
cate our thoughts, and who communicate theirs with us.
(WDO, 8: 145)
In this sense, it is possible to distinguish between different aspects or even

meanings of communicability. Communicability – Mitteilbarkeit – is a more
general term than Mitteilungsfähigkeit. The latter, as we have seen, has to do
with subjective attitudes, the former also with external, objective conditions:
The freedom to communicate one’s thoughts, judgments [and] cog-

nitions – is certainly the only [,] most certain means to test one’s
106
cognitions properly, however, and to verify them. And he who takes

away this freedom is to be regarded as the worst enemy of the exten-
sion of human cognition, indeed, of mankind itself. For just by this
means he takes away from men the one true means they still possess
for ever uncovering, becoming aware of and correcting the frequent
deception of their own understanding and its false steps.
(V-Lo/Blomberg, 24: 150–1)
Notes
1 This research arises in the framework of the project “Conceptual history and
criticism of modernity” (FFI2017–82195-P), directed by Faustino Oncina Coves
(University of Valencia, Spain). Quotations from Kant’s texts are usually from the
Cambridge edition, sometimes modified, with the pagination of the Akademie-
Ausgabe. A first version of the chapter appeared in the journal Studi kantiani,
XXXII, 2019.
2 “Even before ‘public opinion’ became established as a standard phrase in the
German-speaking area, the idea of the bourgeois public sphere attained its theo-
retically fully developed form with Kant’s elaboration of the principle of publicity
in his philosophy of right and philosophy of history” (Habermas 1989: 102).
3 “A communication that meets only with non-interference is from a certain point
of view already a failure” (O’Neill 1989a: 31).
4 This is at least the thesis of the Historisches Wörterbuch der Philosophie: see
Hügli (1980: 1424): “Der deutsche Terminus Mitteilung scheint erst mit Kant
philosophische Bedeutung zu erlangen”.
5 On Kant’s concept of Aufklärung, see at least: La Rocca (2004: 123–38).
6 I discuss in a more detailed way the status of faith in Fonnesu (2015: 361–90).
7 V-Lo/Blomberg, 24:150. On Kant’s logic, see Capozzi (2002, repr. 2013) and
now: Lu-Adler (2018). About Kant’s logical corpus: Hinske (1999).
8 See Capozzi (2002, repr. 2013); Stevenson (2003: 72–101) (a shorter, revised ver-
sion with the same title can be read in: Stevenson 2011: 77–93).
9 On this expression and its historical roots, see Mileti Nardo (2021).
10 “Having an opinion is holding something to be true with the consciousness that
it is subjectively as well as objectively insufficient” (A822/B850).
11 I cannot deal here with the other touchstones proposed by Kant for distinguish-
ing conviction and persuasion: Selbsterforschung (we can try to understand and
explain the deceptive character of our persuasion, analysing our alleged convic-
tion and “developing”, says Kant, its subjective causes) and betting, because we
are looking here for the meaning and importance of communication and com-
municability. See on this A821/B849 and A824/B852.
12 It’s worth mentioning that the notion of holding-to-be-true occurs in Kantian
texts mainly when he deals with the status of moral faith.
13 In fact, the ambiguity does not concern only the Glauben but also the fides and
even the couple faith/belief. It is not possible here to deal with this important and
interesting question.
14 I discuss the Kantian conception of “moral certainty” in Fonnesu (2011: 183–204).
15 This important question cannot be faced here and has several implications. Con-
versations with Claudio La Rocca and Tommaso Piazza convinced me that this
problem cannot be taken as a secondary question. I have to thank them and can
only deal with that problem in a future paper.
107
LUCA FONNESU
16 With curious and interesting exceptions: “Persuasion (of honest people) can be
morally good, although logically bad” (Refl 2699, 16: 475).
17 On this, see also Refl 2541, 16: 410.
References
Capozzi, M. 2002 (2013, repr.). Kant e la logica. Vol. I. Napoli: Bibliopolis.
Chignell, A. 2007. “Belief in Kant”, The Philosophical Review, 106(3): 323–60.
Dörflinger, B. 2004. “Führt Moral unausbleiblich zur Religion? Überlegungen zu
einer These Kants”, in N. Forster (ed.) Kants Metaphysik und Religionsphiloso-
phie. Hamburg: Meiner: 207–23.
Fonnesu, L. 2011. “Kant on Moral Certainty”, in L. Cataldi Madonna and Paola
Rumore (eds.) Kant und die Aufklärung. Hildesheim and New York: Olms: 183–204.
Fonnesu, L. 2015. “Kant on Private Faith and Public Knowledge”, Rivista di filoso-
fia, 106: 361–90.
Gava, G. 2019. “Kant and Crusius on Belief and Practical Justification”, Kantian
Review, 24(1): 53–75.
Grice, P. 1957. “Meaning”, The Philosophical Review, 56: 377–88.
Guyer, P. 1997. Kant and the Claims of Taste. Cambridge: Cambridge University
Press.
Habermas, J. 1984. The Theory of Communicative Action, edited by T. Mc Carthy.
Boston, MA: Beacon Press.
Habermas, J. 1989. The Structural Transformation of the Public Sphere. Oxford: Polity.
Hinske, N. 1999. Illuminismo e critica della ragione. Studi sul corpus logico kan-
tiano. Pisa: Edizioni della Normale.
Hügli, A. 1980. “Mitteilung, Mitteilbarkeit, indirekte Mitteilung”, in J. Ritter et al.
(eds.) Historisches Wörterbuch der Philosophie. Vol. 5. Basel-Stuttgart: Schwabe:
coll. 1424–431.
La Rocca, C. 2003. Soggetto e mondo. Studi su Kant. Venezia: Marsilio.
La Rocca, C. 2004. “Was Aufklärung sein wird. Zur Discussion um die Aktualität
eines Kantischen Konzepts”, in H. Nagl-Docekal and R. Langthaler (eds.) Recht-
Geschichte-Religion. Die Bedeutung Kants für die Gegenwart. Berlin: Akademie
Verlag: 123–38.
Lu-Adler, H. 2018. Kant’s Science of Logic: A Historical and Philosophical Recon-
struction. Oxford and New York: Oxford University Press.
Meier, G. F. 1752. Auszug aus der Vernunftlehre. Halle: Gebauer.
Mileti, N. L. 2021. Forme della certezza. Genesi e implicazioni del Fürwahrhalten
in Kant, Pisa: ETS.
O’Neill, O. 1989a. “The Public Use of Reason”, in O. O’Neill (ed.) Constructions of
Reason: Explorations of Kant’s Practical Philosophy. Cambridge and New York:
O’Neill, O. 1989b. “Reason and Politics in the Kantian Enterprise”, in O. O’Neill
(ed.) Constructions of Reason: Explorations of Kant’s Practical Philosophy. Cam-
bridge and New York: Cambridge University Press: 3–27.
Stevenson, L. 2003. “Opinion, Belief or Faith, and Knowledge”, Kantian Review, 7:
72–101.
Stevenson, L. 2011. Inspirations from Kant. Essays. Oxford and New York: Oxford
University Press.
108
7
KANT’S SPACE OF THEORETICAL
REASON AND SCIENCE
A perspectival reading
Lorenzo Spagnesi
1. Introduction: unity, pluralism, and perspectivism

The unity of science thesis (“monism”) – the idea that unity plays an impor-
tant, if not crucial, role in the investigation of nature – has been severely
criticized from many quarters in recent decades.1 Only a small minority of
philosophers of science would now defend the once-popular metaphysical
claim that science is a unified system which is supposed to reflect the unity
of nature.2 Setting here aside metaphysical questions regarding the unity of
nature and its correspondence to scientific representations, critics have fur-
ther argued that unity is not a welcome hypothesis even from a purely
epistemic point of view. It is on the latter point of view – in particular
with respect to scientific theorizing3 – that I will focus my attention in this
chapter.
Dupré, Cartwright, Chang and many others have in different ways
argued that unity is not a desirable epistemic requirement in the actual
practice of how most of scientific research is conducted. For instance,
Dupré 1996 has argued that theoretical unity is socially and politically
undesirable and Chang 2012 that it is such an unreachable requirement
that we are better off not looking for unitary theories. Even more radi-
cally, it has been argued that such a thesis would detrimentally affect
science, as it would lead us to privilege systematic considerations over
more valuable empirical evidence (Cartwright 1999). Such accounts
are undoubtedly persuasive. It is uncontroversial that the present state
of scientific theorizing is characterized by a plurality of models, theo-
ries and classifications, each revealing different aspects of phenomena.
Pluralism – the view that phenomena cannot be fully explained by a
single theory and, instead, require a plurality of theoretical approaches
for their investigation – seems the meta-scientific principle that is best
suited to guide scientific research.4
DOI: 10.4324/9781003050742-8 109

L O R E N Z O S PAG N E S I
There is a plurality of views about scientific pluralisms, and the descrip-

tion I offer here is only a primer to a very complex topic. But what all
pluralists seem to agree on is that the plurality of approaches that presently
characterizes most areas of scientific investigation does not necessarily rep-
resent a “deficiency” of science (Kellert et al. 2006: x). A recent influential
account – “perspectivism” or “perspectival realism” (Rueger 2005; Giere
2006a; Teller 2011; Massimi 2018) – felicitously captures and motivates this
aspect of pluralism. Multiple approaches and theories are nothing but differ-
ent perspectives from which we investigate phenomena. Given our epistemic
limitations – the point of view we occupy – science is perspectival through
and through.
According to Giere’s perspectivism, not only is scientific observation per-
spectival because of the limited sensitivity of the instruments we use (partial-
ity of the input, non-transparency of the instrument), scientific theorizing
(from data models to scientific principles) is perspectival too: “Newton’s
laws characterize the classical mechanical perspective; Maxwell’s laws char-
acterize the classical electromagnetic perspective; the Schrödinger Equation
characterizes a quantum mechanical perspective” and so on (Giere 2006a:
14). As noted by Massimi, while Giere’s perspectivism focuses on histori-
cal considerations, other types of perspectivism, like the one advocated
by Rueger 2005 and Teller 2011, privilege the cultural situatedness of our
knowledge over the historical one (2018: 168–9). As a result, they focus on
the synchronic coexistence of different theories rather than on their devel-
opment over time. While there are significant disanalogies between the two
forms of perspectivism, they both share the insight that our epistemic limi-
tations fundamentally condition the possibility of scientific knowledge. In
both cases, the plurality of approaches is not to be regarded as a defect of
our knowledge. It is instead the inevitable result of the epistemic situation
we occupy.
If we accept that pluralism characterizes the way we need to think about
nature, there seems to be no room left for a principle of unity in science.
Indeed, as I will show in the next section, most pluralists (within and outside
perspectivism) reject monism as a valuable epistemic principle in science.
Such rejection, however, raises a number of concerns: Does the acceptance
of pluralism inevitably entail the exclusion of unity from a purely epistemic
point of view? Is pluralism sufficient as the only epistemic principle guid-
ing scientific research? Isn’t it possible to redefine scientific pluralism and
monism as non-conflictual principles? In the remainder of this chapter, I
will try to address these questions and suggest a reconciling solution. I will
first problematize the contemporary understanding of the relation between
unity and pluralism in science (Section 2); I will then look at a perspectival
approach to the problem Kant presents in the Critique of Pure Reason (Sec-
tion 3); finally, I will try to explain how Kant’s perspectivism might inspire
the current debate (Section 4).
110
KANT’S THEORETICAL REASON AND SCIENCE
2. Unity vs. pluralism: an “antinomy” that awaits a

solution?
The debate between pluralism and unity I briefly introduced in the previous
section seems to present us with two mutually exclusive epistemic principles.
Scientific research should be guided either by a pluralistic principle or by
a postulated unity of scientific cognition. Since pluralism has proven to be
empirically more suited to scientific research, one should fully endorse it and
reject monism. I contend that such incompatibility is not the inevitable upshot
of two contradictory principles. Rather, it results from a misleading charac-
terization of these very principles. As I will argue, pluralism and monism can
and, indeed, must coexist and complement each other in scientific research.
Let’s first try to better qualify the conflict between pluralism and monism
in the current debate. As previously mentioned, pluralism comes in different
varieties. I will here focus on the variety of pluralism proposed by Kellert,
Longino and Waters in their programmatic volume on scientific pluralism,
namely “empirical pluralism”, or the “pluralist stance” (Kellert et al. 2006:
xiii). This version of pluralism is meant to avoid the limits of weaker and
stronger formulations of pluralism while remaining representative of most
pluralist takes. Empirical pluralism amounts to a “commitment to avoid
reliance on monist assumptions in interpretation or evaluation coupled with
an openness to the ineliminability of multiplicity in some scientific contexts”
(ibid.). With respect to scientific theorizing, this definition implies: (1) (nega-
tively) that an empirical pluralist should not rely on any monist assumption
while elaborating on scientific theories and (2) (positively) that an empirical
pluralist is open to the possibility of a persistent multiplicity of incompat-
ible, yet explanatory, theories describing phenomena.
As can be seen, the incompatibility with monism is built into the very
definition of empirical pluralism. But what kind of monism should a plural-
ist avoid? As Kellert, Longino and Waters put it, scientific monism is the
view that “the ultimate aim of a science is to establish a single, complete,
and comprehensive account of the natural world” (ibid.: x). Crucially, sci-
entific monism assumes that “the nature of the world is such that it can, at
least in principle, be completely described or explained by such an account”
(ibid.). Scientific monism is therefore defined as a metaphysical thesis about
the nature of the world. Pluralists do not, however, support the opposite
metaphysical claim. Instead, they argue that whether such theories might
be unified is an “open, empirical question” (ibid.). I side with pluralists in
thinking that unity cannot be simply stated as a metaphysical truth about
the nature of the world. But I want to problematize the resulting rejection of
monism as such. If we rule out unity as a metaphysical truth, does this mean
that it cannot still play an important role from an epistemic point of view?
I suggest that empirical pluralism, as the sole epistemic principle guid-
ing scientific research, is, at the same time, both too strong and too weak
111
a requirement. It is too strong, for it does not fully account for the process
of unification existing in science: history provides us with brilliant cases of
theory unification (Einstein’s special theory of relativity is paradigmatic in this
sense), and there are fields, like fundamental physics, in which unification still
plays a significant and programmatic function (e.g. in the quest for a theory of
“everything” that unifies the four fundamental forces). It is true that pluralists
do not deny the possibility of unification – they treat it as an open question –
but the process of theory unification remains epistemically unexplained. If we
commit ourselves to avoid monist assumptions, how could we even attempt
to unify theories in the first place? The empirical pluralist’s reply that evidence
should guide us is clearly ill equipped to provide a satisfactory answer. Empiri-
cal evidence is exactly what prompts pluralism, and it is unclear how it can
lead us to unificatory hypotheses without relying on a different set of assump-
tions. Indeed, theory unification is rarely possible on experimental grounds
alone – the construction of the electroweak theory being a clear example of
such evidential insufficiency for unification (see Morrison 2008: 49).
Furthermore, a pluralistic program which is merely satisfied with a plurality
of approaches and strategies seems also too weak a principle, for it does not
really address the question of how scientific research ought to be done. Shall we
regard the multiple approaches and theories presently available as the defini-
tive ones? Or shall we look for ever finer-grained descriptions and explana-
tions? An empirical principle of pluralism leaves us with no definite answer to
such concerns. Indeed, since it rests content with any given plurality of theories,
it seems insufficient to express the very pluralistic urge that motivates it.
Some pluralists have come to explore the possibility of complementing plu-
ralism with some weaker – metaphysically uncommitted – forms of monism.
Giere’s perspectivism is emblematic in this respect. Although perspectivism is
the pluralistic view that there is a plurality of legitimate historically (or cul-
turally) motivated perspectives on phenomena, Giere recognizes that there
are cases in which it is indeed recommendable to unify perspectives into
larger points of view. What we need to do is not introduce a metaphysical
doctrine of the unity of the world. We need instead to complement perspec-
tivism with a mere “methodological presumption”, according to which “we
presume there is a unique causal structure to the world” (Giere 2006a: 34;
see also 2006b: 36). Importantly, this should not be regarded as a neces-
sary requirement of our reasoning: it is an additional desideratum of our
scientific practice that might lead us to further unifications. As Giere puts it:
From a perspectival point of view, one need not be too upset with
the current situation in theoretical physics. Good theoretical science
does not require finding genuinely universal principles. Well-fitting
models, based on a variety of principles, are good enough. And,
indeed, that is all that can be found across most of the sciences.
(2006b: 33)
112
The picture here suggested is of a self-sufficient epistemic principle of plu-

ralism that may be complemented by a principle of unity at a higher level
of inquiry. In other words, the two epistemic principles are compatible at
different levels: while pluralism is necessary for scientific research as such,
monism might turn out to be a useful methodological maxim when it comes
to particularly ambitious scientific projects. I will argue, however, that the
compatibility between pluralism and monism does not merely concern the
possibility of further unification in science. Further unification is, as it were,
the tip of the iceberg of a larger problem. Unification, I suggest, is an ongo-
ing process that is ubiquitous at any level of science and is implied even at
the level of the pluralistic image of models.
The self-sufficiency of pluralism, even at the “basic” level of science, is
indeed questionable. Take two standard examples of “perspectives” in the
history of science: Newton’s theory of motion and Maxwell’s theory of elec-
tromagnetism. It is difficult not to see these two perspectives as themselves
instances of unification. Newton’s theory famously unified celestial and ter-
restrial mechanics. Maxwell’s theory brought together electromagnetism
and optics. Both theories (despite being ultimately incompatible) showed
that different phenomena can be successfully unified with each other. More
generally, to do without a principle of unity at the basic level of science
would be to discourage the very process of theory construction, namely the
hypothetical subsumption of different phenomena under general, abstract
principles. To quote Giere’s own example, it would require “something like
concluding that different samples of the same radioactive isotope had differ-
ent half-lives and that no further explanation of this difference was possible”
(2006a: 34–5). This example clearly goes a bit further than showing that a
mere desideratum of unity might be added to a self-sufficient principle of
pluralism. It seems to suggest, instead, that unification – as a methodological
rule – is an epistemic principle that necessarily complements pluralism even
at the basic level of scientific inquiry.
There seems to be no contemporary account that offers promising solu-
tions to this debate. But following Kitcher 1999 and Breitenbach and Choi
2017, I submit that a solution can be inspired by Kant’s account of theoreti-
cal reason in scientific cognition. Rather than presenting a loosely Kantian-
inspired interpretation of the role of unity in science, however, I will look
into Kant’s text to see whether it directly offers arguments that may be used
in the current debate.5 Crucially, although Kant is generally presented as a
strong advocate of the unity of science thesis, I will argue that his conception
of systematic unity contains – perhaps surprisingly – compelling remarks on
the compatibility between unification and pluralism in the space of reason.
“Insightful men” – Kant notes at the end of the first part of the Appendix
to the Transcendental Dialectic – are “in conflict with one another” as to
whether they should follow a “maxim of the manifoldness of nature” or the
one “of the unity of nature” in empirical investigations (A667/B695).6 As I
113
will explain, the incompatibility between these two maxims results from an
apparent “antinomy” of reason: the antinomy, however, soon disappears as
we recognize the principles that ground the maxims as regulative principles
rather than opposed objective insights. As such, Kant says, “these maxims
can of course be united” (ibid.).
In particular, I suggest that an interesting variation on perspectivism can
be found in Kant’s conception of reason.7 While standard perspectivism
focuses on the plurality of observational points of view, Kant’s “perspectiv-
ism” privileges the common space within which different perspectives can be
taken up. Kant’s space of reason is a place where plurality and unity can and
indeed must coexist with each other as principles, or “axes”, that are both
essential to scientific theorizing. In the following section, I will elaborate on
what I take to be the three crucial features of Kant’s perspectivism ([1] ideas
as foci imaginarii; [2] plurality of points of view; [3] the unified space of rea-
son). In the final section, after some clarification on the difference between
Kant’s treatment of the problem and the contemporary debate, I will explain
how Kant’s perspectivism might enlighten the discussion.
3. Kant’s perspectival space of reason
3.1. The regulative function of reason

Kant addresses the problem of the systematic unity of our empirical cogni-
tion in the very much debated Appendix to the Transcendental Dialectic in
the Critique of Pure Reason.8 In the Appendix, Kant finally presents the
long-promised positive use of reason and its ideas. Ideas are “deceptive” and
“transcendent” when they are mistaken for “concepts of real things” (A643/
B671). The Transcendental Dialectic shows that when we treat mere ideas as
such – when, for example, we treat the idea of God as that of an object and
we even attempt to demonstrate its existence – reason inevitably oversteps
the boundaries of possible experience. Ideas, however, do retain a “good”
and “immanent” use when directed not toward objects but towards the fac-
ulty that, in Kant’s architectonic, directly has to do with objects, namely the
faculty of the understanding or the faculty of concepts (see A126). Reason,
in its positive use, is therefore presented as a second-order faculty. While the
understanding has to do with objects, reason, as Kant puts it, “does not cre-
ate any concepts (of objects) but only orders them and gives them that unity
which they can have in their greatest possible extension, i.e., in relation to
the totality of the series” (A643/B671). As a result, in Kant’s terminology,
the employment of reason cannot be constitutive as that of the understand-
ing. The concepts of the understanding constitute the objects of experience,
but reason, as we saw, is further removed from objects. Reason maintains,
however, an important regulative function: that of guiding the activity of the
understanding by ordering its particular concepts.
114
From these two closely related features of reason (its being a second-order
faculty and having a regulative use), it would seem that its contribution to
empirical cognition is a useful tool or, at best, a welcome desideratum.9 Kant’s
point would be similar to that of those pluralists who are willing to concede
some role to unification in science. Although pluralism has proven efficacious
in most cases, we might keep a methodological principle in favor of unity.
Good science – they argue – does not require unity, yet we might desire or even
look for higher forms of unification in some specific cases. A weak principle
of unity is all that pluralist science may at best need. However, I contend that
this is not a good depiction of Kant’s position and, further, that Kant has good
reasons to think about unity and pluralism in a different way, reasons that – as
I will show in the last section – can be applied to the current debate (with some
caveats). Importantly, Kant characterizes the regulative use of reason not just
as desirable or useful, but as “an excellent and indispensably necessary” one
(A644/B672). As emphasized by Massimi, this indispensably necessary use is
clearly presented in “perspectival” (and “optical”) terms (see Massimi 2017,
2021). Let’s take a closer look at the Appendix in order to understand what
kind of perspectivism is here at stake. I contend that Kant’s perspectivism is
best described by the following three main features.
3.2. First perspectival feature: ideas as foci imaginarii

The first perspectival feature of Kant’s account of reason has already attracted
considerable attention in the literature. Kant presents the necessary function
of ideas of reason in clear optical terms. Drawing upon optical treatises
of his time, Kant defines the ideas of reason as foci imaginarii (imaginary
standpoints) for the rules of the understanding.
[Transcendental ideas] have an excellent and indispensably neces-

sary regulative use, namely that of directing the understanding to
a certain goal respecting which the lines of direction of all its rules
converge at one point, which, although it is only an idea (focus
imaginarius) – i.e., a point from which the concepts of the under-
standing do not really proceed, since it lies entirely outside the
bounds of possible experience – nonetheless still serves to obtain for
these concepts the greatest unity alongside the greatest extension.
(A644/B672)
In describing ideas as imaginary standpoints, Kant is presumably referring

to the VIII axiom of the first book of Newton’s Opticks (1704).10 It is worth
comparing the Kantian example with its source.
According to the VIII axiom, an object (A) reflected by a mirror (mn)
appears to be in the place (a) from where all the rays diverge towards the
observer.11 As the focal point (a) unifies all reflected lines and guides the eyes
of the observer, ideas of reason unify the manifold of concepts and guide the
115
Figure 1 Illustration of the VIII Axiom, from Newton’s Opticks (4th edition, 1730)12
understanding. Kant illustrates this use of reason with a number of examples

taken from a variety of scientific disciplines: chemistry (“pure earth”, “pure
water” and “pure air”; A646/B674), psychology (“fundamental power”;
A649/B677), astronomy (planetary motion; A663/B691), physics (“gravita-
tion”; A663/B691) and biology (“hereditary distinctions between families”;
A667/B695). In each of these cases, reason projects an idea in order to bestow
unity upon the particular cognitions of the understanding. In projecting ideas,
however, reason runs the risk of taking them as concepts of real objects. But
as we saw, ideas cannot be regarded as concepts of objects. As Kant explains:
Now of course it is from this that there arises the deception, as

if these lines of direction were shot out from an object lying out-
side the field of possible empirical cognition (just as objects are
seen behind the surface of a mirror); yet this illusion (which can be
prevented from deceiving) is nevertheless indispensably necessary if
besides the objects before our eyes we want to see those that lie far
in the background, i.e., when, in our case, the understanding wants
to go beyond every given experience (beyond this part of the whole
of possible experience), and hence wants to take the measure of its
greatest possible and uttermost extension.
(A644–5/B672–3; my emphases)
What is the optical “illusion” Kant is referring to? Newton’s axiom may
help us answering this question. The real object (A), which is placed behind
116
the observer (and thus outside her visual field), appears to be in front of
her (that is “behind the surface of a mirror”; mn), in the place a. The illu-
sion is thus created by the fact that the reflected image is almost identical
to the vision we would have if A were really placed in a.13 This seems to
be exactly the optical phenomenon Kant uses in order to explain the ideal
“vision”. The rays between mn and a represent the illusion that enables the
observer to extend the visual field from the space between her and the mir-
ror (given experience: EB, FC, GD) to the space between the mirror and A
(possible experience: AB, AC, AD). The illusion deceives the observer when
she mistakes what is merely a focal point (a; an idea) for a real object (A;
an actual object). By doing so, she turns a mere regulative principle into an
unwarranted metaphysical principle. Reason, however, can fully recognize
the nature of ideas as projections and use them to orient the concepts of the
understanding. It thereby becomes critical: it can eliminate the error and
legitimately hope to extend its visual field.
Let’s take stock here. While the perspectival nature of the ideas of rea-
son as focal points should be clearer now, the status of necessity attached
to them may still raise some perplexity. For Kant seems to formulate such
necessity in hypothetical terms: we need to employ ideas only if we want to
go beyond given experience and attain the greatest unity of cognition. Does
this mean that the regulative function of ideas is a mere additional desid-
eratum of science? In other words, are ideas “necessary” only when we look
for “higher” unifications? I contend that this is not the case. Indeed, Kant
presents this use of reason as hypothetical – but it would be a mistake to read
it as a mere afterthought of scientific inquiry. Kant says:
If reason is the faculty of deriving the particular from the universal,

then: Either the universal is in itself certain and given, and only
judgment is required for subsuming, and the particular is necessar-
ily determined through it. This I call the “apodictic” use of reason.
Or the universal is assumed only problematically, and it is a mere
idea, the particular being certain while the universality of the rule
for this consequent is still a problem; then several particular cases,
which are all certain, are tested by the rule, to see if they flow from
it, and in the case in which it seems that all the particular cases cited
follow from it, then the universality of the rule is inferred, including
all subsequent cases, even those that are not given in themselves.
This I will call the “hypothetical” use of reason.
(A646–7/B674–5)
Reason can be used either apodictically or hypothetically. In the former case,

the universal is given and the particular is determined through it; in the lat-
ter, the particular is given but the universality of its rule is only projected as a
“problematic” concept (i.e. as an idea). Now, to say that the hypothetical use
117
of reason is only necessary for higher unities of empirical cognition would

mean to say that at least some universal concepts are given to us in experi-
ence through concepts of the understanding and empirical intuitions.14 But can
we obtain something like universality or apodictic certainty from experience?
Kant is quite clear that this is not the case:
Empirical concepts, together with that on which they are grounded,

empirical intuition, cannot yield any synthetic proposition except
one that is also merely empirical, i.e., a proposition of experience;
thus it can never contain necessity and absolute universality. . . . No
universally valid, let alone apodictic proposition could ever come
from empirical intuition: for experience can never provide anything
of this sort.
(A47–8/B64–5)
As Kant affirms here and elsewhere, experience can never yield any kind
of “necessary and apodictic propositions” (A721/B749), for all empirical
cognitions are confronted with the problem of induction. If the universal
is never given in experience, then it must be hypothetically postulated for
any empirical concept in accordance with the ideas of reason. The Appen-
dix teaches us that we can employ the regulative ideas of reason in order
to extend the rule that we infer from given particular cases to all possible
experience: this use of reason necessarily contributes to the formation of any
universal empirical concept. And since scientific cognition has to do with
such concepts, it also follows that the hypothetical use of reason is necessary
to transform any given experience into scientific knowledge.
Kant presents the ideas projected by reason in its hypothetical use as per-
spectival focal points of unity. This function of reason – which Kant later
in the text also calls the “logical principle of genera” (A654/B682) or of
“homogeneity” (A658/B686) – is regulative yet indispensable. It is regulative
because reason is farther removed from objects, and ideas cannot stand for
actual things. In other words, it is not a metaphysical principle that consti-
tutes nature as objectively unitary. Nevertheless, reason’s principle of unity
necessarily regulates the conceptualization activity of the understanding. In
particular, it projects the universality of given particulars, thus enabling the
formulation of universal empirical concepts: an activity which is an integral
part of science.
3.3. Second perspectival feature: plurality of points of view

The characterization of ideas as focal points might incline us to think that,
for Kant, there is as much science as there is unity. This is, however, only one
side of the story. Kant’s theory of theoretical reason includes a principle of
unity but is not limited to that. As Kant puts it:
118
To the logical principle of genera there is opposed another, namely

that of species, which needs manifoldness and variety in things
despite their agreement under the same genus, and prescribes to the
understanding that it be no less attentive to variety than to agree-
ment. This principle (of discrimination, or of the faculty of distin-
guishing) severely limits the rashness of the first principle (of wit).
(A654/B682; my emphases)
The principle of “species” (“variety”; A657/B685; or “specification”; A658/

B686) prescribes to look for variety in things despite their being unified
under the same genus. I propose to look at this principle as the second per-
spectival feature of Kant’s theory of reason. While the first feature tells us
that ideas are focal points of unification, this second principle limits the first
one and prescribes to specify each concept in a plurality of points of view.
If this is correct, I contend that Kant’s principle of specification is similar
enough to contemporary empirical pluralism to draw an interesting com-
parison between the respective cases. Recall that empirical pluralism has a
negative part ([1] avoidance of monist assumptions) and a positive part ([2]
admission of plurality of accounts). As concerns (1), to follow the principle
of specification implies avoiding monist assumptions. One may take Kant’s
principle of specification to be subordinated to the principle of unity (as a
species is subordinated to its genus). But note that Kant’s principle of speci-
fication is more radical than that:
Here reason shows two interests that conflict with each other: on
the one side, an interest in the domain (universality) in regard to
genera, on the other an interest in content (determinacy) in respect
of the manifoldness of species; for in the first case the understand-
ing thinks much under its concepts, while in the second it thinks all
the more in them. This expresses itself in the very different ways of
thinking among students of nature; some of whom (who are chiefly
speculative) are hostile to differences in kind, while others (chiefly
empirical minds) constantly seek to split nature into so much mani-
foldness that one would almost have to give up the hope of judging
its appearances according to general principles.
(A654–5/B682–3)
While unification universalizes the domain of a concept, this second prin-

ciple tells us to find variety in the content of a concept. Specification does
not merely express the subordination of species under genera, but it directly
opposes to unification inasmuch as it disunifies the content of any universal
concept. In Kant’s words, “chiefly empirical minds” do not simply specify a
given unitary concept into species and subspecies, but they “split nature into
so much manifoldness that one would almost have to give up the hope of
119
judging its appearances according to general principles”. The two principles

therefore result in interests that conflict with each other – a conflict that
seems very much alive even today.
What about the positive requirement of admitting multiple accounts of
phenomena ([2] previously)? Here there seems to be a difference for, while
Kant’s principle results in the specification of particular phenomena, a con-
temporary pluralist welcomes different accounts of phenomena. Let’s take a
typical example of contemporary pluralism: the concept of species in biol-
ogy. There are different concepts of species: the biological species concept
(based on interbreeding), the phylogenetic species concept (based on ances-
try), the ecological species concept (based on ecological niches) and so on.
Pluralists think there is no best account of “species” and all these concepts
are legitimate representations of particular aspects of the complex “species”
phenomenon. Although Kant’s terminology differs from the one used in the
contemporary debate, I do not see any reason not to regard this and simi-
lar cases as contemporary applications of Kant’s principle of specification.
After all, different concepts of species are nothing but particular aspects of a
complex phenomenon, and ultimately, they are all based on particular phe-
nomena (interbreeding, ancestry, ecological niches, etc.). Empirical pluralists
can therefore be seen as contemporary “chiefly empirical minds” who “split
nature” as much as possible by replacing unitary concepts with a variety of
concepts of particular aspects of phenomena.
There is, however, a lingering worry, namely that the kind of pluralism
resulting from Kant’s principle of specification may be closer to what is
known in the literature as “modest pluralism” than to a genuinely per-
spectival take on scientific investigation (see Kellert et al. 2006: xii–iii).15
According to modest pluralism (of which Mitchell is perhaps the best-
known advocate; see e.g. Mitchell 2003), scientists must indeed engage in a
plurality of accounts, but such accounts are meant to contribute to a single,
integrated picture of a complex phenomenon. For an empirical pluralist, no
such commitment to “integration” is necessary. On the contrary, different
perspectives (for instance, different accounts of species) may be inconsistent
with each other and still enhance our understanding of nature. To answer
this worry, it is important to point out that Kant’s principle of specification
is uncommitted as to whether the “particulars” resulting from it contribute
to an integrated account of a phenomenon. All the principle does prescribe
is, instead, to indefinitely split nature into more and more particular phe-
nomena. But if this is the case, one may wonder if it still makes sense to talk
about an integrated picture of a complex phenomenon rather than just a
plurality of particular accounts. Even for a Kantian “empirical mind”, then,
the investigation of particular phenomena may not contribute to any single,
integrated account of a complex phenomenon.
I have suggested that Kant’s principle of specification can be read in a way
that invites the evolution towards contemporary perspectivism. We should
120
not forget, however, that in Kant’s view, this principle, exactly as the prin-
ciple of homogeneity, can be only regulatively employed. This has two key
implications that I will fully explore in the following section. First, as a
regulative principle, specification does not metaphysically assert that nature
is an irreducible manifoldness or variety of things. It is instead a regulative
principle that is only meant to promote the empirical investigation of nature.
As such, it can be made compatible with an equally regulative principle
of unity: indeed, the two opposite principles mutually complement each
other.16 Second, as a non-objective principle, it can legitimately prescribe
not just plurality of species but indefinite specification of concepts.
Reason demands in its entire extension that no species be regarded

as in itself the lowest; for since each species is always a concept
that contains within itself only what is common to different things,
this concept cannot be thoroughly determined, hence it cannot be
related to an individual, consequently, it must at every time contain
other concepts, i.e., subspecies, under itself.
(A655–6/B683–4)
The process of specification can thus be held as indefinite, or potentially infi-

nite.17 This means that a Kantian “empirical mind” should not be satisfied
with a given plurality of concepts of phenomena. Rather, she should con-
stantly seek to further specify and disunify any postulated universal concept.
Kant concludes his account of principles of reason by introducing a third
principle – “continuity” (A658/B686) or “affinity” (A660/B688) – resulting
from the combination of the first two.
The last arises by uniting the first two, according as one has com-
pleted the systematic connection in the idea by ascending to higher
genera, as well as descending to lower species; for then all manifolds
are akin one to another, because they are all collectively descended,
through every degree of extended determination, from a single high-
est genus.
(A658/B686)
The systematic interconnection of concepts according to unity and disunity

results in continuity of concepts, or “continuous transition from every species
to every other” (ibid.). In other words, reason’s combined presuppositions of
conceptual unification (leading to a highest genus) and of (potentially) infi-
nite specification give rise to a unified picture in which, as Kant puts it, all
different genera are only “partitionings” of a single genus and there cannot
be any “leap” between species (A659/B687). Interestingly, Kant illustrates
this principle with an example from astronomy: the highly elliptical path of
comets. Empirical observation, explains Kant, does not show us their paths
121
in their entirety, yet we guess at a “parabolic course for them since it is still
akin to the ellipse and, if the major axis of the latter is very long, it cannot
be distinguished from it in all our observations” (A662/B690). We thereby
presuppose an affinity of the paths of comets with those of planets. This
is only possible on the basis of the combination of the previous principles.
In this specific case, as Kant explains, we conceive the “world system” as
both “unbounded” (infinitely specified) and “connected through one and
the same moving force” (supremely unified) (A663/B691): such a system
allows us to presuppose “continuous transition” in the paths of celestial
bodies in general. As for the previous principles, continuity should not be
read in constitutive terms either. This last principle does not tell us that the
world is a continuum: no actual continuity, as Kant explains, can be found
in experience (see A661/B689). Continuity is instead a regulative principle
which, however, is “legitimate and excellent” inasmuch as it “points the way
toward systematic unity” (A668/B696).
3.4. Third perspectival feature: the space of reason

I can now elucidate the third perspectival feature of Kant’s system of knowl-
edge. We saw that the principles of unity and disunity are perspectivally
characterized in Kant’s theory of reason. The logical principle of unity
postulates identity under concepts that allows us to project ideas as foci
imaginarii. By contrast, the logical principle of variety postulates a poten-
tially infinite specification of concepts that grounds the opposite rational
interest of specifying ideas into ever finer-grained parts. Now, I further con-
tend that these two principles (together with the third principle that results
from their combination: continuity) make up a perspectival space – Kant’s
space of theoretical reason. Within this perspectival space, it is possible to
pursue unity as well as disunity of cognition. I will first clarify why I think
this is a perspectival feature and then comment on the passages that suggest
this interpretation.
Perspectivism is generally presented as the view that that there are dif-
ferent standpoints from which we “frame” the world. As the term exploits
the metaphor of experientially different perspectives, the emphasis is very
much on the plurality and potential disagreement among different points
of view. It is in this sense that the term “perspectivism” is generally used in
the current debate. But this usage fails to portray another crucial aspect of
perspectivism, namely the fact that different perspectives or points of view
presuppose – and are only possible within – the same space of representa-
tion. It might be instructive to take a brief look at the history of perspective
in the arts. Perspective – once a synonym for optics18 – acquired a specific
artistic meaning during the Renaissance when artists and theorists (first and
foremost Brunelleschi and Alberti) started applying optical and geometrical
studies to the construction of an artistic representation. The art of perspective
122
subsequently spread as allegedly the most accurate representation of reality.

As recognized by several art critics, however, this kind of representation was
based on a precise conception of space, namely a perfectly homogeneous,
infinite and continuous mathematical space.19 The assumption of this space
is the first logical premise in any perspectival construction.
We can apply this consideration to our present case. As perspectival repre-
sentations are only possible within a precise system of geometrical assump-
tions about the topology of space, attempts at conceptual unification and
specification of phenomena are only possible within a perspectival system or
space of representation. Quite remarkably, Kant says that reason “prepares
the field for the understanding” by presupposing unity, variety and continu-
ity of concepts (A657/B685). In other words, it presupposes a systematic
conceptual framework within which scientific knowledge is possible. Kant
goes on to illustrate this conceptual space by using once again specifically
perspectival metaphors:
Systematic unity under the three logical principles can be made

palpable in the following way. One can regard every concept as a
point, which, as the standpoint of an observer, has its horizon, i.e.,
a multiplicity of things that can be represented and surveyed, as it
were, from it. Within this horizon a multiplicity of points must be
able to be given to infinity, each of which in turn has its narrower
field of view; i.e., every species contains subspecies in accordance
with the principle of specification, and the logical horizon consists
only of smaller horizons (subspecies), but not of points that have
no domain (individuals). But different horizons, i.e., genera, which
are determined from just as many concepts, one can think as drawn
out into a common horizon, which one can survey collectively from
its middle point, which is the higher genus, until finally the highest
genus is the universal and true horizon, determined from the stand-
point of the highest concept and comprehending all manifoldness,
as genera, species, and subspecies, under itself.
As the perspectival space is made up of infinite points, each of which potentially

represents a point of view, the logical space of reason is made up of concepts as
points. Every concept can be regarded as a viewpoint with its horizon, that is,
to quote Kant,“a multiplicity of things that can be represented and surveyed, as
it were, from it”. This horizon is also a multiplicity of points, and each of these
points in turn has its own “narrower field of view”. On the one hand, concep-
tual specification is potentially infinite, as there is no such a thing as a species
which includes only an individual within its horizon.20 On the other hand, the
presupposition of unity leads us to a highest genus encompassing all concepts
that must be presupposed as the “universal and true horizon”.
123
This system of concepts is not, to be sure, how scientific knowledge looks

like according to Kant. Indeed, we may never be able to construct such a sys-
tem. It is, however, the ideal framework according to which empirical cogni-
tion can be progressively systematized – that is, unified as well as indefinitely.
It is, as it were, a “blank” space which leaves the content of experience com-
pletely undetermined. Kant specifies that the unity of reason is “in itself unde-
termined in regard to the conditions under which, and the degree to which,
the understanding should combine its concepts systematically” (A665/B693).
Reason cannot constitute the content of cognition, yet it provides an order-
ing template of maximal unity and division of concepts – in Kant’s words, an
“analogue of a schema” (A665/B693) – that is necessary for us to transform
disparate and particular empirical cognitions into scientific knowledge. In
other words, the two principles of theoretical reason (unity and specification)
together with the principle that results from their combination (continuity)
act as the “axes” of an ideal conceptual space within which each “student of
nature” is able to systematize empirical cognition according to her inclina-
tion (A655/B683). Within such a space, as I will argue in the next section, we
can reconsider the contemporary principles of unification and pluralism as
mutually consistent meta-scientific guidelines.
4. A Kantian solution to the “antinomy” between

pluralism and monism
In this section, I want to elaborate on how Kant’s theory of systematic unity
can illuminate the contemporary debate on monism and pluralism. Needless
to say, Kant’s theory is grounded on specific assumptions that might not be
shared by contemporary discussants.21 But even taking into account inevi-
table disanalogies between Kant’s approach and the contemporary status of
the debate, one might doubt the very possibility of a comparison between
Kant’s treatment of scientific rationality and the current debate for at least
two reasons. First, Kant seems to take for granted that systematic unity char-
acterizes scientific knowledge – a characterization that seems to beg the very
question at stake in the debate. I already suggested that this is not fully accu-
rate. Indeed, Kant defines scientific knowledge as systematic cognition.22 But
systematic unity is not mere unity; it is a systematic organization of concepts
in which unity and plurality play equally important roles and express oppo-
site interests of reason. If this is correct, Kant’s systematic unity should not
be merely equated with a regulative ideal of unification or integration of
multiple accounts. Since we do not know whether nature is really a unity
or a plurality, we should not only aim to unify and integrate our theories as
much as possible. Rather, we should pursue both maximal unity and maximal
plurality as epistemic principles leading to systematic cognition.23
Second, one might find Kant’s taxonomic terminology – “gen-
era”, “species”, “transition among species” – ill-suited to be used in the
124
contemporary discussion. While we will probably use neither the image of

an all-encompassing genus to describe the ideal of unity nor the idea of an
indefinitely specified series of species to illustrate the principle of pluralism,
it is not difficult to adapt Kant’s terminology to the present debate. Kant
himself suggests a very broad interpretation of those principles. Not only
does he equivalently use more general terms like “identity” or “variety” to
describe how the principles of genera and species are supposed to work, he
also explicitly specifies that these principles concern “not merely things, but
even more the mere properties and powers of things” (A662/B690). Inter-
estingly, Kant uses the principles of reason to explain how we implement
geometrical and mathematical abstractions in modeling phenomena: for
instance, we employ the principle of genera to infer the elliptical path of
planets, while the principle of continuity is used when “conceiving hyperboli-
cal paths for comets in which these bodies leave our solar system entirely”
(A663/B691). These considerations suggest that Kant’s principles should be
read as meta-scientific norms rather than mere taxonomic guidelines.24 A
principle of genera, for instance, is invoked not only when we classify differ-
ent species of things under the same genus (say, “oak” and “chestnut” under
the genus Fagaceae) but also when we presuppose that the same geometrical
or mathematical abstraction describes a property of various phenomena.
What are, then, the advantages of this comparison and the insights in
Kant’s account of reason we can exploit? In Section 2, I presented the debate
over pluralism. I argued that the contemporary discussion implies that plu-
ralism and monism are incompatible principles. Pluralism is generally inter-
preted as a self-sufficient principle which either excludes any principle of
unity or is complemented by a weak form of monism when it comes to the
desire of unifying different theories. I already suggested that both solutions
are unsatisfactory, for they do not recognize the role unification plays at any
level of science – from “basic” theory formation to “higher” theory unifica-
tion. Moreover, an empirical form of pluralism does not fully express the
prescriptive urge that motivates pluralism itself. Kant provides us with a
useful meta-scientific framework for thinking about these two principles in
a new, possibly more promising way. In what I take to be Kant’s variation
on perspectivism, unity and pluralism do not give rise to an antinomy but
are compatible principles that essentially complement each other. I will now
show how each of these two principles and their combination can be recon-
sidered according to the Kantian framework.
4.1. Regulative unity

We saw that contemporary pluralism tends to neglect the significance of a
principle of unity in science. Pluralists only admit a principle of plurality,
yet they do not go as far as to affirm that unity of nature is impossible. It
is an open question, they argue, whether there is unity in nature. From this
125
insight, however, it does not follow that unity should not play any epistemic
role in science. From a Kantian perspective, we can perfectly agree with the
claim that we do not know whether nature is actually a unity – indeed, this
is why Kant presents unification as a regulative principle. The rejection of
a principle of unity is instead grounded in a misleading characterization
of the very same principle. This characterization is parasitic on what Kant
would call a metaphysical reading of the principle of unity. According to this
reading, the nature of the world is such that it can, at least in principle, be
fully explained by a single account. In Kantian terms, as we saw, that would
mean being deceived by an “optical illusion”. To be guided by a perspectival
principle of unity is not to determine the nature of the world. It is, instead,
to use ideas in order to universalize what is only particularly given to us in
experience. Unity, if thought of along the Kantian lines, is only a regulatively
employed principle and, as such, can be maintained without posing any
threat to a principle of plurality.
The regulative status of the principle should not, however, incline us to
consider unity a mere additional desideratum of science. We saw that some
forms of pluralism – most notably, Giere’s perspectivism – allow weak forms
of unification provided that pluralism remains the only necessary requirement
when we do science. I suggested, however, that the exclusion of a principle of
unity at the “basic” level of science poses an additional challenge. For epis-
temically situated perspectives are themselves instances of unity: Newton’s
theory of motion, Maxwell’s theory of electromagnetism and Einstein’s spe-
cial theory of relativity are all clear examples of perspectives as brilliant unifi-
catory achievements. Although Giere introduces his methodological principle
of unity as an anti-Kantian move (Giere 2006b: 36), Kant would not only
agree with the need of a principle complementing pluralism, he would also
have internal resources to address the previously mentioned challenge.
In Kant’s perspectivism, unity does not merely play an auxiliary function
but is an integral part of the formation of a perspective – each perspective
results from an ideal projection and is itself a form of theoretical unity.
Furthermore, we saw that the presupposition of identity has not merely a
taxonomic import but is equally employed in geometrical and mathemati-
cal representations of phenomena. This different kind of perspectivism can
therefore inspire an integration of unity into the perspectival picture we are
familiar with. If we look at how scientific theorizing works across different
perspectives, unification indeed plays an important role. Scientific theories
typically contain general and abstract principles that are supposed to unify
seemingly different cognitions of phenomena and properties of phenomena.
As Morrison has shown, this is commonly obtained through the mathemati-
cal elaboration of general and abstract frameworks. Such frameworks usu-
ally contain “a theoretical parameter, quantity or concept that ‘represents’
the unifying mechanism – that is, a parameter that functions as the neces-
sary piece of theoretical structure that either facilitates or represents the
126
unification of distinct phenomena” (Morrison 2000: 4). In Maxwell’s elec-

trodynamics, for instance, this theoretical unifier is represented by electric
displacement together with the Lagrangian formalism.
Granted that unification plays a role in the process of scientific theorizing,
one might still ask why that is the case. In other words, what is precisely the
epistemic function of unification? The answer that unifying our cognitions
of phenomena affords a deeper understanding of them is intuitively appeal-
ing but needs further clarification. Kitcher and Friedman have proposed
influential epistemological models according to which unification is essen-
tial to scientific explanation (see Friedman 1974; Kitcher 1981). To put it in
Friedman’s words,
Science increases our understanding of the world by reducing the

total number of independent phenomena that we have to accept as
ultimate or given. A world with fewer independent phenomena is,
other things equal, more comprehensible than one with more.
(Friedman 1974: 15)
Despite their plausibility, these models have been faced with a plethora of
objections in the last decades. Convincing counterexamples suggest a more
cautious distinction between unification and explanation. For example, it
has been shown that in some cases, unification provides us with little or no
explanation – a possible example being the derivation of Mendelian rules of
inheritance within molecular cell biology: even if possible, such a derivation
would not enhance our understanding.25 More generally, it is commonplace
in contemporary science to obtain a deeper understanding of phenomena by
pursuing disunity rather than unity at the explanatory level.
The need to problematize the relation between unification and explana-
tion, however, does not undermine the explanatory value of unifying cog-
nitions and theories. Newton’s mechanics, Maxwell’s electrodynamics and
Einstein’s special relativity all successfully explain phenomena by recog-
nizing them as instances of general principles.26 The challenge is rather to
understand why there is no simple identification between unity and explana-
tion. This – I argue – is perfectly understandable from the Kantian approach
here suggested. It is true that Kant’s principle of unification follows a model
of explanation in which the universal explains the particular. But, crucially,
unity is only a regulative idea that neither determines the content of cogni-
tion nor gives us any assurance that our attempts will succeed. Such an idea
only prescribes to search for any given cognitions, for higher explanatory
principles.27 Scientists might simply not find truly explanatory universal
principles, and, as we saw, universals are only hypothetically postulated –
and therefore revisable through further empirical evidence.
As a result, our epistemic necessity to presuppose unity in order to elabo-
rate general principles is compatible with the admission of a gap between
127
unification and explanation. Indeed, given that regulative unity does not
commit us to a metaphysical thesis of unity nor to the realizability of a
unified system, we have no reason to expect this relation to be otherwise.
Explanatory unification may simply be out of reach, and we may be better
off relying on a plurality of explanations, as in the case of non-derivation of
Mendelian rules within molecular cell biology. Even when available, theo-
retical unities should be best understood as “starting points” for further
explanation: one might, for instance, argue that Einstein’s general relativ-
ity offers a better understanding of gravitation than Newton’s mechanics
(Morrison 2000: 33). In neither case, however, is unity undermined as an
explanatory valuable idea that scientists should strive to realize as much as
possible – from “basic” theory construction to ambitious theory unification.
4.2. Regulative pluralism

Since regulative unity does not posit the existence of a unity of nature, noth-
ing prevents us from also embracing a principle of plurality. Importantly, the
use of this principle should not be interpreted in constitutive terms either.
The principle of specification does not correspond to the metaphysical
assertion that nature is an infinite and irreducible variety or manifoldness.
Nor is it the mere epistemic satisfaction with a given plurality of models
and theories. As we saw, an unqualified form of pluralism does not fully
portray the motivation behind pluralism itself – it is too weak a principle.
Regulative pluralism amounts instead to the presupposition of a maximally
specifiable system of concepts describing phenomena. While following this
principle, scientists are not simply satisfied with a plurality of “local” uni-
ties, as empirical pluralists seem to suggest. Scientists are instead interested
in the progressive, indefinite diversification of phenomena according to ever
finer-grained perspectives.
As mentioned in the previous subsection, the explanatory role of pluralism
can be as beneficial as that of unification. Since the current system of knowl-
edge is incomplete and hypothetical, unified explanations may be unavailable
or too general to be explanatorily satisfactory. This is why it is not uncommon
to find disunity at the level of explanation in many fields of contemporary
science. We can generalize from this fact. Given that unification is an ongo-
ing process that does not metaphysically reduce particulars to universals, we
necessarily need to rely on specific theories to explain particular phenomena
or particular aspects of phenomena. To regulatively presuppose pluralism
of cognition amounts to the demand to seek ever more specific theoretical
approaches. And such prescription allows us to account for the explanations
of phenomena that unified theories are not able to provide us with.
Thinking about unity and pluralism in regulative terms finally answers
our initial question on the apparent incompatibility between these two prin-
ciples in contemporary debates. As Kant argues, an antinomy is taking place
128
between unity and pluralism only when these principles are interpreted as
“constitutive” or “objective” principles:
If merely regulative principles are considered as constitutive, then

as objective principles they can be in conflict; but if one considers
them merely as maxims, then it is not a true conflict, but it is merely
a different interest of reason that causes a divorce between ways
of thinking. . . . In this way the interest in manifoldness (in accor-
dance with the principle of specification) might hold more for this
sophistical reasoner, while unity (in accordance with the principle
of aggregation) holds more for that one. Each of them believes that
his judgment comes from insight into the object, and yet he grounds
it solely on the greater or lesser attachment to one of the two prin-
ciples, neither of which rests on any objective grounds, but only on
the interest of reason, and that could better be called “maxims” than
“principles”.
As soon as we stop regarding these principles as insights into the nature of

objects, the antinomy between them disappears. We are left instead with dif-
ferent “maxims” that, although in conflict with each other, can be reunited as
compatible “ways of thinking” in the larger framework of reason. One rea-
soner may be inclined towards finding unity in nature; another may instead
be interested in empirical variety. Although each of them believes that her
judgment is objectively motivated – that is, constitutive of nature – their
conflict does not result from rational incompatibility. It only regards their
“attachment to one of the two principles”. I contend that the present debate
on pluralism and unity presents us with a similar “antinomy”. Monism and
pluralism are both legitimate and compatible “axes” of the same conceptual
space of scientific thinking. Conflicts arise only when at least one of these
two principles is held as an objective insight. Classic metaphysical monism
holds unity as an objective truth about the natural world. Empirical plural-
ism does not fare better than the classic view, for it recognizes unity only as
an objective claim about the world and rejects it altogether. These judgments,
however, are mistaken about the very rationale of these two principles. As
we saw, both principles, if regulatively employed, are needed to successfully
promote the scientific investigation of nature.
4.3. The space of theoretical reason

The resolution of the antinomy, however, still leaves us with a lingering
problem. For if monism and pluralism are regarded as compatible subjec-
tive maxims, we seem to have lost the necessary status Kant also attaches to
these epistemic principles. Indeed, several interpretations have been misled
129
by Kant’s terminology and suggest a weak reading of the principles of rea-

son as mere maxims (e.g. Guyer and Walker 1990: 227–8; Pickering 2011).
As we saw, however, unity and pluralism are not just welcome epistemic
rules; they are necessary, meta-scientific principles. Together they make up
a perspectival space in which scientific knowledge is possible. Are scientific
maxims compatible with Kant’s overall picture? And if this is the case, how
can this inspire the contemporary debate? First, Kant’s language should not
mislead us. The previous passage seems to suggest that objective principles
should be replaced by merely subjective, individual maxims. In this way, an
antinomy turns out to be a mere conflict between opposed interests. This is
correct as far as it goes. However, Kant clarifies:
Reason has in fact only a single unified interest, and the conflict
between its maxims is only a variation and a reciprocal limitation
of the methods satisfying this interest.
(A666/B694)
I take reason’s “single unified interest” to mean the systematization of

empirical knowledge that results from the combination of the principles
of reason.28 Maxims are not mere synonyms for the “indispensably neces-
sary” regulative principles, as the previous passage may seem to suggest.
Instead, they only express the individual attachment each scientist has to
one principle rather than to the other. They are therefore merely alterna-
tive, yet compatible “methods” and together satisfy the only real interest of
reason, namely the construction of the system of knowledge. If this is cor-
rect, the fact that scientists may be individually interested in following just
one maxim does not undermine the necessity of both unity and disunity as
epistemic principles. Although each principle can be isolated as a maxim,
systematic unity requires the presupposition of both unity and plurality
of cognition for its construction. These principles make up the conceptual
space within which any individual attempt to maximize either unity or dis-
unity of cognition is possible.
We can finally apply this insight to the contemporary debate. Pluralism
and unification should not be regarded as merely compatible methodologi-
cal maxims. “Pluralists” are more interested in empirical variety and privi-
lege the axis of plurality; “monists” are more inclined towards rational unity
and therefore pursue unification of cognition. Despite individual conflicts
of interests, however, they ultimately share the same meta-scientific concep-
tual space of investigation. Perspectivism can therefore be reconsidered not
just as the place of conflict among perspectives nor as a form of pluralism
opposed to monism but as the common space where different perspectives
can always be further unified and disunified. The space of theoretical reason
does not, to be sure, prescribe or determine the content of phenomena. Yet
it provides scientists with a template of maximal systematicity of knowledge
130
which grounds the possibility of different, compatible perspectives – either

aimed at unity or at disunity of scientific cognition.
Conclusion
I suggested that the current conflict between pluralism and monism is based
on a misrepresentation of those epistemic principles. The conflict ultimately
resides in a metaphysical characterization of unity in science that does not
fully capture the epistemic significance of monism in scientific theorizing.
Looking at Kant’s perspectivism in the Critique of Pure Reason not only
allows us to resolve this apparent antinomy but also to rethink unity and
pluralism as necessary regulative principles. These principles together make
up a meta-scientific space of reason, within which each scientist can follow
her inclination towards unity or disunity of cognition.
There are several important issues this chapter leaves open. Let me con-
clude by mentioning one which is particularly pressing. I focused on the
epistemic function of unity and pluralism within a perspectival framework.
I therefore bracketed the question of whether perspectivism may deliver a
form of realism in science (and if so, what kind of realism). The two prob-
lems are, however, very tightly connected. The problem of realism is indeed
prominent in the work of many proponents of perspectivism (see e.g. Rueger
2005; Massimi 2018). And it is also present in Kant’s discussion of systema-
ticity of science. To investigate this issue from a Kantian perspective would
require looking more closely at the transition from logical to transcendental
principles of systematic unity Kant presents in the Appendix. Kant has a
fairly complex story regarding whether and how this transition might be
justified, a story that would well be worth investigating in parallel with cur-
rent debates in contemporary science. As such, however, it falls outside the
scope of this chapter.29
Notes
1 Early criticism of the unity of science thesis can be found in Suppes (1978). More
recent criticism can be found in Dupré (1996), Cartwright (1999) and Chang
(2012), discussed subsequently.
2 Proponents of this claim include, for example, Paul Oppenheim and Hilary Put-
nam; see their paper, “The Unity of Science as a Working Hypothesis” (1958).
3 I use “scientific theorizing” to refer to the construction and employment of theo-
ries in science, broadly construed.
4 For the distinction between plurality and pluralism, see Kellert et al. (2006: ix–x).
5 Kitcher presents a Kantian-inspired “modest unificationism” that tolerates mul-
tiple accounts but still aims to reduce their number. Inspired by Kant’s notion
of systematic unity, Breitenbach and Choi (2017) make a convincing case that
the unavoidable plurality of accounts should not discourage our attempt to
integrate them in a unified system. While my chapter greatly benefits from
both contributions, I will present Kant’s systematic unity not merely as an ideal
131
of unification but as a conceptual space that results from both unification and
pluralism.
6 References to the Critique of Pure Reason follow the standard A and B pagina-
tion, citing Kant (1998). All other Kant citations refer to the Akademie edition
and to the English translation used.
7 A similar proposal (with respect to realism in science) can be found in Massimi
(2017). This chapter is, however, noncommittal to the realism of scientific theo-
ries and focuses instead on the epistemic features of perspectivism.
8 Particularly in its first part (A642/B670–A668/B696).
9 For example, Guyer has famously argued for this position: “Systematicity is not
a factor which enters into understanding’s constitution of empirical knowledge
itself, but only an additional desideratum which reason seeks to find or construct
in the empirical knowledge produced by understanding” (1990: 28).
10 See, for example, Grier (2001: 37) and Allison (2004: 425). Massimi (2021)
notes that Kant does not mention Newton in the Appendix nor in the Dreams
of a Spirit-Seer (1766), where the metaphor appears for the first time. She then
argues that Kant’s source is actually Descartes’s Treatise on Man (which Kant did
mention in 1766). While it is plausible that Kant inherited the metaphor from
Descartes, I also maintain that Kant was at least partly inspired by Newton’s
Opticks. In the Dreams, just after mentioning the focus imaginarius for the first
time, Kant explains this analogy by using the example of an object reflected by a
concave mirror – interestingly, the same image appears in the second part of the
VIII axiom (see Newton 1952: 19 and Ak. 2:344; Kant 1992b: 331–2). Further-
more, as I will argue in the following, the “optical illusion” Kant is talking about
perfectly matches Newton’s description of the illusion occurring when an object
is reflected by a mirror.
11 “AX. VIII. An Object seen by Reflexion or Refraction, appears in that place
from whence the Rays after their last Reflexion or Refraction diverge in falling
on the Spectator’s Eye. If the Object A be seen by Reflexion of a Looking-glass
mn, it shall appear, not in its proper place A, but behind the Glass at a, from
whence any Rays AB, AC, AD, which flow from one and the same Point of the
Object, do after their Reflexion made in the Points B, C, D, diverge in going from
the Glass to E, F, G, where they are incident on the Spectator’s Eyes. For these
Rays do make the same Picture in the bottom of the Eyes as if they had come
from the Object really placed at a without the Interposition of the Looking-
glass; and all Vision is made according to the place and shape of that Picture”
(Newton 1952: 18).
12 Reprinted in Newton (1952). Permission to reproduce the image granted by
Dover Publications.
13 The image is still reversed in the direction perpendicular to the mirror surface.
14 It is worth noting that, for Kant, all concepts are universal (“In kind, all concepts
are universal and can always hold of other things in a certain way”; see Vienna
Logic, Ak. 24: 909 [Kant 1992a: 352]). Empirical concepts are no exception.
Kant often uses “empirical concept”, “common concept” and “universal con-
cept” interchangeably in his lectures on Logic: see, for example, Blomberg Logic,
Ak. 24: 269 (Kant 1992a: 208). In the Appendix, Kant also reminds that “each
species is always a concept that contains within itself only what is common to
different things” (A655/B683).
15 Thanks to Luigi Caranti for pressing me on this point.
16 The concepts of genus and species are mutually related. See Jäsche Logic, Ak.
9: 97: “Like higher and lower concepts, genus and species concepts are distin-
guished not as to their nature, then, but only in regard to their relation to one
another (termini a quo or ad quod) in logical subordination” (1992a: 594).
132
17 Kant, however, explicitly rejects the Leibnizian idea of “actual infinity”, for, he
explains, that would determine the conceptual sphere of division. The logical
principle of species only asserts the “indeterminacy” of conceptual specification
(A656/B684).
18 The Latin word perspectiva comes from perspicere, meaning “to see clearly”.
It thus amounts to a literal translation of the Greek word optikḗ; See Panofsky
(1991: 75–6).
19 See, for example, Panofsky (1991: 31). According to Panofsky, perspective is a
kind of representation in which “bodies are absorbed into a homogeneous and
infinite system of dimensional relationships”, that is a “quantum continuum”
(ibid.: 44). Similar remarks are easy to find throughout the secondary literature
(see e.g. Belting 2011).
20 The nature of a concept is such that it can only contain what is common to dif-
ferent things.
21 One, for instance, might question the nature of the Kantian faculties or the logi-
cal approach to the investigation of nature to begin with.
22 “Systematic unity is that which first makes ordinary cognition into science, i.e.,
makes a system out of a mere aggregate of it” (A832/B860).
23 See Breitenbach and Choi (2017) on this point. They argue that “ideal science”
may turn out to be a single theory of everything or a “unity constituted of het-
erogeneous parts”; “we have to wait for nature to tell us” (ibid.: 398). Note,
however, that empirical pluralism denies that scientists should aim for the inte-
gration of heterogeneous accounts (see 3.3). Such alternative may therefore be
too “unificationist” for the empirical pluralist. In my reading, the alternative is
not between a single and an integrated unity of nature but between a unified and
a disunified (i.e., non-integrated) account of nature.
24 Guyer (2017: 56–7) makes a similar point.
25 Example taken from Kitcher (1999: 337).
26 For a discussion of how theory unification does not necessarily imply loss of
explanation, see Rueger (2005).
27 Following Willaschek (2018: 66–7).
28 For a similar suggestion, see Watkins (2017: 26).
29 I wish to thank Michela Massimi, Alix Cohen and the audience at the Multilat-
eral Kant Colloquium held in Catania in October 2018 for helpful comments
on earlier versions of this chapter. Special thanks go to Luigi Caranti and Ales-
sandro Pinzani for organizing this volume and the opportunity to submit my
contribution. This chapter is part of a project that has received funding from
the European Research Council under the European Union’s Horizon 2020
research and innovation program (European Consolidator Grant H2020-ERC-
2014-CoG 647272, Perspectival Realism: Science, Knowledge, and Truth from a
Human Vantage Point).
References
Allison, H. E. 2004. Kant’s Transcendental Idealism. New Haven: Yale University Press.
Belting, H. 2011. Florence and Baghdad: Renaissance Art and Arab Science. Cam-
bridge, MA: Belknap Press of Harvard University Press.
Breitenbach, A. and Choi, Y. 2017. “Pluralism and the Unity of Science”, The
Monist, 100: 391–405.
Cartwright, N. 1999. The Dappled World: A Study of the Boundaries of Science.
Cambridge: Cambridge University Press.
133
Chang, H. 2012. Is Water H2O? Evidence, Pluralism and Realism. Dordrecht:

Springer.
Dupré, J. 1996. “Metaphysical Disorder and Scientific Disunity”, in P. Galison and
D. J. Stump (eds.) The Disunity of Science: Boundaries, Context and Power. Stan-
ford: Stanford University Press: 101–17.
Friedman, M. 1974. “Explanation and Scientific Understanding”, Journal of Phi-
losophy, 71: 5–19.
Giere, R. N. 2006a. Scientific Perspectivism. Chicago: University of Chicago Press.
Giere, R. N. 2006b. “Perspectival Pluralism”, in S. H. Kellert, H. E. Longino,
and C. K. Waters (eds.) Scientific Pluralism. Minneapolis: University of Minnesota
Press: 26–41.
Grier, M. 2001. Kant’s Doctrine of Transcendental Illusion. Cambridge: Cambridge
University Press.
Guyer, P. 1990. “Reason and Reflective Judgement: Kant on the Significance of Sys-
tematicity”, Noûs, 24: 17–43.
Guyer, P. 2017. “Imperfect Knowledge of Nature. Kant, Hume, and the Laws of
Nature”, in M. Massimi and A. Breitenbach (eds.) Kant and the Laws of Nature.
Cambridge: Cambridge University Press: 49–67.
Guyer, P. and Walker, R. 1990. “Kant’s Conception of Empirical Law”, Proceedings
of the Aristotelian Society, 64: 221–58.
Kellert, S. H., Longino, H. E., and Waters, C. K. (eds.). 2006. Scientific Pluralism.
Minneapolis: University of Minnesota Press.
Kitcher, P. 1981. “Explanatory Unification”, Philosophy of Science, 48: 507–31.
Kitcher, P. 1999. “Unification as a Regulative Ideal”, Perspectives on Science, 7:
337–48.
Massimi, M. 2017. “What Is This Thing Called ‘Scientific Knowledge’? Kant on
Imaginary Standpoints and the Regulative Role of Reason”, Kant Yearbook, 9:
63–84.
Massimi, M. 2018. “Perspectivism”, in J. Saatsi (ed.) The Routledge Handbook on
Scientific Realism. London: Routledge.
Massimi, M. 2021. “Points of View. Kant on Perspectival Knowledge”, Synthese,
198: 3279–3296.
Mitchell, S. 2003. Biological Complexity and Integrative Pluralism. Cambridge:
Cambridge University Press.
Morrison, M. 2000. Unifying Scientific Theories: Physical Concepts and Mathemat-
ical Structures. Cambridge: Cambridge University Press.
Morrison, M. 2008. “Reduction, Unity and the Nature of Science: Kant’s Legacy?”,
in M. Massimi (ed.) Kant and Philosophy of Science Today. Cambridge: Cam-
bridge University Press: 37–62.
Newton, I. 1952. Opticks, or a Treatise of the Reflections, Refractions, Inflections &
Colours of Light. New York: Dover Publications.
Oppenheim, P. and Putnam, H. 1958. “The Unity of Science as a Working Hypoth-
esis”, Minnesota Studies in the Philosophy of Science, 2: 3–36.
Panofsky, E. 1991. Perspective as Symbolic Form. New York: Zone Books.
Pickering, M. 2011. “The Idea of the Systematic Unity of Nature as a Transcendental
Illusion”, Kantian Review, 16(3): 429–48.
Rueger, A. 2005. “Perspectival Models and Theory Unification”, British Journal for
the Philosophy of Science, 56: 579–94.
134
Suppes, P. 1978. “The Plurality of Science”, in P. Asquith and I. Hacking (eds.) Pro-
ceedings of the 1978 Biennial Meeting of the Philosophy of Science Association.
East Lansing: Philosophy of Science Association: 3–16.
Teller, P. 2011. “Two Models of Truth”, Analysis, 71: 465–72.
Watkins, E. 2017. “Kant on the Unity and Diversity of Laws”, in M. Massimi and
A. Breitenbach (eds.) Kant and the Laws of Nature. Cambridge: Cambridge Uni-
versity Press: 11–29.
Willaschek, M. 2018. Kant on the Sources of Metaphysics: The Dialectic of Pure
Reason. Cambridge: Cambridge University Press.
135
8
CAN PHYSICS EXPLAIN PHYSICS?
ANTHROPIC PRINCIPLES AND
TRANSCENDENTAL IDEALISM
Robert Hanna
Science is a dialogue between [hu]mankind and nature, the

results of which have been unpredictable. At the beginning of
the twentieth century, who would have dreamed of unstable
particles, an expanding universe, self-organization, and
dissipative structure? But what makes this dialogue possible?
A time-reversible world would also be an unknowable world.
Knowledge presupposes that the world affects us and our
instruments, that there is an interaction between the knower
and the known, and that this interaction creates a difference
between past and future. Becoming is the sine qua non of
science, and indeed, of knowledge itself.
(Prigogine 1997: 153)
1. Introduction
Can contemporary physics, understood as a natural science that is commit-
ted to the Standard Models, explain itself? Yes, but only if it supplements
the Standard Models by affirming a philosophically defensible and scientifi-
cally respectable version of the Anthropic Principle that is also equivalent to
a suitably weak version of Immanuel Kant’s metaphysics of transcendental
idealism. Or so I will argue in this chapter.
2. The standard models and the anthropic principle:

weak vs. strong
By the Standard Models in contemporary physics, I mean the current Stan-
dard Model of Cosmology (SMC), together with a proper sub-part of SMC,
the current Standard Model of Particle Physics (SMPP), per this informative
non-technical gloss:
136 DOI: 10.4324/9781003050742-9

The current Standard Model of Cosmology (SMC), also called

the “Concordance Cosmological Model” or the “ΛCDM Model,”
assumes that the universe was created in the “Big Bang” from pure
energy, and is now composed of about 5% ordinary matter, 27%
dark matter, and 68% dark energy.1
[T]he SMC is based primarily upon two theoretical models: (1) the
Standard Model of Particle Physics (SMPP),2 which describes the phys-
ics of the very small in terms of quantum mechanics and (2) the
General Theory of Relativity (GTR),3 which describes the physics of
the very large in terms of classical mechanics; it also depends upon
several additional assumptions.
The main additional assumptions of the SMC are: (1) the uni-
verse was created in the Big Bang from pure energy; (2) the mass
energy content of the universe is given by 5% ordinary matter, 27%
dark matter, and 68% dark energy; (3) the gravitational interactions
between the above three components of the mass energy content
of the universe are described by the GTR; and (4) the universe is
homogeneous and isotropic on sufficiently large (cosmic) scales.
Unfortunately, both the SMPP and the GTR are considered to be
incomplete in the sense that they do not provide any understand-
ing of several empirical observations. The SMPP does not provide
any understanding of the existence of three families or generations
of leptons and quarks, the mass hierarchy of these elementary par-
ticles, the nature of gravity, the nature of dark matter, etc.4 The
GTR does not provide any understanding of the Big Bang cosmol-
ogy, inflation, the matter-antimatter asymmetry in the universe, the
nature of dark energy, etc.
Furthermore, the latest version of the SMC, the ΛCDM Model is
essentially a parameterization of the Big Bang cosmological model in
which the GTR contains a cosmological constant, Λ, which is asso-
ciated with dark energy, and the universe contains sufficiently mas-
sive dark matter particles, i.e., “cold dark matter.” However, both dark
energy and dark matter are simply names describing unknown entities.
(Robson 2021)
Correspondingly, here are two informative non-technical glosses on the

Anthropic Principle.
The anthropic principle is a group of principles attempting to deter-

mine how statistically probable our observations of the universe
are, given that we could only exist in a particular type of universe to
start with. . . . In other words, scientific observation of the universe
would not even be possible if the laws of the universe had been
137
ROBERT HANNA
incompatible with the development of sentient life. Proponents of

the anthropic principle argue that it explains why this universe has
the age and the fundamental physical constants necessary to accom-
modate conscious life, since if either had been different, we would
not have been around to make observations. Anthropic reasoning
is often used to deal with the fact that the universe seems to be fine
tuned.
There are many different formulations of the anthropic principle.
Philosopher Nick Bostrom counts them at thirty, but the principles
can be divided into “weak” and “strong” forms, depending on the
types of cosmological claims they entail. The weak anthropic prin-
ciple (WAP) such as the one defined by Brandon Carter, states that
the universe’s ostensible fine tuning is the result of selection bias
(specifically survivorship bias). Most often such arguments draw
upon some notion of the multiverse for there to be a statistical pop-
ulation of universes to select from. However, a single vast universe
is sufficient for most forms of the WAP that do not specifically deal
with fine tuning. The strong anthropic principle (SAP), as proposed
by John D. Barrow and Frank Tipler, states that the universe is in
some sense compelled to eventually have conscious and sapient life
emerge within it. (Wikipedia 2021a)
We could not possibly have existed in conditions that are incom-
patible with the existence of observers. The famous weak anthropic
principle (WAP) . . . suggests that this apparently trivial point may
have important consequences:
[W]e must be prepared to take account of the fact that our
location in the universe is necessarily privileged to the extent of
being compatible with our existence as observers.
(Carter 1974: 293, emphasis in the original)
Our methods of empirical observation are unavoidably biased
towards detecting conditions which are compatible with the exis-
tence of observers. For example, even if life-hostile places vastly
outnumber life-friendly places in our universe, we should not be
surprised to find ourselves in one of the relatively few places that
are life-friendly and seek an explanation for this finding, simply
because – in virtue of being living organisms – we could not possibly
have found ourselves in a life-hostile place. Biases that result from
the fact that what we observe must be compatible with the exis-
tence of observers are referred to as observation selection effects.
The observation selection effects emphasized by the weak anthropic
principle with respect to location in the universe are emphasized by
what Carter dubs the strong anthropic principle (SAP) with respect
to the universe as a whole:
138
[T]he Universe (and hence the fundamental parameters on

which it depends) must be such as to admit within it the cre-
ation of observers within it at some stage.
(Carter 1974: 294)
Carter’s formulation of the SAP has led some authors, most influen-
tially Barrow and Tipler (Barrow and Tipler 1986), to misinterpret
it along teleological lines and as thereby categorically different from
the WAP.
(Simon 2018)
The philosophical upshot, so far, is threefold.

First, both versions of the Anthropic Principle – that is, weak (WAP)
and strong (SAP) – go significantly beyond what’s contained in the Stan-
dard Models per se, by requiring an asymmetric or irreversible, forward-
directed arrow of time and therefore also requiring the manifest reality of
time (Prigogine 1997; Smolin 2014). Moreover, if, as per the epigraph of this
chapter, Prigogine is correct (and I think that he is), then postulating the man-
ifest reality of an asymmetric or irreversible, forward-directed arrow of time
is a necessary condition of adequately explaining human scientific knowledge
Second, although there are various possible interpretations of WAP, I’m
going to interpret it as the thesis that our nomological possibility as a bio-
logical species is built into in the natural world from the Big Bang forward.
It is especially to be noted that WAP is consistent with various versions
of metaphysical materialism or physicalism, universal natural determinism,
universal natural indeterminism and universal natural mechanism (Hanna
and Maiese 2009; Hanna 2018a).
Third, although there are also various possible interpretations of SAP,
some of which take SAP to be consistent with WAP, I’m going to interpret
SAP as the categorically distinct thesis that the nomological or metaphysical
necessity of our existence as sentient, sapient animals is built into the natural
world from the Big Bang forward. It is especially to be noted that WAP is
consistent with classical theism and the existence of a “3-O” creator-God:
omnipotent, omniscient and omnibenevolent.
3. The moderate anthropic principle and weak

transcendental idealism
Given
1 that we actually exist,

2 that we also know a priori that our actual existence logically follows
from the fact of our conscious and self-conscious thinking (the Carte-
sian Cogito: necessarily, if I’m thinking, then I actually exist),
139
ROBERT HANNA
3 that the natural world actually exists, and

4 that the natural world actually existed long before rational “human, all-
too-human” minded animals began to exist as a biological species,
then since actuality entails possibility, it follows that we cannot be either

logically, metaphysically or nomologically impossible. Or otherwise and,
now positively put, necessarily, given that we actually exist and also know
a priori that our actual existence follows from the fact of our conscious
and self-conscious thinking, then the natural universe, from the Big Bang
forward, structurally contains our real possibility as rational “human, all-
too-human” minded animals. This is what I call the Moderate Anthropic
Principle, a.k.a. MAP.
MAP is categorically stronger than WAP, because MAP entails the denial
of metaphysical materialism or physicalism, universal natural determinism,
universal natural indeterminism and universal natural mechanism alike. But
MAP is also categorically weaker than SAP, since MAP rules out any ver-
sion of classical theism. Indeed, and perhaps surprisingly, MAP is necessarily
equivalent to a suitably weak version of Kant’s metaphysics of transcenden-
tal idealism. Let me explain.
What I’m talking about is what I call Weak Transcendental Idealism, a.k.a.
WTI, which is my interpretation of a doctrine that Kant presented and defended
in the Critique of Pure Reason in order to explain the real possibility of a
priori or non-empirical necessary truth and knowledge in logic, mathematics,
metaphysics and natural science, as well as the real possibility of a posteriori
or empirical/observational contingent truth and knowledge in natural science.
According to Kant, a mental representation is transcendental when it is
either part of, or derived from, our non-empirical (hence a priori) innately
specified spontaneous cognitive capacities. Then transcendental idealism can
be stated as a two-part philosophical equation: transcendental idealism =
representational transcendentalism + cognitive idealism, per the following
definitions.
1 Representational Transcendentalism: Necessarily, all the forms or struc-

tures of rational human cognition are generated a priori by the empiri-
cally triggered, yet stimulus-underdetermined, activities of our innately
specified spontaneous cognitive capacities, that is, cognitive compe-
tences, cognitive faculties or cognitive powers (A11/B25 and Prol 4:
373n.).
2 Cognitive Idealism: Necessarily, all the proper objects of rational human
cognition are nothing but sensory appearances or phenomena, that
is, mind-dependent, spatiotemporal, directly perceivable, manifestly
real objects and never things-in-themselves or noumena, that is, mind-
independent, non-sensible, non-spatiotemporal, real essences constituted
by intrinsic non-relational properties. (A369 and Prol 4: 293–4, 375)
140
Now 1 + 2 also = Kant’s “Copernican revolution” in metaphysics:
Up to now it has been assumed that all our cognition must conform
to the objects; but all attempts to find out something about them a
priori through concepts that would extend our cognition have, on
this presupposition, come to nothing. Hence let us once try whether
we do not get farther with the problems of metaphysics by assum-
ing that the objects must conform to our cognition, which would
agree better with the requested possibility of an a priori cognition of
them, which is to establish something about objects before they are
given to us. This would be just like the first thoughts of Copernicus.
(Bxvi)
which I will rationally reconstruct as the Conformity Thesis:
It is not the case that rational human minds passively conform to the
objects they cognize, as in classical Rationalism and classical Empiri-
cism. On the contrary, necessarily, all the proper objects of rational
human cognition conform to – that is, they have the same form or
structure as, or are isomorphic to – the forms or structures that are
non-empirically generated by our innately specified spontaneous
cognitive capacities. So, necessarily, the essential forms or structures
of the manifestly real world we cognize are mind dependent.
In this way, all versions of transcendental idealism hold that the manifestly real
world we directly perceive conforms to the non-empirical forms or structures
of our innately specified cognitive capacities in some modally robust sense.
Many Kantians are also committed to Strong Transcendental Idealism,
a.k.a. STI, which says:
1 Things-in-themselves (a.k.a. “noumena”, or Really Real things, that is,

things as they could exist in a “lonely” way, altogether independently of
rational human minds or anything else, by virtue of their intrinsic non-
relational properties) really exist and cause our perceptions, although
rational human cognizers only ever perceive mere appearances or sub-
jective phenomena,
2 Rational human cognizers actually impose the non-empirical structures
of their innate cognitive capacities onto the manifestly real world they
cognize; that is, necessarily, all the essential forms or structures of the
proper objects of human cognition are literally type-identical to the a
priori forms or structures that are non-empirically generated by our
innately specified spontaneous cognitive capacities, and
3 Necessarily, if either all rational human cognizers went out of existence
or all minded beings of any kind went out of existence, then so would
141
ROBERT HANNA
the manifestly real world they cognize, and if either no rational human
cognizers had ever existed or no minded beings of any kind had ever
existed, then the manifestly real world would never have existed.
But I think that STI is clearly objectively false. More specifically, I think that
it is clearly objectively false that if either all actual human minds, including
mine, or all other kinds of minds, went out of existence, then the manifestly
real world would necessarily go out of existence too. I think that it is clearly
false that, for example, the actual existence of Pike’s Peak (a 14,000-foot
mountain near Colorado Springs, CO, USA, with a cog railway that runs right
to the summit [Wikipedia 2021b]) strictly depends on the actual existence of
human minds, including mine, or on the actual existence of any other kinds
of minds. Clearly, I think, Pike’s Peak can exist even if everyone, and every
minded being, including myself, does not actually exist, and in fact, I think
that Pike’s Peak actually existed millions of years before any conscious minds
of any kind existed, including of course the conscious minds of all rational
human animals, obviously including mine. In this way, clearly, a great many
things, including mountains like Pike’s Peak, exist objectively – for example,
shoes, ships, sealing wax, cabbages, kings, seas that do not boil and pigs
without wings. They are, all of them, neither subjective (i.e. strictly dependent
on individual minds of any kind) nor relative (i.e. strictly dependent on cul-
tures or societies of any kind). It is clear that they are all objectively manifestly
real and, to that extent, mind independent. So STI is clearly objectively false.
QED.
On the contrary, then, I believe that Kantians should be committed instead
to the objective truth of Weak Transcendental Idealism, which says:
1 Things-in-themselves/noumena are logically possible, but at the same

time, it is knowably unknowable and unprovable whether things-in-
themselves/noumena exist or not; hence, for the purposes of an ade-
quate anthropocentric or “human-faced” metaphysics, epistemology
and ethics, they can be ignored (= radical agnosticism and methodologi-
cal eliminativism about things-in-themselves/noumena),
2 Necessarily, all the proper objects of rational human cognition have
the same forms or structures as – that is, they are isomorphic to – the
forms or structures that are non-empirically generated by our innately
specified spontaneous cognitive capacities, but at the same time those
manifestly real worldly forms or structures are not literally type identi-
cal to those a priori cognitive forms or structures (= the isomorphism-
without-type-identity thesis),
3 It is a necessary condition of the existence of the manifestly real world
that if some rational human animals were to exist in that world, then
they would know that world a priori and also a posteriori, at least to
some extent (= the counterfactual knowability thesis), and
142
4 The manifestly real world has at some earlier times existed without
rational human minded animals, or any other minded beings, to know
it, and could exist even if no rational human minded animals, or any
other minded beings, ever existed to know it, even though some ratio-
nal human animals now actually exist in that manifestly real world (for
example, I [R.H.]), who do in fact know it a priori (for example, by my
knowing some simple necessary truths of logic and mathematics and also
by my knowing that necessarily, if I’m thinking, then I actually exist) and
also a posteriori, at least to some extent (= the existential thesis).
Here’ s a slightly more precise formulation of WTI’s crucial thesis (3), the
counterfactual knowability thesis:
Ap □ (∀x) (∃y) [MRWx → {(RHAy & MRWy) □→ Kyx}]
Definitions:
Ap □ = a priori necessarily
P □→ Q = If P were the case, then Q would be the case
MRWx = x belongs to the manifestly real world
MRWy = y belongs to the manifestly real world
RHAy = y is a rational human animal
Kyx = y knows x a priori and also a posteriori, at least to some extent
Natural Language Translation:

A priori necessarily, anything that belongs to the manifestly real world
is such that if some rational human animals were to exist in that world,
then they would know that thing and that world a priori and also a
posteriori, at least to some extent.
And here are two crucial implications of this thesis. First, the counterfactual
knowability thesis holds even if no rational human minded animals, or any
other minded beings, actually exist, or ever actually existed.5 Second, if any-
thing is such that rational human minded animals are unable to know it, at
least to some extent – for example, things-in-themselves or noumena – then
that thing does not belong to the manifestly real world. The first crucial
implication conveys the weak mind-independence and ontic integrity of the
manifestly real world. The manifestly real world is what it is, even if no minds
actually exist or ever actually existed. And the second crucial implication con-
veys the weak mind-dependence and inherent knowability of the manifestly
real world. The manifestly real world is what it is, only in relation to actual
or possible rational “human, all-too-human” animal minds like ours. The
single upshot of the two crucial implications is that the manifestly real world
is as real as anything can ever possibly be, on the reasonable assumption that
some epistemic-luck-resistant, global-new-evil-demon-skepticism-resistant
143
ROBERT HANNA
rational human knowledge of that world is actual or really possible. Or,

more precisely, and perhaps most surprisingly of all for anti-Kantians: any
epistemically tenable realism – that is, any realism that is truly capable of
avoiding epistemic luck (i.e. the merely accidental connection between truth
and conscious belief) and global-new-evil-demon-skepticism (i.e. the worry
that in a nearby possible world, we’d have the very same beliefs and the same
apparent evidence for holding those beliefs, but they would all be false; hence,
we aren’t rationally justified in holding those beliefs in the actual world) –
requires WTI (Hanna 2015: chs. 6–8).
4. Conclusion
The Moderate Anthropic Principle, which, as we have seen, is equivalent to
Weak Transcendental Idealism, falls between the Weak Anthropic Principle
and the Strong Anthropic Principle and explains the real possibility of contem-
porary physics as a natural science. MAP is therefore a theoretically unique
anthropic principle (i.e. categorically it is neither WAP nor SAP) that is not
only philosophically defensible in that it equally avoids the Scylla of material-
ism/physicalism, universal natural determinism/indeterminism and universal
natural mechanism and the Charybdis of classical theism but also scientifi-
cally respectable in the strong sense that it extends the Standard Models in a
way that enables physics to explain physics as a natural science that includes
human a priori knowledge of necessary truths of logic and mathematics,
human knowledge of natural laws6 and also human a posteriori knowledge of
contingent empirical/observational facts. Obviously, if physics cannot explain
physics, then it is explanatorily incomplete. So the Moderate Anthropic Prin-
ciple completes physics to that extent.7
Notes
1 Ade PAR et al. (Planck Collaboration). Planck 2013 Results. I Overview of Prod-
ucts and Scientific Results. Astronomy and Astrophysics. 2014;571:A1, 48pp.
(Footnote 1 in Robson 2021).
2 Gottfried K, Weisskopf VF. Concepts of Particle Physics. Vol. 1. New York: Oxford
University Press; 1984. 189pp. (Footnote 2 in Robson 2021).
3 Einstein A. The Basics of General Relativity Theory. Annals of Physics.
1916;49:769–822. (Footnote 3 in Robson 2021).
4 Robson BA. The Generation Model of Particle Physics. In: Kennedy E, ed. Particle
Physics. Rijeka: InTech; 2012. pp. 1–28. (Footnote 4 in Robson 2021).
5 It is generally believed that all Kantians must accept STI. For example, Jerrold
Katz claims that “however Kant’s transcendental idealism is understood, it locates
the ground of [real] facts within ourselves in at least the minimal sense that it
entails that such facts could not have existed if we (or other intelligent beings) had
not existed” (1998: 9). But, in fact, that is not correct: although this claim is true
of STI, it’s false of WTI.
6 Kant holds that our knowledge of natural laws is a priori knowledge of a certain
specially restricted kind, and I not only agree but also have argued for that in detail
144
elsewhere. For the purposes of this chapter, however, I will leave it open whether
our nomological knowledge is a priori or a posteriori. See (Hanna 2009, 2018b).
7 In my opinion, there are also some other ways in which contemporary physics is
explanatorily incomplete. See, for example, Hanna (2022: ch. 4).
References
Barrow, J. and Tipler, F. 1986. The Anthropic Cosmological Principle. Oxford:
Oxford University Press.
Carter, B. 1974. “Large Number Coincidences and the Anthropic Principle in Cos-
mology”, in M. S. Longair (ed.) Confrontation of Cosmological Theory with
Observational Data. Dordrecht: Reidel: 291–8.
Hanna, R. 2015. Cognition, Content, and the A Priori: A Study in the Philoso-
phy of Mind and Knowledge. The Rational Human Condition. Vol. 5. Oxford:
Oxford University Press. Also available online in preview at URL = <www.
academia.edu/35801833/The_Rational_Human_Condition_5_Cognition_Content_
and_the_A_Priori_A_Study_in_the_Philosophy_of_Mind_and_Knowledge_
OUP_2015_>.
Hanna, R. 2018a. Deep Freedom and Real Persons: A Study in Metaphysics. The
Rational Human Condition. Vol. 2. New York: Nova Science. Also available online
in preview at URL = <www.academia.edu/35801857/The_Rational_Human_Con
dition_2_Deep_Freedom_and_Real_Persons_A_Study_in_Metaphysics_Nova_
Science_2018_>.
Hanna, R. 2018b. “Kant’s Neo-Aristotelian Natural Power Grid: On Kant and the
Laws of Nature”, Critique 2018. Also available online in preview at URL = <www.
academia.edu/37619661/Kants_Neo_Aristotelian_Natural_Power_Grid_On_
Kant_and_the_Laws_of_Nature_Published_version_in_Critique_2018_>.
Hanna, R. 2022. The Philosophy of the Future: Uniscience and the Modern
World. Unpublished MS. Available online at URL = <https://www.academia.
edu/62653411/THE_PHILOSOPHY_OF_THE_FUTURE_Uniscience_and_the_
Modern_World_2022_version_>.
Hanna, R. and Maiese, M. 2009. Embodied Minds in Action. Oxford: Oxford Uni-
versity. Also available online in preview at URL = <www.academia.edu/21620839/
Embodied_Minds_in_Action>.
Katz, J. 1998. Realistic Rationalism. Cambridge, MA: MIT Press.
Prigogine, I. 1997. The End of Certainty: Time, Chaos, and the New Laws of Nature.
New York: Free Press.
Robson, B. A. 2021. Introductory Chapter: Standard Model of Cosmology. Avail-
able online at URL = <www.intechopen.com/books/redefining-standard-model-
cosmology/introductory-chapter-standard-model-of-cosmology>.
Simon, F. 2018. “Fine-Tuning”, in E. N. Zalta (ed.) The Stanford Encyclopedia of
Philosophy (Winter 2018 edition). Available online at URL = <https://plato.stan
ford.edu/archives/win2018/entries/fine-tuning/>.
Smolin, L. 2014. Time Reborn: From the Crisis in Physics to the Future of the Uni-
verse. New York: Mariner/Houghton Mifflin Harcourt.
Wikipedia. 2021a. “Anthropic Principle”. Available at: https://en.wikipedia.org/wiki/
Anthropic_principle, accessed on March 3, 2022.
Wikipedia. 2021b. “Pikes Peak”. Available at: https://en.wikipedia.org/wiki/Pikes_
Peak, accessed on March 2022).
145
INDEX
Alberti, L. B. 122 Chignell, A. 99

Allais, L. 10, 18–22, 91–2 Choi, Y. 39, 113, 131, 133
Allison, H. E. 4–5, 21, 23 cognition (scientific) 4, 39, 111, 118,
Altmann, S. 6 131
analytic 9, 12, 16, 22, 46–7, 51, 96 communication 94–8, 101, 103–6
animal(s) 30, 35, 51, 55–9, 62, 67–72, concept (of) 14–16, 18–19, 22, 30–2,
74–5, 96, 139–40; human 6, 60, 35–6, 41, 49, 90; Aristotelian 30;
142–3; non-human 6, 55–7, 68–75 causality 51, 53; communicability
anthropic principle 7, 136–40, 144–5 102; continuity 82–4, 88, 92;
a posteriori 85–6, 140, 142–5 empirical 68–9, 118; Praussian 83;
appearance(s) 10–11, 13–19, 22, 47–8, purposiveness 40, 50; species 120;
63–4, 68, 71, 77, 86–7, 91, 119–20, universal 119, 121, 132
140–1 consciousness 13–14, 21, 55–62, 64–78,
apperception 14, 17, 56, 60–2, 64–8, 86–7, 97, 107
72–5, 77–8 Copernican revolution 9, 45, 141
a priori 11, 15–16, 19–22, 25–7, 30, Copernicus 141
45, 47, 48, 51, 59, 64, 80, 85–6, 91, Critique of Practical Reason 25,
139–44 27, 31–3, 36–8, 41–2, 56, 96,
autonomy 52 100, 103, 108
Critique of Pure Reason 4, 6, 25, 28,
Barrow, J. D. 138–9 31, 33, 51, 58, 69, 70, 88, 97,
belief(s) 1–2, 99–101, 144 131–2
Big Bang 137, 139–40 Critique of the Faculty of Judgment
Bird, G. 3, 8 30–2, 42, 47–51, 58, 69, 71,
Bonnet, C. 35 101–3, 106
Bostrom, N. 138
Brandom, R. 44–6 de Boer, K. 2
Breitenbach, A. 25, 39 Descartes, R. 1, 5, 12, 19, 33, 35, 132
Brunelleschi, F. 122 Die Metaphysik der Sitten see
Buchdahl, G. 26, 53, 54 Metaphysics of Morals
Buffon, G. L. L. 35 Die Metaphysik Volckmann 56, 57, 74
Die Metaphysische Anfangsgründe der
Caranti, L. 5, 7 Naturwissenschaft see Metaphysical
Carter, B. 138–9 Foundations of Natural Science
Cartwright, N. 109 Doctrine of Right 107, 142
Cassam, Q. 5, 10, 12–14, 17, 21, 23 Doctrine of Virtue 96, 105
category 16 Dörflinger, B. 6, 76, 101
Chang, H. 109, 131 Dupré, J. 109, 131
146
INDEX
Einstein, A. 1, 2, 91, 112, 126–8, 137 knowledge 9, 11, 16, 18–19, 25–30,
electrodynamics 127 32–4, 37–40, 45–6, 50, 55, 78, 98,
Euclid 2–4, 91 105, 110, 118, 122–4, 128, 130,
132–3, 136, 139–40, 144; scientific
Fonnesu, L. 6, 7 30, 34, 118, 123–4, 127, 130, 139
forms of sensibility 20–2, 47, 55, 72, Kreines, J. 26
81, 87, 90, 100, 105, 107, 125, 128, Kritik der praktischen Vernunft see
130–1, 134 Critique of Practical Reason
Friedman, M. 3–4, 26, 91–2, 127 Kritik der reinen Vernunft see Critique
of Pure Reason
Gauß, C. F. 91 Kritik der Urteilskraft see Critique of
Gava, G. 97 the Faculty of Judgment
geometry 2–4, 80, 89, 91
Giere, R. 110, 112–13, 126 Land, T. 75
Ginsborg, H. 27, 91 La Rocca, C. 102
God 35, 96–7, 100, 101, 114 law (scientific) 2–7, 25, 31
Grice, P. 96 Leibniz, G. W. 22, 30, 35, 41, 77, 79, 133
Groundwork of the Metaphysics of Leland, P. R. 60, 69, 76–7
Morals 96 logic 5, 46, 58, 70, 76, 78, 97, 100, 105,
Grundlegung der Metaphysik der Sitten 107, 132, 140, 143–4
see Groundwork of the Metaphysics
of Morals Maimon, S. 73
Guyer, P. 23, 102, 132 Maimonian 73
Malebranche 35
Habermas, J. 94, 96 Massimi, M. 25, 110, 115, 131
Haller, A. V. 35 mathematics 2, 4, 33, 88, 91–2, 140,
Hanna, R. 1, 7, 53, 91, 136, 139, 144 143–4
happiness 100 Maxwell, J. C. 110, 113, 126, 127
Harrison, R. 5, 10–12, 14–15, 17, mechanics 3, 51, 113, 127–8, 137
21, 23 Meier, G. F. 97–8, 105, 108
Herz, M. 72 Mendelian rules 127–8
human(s) 1, 9, 15, 17, 26–7, 34–41, 44, Metaphysical Foundations of Natural
47, 50–2, 55, 60, 62, 68–70, 73–6, Science 3
78, 95–6, 99, 103–4, 106–7, 133, Metaphysics of Morals 96
139–44 Milky Way 62–3
Hume, D. 17–18 Mitchell, S. 120
Huneman, P. 35 Morrison, M. 126
idea 9, 29–34, 36–42, 44, 53, 62, 82, naturalism 1–3, 5, 7

92, 94, 96–7, 102, 106, 107, 109, Newton, I. 2–4, 35, 51, 53, 110, 113,
114–17, 121, 125, 127–8, 133 115–16, 126–8, 132, 134
idealism 9–14, 18–23, 78, 136, 139–42, normativity 6, 45, 46–7, 49, 51
144 noumenon(a) 47, 49, 53, 140–3
illusion 4, 40, 99, 116–17, 126, 132,
134 O’Neill, O. 6, 94–5, 98, 102–3
imagination 14, 16, 17, 48, 66, 74–5, Opticks 113–32
78, 88 organism (metaphor) 28, 32, 34–6
Kauark-Leite, P. 6, 51 perception(s) 16–17, 23, 48, 56, 64–5,

Kellert, S. H. 111, 131 68–71, 75, 77–8, 85–7, 141
Kierkegaard, S. 23, 100 perspectivism 7, 109–10, 112, 114–15,
Kitcher, P. 26, 127, 131, 133 120, 122, 125, 126, 130–2
147
INDEX
persuasion 97–9, 104–5, 107, 108 self-representation 65–6, 75

phenomenon(a) 18, 26, 36–7, 39, 46, sensation(s) 2, 20, 59, 61–2, 71, 85, 87,
51, 77, 109–13, 117, 120–1, 123, 102, 104
125–8, 130, 140–1 sensibility 20–2, 38, 52, 55, 57, 59, 61,
physics 2, 3, 6, 7, 33, 51–3, 81, 91, 112, 67, 70, 72, 74–5, 77, 85, 91
116, 136–7, 144–5 Shakespeare, W. 23
Pickering, M. 130 skeptic see sceptic
Pike’s Peak 142 skepticism see Descartes, R.
pleasure 15, 76, 102 Smolin, L. 139
pluralism (scientific) 7, 110–11 space 4, 6, 19, 77, 80–92, 105, 109,
Pollok, K. 47, 54, 79 113–14, 117, 122–4, 129–32
Prauss, G. 6, 80–4, 86–92 Spagnesi, L. 7
Prigogine, I. 136, 139 Spinoza, B. 33
principle(s) 1–4, 6, 7, 25–37, 39–41, standard model(s) 7, 136–7, 139, 144
46–7, 50–3, 56–7, 63, 80, 107, Stern, R. 10, 25
109–15, 117–31, 133, 136–40; Strawson, P. F. 2, 5, 9–13, 19, 21, 23
scientific 110–14, 116–18, 120, Stroud, B. 2, 12
123–4, 126–7, 129–30, 132, 134 synthetic 1, 2, 4, 11, 17, 46–7, 48–9,
progress (scientific) 5 52, 72–3, 118; a priori 1, 2, 4, 11,
Prolegomena to Any Future 47, 52
Metaphysics 17, 19, 29–30, 140
Prolegomena zu einer jeden künftigen Teller, P. 110
Metaphysik see Prolegomena to Any theories (scientific) 26–9, 31, 33, 37,
Future Metaphysics 51–3, 75, 109–10, 112–14, 126–7,
131–2, 139
quantum 84, 110, 133, 137 theory of relativity see Einstein, A.
Quine 2, 3 things in themselves 11, 15, 19, 140–2
time 4, 18–19, 29, 34, 37–8, 42, 48–9,
realism 5, 99, 110, 131–3, 144 58, 61, 81, 83, 88–91, 96–7, 100,
reason 25, 27–34, 36–42, 49–53, 57, 102, 110–11, 115, 121, 132, 139,
59, 70, 73–4, 80, 94, 98–100, 102, 142
104–5, 109, 113–25, 128–32, 140 Tipler, F. 138–9
Refutation of Idealism see idealism transcendental 1–5, 7, 9–14, 18–23, 27,
Reichenbach, H. 2 31, 41, 46, 47, 49–53, 66, 68, 73, 77,
representation(s) 1, 6, 15–16, 19, 21, 80–1, 87, 91, 98, 113–15, 131, 133,
34, 53, 55–78, 81–90, 92, 109, 120, 136, 139–42, 144
122–3, 126, 133, 140 transcendental deduction 5, 9, 10, 12,
research: enquiry 27–8, 29, 33, 40; 14, 18, 66, 68, 73, 77
investigation 27, 95, 101, 109–10, transcendentalism 140
113, 120–1, 129–30; scientific 7, truth 2, 11–13, 23, 53, 97–9, 111, 129,
109–13 133, 140, 142–4
Riemann, B. 91
Rueger, A. 110, 131–3 understanding 4, 6, 13, 14, 17–18,
21–2, 25–6, 31–2, 34, 36–7, 42,
sceptic 10, 12–13, 23 45–50, 51–2, 55–7, 59, 61–2, 66–8,
Schrödinger Equation 110, 140 70, 72–5, 77, 84, 88, 91, 97–100,
science 1–5, 7, 33–4, 38–41, 47, 50–1, 103, 105, 107, 110, 114–20, 123–4,
105, 109–13, 115, 117–18, 125–8, 127–8, 132, 137
131–3, 136, 140, 144
self-attribution 6, 55–6, 60–6, 73, 75, 77 Walker, R. 130, 134
self-consciousness 6, 56, 60–11, 139–40 Watkins, E. 133
self-knowledge 2, 61 Williams, B. 10, 43
148
INDEX
Wolff, M. 91, 105 real 45, 48, 141–3; spatio-temporal 2;

world: empirical 19; external 2, 19, system 122
23, 45, 80, 87–8; natural 46, 111, worldview (scientific) 3
126, 129, 139–40; objective 45;
phenomenal 40; physical 6, 45; Ypi, L. 5–6, 28, 36
149

Kant and The Problem of Knowledge - Sanet.st

Uploaded by

Copyright:

Available Formats

Kant and The Problem of Knowledge - Sanet.st

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

Kant and The Problem of Knowledge - Sanet.st

Uploaded by

Copyright:

Available Formats

What is the book about?

What is the book about?

What are some of the main ideas discussed in the book?

What are some of the main ideas discussed in the book?

KANT AND THE PROBLEM

This book examines Kant’s contributions to the theory of knowledge and

Luigi Caranti is Professor of Political Philosophy at the Università di

Alessandro Pinzani is Professor of Ethics and Political Philosophy at the

Edited by Luigi Caranti and

ISBN: 978-0-367-90316-9 (hbk)

List of contributors vii

1 The transcendental deduction and transcendental

2 Kant on scientific laws 25

3 Normativity and reflective casual inference 44

4 Consciousness as representation of representation: Kant on

5 Critique and development of the Kantian theory of space

7 Kant’s space of theoretical reason and science:

8 Can physics explain physics? Anthropic principles

Henry E. Allison is Professor Emeritus at the University of California, San

Contemporary Kantian Philosophy Project and also Editor of two jour-

References to Kant’s works follow the sigla reproduced in the following.

FRT Fragment einer späteren Rationaltheologie (AA 28)

NTH Allgemeine Naturgeschichte und Theorie des Him-

VAKpV Vorarbeit zur Kritik der praktischen Vernunft (AA 23)

V-Mo/Collins Moralphilosophie Collins (AA 27)

Luigi Caranti and Alessandro Pinzani

How does Kant contribute to today’s epistemology? Answers legitimately

the general epistemological orientation after relativity was largely inspired by

A more recent representative of soft naturalism is Michael Friedman. He

at some passages of the Critique of Judgment, Fonnesu uncovers the episte-

Does the acceptance of pluralism inevitably entail the exclusion

In the attempt to suggest a positive answer, Spagnesi looks at a perspec-

The publication of P. F. Strawson’s The Bounds of Sense in 1966 was a major

Strawson’s work has spurred ongoing debates regarding the nature of

1. Kant after Strawson

significant amount of attention on Kant, whom he considers the philosopher

Because appearances [for Kant] depend on us there is not a crucial

Harrison’s proposal for saving transcendental arguments from the fatal

nature of the world. He suggests that without idealism, there appear to be

Cassam tests his proposal by applying it to the A-version of the Transcen-

The Satisfaction Component then makes a claim about how the

Correlatively, the Satisfaction Component is that “appearances must display

It is quite sufficient to say that what a transcendental argument

Having dismissed any need for transcendental idealism, Cassam attempts

of idealism, because it is only on the assumption that the requisite order-

It is enough . . . that appearances do conform to the conditions of the

Appearances could . . . be so constituted that the understanding

and effect, so that the concept would therefore be entirely empty,

It is crucial to recognize that Kant regarded this as a genuine worry that

We find . . . that our thought of the relation of all cognition to

despite his clear insistence on the contingent nature of empirical propositions,

Although Kant is referring to the conditions of the possibility of reproduc-

because reproducibility is a necessary (though not a sufficient) condition of

“extreme” metaphysical interpretations, which classify Kant as either a nou-

of Kant’s arguments as not dependent on, leading to, otherwise intimately

I think that we should take transcendental idealism seriously, and it

Accordingly, the question becomes exactly what arguments and epistemo-

suspect, it is to be understood as a problem in normative epistemology, then

A serious objection to making the argument of the Deduction