Kant's Aesthetic Epistemology - Form and World (2007)

Fiona Hughes is Lecturer in
Philosophy at the University KANT’S KANT’S AESTHETIC
KANT’S AESTHETIC EPISTEMOLOGY

of Essex. She is the author of
EPISTEMOLOGY
published articles on Kant and
aesthetics and on Nietzsche, AESTHETIC FORM AND WORLD
aesthetics and hermeneutics.
She is currently writing a book
on aesthetics and the arts.
EPISTEMOLOGY Fiona Hughes
‘A brilliantly executed presentation of a
case for a new and fruitful way of
perceiving Kant’s Critical project.’
FORM AND WORLD Drawing on resources from both
the Analytical and Continental
John Llewelyn, Reader in Philosophy, traditions, Fiona Hughes argues that
University of Edinburgh a comprehension of Kant’s aesthetics
is necessary for grasping the scope
and force of his epistemology.
She draws on phenomenological

and aesthetic resources to bring out
the continuing relevance of Kant’s
project. One of the difficulties
faced in reading The Critique of
Pure Reason is finding a way of
reading the text as one continuous
discussion. This book offers a
FIONA HUGHES
reading at each stage of Kant’s
epistemological argument, showing
how various elements of Kant’s
argument, often thought of
as extraneous or indefensible,
Cover image: Ian Hamilton Finlay, Little Sparta
can be integrated.
Photograph by Andrew Lawson
Cover design: Cathy Sprent
This incisive study, arguing for the
Edinburgh University Press
22 George Square
centrality of aesthetics in philosophy,
Edinburgh EH8 9LF and within experience in general,
www.eup.ed.ac.uk challenges a blind spot in the Anglo-
ISBN 978 0 7486 2122 4
American tradition of philosophy
Edinburgh
and will contribute to a growing
interest in the general signiﬁcance
of aesthetic culture.
Fiona Hughes
M710 HUGHES PRE M/UP.qxd 29/3/07 8:11 AM Page i Gary Gary's G4:Users:Gary:Public:Gary's Job
Kant’s Aesthetic Epistemology

M710 HUGHES PRE M/UP.qxd 29/3/07 8:11 AM Page ii Gary Gary's G4:Users:Gary:Public:Gary's Jo
‘Music . . . is not an expression of what life is, but an expression

of what life could be, or what it could become.’
Daniel Barenboim. Reith Lectures 2006
In memory of George Elder Davie (1912–2007), student of Kemp Smith

and scholar of the Scottish (and Irish) Enlightenment.
M710 HUGHES PRE M/UP.qxd 29/3/07 8:11 AM Page iii Gary Gary's G4:Users:Gary:Public:Gary's J
KANT’S AESTHETIC
EPISTEMOLOGY
FORM AND WORLD

2
Fiona Hughes
Edinburgh University Press

M710 HUGHES PRE M/UP.qxd 29/3/07 8:11 AM Page iv Gary Gary's G4:Users:Gary:Public:Gary's Jo
© Fiona Hughes, 2007
Edinburgh University Press Ltd

22 George Square, Edinburgh
Typeset in Sabon
by Servis Filmsetting Ltd, Manchester, and
printed and bound in Great Britain by
Biddles Ltd, King’s Lynn, Norfolk
A CIP record for this book is available from the British Library
ISBN 978 0 7486 2122 4 (hardback)
The right of Fiona Hughes

to be identified as author of this work
has been asserted in accordance with
the Copyright, Designs and Patents Act 1988.
M710 HUGHES PRE M/UP.qxd 29/3/07 8:11 AM Page v Gary Gary's G4:Users:Gary:Public:Gary's Job
Contents
Acknowledgements vi
References to Kant’s Works viii
Introduction 1
1 The Centrality of the Problem of Formalism 8
2 Formalism and the Circle of Representation 49
3 Formal Idealism and the Aesthetic Condition

of Experience 86
4 The Deep Structure of Synthesis 112
5 The Completion of the Subjective Deduction in

the Deductions of the Critique of Judgement 169
6 A Priori Knowledge as the Anticipation of a Material

Given and the Need for a Spatial Schematism 207
7 Empirical Systematicity and its Relation to

Aesthetic Judgement 248
8 Aesthetic Judgement’s Exemplary Exhibition

of Cognition 277
Afterword 311
Bibliography 316
Author/subject index 321

M710 HUGHES PRE M/UP.qxd 29/3/07 8:11 AM Page vi Gary Gary's G4:Users:Gary:Public:Gary's Jo
Acknowledgements
This project has been a long time in gestation and I owe many thanks
to particular people and the creative milieu in a number of different
institutions. My acknowledgements have to reach back at least as far
as my days as an undergraduate at the University of Edinburgh and
to a circle of friends. Later, at the University of Oxford I gained
another informal research group, some philosophers like Sally (Sarah)
Gibbons, some not. My supervisor, Michael Rosen, was unfailingly
helpful. At Oxford I found further resources to help me situate my
philosophical interests within a more general framework. The moti-
vation to do so – traces of which I hope are still visible at the margins
of this book – arose in great measure from the intellectual culture in
Edinburgh and the influence of George Davie’s The Democratic
Intellect. From my appointment at Essex as a new lecturer, I had the
great fortune to expand my horizons through exposure to the style of
thinking of the philosophical art historian, Michael Podro, who
became a true colleague and friend. For many years, we ran joint sem-
inars together and if the students enjoyed them and learnt as much as
I did, I am satisfied. Over the years, my students at Essex – mainly in
the philosophy department, but also those working in other disci-
plines – have been an ongoing source of inspiration.
I am particularly grateful to a number of people who gave their
valuable time to read and comment on parts or the whole of my man-
uscript. Needless to say, any remaining problems are my own, not
theirs. These include Neil Cox, Douglas Burnham, Sebastian Gardner,
Dana MacFarlane, Wayne Martin and Michael Podro. I want to espe-
cially thank John Llewelyn. He has been an inspiration and a great
support since my Edinburgh days. I cannot adequately capture his
scrupulous attention to detail, combined with unstinting generosity –
not only in reading my manuscript, but also in philosophy and in life
in general. Other friends helped on specific points and kept my spirits
up when the going was tough.
In 2003 the Arts and Humanities Research Board (now AHRC)
awarded me a Research Leave Scheme Grant allowing one term’s
vi
M710 HUGHES PRE M/UP.qxd 29/3/07 8:11 AM Page vii Gary Gary's G4:Users:Gary:Public:Gary's J
Acknowledgements
leave from teaching. In conjunction with sabbatical leave from the

University of Essex, this allowed me to complete a first draft of
my manuscript. Many individual members of staff at Edinburgh
University Press have been very helpful. I must mention, in particular,
Jackie Jones and Carol Macdonald, my editors, who, without fail,
have been both constructive and insightful.
I am grateful to Nicola Gray, whose perceptive interventions were
extremely helpful in the preparation of my manuscript for publica-
tion, and to James Corby for preparing the index. I would also like to
especially acknowledge here the great support given me by my mother
and also by my father, when he was alive. From early on, they both
were prepared to make many sacrifices in support of my education.
Finally, thanks to John Walshe, who gives so much encouragement
and, refreshingly, not necessarily in ways I might expect!
vii
M710 HUGHES PRE M/UP.qxd 29/3/07 8:11 AM Page viii Gary Gary's G4:Users:Gary:Public:Gary's
References to Kant’s Works
All references to Kant’s works are to Kant’s Gesammelte Schriften

herausgeben von der Königlich Preußischen Akademie der
Wissenschaften [Berlin 1902– ], commonly referred to as the
‘Akademieausgabe’ [ AA from now on].
When necessary I offer my own translations. Otherwise, I make use
of the following standard versions in English:
[1781 and 1787] (1933, 2nd edn) Critique of Pure Reason, trans.
N. Kemp Smith, London and Basingstoke: Macmillan [ CPR].
Page references to the first edition of 1781 are preceded by A, while
those to the second edition of 1787 are preceded by B.
Material is reprinted by permission of Macmillan Publishers Ltd.
The A edition appears in Volume IV of the Akademie edition, while
Volume III comprises the B edition.
[1790] (1987) Critique of Judgment, trans., W. Pluhar, Indianapolis,
IN: Hackett. All references to this text are preceded by AA.
Material is reprinted by permission of Hackett Publishing
Company, Inc. All rights reserved.
The main text appears in Volume V of the Akademie edition. An
alternative, unpublished version of the Introduction, commonly
referred to as the ‘First Introduction’ appears in Volume XX.
References to the earlier and longer version will be distinguished by
a prime mark, for example, AA 191′.
Other translations referred to:
(2004) Metaphysical Foundations of Natural Science, ed. M.
Friedman, Cambridge: Cambridge University Press.
(1963) On History, ed. L. W. Beck, Indianapolis, IN: Bobbs-
Merrill for ‘Perpetual Peace’ and ‘The End of All Things’.
(1991) Political Writings, ed. H. S. Reiss, Cambridge: Cambridge
University Press.
viii
M710 HUGHES TEXT M/UP.qxd 29/3/07 8:12 AM Page 1 Gary Gary's G4:Users:Gary:Public:Gary's Jo
Introduction
In ‘Little Sparta’, a garden in the Pentland hills outside of Edinburgh,

two planks stretch across a stream. On each is inscribed: ‘That which
joins and that which divides is one and the same.’ The inscription on
one faces that on the other. At first sight we see a bridge, but then
we see two planks divided from one another; looking again we see
the two planks as one structure, realising that each bears the same
inscription. The inscription stretching the length of each plank takes
time to read and there is no one perspective from which both can be
apprehended. Our realisation of the similarity of the two planks is
only achieved through a process of recognition, for after an initial
impression of unity, they appear as disrupting one another. Our aes-
thetic response is of unity achieved only through a process linking two
distinct components.
Much more could be said about this contemporary artwork by
the Scottish artist Ian Hamilton Finlay, his use of a fragment from
Heraclitus and the relation in which the work stands to the philo-
sophical ideas that follow. In general my point is that through an aes-
thetic presentation – in this case, visual – a train of thought is
instigated. What we see makes us think, and what we think makes us
look further, in turn making us think more. While a whole philosoph-
ical treatise may, and indeed will in what follows, attempt to express
a thought, this artwork quickly and effortlessly brings us into the
centre of a complex set of associations. We are led in by what we see,
just as we might be led across the stream by the bridge. We find that
we are already in the midst of a way of seeing the world and of ori-
enting ourselves within its perspective. Both artworks and aesthetic
experiences of nature have the capacity to bring about such an insight,
which counts as a form of reflection on the world and our place in it.
This, I believe, is what Kant discovered in his Critique of Aesthetic
Judgement. More specifically, I will argue in what follows that he came
to the conclusion that a harmonious combination of mental capacities,
characteristic of aesthetic judgement, allows for a reflection on a more
general cooperation that is necessary for any experience whatsoever.
1
Finlay’s work is particularly pertinent to establishing this role for

aesthetics, for the cooperation of orientations in question is not one
of bland uniformity. We are complex beings bound to the world by a
sensory or aesthetic capacity to take things in, namely, receptivity. But
we are also able to stand back from our environment in a reflective
stance. That which joins aesthetic receptivity with reflection is a third
capacity, Kant tells us, imagination. Imagination makes possible a
unification of what is ineliminably different – that is, our capacity to
be affected and our capacity to reflect on the world. The mediating
faculty does not achieve this by eliminating the distinctiveness of the
two faculties, but by finding a process that allows for a relation
between one side of our disposition and the other, just as the bridge
at ‘Little Sparta’ makes possible a crossing from one side of the stream
to the other. In the philosophical case under consideration in this
book, imagination effects a crossing by making possible synthesis, a
process of connecting something given to our senses with a concept
that aims to make sense of it. I will argue that aesthetic phenomena
allow us to reflect on the process of synthesis that is the condition for
subjects such as ourselves having access to an external world, which
is ours, as our environment, but is not in our possession. This
complex relation is also expressed in Finlay’s heraclitean reflection.
The relation in which aesthetic receptivity stands to reflection is
thus one that occurs across a gap. Synthesis in general and aesthetic
harmony, in particular, are not in plainsong; their harmony is not
uniform. The unity that is a result of synthesis and the harmony char-
acteristic of aesthetic judgement both have difference as a constitutive
moment, just like the bridge at ‘Little Sparta’. Unity and harmony are
constructed out of distinctive ways in which we relate to the world
and yet these orientations could not prevail in isolation from one
another. We can only be receptive and we can only be reflective
insofar as we are both at one and the same time.
The background problem, out of which the particular arguments
in this book emerge, is one of the place of aesthetics within philo-
sophical practice and its relevance for experience in general. The way
in which this problem is viewed can be captured as giving rise to two
responses. In – for want of a better title – the Continental idiom of
practising philosophy, it is widely accepted that aesthetics shares phi-
losophy’s project and has no small part to play within it. Meanwhile
– and, again, for want of a better expression, for there are many dif-
ferent types of philosophical analysis – in Analytical philosophy we
find more of a hesitation in granting a central role to aesthetics.
2
Introduction
Aesthetics or the philosophy of the arts is practised seriously and to

great local effect, but there is little confidence that aesthetics is any-
thing more than a regional activity, secondary to the more essential
sub-disciplines of epistemology, philosophy of mind and moral phi-
losophy. The disciplinary constitution of the aesthetic is sometimes
reinforced by a more general suspicion that aesthetic appreciation is,
after all, an optional activity and may even be a distraction from the
central questions of philosophy and of life.
While I do not pretend that I will be able to answer this general
problem of the status of aesthetics in one book on Kant’s epistemol-
ogy and aesthetics, I do hope to make some progress by establishing
the importance he gave to the relation between aesthetic judgement
and epistemology. Many other writers on Kant – a number of whom
I will discuss – have made significant progress in uncovering the role
played by aesthetic judgement within his philosophical system as a
whole; however, I believe that, as yet, there is no convincing account
of how the systematic link between cognitive and aesthetic judgement
might be defended in a thoroughgoing fashion. My hope is that this
work will help secure the commitment to aesthetics held by many
Continental philosophers. Meanwhile, the way in which I pursue my
account of the link between epistemology and aesthetics may persuade
some of an Analytical persuasion to venture onto the bridge. Ian
Hamilton Finlay’s visual metaphor is useful here, for there is no middle
ground and no well-established territories on either side. ‘Analytical
philosophy’ is a plurality of practices and so too is ‘Continental phi-
losophy’. We fool ourselves to think that one well-defined practice
opposes another. But what we can hope for is that there may be cross-
ings or mediations of the stream. In a heraclitean frame of mind, I am
inclined to think that the different ways of engaging in philosophy
occur on the bridge, while philosophy is the stream.
Why is it important to establish a role for aesthetics? I believe it is
because aesthetics allows for an insight into the plural constitution of
our experience in general and that once we lose sight of this, dogma-
tism threatens. If aesthetics does this – something I will argue through
my interpretation of Kant in this book – it does so indirectly in the
course of a preoccupation with an aesthetic phenomenon or event. I
will argue that the indirectness of the reflection is a necessary one.
There are two ways in which the term ‘aesthetic’ will be used in
this book. The first sense is the common one of a pleasurable experi-
ence arising from the senses – be the event artistic or natural – and
giving rise to a judgement of taste. The second is a more primary one
3
referring to any sensory experience whatsoever. In this book I will

seek to show how, for Kant, the two senses of aesthetics are neces-
sarily intertwined.
The second sense of ‘aesthetic’, which figures directly in the title of
this book, marks the status of knowledge from Kant’s point of view.1
Knowledge emerges in space and time, these being the context within
which any sensory affect arises. Knowledge requires not only con-
cepts, but also sensory or aesthetic givens, which he calls intuitions. I
will argue that the wider ramifications of this stance, which is well
known in the philosophical world at least, is that knowledge arises as
a project undertaken over time and across space. Knowledge is not
simply a fait accompli, but a process that marks any achievement that
is reached.2 I will further argue that Kant comes to the view that aes-
thetic judgement – that is, ‘aesthetic’ in the first or common sense of
the word – has the power to reveal this deeper truth about knowledge.
The picture that emerges is one of a subject bearing certain broadly
mental capacities with which he or she is able to take up and make
sense of an environment or world of objects. I thus argue that the
attention Kant pays to what he calls the ‘faculties’ or subjective capac-
ities is not a hindrance for his intent of establishing the possibility of
objective knowledge, but rather the means by which he achieves it.
The engagement with Kant that emerges in the pages that follow
is at times closely textual, but it is not a work in pure descriptive inter-
pretation. I read Kant because he can help me make sense of prob-
lems that exist outside of the technical content that he himself
developed. I believe that Kant also viewed philosophy in this way and
that is why he has a continuing fascination for us. Although it is
impossible to read Kant without entering into a series of technical
debates about competing interpretations, the problem is institution-
ally elaborated, not institutionally constituted.
In what follows, I will give an overview of what may be expected.
I will not enter into the details of my arguments; there will be time
enough for that later. Although my goal is to establish the importance
of Kant’s aesthetics for his epistemology, my initial investigations are
focused strictly on his theory of knowledge. More specifically, I begin
by situating the problems I will address in later chapters within the
context of the very rich interpretative reception of Kant’s epistemol-
ogy that has developed over the last forty or so years. In later chap-
ters I will argue that Kant’s aesthetic formalism – despite it often being
the subject of criticism – provides resources for establishing the char-
acter of his less familiar formalist account of knowledge.
4
Introduction
Chapters 1 and 2 establish the centrality of cognitive formalism for

both critics and defenders of Kant’s epistemology alike. A critical
assessment of some of the major contributions to current debates pro-
vides a backdrop against which I will situate my own account of
Kantian formalism. In Chapter 1 I am concerned with the way in
which Kant has been accused of ‘impositionalism’, that is, the posi-
tion that the mind imposes form or order on the world. The problem
that seems to ensue is that, according to this view, knowledge would
be not so much of things in the world, but rather a merely subjective
projection of mind’s own structures. I examine how this criticism
emerges from the readings of Robert Pippin, Paul Guyer, Dieter
Henrich and Peter Strawson. In Chapter 2 I turn to some of Kant’s
major defenders – namely, Gerd Buchdahl, Henry Allison and
Béatrice Longuenesse – and show how they have argued that formal-
ism does not lead to a subjective status for knowledge. While sug-
gesting the beginnings of a way of defending Kant against the charge
of impositionalism, I argue, nonetheless, that his defenders have not
supplied an account of formalism that convincingly rebuts this
charge.
In Chapter 3 I begin my interpretation of formalism, starting
from Kant’s own distinction between formal and material idealism. I
suggest that the commitment to a material correlate of formal ideal-
ism is to be found in the initial paragraphs of the ‘Transcendental
Aesthetic’. This is the material dimension of the aesthetic or sensory
side of the dualist project. In Chapter 4 I turn to the reflective or con-
ceptual side of dualism, as the condition of the possibility of synthe-
sis. I argue that synthesis arises as a cooperation of a plurality of
faculties and offer the ‘Transcendental Deduction’, in both its edi-
tions, as case studies. I conclude that only aesthetic judgement reveals
the deep structure of the cooperative structure of synthesis within
experience. In Chapter 5 I reassess the importance of the subjective
deduction, first mentioned in the Preface to the A edition of the
Critique of Pure Reason. I argue that the subjective deduction is the
necessary corollary of the objective deduction of the categories and
that the full account of the subjective side of cognition can only be
found in certain passages in the Critique of Judgement.
Establishing the relation between knowledge and aesthetic judge-
ment cannot be achieved in a linear fashion. Just because the relation
is reciprocal, there is no singular point of origin out of which the other
term can be deduced. In Chapter 6 I turn back to the cognitive side of
the relation, looking in more detail at the objective deduction, which
5
was addressed in nuce in Chapter 4. Having established the role of

synthetic process in the preceding chapters, I argue that knowledge
arises out of a process of subjective anticipation of a material given in
experience. In Chapter 7 I take up the question of the purposiveness
or systematicity of nature, introduced by Kant in the Critique of
Judgement. The latter addition is important because, as I will show, it
is necessary for empirical cognition. I argue that we can make sense of
Kant’s claims that there is a link between systematicity and aesthetic
judgements, once we distinguish between two levels at which purpo-
siveness operates, while, at the same time, recognising that it combines
a subjective and an objective orientation. I conclude that aesthetic
judgements are exemplary for a general purposiveness of nature. In
Chapter 8 I fine-tune this account, characterising the exemplarity of
aesthetic judgements as contrastive or, as I call it, ‘contrapuntal’.
Aesthetic judgement is not in any sense cognitive; nevertheless, in our
aesthetic appreciation of artworks and of nature we are able to engage
in an implicit reflection on the relation in which we stand to the world.
This relation of subject and object is the fundamental engagement out
of which knowledge emerges.
In my Afterword I make some suggestions as to how I would
further explore the relation in which the harmony characteristic of
aesthetic judgements stands to the Kantian sublime. I return to Ian
Hamilton Finlay’s heraclitean work at ‘Little Sparta’ to explain why,
although the sublime has not been directly addressed in the body of
this book, it still has a necessary role to play. The harmony of the
faculties arises across difference; thus, the possibility of disruption
is always already implicit. The beautiful marks a moment when the
senses make sense of something in the world, while the sublime
arises when the senses fail us. The beautiful is singular and, thus,
indirectly reveals the fragility of our sensory grasp of the world or
the possibility that sense might break down. The beautiful and the
sublime cannot be articulated or experienced without the trace of
the other.
Notes
1. Throughout this book I will refer to ‘knowledge’ rather than to ‘cogni-
tion’, which is the more accurate translation of Erkenntnis. Kant’s
project in the Critique of Pure Reason is to establish the possibility of
objectively valid cognition, which would qualify as knowledge. He is
not concerned with cognitions such as opinions or beliefs. So although
6
Introduction
Erkenntnis has a much broader range than ‘knowledge’, I think it is jus-

tifiable to use a term that is more intuitive and requires less qualifica-
tion than the more technical ‘cognition’.
2. My primary concern will be to establish the anticipatory status of
formal, a priori cognition.
7
The Centrality of the Problem of Formalism
What is the nature of Kant’s epistemological formalism, as his critics

see it? I will argue that there are several versions, but that in all cases
the criticism of formalism amounts to the charge that the outcomes
of Kant’s method are ultimately subjective. Critics of Kant claim that
his insistence that the form of experience arises from our minds,
finally makes the empirical world of objects dependent on a form of
subjectivity. If our minds introduce the form of experience, so the
argument goes, then the objects we experience are only objects ‘for
us’ and there is no access to a genuinely extra-mental world. In reply-
ing to this in later chapters I will argue that Kant’s formalism is an
attempt to systematically link the subjective and objective conditions
of experience. So while Kant’s critics are right to link formalism with
a turn to the subject, they are telling only one side of the story. The
project of the Copernican revolution is exactly that of showing how
the turn to the subject will secure the possibility of knowledge of an
objective world. But the full version of this project does not involve
reducing objectivity to the subjective conditions of experience, which
are necessary and not sufficient. Later I take up the problem of the
empirical applicability of the subjective forms of thought by empha-
sising the aesthetic dimension of Kant’s critical turn.
In this chapter I will show how formalism has been seen as a
central issue for Kant’s epistemology. All of the authors I examine in
this chapter conclude that Kant’s position is a formalist one. Paul
Guyer and Peter Strawson argue that formalism results in the posi-
tion that mind imposes order on objects.1 This is the thesis that is
referred to as ‘impositionalism’. Robert Pippin and Dieter Henrich
resist this conclusion, although I will show that ultimately both fail
to provide a sufficiently robust defence of Kant’s position.
Paul Guyer has been the most robust defender of the view that
Kant’s epistemology portrays mental formative activity as an imposi-
tion on matter. As I will discuss in detail later, Guyer distinguishes
between two models of the relation between mind and the material
given. If the conformity between mind and reality is always already
8
guaranteed – a thesis to which he believes Kant is committed – then

this can only be explained by:
the mind’s ability to impose its rules on a pliable or formless material, or
at the very least on a material any intrinsic form of which is inaccessible
to the mind and can be replaced by a form of the mind’s own making.2
Guyer contrasts this to the view that:

it is possible to think that we can know in advance of successful experi-
ence what objects must be like if we are to experience them but that we
have no special power to make objects be like that, thus to make experi-
ence possible.3
Thus the impositionalist interpretation of Kant attributes to him the

view that mind imposes form on a material given and that this results
in our being able to produce the objects of experience.4
Now why should attributing impositionalism to Kant lead to dif-
ficulties in his account of knowledge? Although none of the inter-
preters discussed in this chapter would deny that the subject plays a
necessary constitutive role in knowledge acquisition, the worry is
that, from the perspective of the formalist method, the subject is faced
only with our own mental structure, in the absence of any real
encounter with an object.
Robert Pippin’s Kant’s Formalism, which I discuss in the first
section of this chapter (pp. 10–17), still counts as seminal for this
debate. His critical, but sympathetic and subtle, reconstruction of
Kant’s epistemology sets the standard for subsequent discussions. His
eventual rejection of Kant’s project arises from the latter’s aim of
establishing an a priori framework for knowledge that relies on a
questionable account of pure intuition. An important aspect of his
account is the claim that reflective judgement amounts to a mental
projection that reinforces the subjectivist status of Kant’s epistemo-
logical project.
Paul Guyer, whom I discuss in the second section (pp. 17–25), is,
in general, critical of the formalist dimension of Kant’s method.
However in his assessment of the latter’s epistemology, his criticism
is, more precisely, of Kant’s ‘impositionalism’. Guyer’s aim is to reha-
bilitate a non-formalist core ultimately to be found only in certain
manuscripts not prepared for publication by Kant. His view is that
Kant often resorts to arguing that the mind imposes form on reality.
In contrast to this, Guyer aims to unearth a ‘realist’ Kant who eschews
the assumptions of transcendental idealism and is committed to the
9
M710 HUGHES TEXT M/UP.qxd 29/3/07 8:12 AM Page 10 Gary Gary's G4:Users:Gary:Public:Gary's J
independent existence of an extra-mental world accessed by a causal

perceptual process. Like Pippin’s, this criticism amounts to the charge
that Kant’s formalism leads him into an unwarranted subjectivism vis
à vis knowledge.
Dieter Henrich, whom I discuss in the third section (pp. 25–35),
has been hugely influential in the reception of Kant both in the
German and English speaking worlds. His reconstructive argument
has greatly encouraged Guyer and others for whom the absolute
Cartesian certainty of Henrich’s starting place entails impositional-
ism. It is for this reason that I include a discussion of Henrich at this
stage of my account, even though he defends Kant and denies that his
interpretation entails impositionalism.
An earlier critique of Kant’s formalism is to be found in the work
of Strawson, whom I discuss in the final section. He sees the domi-
nant problem with Kant’s approach as his dependence on ‘faculty
talk’. Instead of focusing exclusively on the objective empirical con-
tents of experience, Kant repeatedly expresses his position in terms
of the relations between mental capacities. The danger is that the
critique of knowledge falls into psychologism. Importantly, Strawson
also accuses Kant of impositionalism. The supposed problem, once
again, is that Kant’s formalism – now specifically the claim that
experience is dependent on certain formal mental capacities – results
in knowledge arising from the mind enforcing its patterns onto
reality.
In the course of this chapter we will see that formalism takes on a
specific shape for each of these authors. Interestingly, all of them,
except Strawson, entertain the possibility that there may be an alter-
native version of formalism that does not fall into the impositionalist
trap.
I Pippin’s Critique of Kant’s Formalism

Robert Pippin’s Kant’s Theory of Form offers a thoroughgoing and
systematic account of Kant’s epistemological project. While Pippin is
critical of Kant’s formalism for resulting in a merely subjective legiti-
mation of knowledge, he nevertheless resists interpretations that
accuse Kant of impositionalism, the thesis that order arises in experi-
ence due to the imposition of mental forms.
Kant, as interpreted by Pippin, fails to fulfil his deepest held under-
taking to achieve a formal epistemology applicable to empirical
experience, despite his commitment to ‘empirical guidedness’, that is,
10
the thesis that the forms of experience must be applied to empirical

matter.5 Following the transcendental turn, empirical matter guides
experience strictly in accordance with rules we introduce ourselves.6
Consequently, Kant’s position, which combines transcendental ideal-
ism with empirical realism, can only be ‘a comprehensive attempt to
determine what can count as experience of objective reality “for
us” ’.7 While Kant’s idealism is, Pippin believes, defensible, his for-
malism is ultimately subjectivist.
From the outset Pippin disagrees with those critics for whom Kant’s
epistemology amounts to the view that the subject projects or imposes
form or structure onto a material world that, in itself, has no poten-
tial meaning. After the transcendental turn, matter can only have
meaning in relation to conceptual form, but at the same time concepts
are only meaningful insofar as they apply to something given in experi-
ence. This is the force of Pippin’s concern with empirical guidedness,
which explains why Pippin finds himself in opposition to Norman
Kemp Smith, Hans Vaihinger and others who he considers have too
mentalist an interpretation of Kant.8 For Kemp Smith, consciousness
confers meaning and indeed the objects of consciousness are, strictly
speaking, other mental states.9 In particular, necessity and universal-
ity are ‘imposed by the mind’.10 Vaihinger insists that the mind pro-
jects order on the world.11 In contrast, Pippin is committed to the view
that, for Kant, extra-mental reality is not simply the passive recipient
of projections on the part of the subject. The world is given to us,
admittedly in an indeterminate and indirect way. It is indeterminate
because it is given only in sensation and indirect because sensations
without concepts are blind.12 For Pippin, this conclusion is necessar-
ily drawn from Kant’s dualism.
Pippin is equally critical of Guyer’s impositionalist interpretation,
which he sees as rooted in what he calls a literal interpretation of
pure a priori synthesis. Guyer concludes that the latter is responsible
for the subjectivism of Kant’s formalist position. Pippin’s reply is
restricted to consideration of an article published in 1980, which sets
the scene for Guyer’s systematic interpretation of Kant published as
Kant and the Claims of Knowledge seven years later.13 Pippin argues
that although Kant seems to commit himself to an independent syn-
thesis prior to empirical synthesis, this is misleading.14 In contrast,
Pippin reads Kant’s use of the a priori as an adverb.15 The point
Pippin makes is that to talk of an a priori synthesis is not to talk of a
synthesis that literally comes before experience, but rather of a syn-
thesis that is independent of experience and yet serves as a ground for
11
our knowledge of it. Elsewhere Pippin further counters the challenge

to Kant’s formalism saying that to assert that the unity of appercep-
tion is ‘formal’ means to assert merely that judgement, cognitive
activity or experience are only possible if the contents of conscious-
ness are brought to a unity.16 He also stresses that formal talk refers
to the activity of mind and not to some special non-material object.17
Pippin agrees with Guyer that the way in which we ultimately
understand what Kant means by a priori synthesis depends on how
we understand the idea of ‘mind-imposed unity’.18 If we read the
latter phrase literally then we are inevitably led to interpret a priori
synthesis as an independent act occurring prior to empirical experi-
ence. But if we adopt Pippin’s adverbial reading, we need only inter-
pret the a priori as the formal condition of any empirical experience
and not as a distinctive act.19 We will see, however, that Pippin retains
worries about the status of a priori formal intuition.
Pippin argues that the impositionalist interpretation cannot be
right because Kant does not simply claim that the categories of the
understanding – the most pure concepts with which we operate in
unifying any object of experience – establish how an object appears
for us, but says that any object that appears for us must appear under
certain formal conceptual conditions.20 This is the notion of ‘com-
prehensiveness’ and is the key to Pippin’s rejection of the imposition-
alist reading. Kant starts by trying to establish our understanding of
the general structure of reality, but he also shows that this is the only
way in which experience could be understood by human subjects. He
thus establishes the necessity of the formal structure we in fact adopt.
‘This formal unity is not imposed; it just is the only type of unity we
could understand in our experience.’21 Pippin thus distinguishes
between a merely subjective imposition of the structure or form of
reality and a comprehensive account that would establish the neces-
sity of this formal structure for our experience. In the end Pippin’s
judgement is that the comprehensiveness claim is not made good, thus
undermining Kant’s commitment to empirical guidedness. But the
fact that Kant aims at the latter goal counts as the major reason for
Pippin’s opposition to readings like Guyer’s.
Importantly, in his defence of Kant, Pippin contrasts ideas and con-
cepts insisting that the latter are not subjectively imposed.22 However,
if concepts are to qualify as empirically guided, then this must be dis-
coverable in their empirical application. Even if a priori concepts are
not subjective, this will count for little if empirical concepts are. The
aim of empirical guidedness will fail, if empirical concepts turn out to
12
be instances of subjective projection and this is just what Pippin

concludes, as I will discuss below.
Pippin also disagrees with those interpreters for whom Kant’s
major failing is his insistence on the role of subjective faculties. His
argument here is principally addressed to Strawson’s critique of
Kant’s idiom of the faculties. For Strawson, Kant’s supposed psy-
chologism amounts to the claim that the mind imposes form on
matter.23 While Pippin concurs that Kant’s position is ultimately sub-
jectivist, he insists that it is a concern with the ‘activity of the subject’
and does not amount to psychologism.24 In making this argument he
insists that faculty talk is best understood as a consideration of ‘forms
of our conceiving’.25 However a suspicion of Kant’s faculty talk re-
emerges in his discussion of a priori intuition, especially with refer-
ence to ‘The Schematism of the Pure Concepts of Understanding’.26
Now, despite the distance at which Pippin stands to Guyer and to
Strawson he nevertheless concludes that Kant falls into a subjectivism
that should be ruled out by his commitment to empirical guidedness.
How does this come about?
The crux of Pippin’s critique of Kant is the latter’s formalism, which
he believes counts as an ‘independent analysis of the “rules” for knowl-
edge, prior to and divorced from experience of all kinds’.27 The aim of
achieving comprehensiveness and therefore of avoiding impositional-
ism fails because Kant attempts to secure the legitimacy of claims to
knowledge and their applicability to experience within the sphere
of subjectivity alone. His objective of comprehensiveness, that is, of
showing that knowledge is not simply for us but reveals the necessary
conditions of all possible objects of experience, depends on achieving
formal comprehensiveness as we have seen.28 Pippin does not immedi-
ately conclude that the formality of this move entails impositionalism.
Indeed, on the contrary, he says that the formal status of the compre-
hensiveness claim means ‘not just that we impose’ order, but that this
is the only possible ordering.29 However the distinction between sub-
jective and objective ordering that this claim to comprehensiveness
implies is not delivered: ‘Just how are we to understand this notion of
a subjective unity of consciousness “depending” on, being “derivable”
from, or being “grounded” in an objective unity?’30
Pippin’s adverbial defence of Kant’s talk of a priori synthesis is, as
we have already seen, that it amounts to the claim that there are
certain epistemic conditions for experience. This, he holds, should
not be read as evidence of either impositionalism or psychologism,
but rather reveals the formal, methodological or meta-level status of
13
Kant’s project.31 But the solution Pippin offers us now turns out to be
problematic: how do we get from a comprehensive system of the sub-
jective conditions of experience to securing the application of those
epistemic conditions in empirical experience? As he puts it in an
earlier pithy statement ‘If that formal idealism is to be successful, we
must be able to understand its connection with this “material” or
empirical realism . . .’.32 Elsewhere he says that if Kant is to avoid
constructivism, by which he means subjectivism or impositionalism,
he needs a ‘less metaphorical interpretation’ of the relation between
the forms of experience and ‘empirically apprehended “matter” ’.33
The epistemic solution has reintroduced subjectivism by the back
door.
Pippin believes that Kant’s theory of judgement could have pro-
vided a solution to the problem. In the ‘Schematism’ Kant tries to
escape from a strictly formal level of analysis, but fails due to his con-
tinuing to operate at an a priori level.34 It is his reliance on ‘his even
more obscure theory of pure intuition (which itself stresses all over
again a more constructivist theory of phenomenal unity)’ that under-
mines the potential for eventually achieving an adequate account of
empirical guidedness through the strategy of comprehensiveness.35
Thus an attempt at showing how the epistemic subjective forms of
experience genuinely grasp material reality, a problem that was
promised specific attention in the ‘Schematism’, falls back into sub-
jectivism by relying on an unexplained claim that our subjective
power of intuition is capable of constructing reality. Pippin believes
that the ‘Schematism’ chapter is further flawed insofar as it reveals a
‘serious instability in the understanding/sensibility distinction so very
essential to the Critique’.36
While, as I have already mentioned, formalism does not necessar-
ily entail impositionalism in Pippin’s eyes and indeed signals Kant’s
intent of providing an alternative to the latter, a priori formalism is
finally the problem. Thus Pippin says that for Kant ‘the homogeneity
of the manifold is just due somehow to the demand of thought’.37 In
other words it would appear in the end that the charge of imposi-
tionalism is correct. Combination of sensory input comes from
thought alone and there is no effective empirical guidedness from the
side of sense. He then immediately goes on to say: ‘And, if transcen-
dental philosophy is to remain formal and a priori, it is hard to see
how the situation could be otherwise.’38 A formalism that was not a
priori might avoid the fate of impositionalism, but Kant’s version
cannot do so on Pippin’s reckoning:
14
Once Kant argued for the ‘spontaneous’ nature of the understanding, and
the indeterminate nature of the material of sensibility, we naturally wanted
to know ‘by what right’ the understanding can so spontaneously, a priori,
and with certainty and authority of necessity, legislate, and we are here to
receive Kant’s unique answer: that the ‘ground’ or basis for this legislation
rests in the requirements for a unified, self-identical subject.39
Pippin adds that the unity of the subject must be understood for-
mally.40 Formalism ultimately situates the claimed legitimacy of
the categories in subjectivity and, more precisely, in the unified self-
identical subject.41 It would appear that the distinction between con-
cepts, which are not subjective and ideas, which are, has collapsed.42
Empirical guidedness turns out to be a straw man, insofar as it only
arises in accordance with formal structures that we introduce a priori.
One of the most important questions that Pippin raises is that of
the legitimation or ground of empirical concepts. This point is central
to his assessment of Kant’s success in establishing empirical guided-
ness.43 If empirical concepts cannot be shown to arise at least in part
from conditions in extra-mental reality, then their fate is no better
than that of pure a priori concepts. We have already seen that Pippin
distinguishes concepts from ideas on the grounds that the former are
not subjectively imposed while ideas are. However we have also seen
that, on Pippin’s reading, while Kant intended to avoid the conclusion
that concepts impose form on matter, he finally did not succeed in this
aim. Thus, concepts can, in the end, do nothing other than subjec-
tively impose a priori form on the external world. The remaining
question is whether empirical concepts have a ground independent of
pure concepts.
Pippin considers various options; most importantly, for our dis-
cussion, that reflective judgement provides the ground for empirical
concepts. In the Critique of Judgement Kant introduces a distinction
between determining judgement that subsumes a particular instance
or intuition under a universal rule, principle or law and reflective
judgement that seeks a rule or concept for a given particular.44 Kant
argues in the Introductions to the third Critique that reflective judge-
ment provides the principle necessary for the systematic employment
of empirical judgement and, arguably, also for empirical concepts.45
But the problem, as Pippin sees it, is that reflective judgement is a
development of the regulative use of ideas in the Critique of Pure
Reason. As such, reflective judgement, like regulative ideas, is pro-
jected by the mind onto nature. Pippin thus rejects the possibility that
reflective judgement might be the sought-for autonomous ground of
15
empirical concepts, as it is simply too subjective to render it capable

of securing the necessary reference to objectivity. ‘While the most
general explanation for our association of markers into concepts,
species, and so forth, it [reflective judgement] is also the furthest
removed from any connection with any “objective” or empirically
based origin.’46
Thus, Kant’s account of reflective judgement could, or even should,
have given us the means for understanding how empirical concepts
(and, we might add, perhaps even indirectly pure a priori ones) are
capable of being empirically guided. This, however, is impossible
because reflective judgement is a subjective projection. Both in his
book and in a roughly contemporary article, Pippin does not argue
for the premise that secures this conclusion, citing only Buchdahl in
defence of it.47
In order to show how far Pippin’s final verdict lies from his initial
defence of Kant, I must cite him again at some length:
In sum, however much Kant intends to avoid the ‘imposition’ interpreta-
tion, it is still the case that, to use his own favourite metaphor for his for-
malism, reason legislates to nature, or even commands to nature, so
formally, so independently of any ‘material’ reflection about nature, that
the very possibility of ‘obedience’ to such laws, or the role of reason’s ‘sub-
jects’ in formulating them seem considerations much too abstractly
excluded. It is not incidental, I would suggest, that something of the same
problem can be said to emerge in Kant’s moral, aesthetic, and teleological
formalism.48
So while Pippin does not state that Kant’s position is impositionalist,

it is clear that, in Pippin’s view, Kant ultimately falls into this trap.
Pippin goes on to say that Kant would have needed ‘a more
complicated account of the relation between form and a perhaps
metaphysical “content” than his transcendental methodology allows
. . .’.49 Pippin’s own alternative, he admits, is only glimpsed at the
margins of his critique of Kant. The principal elements of his own
position are a comprehensiveness that does not fall back into subjec-
tivism;50 an account of subjectivity that is not indeterminate;51 and a
theory that looks for its confirmation within actuality.52 These fea-
tures suggest a broadly Hegelian programme, especially when com-
bined with Pippin’s contrast between idealism and formalism.
Pippin admits the limitations of his own alternative to Kant:
Admittedly, of course, the idea of ‘breaking down’ Kant’s sensibility/
understanding dualism, or ‘opening the door’ to a fuller explanation of the
16
substantive or metaphysical foundation for pure intuition, concept,

schema, and idea, or even just recognizing the necessity for a substantive
account of the subject which thinks rather than just the rules of its think-
ing, are all fairly abstract conclusions to draw from the problems pre-
sented here.53
I will answer Pippin’s critique by supplying an alternative account
of formalism. We will then be in a position to recognise that Kant is
closer to Pippin than Pippin’s interpretation of Kant would suggest. I
will argue that Kant’s conception of a formal system does not neces-
sarily fall back into subjectivism and that his account of the ‘activity’
of the subject is less rigid than Pippin assumes. I will show how form
arises in anticipation of matter and this allows us to construe empir-
ical guidedness as more than an empty promise. I will not, however,
argue that Kant offered a more substantive metaphysical foundation
for the thinking subject. That would be to go beyond formal idealism
towards the absolute idealism of Hegel. The account of the subject
offered here will, rather, situate the formal framework for thought
within the empirical world.
II Guyer’s critique of impositionalism

Guyer believes that he can uncover within Kant’s writings an alterna-
tive to his official position. Guyer has great sympathy for this unofficial
position, which he believes counts as a transcendental theory of experi-
ence. It is to be found in unpublished manuscripts and in a nascent form
in the ‘Refutation of Idealism’ in the Critique of Pure Reason.54 While
the unofficial position avoids metaphysical dogmatism, the official one
does not and counts as impositionalist. Guyer’s extensive analysis of
Kant’s epistemological project appeared in his book Kant and the
Claims of Knowledge.55 Its core distinction was established in his earlier
article to which, as we saw in the last section, Pippin responded.56
Central to Guyer’s reading of Kant’s epistemology is a distinction
between two orders of necessity. He expresses the first as follows: a
conditional necessity would arise were Kant claiming that if I experi-
ence an object, this requires that I am in fact aware of regularity
among my representations of it.57 All this necessitates is, if I am to
experience an object, it must in fact be of such a form as to be acces-
sible to my means of grasping it cognitively.58 In contrast, absolute
necessity would be required if experience of an object entails aware-
ness of a necessary regularity among my representations of it.59 The
absolute necessity of this condition can only be ensured if the mind
17
imposes this form.60 This is contrasted to what Guyer considers a dis-

tinctive metaphysical view of the relation between mind and reality
where the conditions of the possibility of experience restrict rather
than produce what can be experienced.61
Guyer reassesses the so-called overlooked alternative discussed in
the nineteenth century by Adolf Trendelenburg:
Kant assumes that space and time must be features either of our repre-
sentations or of objects but not both, rather than that space and time may
be either properties of objects or both necessary constraints on our per-
ception of objects and genuine features of the objects we do succeed in per-
ceiving.62
Guyer argues that, from Kant’s perspective, space and time could not
be features of both subjectivity and of objects, because in the latter
case they would not deliver the absolute necessity that is required.
Only if space and time are subjective can they give rise to certain
knowledge of appearances in space and time. Thus, the alternative
was not so much overlooked as excluded by Kant.63
I will now sketch Guyer’s argument as presented in his earlier article
before examining in more detail some of its principal claims.64 He con-
tends that Kant is committed to an a priori synthesis of objects in all
cases of knowledge, not just in mathematics.65 This a priori synthesis
amounts to constitution, insofar as it imposes order on the manifold
of empirical intuition.66 Anticipating attempts to interpret the priority
of synthesis as the claim that all empirical manifolds are merely con-
strained by certain a priori conditions, Guyer identifies further pas-
sages to support his reading that a pure a priori synthesis precedes
empirical syntheses. If empirical syntheses were simply constrained by
a priori rules, then we would have the conditional necessity of Guyer’s
preferred transcendental theory of experience. This would mean that
our minds were restricted to a certain range of accessible objects. In
contrast to this, he finds that Kant is committed to there being a prior
pure synthesis that produces its own objects by imposing form upon
matter. This counts as impositionalism and is a consequence of
attributing certainty to the transcendental unity of apperception. He
goes on to argue that Kant’s thesis that we are certain about the unity
of apperception can only be construed as a synthetic and empirical
claim, not analytical and a priori as Kant suggests, and, finally, that it
rests on a conflation of consciousness and self-consciousness.67
Guyer’s alternative strategy is to argue that Kant could have proposed
a transcendental, but not a priori formal theory.
18
Despite my disagreement with Guyer’s diagnosis of the identity of

the ‘official’ Kant, my interpretation shares some features with
Guyer’s unofficial transcendental theory of experience, being one in
which mind is responsive to an external world not of its own making.
In this regard, I would agree with Guyer that there is no necessary
conflict between formal idealism and a version of realism. The differ-
ence between our positions is that I do not believe we have to divest
Kant of transcendental idealism and most of his published works in
order to be able to establish this. In later chapters I will show that
there is a way of reading the a priori and formalism that illuminates
the relationship between mind and object, showing their necessary
relation without attributing impositionalism to Kant.
Guyer admits that there are passages that could be interpreted
according to a more modest interpretation of the activity of the mind.
These would be in line with what in his book he calls conditional neces-
sity. In his article he describes this tendency thus: ‘the purely hypo-
thetical and analytic assertion that if the ascription of a representation
to an identical self is possible, then it must also be possible to connect
that representation to others ascribed to that self by a certain kind of
synthesis’.68 But there are other passages that he believes cannot be so
interpreted. These commit Kant to the absolute necessity of trans-
cendental apperception and amount to the ‘thoroughly existential and
synthetic claim that all of one’s representations can be ascribed to
oneself’.69 Only this position requires a theory of imposition.
Is there really a clear distinction between these two sorts of pas-
sages? Guyer supplies three examples of passages that, he concedes,
commit Kant only to the claim that a priori syntheses bear conditional
necessity. The first example Guyer says is concerned with a priori syn-
thesis of the pure intuitions of mathematics, which he concedes
implies no imposition on experiential objects as such.70 Meanwhile,
the second and third examples are concerned with experience yet do
not commit Kant to an imposition of pure synthesis. The second
passage Guyer cites claims that appearances ‘must stand under a
priori rules of their synthetic unity’,71 while in the third Kant says that
appearances ‘are subjected to a priori conditions, with which their
synthesis . . . must thoroughly accord’.72 Guyer is ready to concede
that such passages do not suggest that there is a pure synthesis prior
to empirical syntheses, but rather that certain a priori constraints
govern the latter.
However he insists that there is plenty of evidence for the stronger
claim for absolute rather than conditional necessity. He begins with
19
the synthesis of reproduction that Kant says counts as one of the

‘transcendental actions of the mind’ and is ‘grounded prior to all
experience’.73 Guyer then refers to another passage from the first
edition of the ‘Transcendental Deduction’, or A ‘Deduction’ as
opposed to the second edition or B ‘Deduction’. Here Kant says that
apperception is the source of all synthetic unity and that it presup-
poses a productive synthesis of imagination, which is ‘prior to apper-
ception’ and is ‘the ground of all cognition’.74 Guyer sees this as
compelling evidence for Kant’s commitment to a pure synthesis prior
to any empirical synthesis. Finally Guyer turns to a passage from the
B ‘Deduction’ where we are told that the synthetic unity of the man-
ifold ‘precedes a priori all my determinate thought’.75 He does not
comment on the complex relation between the faculties that emerges
in the second of these passages. He is intent simply on insisting that
Kant is committed to certain pure operations of the mind that are
prior to all experience.76
What exactly does Guyer mean by prior in these cases? He clearly
does not mean temporally prior. Nor can he mean apperception is
simply logically prior, for if he did there would be no reason to make
the distinction between passages that express a restriction on our
mental activity and those that supposedly express an imposition of
form on matter. Logical priority would only entail that apperception
was a necessary condition for the possibility of experience of objects,
as we have already seen in our discussion of Pippin’s adverbial
reading. This would be quite consistent with Guyer’s preferred con-
ditional necessity and would not entail that apperception was the suf-
ficient condition of that experience, insofar as it imposed form on
appearances. For Guyer the priority at issue is synonymous with
absolute necessity. This is the real crux of the matter and lies beneath
what Pippin calls Guyer’s literal reading of the priority claim. A pure
synthesis is prior in the sense that apperception produces its own
objects by imposing forms on nature. Thus mind produces forms that
are not only necessary, but also sufficient conditions for the order of
nature.
Is Guyer’s literal reading of the priority of pure synthesis proven
by the textual references he supplies, or is it rather that his indepen-
dent conviction that Kant is committed to impositionalism leads him
to read these passages as literally as possible? The latter view is
encouraged by the next move in Guyer’s argument. He says that even
were the passages initially discussed in his article inconclusive, it
cannot be doubted that Kant is committed to the view that the mind
20
constitutes ‘the objective affinity of nature as underlying all particu-

lar empirical investigation’.77 After all, Kant has said that it is we who
introduce order and regularity into appearances or nature.78 How,
asks Kant, ‘should we be able to get such a unity started a priori, if
subjective grounds of such unity were not contained in the original
sources of cognition in our minds?’79 Admittedly, Kant’s account at
this stage of the first edition ‘Deduction’ raises some problems. How,
crucially, is he sure that the unity achieved will be an ‘a priori certain
unity of the connection of appearances’?80 Moreover, the statement
that this certainty is somehow contained in the original sources of
cognition seems to support Guyer’s belief that Kant is committed to
an absolute necessity imposed transcendentally by the faculties on
appearances and ultimately on matter.
Guyer thinks that the evidence is incontrovertible and that we must
read Kant as an impositionalist. This leads him to miss a possibility
latent in the passages just mentioned. Might there not be a tacit dis-
tinction between the subjective and objective grounds of the same
unity? If this suggestion is taken up, then Kant may only be commit-
ted to the view that the subjective conditions of cognition supply nec-
essary, though insufficient, conditions of the formal structure of
nature. The subjective grounds would then only ‘get things started’
[auf die Bahn bringen]: they initiate unity, but they do not achieve it.81
The grounds of necessary unity are contained in our minds only in an
initial fashion.
The original German offers some potential support for my sugges-
tion. The crucial two sentences read:
For this unity of nature should [soll] be a necessary, that is an a priori

certain unity of the connection of appearances. How should we be able to
get an a priori synthetic unity started [auf die Bahn bringen literally means
‘to bring onto the track’82], if there were not subjective grounds of such
unity a priori in our original cognitive powers and if these subjective
powers were not objectively valid insofar as they are the grounds of the
possibility of knowing any object whatsoever in experience.83
The modal verb sollen suggests a task, even an epistemic duty. Kemp
Smith translates ‘soll’ as ‘has to be’. This expression also bears a
future connotation, although it can be mistaken for a simple state-
ment of necessity. The modal verb soll could also be translated as ‘is
to be’, resulting in the claim that the unity of appearances ‘is to be’
constructed as the result of a project. Thus it would not be contained
already, other than in principle, in the initiating conditions, that is, in
21
the power of synthesis. And although translating ‘auf die Bahn

bringen’ by ‘to establish’, as Kemp Smith does, is accurate, this
English expression eliminates the future nuance of the German
phrase, which could suggest a task that is only just beginning.
Focusing on the future orientation in Kant’s expression allows us to
reassess the status of the immediately preceding claim that ‘the order
and regularity in the appearances, which we entitle nature,we our-
selves introduce [bringen wir selbst hinein]’.84 While we introduce
order, we may not construct or impose it. So when Kant says that
‘[t]he synthesis of the manifold through pure imagination, the unity
of all representations in relation to original apperception, precede[s]
all empirical knowledge [gehen aller empirischen Erkenntnis vor]’,
the priority in question need only be the initiating condition of neces-
sity.85 While the hints afforded by the text support, but are not con-
clusive as to the correctness of my interpretative suggestion, they
should at least make us pause for thought before concluding that the
priority that Kant seeks to establish entails that the mind imposes
order on empirical matter. And if this is so, these passages need not
commit Kant to impositionalism, although they certainly aim to
secure more than a merely conditional or factual coincidence between
mind and objects.
Guyer’s suspicion is that the only way such an a priori certain unity
could be achieved would be if the mind were capable of imposing its
own necessary structures on objects. At this stage I will simply sketch
a possible alternative solution, which I will develop in greater detail
in later chapters. The key here is to distinguish between two possible
levels at which nature may be unified. When Kant says that there
should be an a priori certain unity of appearances, he may mean that
there must be an a priori framework that guarantees the possibility of
connection among empirical appearances. While the framework is
certain at the a priori level, this does not entail that there is such cer-
tainty with regard to empirical experience. We will see later that in
both editions of the ‘Deduction’ Kant insists that the categories do not
fully determine empirical laws and thus reveals that there is a gap
between the a priori framework and empirical nature.86 If I am right,
Kant is only committed to certainty at the transcendental level as a
framework for the possibility of experience, while at the empirical
level, certainty is an ongoing project rather than a fait accompli.
Are there significant developments in the arguments Guyer pre-
sents in his later book? At the outset of his later account Guyer fails
to mention that Kant’s supposed problem arises from a confusion
22
between consciousness and self-consciousness, but otherwise his

argument is the same.87
[Kant] simply assumes that we know certain propositions as universal and
necessary truths; he then argues that such claims to knowledge of neces-
sary truth can be explained only by our antecedent possession of certain
conceptions and capacities which we must, in turn, be able to impose upon
a reality which does not itself, even contingently, conform to these condi-
tions . . . 88
Toward the end of the book he reconstructs the ‘error’ in three stages.
Once again Kant is found to have presupposed the validity of a claim
to knowledge of a universal and necessary truth. Second, we are told
that ordinary experience can never justify such a claim, which must
therefore have an a priori basis. Third, Kant concludes that only
certain forms of intuition, concepts or principles of judgement could
satisfy this requirement.89 This is a more nuanced account than was
provided in the earlier article, but the nub is still the same: namely,
Kant assumes that we are capable of a priori certainty about the unity
of appearances and the only way this can be explained is to posit an
a priori capacity or synthesis that produces such certainty by project-
ing its unity onto the manifold in apprehension. This claim is nothing
other than the impositionalist thesis.
In Kant and the Claims of Knowledge, Guyer declines to answer
critics of his earlier article and presents instead a more detailed
account of Kant’s argument.90 Following an analysis of Kant’s argu-
ments in which he attempts to deduce the validity of the categories
starting from concepts of the object or of judgement, he now returns
to the argumentative territory of his earlier account in analysing those
arguments that start from the self:
[H]e discovered what he thought was a simple but powerful argument
from what he took to be our a priori knowledge of the necessary unity or
identity of the self throughout all of our experiences to the need for an a
priori synthesis of the manifold of intuition, conducted in accord with a
priori concepts.91
Thus the certainty a subject has in its identity throughout experience
entails that it must be able to achieve an equally certain synthesis of
the manifold of apprehension.92 It does so, as we have already seen in
his earlier article, by imposing the forms of intuition onto that man-
ifold in accordance with the categories.
Guyer goes on to say that this ‘supplement[s] his [i.e. Kant’s] in any
case considerable confidence in the deduction of the categories from
23
the conceptions of object and judgement’.93 So it sounds as if this

‘powerful’ argument is not an alternative to those other strategies, but
rather complementary to them. Yet in what follows it becomes clear
that, just as Guyer held in his earlier article, what he considers to be
a new argument is, in his opinion, Kant’s primary one.
The linchpin of Guyer’s argument is once again Kant’s claim that
‘the understanding is itself the source of the laws of nature’.94 This is
what convinces Guyer that Kant’s position must be that ‘the mind can
impose an “affinity” on all appearances’ for only this secures Kant’s
conclusion that they must be ‘associable in themselves and subjected
to universal rules of thoroughgoing connection’.95 This, he thinks, is
‘the only possible explanation of the premise of this new argument,
the assumption that unity of apperception is a priori certain or guar-
anteed to obtain under all possible circumstances’.96 Thus only if
there is an a priori synthesis that unifies all experiences, can the
unity of apperception in which we are already confident be validated.
The unity apperception introduces into experience is, moreover, an
absolute necessity. This is so because our minds are capable of synthe-
sising or determining all appearances that arise for us.
I have argued that Guyer’s evidence for claiming that Kant is com-
mitted to the absolute necessity of apperception is not conclusive. He
certainly has evidence that Kant holds that the transcendental unity
of apperception is a necessary condition of any empirical knowledge,
but does he have the further proof that the latter arises from a pure a
priori synthesis that imposes form on matter? In 1987 Guyer’s evi-
dence is similar to that presented in his earlier article. He still insists
that it is Kant’s ‘unequivocal view that we can impose our categories
on any data of sensibility whatever’.97 In other words, we simply must
read Kant’s statements about a priori synthesis literally, because the
passages and the argument would not bear any other interpretation.
For Guyer, this is the crux of transcendental idealism, for it is only the
transcendental ideality of space – that is, its status as a mental form –
that ensures the transcendental affinity of order among objects qua
appearances necessary for the certainty to which he believes Kant is
committed.98
Interestingly, in 1987 Guyer does not oppose what he calls formal
idealism, which he identifies with Kant’s mature position.99 The latter
is ‘compatible with a realistic interpretation of the intended con-
clusion of the refutation of idealism’.100 What he objects to is the
interpretation of form as imposed on objects of experience. This inter-
pretation entails that transcendental idealism counts not as a doctrine
24
of epistemological modesty pace Henry Allison, but rather as a meta-

physical dogma.101 Hence Guyer is opposed to the formalism correl-
ative to transcendental idealism, which he takes to be the thesis that
form is imposed on objects by the mind. He remarks that Kant’s
mature position commits him only to the ideal status of the forms and
not of the existence of the objects of intuition.102 I hope to have shown
in a provisional way that even in the admittedly difficult territory of
the ‘Transcendental Deduction’ there is no conclusive reason why
Kant should be consigned to impositionalism. Objects have an ideal
dimension insofar as they are formed, but this does not necessarily
entail that form is imposed on matter. I have suggested that Kant’s for-
malism can be seen as the initiation of order through the synthetic
activity of our subjective cognitive faculties. In later chapters I build
on this suggestion.
III Henrich’s Cartesian Certainty – A Basis for

Impositionalism?
While Dieter Henrich, one of Kant’s most influential contemporary
German interpreters, ultimately aims to defend Kant against the
charge of impositionalism, Kant’s critics have found encouragement
in the interpretation he offers in Identity and Objectivity.103 Guyer, in
particular, relies on Henrich’s reading, something I will establish by
an examination of a review written by Guyer just before his publica-
tion of the article considered in the previous section of this chapter.104
Thus, while it may seem odd to include Henrich among Kant’s critics,
it is helpful to include a discussion of his highly reconstructive reading
at this point.
Guyer’s view that for Kant apperception is apodeictically certain
reveals his debt to Henrich, for whom apperception counts as a sort
of Cartesian awareness of what he calls the numerical identity of the
self.105 Against this Allison argues that knowledge requires only the
awareness of ‘the “fact” that this identity must be presupposed as a
necessary condition of knowledge’, in other words, it requires only
‘the necessity of a possibility’, that is, the possibility of being aware
of numerical identity ‘but not its actuality or necessity’.106 As Allison
also puts it: ‘The consciousness of this act . . . is . . . the conscious-
ness of the form of thinking’.107 Importantly he says that Kant’s
account is ‘a formal model or schema for the analysis of the under-
standing and its “logical” activities’ and concludes: ‘As such, it is
neither a bit of introspective psychology nor an idealistic ontological
25
thesis concerning the manner in which the mind “creates” the phe-
nomenal world by imposing its forms upon the given sensible data.’108
Despite the apparent difference between Allison’s formal reading
and Henrich’s Cartesian reading of apperception, we will see later in
this section that the contrast is quite hard to draw.109
Henrich argues that, for Kant, the certainty that we have in the
numerical identity of the self depends on our ability to employ certain
forms of judgement. Guyer agrees with this entailment and the con-
sequent denial of an immediate certainty of self in detachment from
our judgemental capacity. However, while Henrich argues that Kant
is not, or at least should not be, committed to ‘constitution-theoretic
talk’ – that is, to impositionalism – Guyer argues that the latter is a
necessary corollary of Henrich’s correct diagnosis of the basic premise
of Kant’s deduction, namely, certainty about numerical identity:
In fact, we could be certain a priori that we will be conscious of our con-
tinuing identity so long as we are conscious of any representations at all
only if we can always process or force our representations to conform to
the conditions of our consciousness of identity, which is, I take it, the gist
of any theory of constitution.110
So whereas Henrich believes that Kant only slips into impositional-

ism because of an unnecessary confusion, Guyer insists that it is the
unavoidable corollary of the certainty of apperception.111 It is because
the numerical identity of the self can be established only if it is dis-
covered in all appearances, that the former as the transcendental iden-
tity of apperception must impose unity on the latter. Therefore
impositionalism leads on directly from Henrich’s insistence on the
certainty of apperception.
Significantly, Guyer’s core distinction is first aired in response to
Henrich’s interpretation. The theory of constitution arises from what
Guyer calls Henrich’s key argument, that is, ‘that we know a priori
that we will be conscious of our numerical identity in whatever rep-
resentations we will have’ in contrast to the weaker claim that ‘we can
be conscious of our identity in any given sequence of representations
only if they conform to the necessary conditions for our conscious-
ness of identity’.112 This is the by now familiar distinction between
absolute and conditional necessity.113
As we saw, Guyer concludes that Kant’s impositionalism arises
from a confusion between consciousness and self-consciousness.114
Henrich’s insistence that the legitimacy of the categories requires
a priori knowledge of the self – and not just, as Allison argues,
26
consciousness of the form of thinking – once again is a point d’appui

for Guyer’s critique.115
It appears that Henrich has set the stakes too high. He concludes
that the presupposition of Kant’s thought is ‘wholly original’, ‘highly
productive’ and ‘exceptionally convincing’ and sums it up thus: ‘one
sees in the notion of the ego [im Ich-Gedanken] a consciousness
which demonstrates the highest degree of certainty and which is, so
to speak, a priori superior to everything.’116
It is not possible to enter into a detailed reconstruction and assess-
ment of Henrich’s argument here. Nevertheless we must consider the
way in which Henrich rejects the impositionalist conclusion Guyer
draws from his Cartesian premise.
Henrich’s rejection of impositionalism emerges principally from
the important distinction he draws between the identity of self-
consciousness and the identity of the act.117 In Henrich’s view, Kant’s
tendency to focus on the act serves as an ‘unfavourable influence’ on
the development of his proof.118 Henrich interprets this development
in Kant’s account as entailing that the act of the mind ‘subjects
appearances to rules’ and concludes that this commits Kant to a thesis
of ‘real constitution’ rather than simply logical derivation.119 Henrich
is referring to the passage from the first edition of the ‘Deduction’
where Kant famously states:
For the mind could never think its identity in the manifoldness of its rep-
resentations, and indeed think this identity a priori, if it did not have
before its eyes the identity of its act which subjects all synthesis of appre-
hension (which is empirical) to a transcendental unity and thereby renders
possible its connection according to a priori rules.120
Henrich comments that this implies a distinction between particular

acts of synthesis and the act of synthesis itself that generates the a
priori rules necessary for any synthesis whatsoever.121 His concern is
that Kant suggests that we can move directly from the identity of the
subject to the act of synthesis and this in turn suggests that the self
‘subjects appearances to rules’.122 I am not convinced that it is neces-
sary to make the distinction Henrich suggests, as I read this passage
as a not yet well-formulated expression of Kant’s view that the tran-
scendental synthesis of apperception entails some awareness of its
own activity. If so, Kant may not be guilty of falling into imposition-
alism as charged.
In my view the real problem lies not in Kant’s focus on the act of
apperception, but rather in the speed with which he thinks he can
27
move from a priori synthesis to empirical determination. Kant is not

clear about the level at which his argument operates. This is evident
not only in the sentence Henrich holds up for criticism, where he
already talks about empirical apprehension, but in the preceding one
where Kant claims that the ‘Deduction’ achieves the determination
of objects, without making clear whether he is talking about empir-
ical objects or an object ‘in general’, that is, the conditions of possi-
bility for objectivity.123 There is, however, a way of decelerating his
claim. We can take it to imply that the synthesis of apprehension ulti-
mately falls within the general framework of transcendental unity,
but not that the latter imposes form on empirical objects, there
being further conditions of empirical unity yet to be established.
Nevertheless, it seems probable that at the time of writing the first
edition ‘Deduction’ Kant thought he could conclude his argument
much more quickly than turned out to be the case.124
Henrich responds to the tendency in Kant’s position associated
with the talk of an ‘act’ and which we can identify as impositionalist,
as follows:
Now the subject is certainly the agent [Akteur] of a number of activities.
It can invariably reflect upon itself and initiate syntheses. . . . it still does
not always follow that it is in the subject and its acts that the ground must
be sought of all the conditions without which the concept of the subject
cannot be thought and without which the subject could have no knowl-
edge of itself.125
It is crucial here that not only self-knowledge, which would imply

empirical intuition, but also the very thought of the concept of the
subject, the ‘I’, may require conditions beyond the subject. In an
earlier passage Henrich insists that it is not the case that the subject
autonomously guarantees its own unity through its activity of syn-
thesis.126 This is because ‘the continuance of its activity’ is not depend-
ent on itself.127 What other grounds are there for the conditions of
self-consciousness and self-knowledge? One promising suggestion
may be found in a yet earlier passage where he explains why Kant’s
commitment to the subject’s dependence on what is given to it in intu-
ition, results in the fact that he could not have adopted Leibniz’s strict
conception of identity.128 Drawing out Henrich’s general position in
this way allows us to clarify his distinction between real constitution
and logical derivation. The first of these would amount to the posi-
tion that appearances are directly extracted from subjectivity. The
second would coincide with the more complex thought that while the
28
transcendental unity of apperception is the ground of the connection

between subject and world, self-consciousness necessarily stands in
relation to a given and ultimately to objects in the world.
What Henrich means by logical derivation is clarified by the dis-
tinction he draws between numerical identity and the simplicity of the
subject. The deduction of the categories – that is, the proof of their
legitimacy for all appearances – only succeeds through a ‘complex
proof-procedure . . . which must make reference first to the identity
of the subject, but then, necessarily to its simplicity as well’.129
Henrich’s view is that Kant tends to concentrate on simplicity at the
expense of identity. This is what leads Kant to move so quickly from
‘the subject as a principle of identity to the assumption of a priori
rules which apply to all apprehensions’.130 A more mediated relation
between subject and rules is required if Kant is not to fall into a theory
of ‘real constitution’. This supports my suggestion that Henrich is
committed to the subject’s numerical identity arising from a mediate
relation to objectivity. For, surely, the rules are ones that put the
subject in a necessary relation to the possibility of experiencing
objects. However, Henrich’s analysis remains oriented towards sub-
jectivity. Moreover, as we will see, he undermines his commitment to
objectivity, which I do not doubt he holds, by insisting that the rules
that secure numerical identity are pre-temporal. Henrich’s commit-
ment to the link between apperception and objectivity thus remains
problematic.
These passages reveal that, for Henrich, impositionalism is
avoided principally insofar as Kant establishes complexity within
subjectivity. Self-consciousness is certain strictly in the sense that it is
underived from any other consciousness.131 This entails only that self-
consciousness is the ineliminable starting point or origin of all knowl-
edge, but not that it contains the sufficient conditions of the latter.
The crux of Guyer’s dissatisfaction with Henrich’s account and, at
the same time, its productivity for his own attribution of imposition-
alism to Kant, lies at the other pole of the subject–object relation. This
is the implication that we can know a priori that all our representa-
tions will be synthesisable under rules that are the conditions of trans-
ition between different states of the subject.132 The problem, Guyer’s
argument goes, is that this commits Kant to absolute necessity, that
is, to the position that anything we experience is necessarily condi-
tioned by rules that arise from self-consciousness. But the question
here is surely, how maximal or minimal is the condition? Is the
claim that the categories provide the broad framework within which
29
experience is possible, or that they determine the shape or form of

any empirical experience whatsoever? Given that Guyer argues that
Henrich’s position entails impositionalism, he clearly believes that
Henrich is committed to the latter of these two options.
In attributing a commitment to absolute necessity to Kant as inter-
preted by Henrich, Guyer represents Henrich’s position as entailing
that ‘all of our representations can be known a priori to be subject to
synthesis into a “nature” or world of objects themselves completely
interrelated by these rules’.133 In the first passage referred to by Guyer,
Henrich states that all objects are subject to universal laws insofar as
they belong to nature, while in the second he claims ‘It is in relation
to this manifoldness of fundamental forms of judgement that the col-
lective unity of objects in One Nature must be thought’.134 In the
latter passage Henrich is explicitly concerned with the question of
objectivity, however there is no reason to think that he is talking
about empirical objectivity. The nature he has in view in both pas-
sages is surely that of the most general order of things in relation to
the transcendental unity of apperception. Nature at this level is not
yet equivalent to the system of empirical nature. This leaves open the
possibility that while nature in general is subject to a priori synthesis,
empirical nature in its particular detail is not. Were Henrich speaking
of nature at the empirical level, this would follow, but we need not
conclude that he is doing so when he has made no mention of the
empirical. My suggestion gains additional plausibility from Henrich’s
insistence that the subject is the underivable starting place for any
knowledge, not its sole condition and from his claim that it is not nec-
essary to read Kant as committed to impositionalism.
Nevertheless, a statement of the limitations of the argument of the
‘Transcendental Deduction’ is sorely missing in Henrich’s account.
My suggested defence of his account in the face of Guyer’s critique
entails attributing to him the position that the ‘Transcendental
Deduction’ is only the beginning of the project of legitimating the cat-
egories and will be supplemented by further arguments showing how
they apply to empirical objects. His insistence on the ‘Deduction’ as
the sole locus of legitimation for the categories tells against such an
extension of his account. As Karl Ameriks reports an observation by
Ralph Walker, Henrich’s reading of the ‘Deduction’ may establish
nothing more than the necessary application of categories to ‘a blue
qualia (sic) that is square’.135 And without an account of the broader
epistemological project within which the ‘Deduction’ arises, Guyer’s
interpretation is far from unreasonable.
30
But, what of Henrich’s relationship to formalism? The point at

which Henrich most clearly adopts a formalist position comes in the
course of a discussion of the way in which the mind ‘can ascertain its
identity solely in view of a regulated synthesis’.136 Henrich favours
moderate numerical identity in contrast to Leibniz’s strict identity.
Moderate numerical identity allows for a change in the states of a
thing.137 Henrich argues that it must be possible to know something
about transitions taking place under the subject’s identity conditions
‘without being able, in this knowledge, to refer to any definite trans-
itions in which the subject has found itself’.138 What Henrich means
is that there is a formal knowledge of transitions or changes prior to
any identification of change in the subject at the empirical level. The
functions governed by the categories count as modes of transition
which: ‘must be constant, because it is only in this way that they are
suitable for rendering possible, independently of all experience and in
unconditional universality, the knowledge of the subject’s identity,
which is at all times possible.’139
Henrich concludes that knowledge of such modes of trans-
ition or functions is a necessary condition of the subject’s a priori
awareness of itself as identical. Thus self-certainty will only be con-
firmed as apodeictic insofar as it can be guaranteed that it perseveres
throughout such modes of transition. This process is not to
be understood as empirical, but rather as prior to experience.140
Henrich next considers the objection that too much may have been
derived from a ‘completely formal principle’ and that the mere
thought of a sequence of thoughts in time would suffice.141 The
objection under consideration is that the temporality of consci-
ousness would be sufficient for establishing the numerical identity
of the self. Henrich replies that the transitions he is analysing are
prior to temporality and count as analogous to rules of logical
derivation.142
It might appear that Henrich’s method is formalist in the particu-
lar sense that it is modelled on logic. Yet we have already seen that
logical derivation implies a complex derivation of numerical identity
as opposed to the claim that knowledge is simply entailed in the
subject per se.143 Moreover, Henrich distances himself from ‘a theory
of purely formal principles of knowledge (which can also be called a
logical theory of experience)’.144 He insists on the need to combine
the logical and the real trajectories of Kant’s argument, instead of
choosing between them as, he claims, the latter’s commentators have
done.145 It is only by combining both dimensions in Kant that the
31
categories can be grounded and, significantly, Henrich suggests that

this will entail a formal element:
Starting from the fundamental notion of the spontaneous and reflectively
acting subject, one can ground the validity of necessary rules for all
appearances only by means of the other basic notion of the formal and a
priori principle of subject identity.146
There are therefore two basic notions: the real one of the con-
sciousness of the ‘I think’, and the formal or logical one of the numer-
ical identity of the subject over time. This reveals that Henrich’s
position is a hybrid one, combining form with the particularity or
reality of the Cartesian subject. But it is only numerical identity that
secures the validity of the categories. Interestingly, this shows his com-
mitment, once again, to the view that the capacity for self-reflection is
not merely formal, as Allison would have it, but also real. This is but
one of the many complexities of the position he adopts. The interest
of this for my own account is the way in which his, admittedly under-
developed, account of formalism challenges the prevalent equation
between the latter and logicism or mere subjectivism.
In a later article, Henrich distances himself from logical formalism
such as the neo-Kantian project to establish ‘a merely formal property
of thoughts which can itself never become an instance of conscious-
ness’ in contrast to an actual thought of self-identity.147 The position
he favours insists once again that, although formal, self-consciousness
is a ‘real act’ of consciousness.148 But this is not to say that it is actual
at all times: ‘it is always possible for self-consciousness to become
actual, it does not have to be actual’.149 The form of consciousness is
constant and I can be certain of it a priori. The fact that I achieve this
level of self-consciousness is, however, not known a priori.150 It is now
clear that Henrich’s Cartesian certainty is the formal, but not merely
formal, possibility that I can be conscious of my identity in relation to
any representation.
How great a distance, then, is there between Allison’s account of
formal possibility and Henrich’s Cartesian position? We have seen
that in Henrich’s account of the apodeictic certainty of apperception,
he insists that a Cartesian reflection must be possible on at least some
occasions. Allison, we have seen, states that what is at issue is in no
sense introspective, being only a form of thought. But it is arguable
that self-reflection on the activity of thinking is not introspection. And
now it seems that Henrich would be happy with Allison’s claim for
the necessity of a possibility of such self-consciousness, for, as we have
32
just seen, the former holds that we are a priori certain only of the
form of consciousness. Henrich insists that actual self-consciousness
cannot remain a mere possibility and must be capable of actualisa-
tion. But does not Allison’s claim for the necessity of a possibility
commit him to the same position? To say this is surely to say more
than that only a formal awareness is possible and to commit to the
actualisation of that form as a possibility. A fine distinction remains
and it seems to be one of emphasis: while Henrich insists the actual-
isation of the possibility must be possible, Allison says only that the
possibility and ‘not its actuality or necessity’ is necessary.151 The final
distinction is not as clear-cut as Allison seems to think and is very dif-
ficult to pin down, but we can say that, for Henrich, consciousness is
referenced to self-consciousness more explicitly than it is for Allison.
An assessment of the structure of Henrich’s Cartesianism goes
beyond the scope of this discussion. Nevertheless, we have seen that
the formal moment in Henrich’s approach aims to establish a set of
rules prior to experience and to time. In my opinion this is where the
real problem lies. The formal Cartesian status of apperception may
not in itself lead to impositionalism, but a-temporal a priori rules
appear to do so. While the subject’s identity can only be established
in relation to a range of possible judgements, this occurs prior to any
actual application within experience. Now in a sense, from a Kantian
position, this is uncontroversial. The categories are the formal frame-
work for experience and as such are prior to any actual experience.
But the question is: is this a set of, in principle, formal moves that will
only be fully worked out in relation to experience, or are they fully
formed prior to any application? Their being a priori leaves open both
options, but in the second case the order in intuition is imposed. There
are two ways in which we can develop the idea that the rules operate
in principle. In the first case, the categories, although a-temporal,
require temporal schematisation if their full objective validity is to be
established.152 But this would mean that the forms as presented in the
‘Transcendental Deduction’ are only provisionally stated. Henrich’s
insistence on the primacy of the ‘Deduction’ does not suggest he
would wish to adopt this strategy. A stronger position would involve
arguing that even apperception stands in some relation to temporal-
ity, which is fine-tuned in the ‘Schematism’ and ‘Principles’ chapters
of Kant’s Critique of Pure Reason. Thus the categories, while a priori,
are temporal at a formal level and are only the first stage of the deter-
mination of the rules for the unity of apprehension. It is clear that
Henrich would not want to accept that apperception is in any sense
33
temporal, so this option is definitely not open to him. Once again, it

would appear that the additional moves that Henrich could make in
order to avoid impositionalism are not ones he could or would take.
But is Henrich’s insistence on the pre-temporal status of the rules
that express the possible modifications of numerical identity, in effect
the categories, not wholly faithful to Kant? After all, is it not the case
that, for Kant, apperception is distinct from temporality? It would not
be possible to supply a thorough answer to this complicated question
here, but I will provide a sketch developing the second interpretative
option just mentioned. The transcendental unity of apperception is
the form of thought and thus is peculiarly linked to the understand-
ing. However, apperception is a synthetic unity and what it synthe-
sises is the manifold in intuition. All intuition, without exception, is
temporal. Thus apperception is the exercise of the understanding in
relation to intuition. The initial condition of apperception lies on the
intellectual side of dualism, which taken in abstraction from the
activity of synthesis counts as a-temporal. But this initial condition
initiates a synthesis that is necessarily temporal. The transcendental
unity of apperception is the synthetic unity of the temporal achieved
by the understanding. This synthesis operates at a formal level and
thus entails no empirical time consciousness, but it counts as the form
of temporality that necessarily precedes any particular temporal
determination.153
We can conceive of a version of this account more closely modelled
on Henrich’s interpretation. Self-certainty is only achieved in relation
to a number of rules that govern appearances, including, ultimately,
empirical ones. These rules are nothing other than the rules for syn-
thesis of the manifold of intuition. Certainty attaching to appercep-
tion thus arises only in relation to a synthesis that necessarily has
intuition as one of its terms. In effect, as in my original account of this
version, apperception entails the exercise of understanding on intu-
ition. The certainty of apperception cannot be established in detach-
ment from this synthesis.
It is neither Henrich’s intention, nor is it a necessary corollary of his
Cartesian interpretation of Kant, that his account results in imposi-
tionalism. We have seen, however, that insufficient attention to the
levels at which Kant is concerned with nature encourage interpreta-
tions such as Guyer’s. I have argued, nevertheless, that the real danger
of impositionalism derives from Henrich’s insistence that experience is
based on an a-temporal a priori set of rules. Apperception is thus set
outside or at the limits of the spatio-temporal world. In a more recent
34
publication he goes so far as to suggest a reciprocal relation between

subject and world: ‘But the source from which a world originates is
equally dependent on that world.’154 Both mind and world are co-
originary, yet Henrich does not sufficiently develop the object pole of
the subject–object relation. The result has been that his Cartesian posi-
tion has been able to serve as a basis for impositionalist readings of
Kant despite Henrich’s strong intentions to the contrary.
IV Strawson’s critique of transcendental psychology

I now turn to consider the interpretation of Peter Strawson, whose
Bounds of Sense (1966) has been so influential within the English-
speaking world. The distinctive contribution of his interpretation is
his diagnosis of faculty talk as characteristic of Kant’s formalism and
as the source of the latter’s ‘impositionalism’.
For Strawson, impositionalism and formalism are inextricably
entwined. Kant’s epistemology rests on the distinction between a
world of appearances and a world of things in themselves, while his
account of our knowledge of the world of appearances depends on a
metaphysics of mind, that is, a faculty theory. The latter is what
Strawson calls ‘the idiom of the faculties’.155 According to Strawson,
it is the second thesis that does all the argumentative work. It is
because Kant insists that the world is only accessible to us through the
apparatus of the faculties that he relies on a ‘two-worlds’ metaphysics.
Strawson adopts an approach that has become common within
Anglo-American approaches to Kant and other systematic thinkers.
Some supposed kernel of critical philosophy is defended, while an
array of other, apparently central, theses are rejected. A general sus-
picion of ‘systematicity’ supports this approach. Strawson intends to
retain Kant’s commitment to ‘the principle of significance’. This is the
maxim that ‘there can be no legitimate, or even meaningful, employ-
ment of ideas or concepts which does not relate them to empirical or
experiential conditions of their application’.156 According to this
principle, transcendent metaphysics is rejected because of its lack of
empirical significance. It is therefore replaced by a transcendental
‘investigation of that limiting framework of ideas and principles the
use and application of which are essential to empirical knowledge,
and which are implicit in any coherent conception of experience
which we can form’.157 Strawson claims that empiricism is in broad
agreement with the principle of significance and with its rejection of
transcendent metaphysics.
35
However Strawson believes that the search for an a priori frame-

work leads Kant to straddle the boundary that he has shown to be
impassable. Operating at an a priori level would entail exercising
reason beyond the bounds of sense, thus entering into the transcen-
dent metaphysical area Kant himself proscribed. While Strawson
maintains that Kant does not slip back into a traditional form of tran-
scendent metaphysics, he claims that a new and equally illicit meta-
physical position emerges as ‘transcendental psychology’. If there is
no transcendent beyond, there is instead a transcendent mind that
imposes structures on nature. This runs very much against the grain
of Strawson’s own empiricism:
Is it not, after all, easy to read the very formulation of the programme –
‘the determination of the fundamental general structure of any conception
of experience such as we can make intelligible to ourselves’ – in such a way
as to suggest the Kantian-seeming thought that any necessary limits we
find in such a conception are limits imposed by our capacities?158
Strawson’s impositionalist conclusion arises from failing to distin-
guish two ways in which subjectivity could enter into Kant’s argu-
ment. Subjective forms could count as necessary conditions of
determination – this is in fact what Kant holds and it is a highly defen-
sible position – or they could be the necessary and sufficient condi-
tions of determination.159 I have suggested in my response to Guyer
that Kant’s claim is that mind initiates order, but not that the latter is
imposed. According to Strawson, Transcendental Idealism necessar-
ily entails the stronger claim. He concludes that this is not just a pos-
sible reading of Kant, but the one he favours.160 Kant’s mistake, in
Strawson’s eyes, is to make experience wholly dependent on the exer-
cise of mental faculties that impose limits or forms. Thus this inter-
pretation, which started by distinguishing the transcendent from the
transcendental, ends up accusing Kant of muddying the divide. The
transcendental faculties that impose form or limitation on matter
have become the new transcendent and reopen a ‘two-worlds’ ontol-
ogy despite the Copernican turn, which sought to eliminate all objects
not capable of being experienced by us. As Strawson crucially
remarks: ‘The doctrine is not merely that we can have no knowledge
of a supersensible reality. The doctrine is that reality is supersensible
and that we can have no knowledge of it.’161
Strawson sums up the supposed difficulty thus:
The doctrine of transcendental idealism, and the associated picture of the
receiving and ordering apparatus of the mind producing nature as we
36
know it out of the unknowable reality of things as they are in themselves,

are undoubtedly the chief obstacles to a sympathetic understanding of the
Critique.162
He proceeds to consider, briefly, that there may be other ‘weakened

interpretations of these doctrines’. Maybe this talk of faculties is only
an expository device or ‘a device for presenting an analytical or con-
ceptual inquiry in a form readily grasped by the picture-loving imag-
ination’.163 He dismisses both options as imputing a ‘half-conscious
irony’ foreign to Kant.164 One cannot but agree that Kant must have
intended more than expository clarity by such an intricate device.
And although unconscious irony cannot be ruled out, half-conscious
irony indeed seems alien to the generally sober tone of the Critique.
Another possibility is that the faculty idiom does some real transcen-
dental work in Kant’s account.
Strawson’s lack of sympathy with Kant’s systematic project is
transparent in the ‘General Review’ of The Bounds of Sense, where
‘the story of synthesis’ is rejected on the basis that ‘we can claim no
empirical knowledge of its truth’.165 He goes on to raise the very per-
tinent issue of what resources this leaves for addressing the question:
‘when we speak of the necessary unity of experiences, what are the
items so unified, and in what does their necessity consist?’166 His
answer is as follows:
First, the unified items are just the experiences reported in our ordinary
reports of what we see, feel, hear, etc. . . . Second, the unity of these expe-
riences under the rules embodied in the concepts of objects is just what is
exemplified in the general coherence and consistency of our ordinary
descriptions of what we see, hear, feel, etc. The employability of such con-
cepts as these, hence the objectivity of experience in general, is necessar-
ily bound up with the fulfilment of this requirement of consistency or
unity.167
According to this view, the necessary unity within experience does not
imply a priori forms generated by a spontaneous mind. Rather, our
ordinary language descriptions pick out the structure of experience,
which is now to be understood as amounting to its ‘coherence and con-
sistency’. As Strawson later remarks, ‘no high doctrine’ – that is, theory
of a priori form – is necessary.168 Kant’s insistence that only a tran-
scendental form of experience will vindicate claims to validity within
empirical experience, has been replaced with the claim that empirical
experience itself – or at least our ordinary linguistic descriptions of it –
exhibits its validity insofar as it is consistent and coherent. We no
37
longer find a transcendental theory of knowledge but, rather, empirical

coherentism.
I agree with Strawson that dualism and faculty talk are two
sides of one coin, but I understand both of these and the relation
in which they stand to one another in a quite different fashion. I
will argue in what follows that Kant’s commitment to an investiga-
tion of the internal complexity of mind does not entail his giving
up on a commitment to the empirical significance of those struc-
tures. Kant transformed the ‘faculty talk’ already current in the
philosophical vocabulary.169 His faculty talk allows him to express
a new model of reason as a dynamic interaction between a plurality
of mental capacities or orientations. It also allows him to present
experience as an ongoing task, rather than as a fait accompli.
Subjective conditions initiate the task of knowledge, but they do
not complete it. This positive account of ‘faculty talk’ distinguishes
my position not only from those such as Strawson and Guyer who
criticise it, but also from others like Allison, Pippin and Buchdahl
who seek to replace it with a wholly logical concern with ‘epistemic
conditions’.
In the same vein, I take issue with Strawson’s understanding of
Kant’s concern for limits.170 While Strawson follows Locke’s stricture
that philosophy should practise a hygiene of self-regulation, the
Copernican revolution introduces a radical reinterpretation of the
role of limit in philosophy.171 Locke’s version expresses a pragmatic
interest and intellectual modesty, whereas Kant moves the discussion
to a metaphysical level insofar as he is concerned with limits as con-
ditions of possibility for physis or nature.172 Where Strawson thinks
that limits should only be abstracted from experience, Kant holds that
limits are necessary conditions for the possibility of experience and
thus count as a priori. Transcendental Idealism insists that empirical
realism will only be secured if we are able to think the possibility of
experience, not just its actuality. This entails thinking of limits in a
more radical way than is allowed by a Lockean economy and leads to
a fundamental disagreement about the role of form in experience. The
a priori framework is the limit beyond which knowledge is impossi-
ble, but it is also the condition of the possibility of knowledge and
experience. Our subjective cognitive faculties are the condition for
any experience whatsoever and as such count as the limiting frame-
work of that experience. However, as I have already suggested and
will argue in what follows, these conditions initiate and do not
impose order on nature.
38
Strawson’s critique of ‘faculty talk’ still has a significant influence

on how Kant is read, despite the ways in which contemporary
commentators distance themselves from his interpretation. It is likely
that it is because of the way in which the ‘idiom of the faculties’ has
been tainted that Kant’s supporters try to avoid the issue if at all pos-
sible. Meanwhile we have found that his critics return to Kant’s talk
of the faculties.
We have seen that Pippin rejects Strawson’s critique, preferring a
metaphysically neutral reading over a psychological one. Pippin’s
analysis of Kant’s supposed failure focuses on the latter’s aim of estab-
lishing an ‘independent analysis of knowing’ and not on faculty talk
per se. However, it is faculty-talk that prevents Kant from escaping
impositionalism via the theory of judgement. The reliance on an
‘obscure theory of pure intuition’ undermines the project of achiev-
ing a theory of knowledge that is comprehensive and at the same time
‘empirically guided’.173 And while Guyer seeks to distance his account
from Strawson’s, I have suggested that his distinction between
absolute and conditional necessity is more reliant on a concern with
the ‘idiom of the faculties’ than he admits. Passages Guyer relies on
for establishing Kant’s commitment to the absolute necessity refer to
the faculties, while others that Guyer concedes display only a condi-
tional necessity use the more neutral terminology of ‘conditions’.
While Guyer’s criticism of Kant here is not in direct lineage with
Strawson’s, it is nevertheless Kant’s concern with the internal struc-
ture of the mind and its status as a ground for experience that leads,
in both their views, to impositionalism.
Conclusion
In this chapter we have seen that all four interpreters considered
present Kant’s position as formalist. The precise nature of the for-
malism attributed to him is varied, but in every case the latter is seen
as resulting in the subjectivism of his epistemological project.
Moreover for all of these authors, Kant’s project offers not just an
investigation of the subjective necessary conditions of experience,
but also the claim that its foundational ground lies in the subject
alone. This leads Guyer and Strawson to conclude that Kant is
impositionalist. While Pippin and Henrich resist this view, I have
argued that their interpretations ultimately invite the conclusion
that Kant’s formalism results in an overly subjective account of
knowledge.
39
Notes
1. For now I will leave open the ontological status of objects, which could
be understood in a realist or phenomenalist manner. In what follows,
I hope to establish that these are not the only options.
2. Paul Guyer, Kant and the Claims of Knowledge, p. 55.
3. Ibid., For an extended account of Guyer’s contrast between these two
positions, see pp. 53–61 and my discussion below, pp. 17–18.
4. I will investigate how we might understand this claim in what follows.
5. Robert Pippin, Kant’s Theory of Form ( KTF), p. 223. See also
pp. 29–30 on the notion of ‘guidedness’ and pp. 46–53.
6. Ibid., p. 51.
7. Ibid., p. 218. See also pp. 46 ff. on the problem of Angewiesenheit or
guidedness and on ‘affection’.
8. Ibid., p. 219, n. 5. See Norman Kemp Smith, Commentary to Kant’s
Critique of Pure Reason, and Hans Vaihinger, Kommentar zu Kants
Kritik der reinen Vernunft.
9. This is the core of Kemp Smith’s phenomenalist reading of Kant, which
is, he believes, equivalent to an objective idealism.
10. Kemp Smith, Commentary, p. xxxiii.
11. See, for instance, Vaihinger, The Philosophy of ‘As if’, pp. 30–1, where
the categories are characterised as ‘analogical fictions’ that do not give
rise to knowledge of reality. Interestingly, Pippin uncritically adopts
Gerd Buchdahl’s characterisation of reflective judgement as a mental
projection. Buchdahl’s account can be seen as a much more sophisti-
cated reworking of Vaihinger’s interpretation. See discussion of
Buchdahl in Chapter 2 (pp. 59–60).
12. See Pippin, KTF, pp. 34–5 on indeterminacy and pp. 29–30 on indi-
rectness.
13. See Paul Guyer, ‘Kant on Apperception and A Priori Synthesis’; and
Kant and the Claims of Knowledge. See Pippin, KTF, p. 102, n. 16.
14. Pippin, KTF, p. 102.
15. Ibid., p. 103.
16. Ibid., p. 219.
17. Ibid., p. 219.
18. Ibid., p. 102, n. 16.
19. See, for instance, Pippin, KTF, p. 100, n. 13, for the agreement in
which Pippin stands to Melnick for whom ‘categories are the features
objects must have if they are to be subject to our forms of thought’.
Melnick uses the expression ‘epistemic concepts’ to capture the formal
characteristic of the categories. Pippin refers to Melnick’s Kant’s
Analogies of Experience, pp. 40 ff. especially p. 45. This account
seems very close to Allison’s characterisation of the categories as epis-
temic conditions. See my discussion of Allison in the next chapter. The
40
categories provide the form of experience, rather than imposing form

on the latter. See Pippin, KTF, p. 102: ‘The question the deduction will
pose is thus not: Can we discover in the understanding (as birthplace)
concepts which lie there (like seeds) prior to any actual experience?
but: Can we identify and justify a use of concepts which establishes a
relation to all possible objects of experience which does not justify that
use by appeal to what we have experienced?’
20. See Pippin, KTF, p. 185.
21. Ibid., p. 221.
22. Ibid., pp. 209–10, n. 38. See also p. 221.
23. See the final section of this chapter (pp. 35–9).
24. See, for instance, Pippin’s statement that Kant has a ‘metaphysically
neutral notion of intellectual activity’, KTF, p. 221. See also p. 76.
25. Ibid., p. 92 (Pippin’s emphasis).
26. Critique of Pure Reason, A 137–47, B 176–87.
27. Pippin, KTF, p. 87. Pippin’s restriction of Kant’s project to a system of
rules shows, I believe, his agreement with Allison’s reading of tran-
scendental idealism as a theory of epistemic conditions. See Chapter 2
of this volume (pp. 62–3). However, as we will see later in this chapter,
Pippin’s account is also reminiscent of Henrich’s reading of Kant,
where pre-temporal rules provide a formal framework. Whereas both
Allison and Henrich hold that a formalist interpretation secures the
success of Kant’s epistemic project, Pippin believes that Kant’s formal-
ism is the source of its failure.
28. See also Pippin, KTF, pp. 154 ff.
29. Ibid., p. 185.
30. See ibid., p. 171.
31. The methodological is equated with the formal ibid., p. 196. The latter
is equated with a meta-level analysis on p. 217.
32. Ibid., p. 25.
33. Ibid., p. 223.
34. Ibid., p. 142; see also pp. 221–2. Pippin also suggests that Kant’s
theory of judgement could have provided a solution to the problem of
applicability. I hope to show in what follows that judgement is indeed
the clue to the possibility of applicability.
35. Ibid., p. 222.
36. Ibid., p. 222. In Chapter 4, pp. 113–16, I argue that dualism is less
rigid than Pippin suggests and entails a plurality of the faculties.
37. Ibid., p. 142 (Pippin’s emphasis). The tension in Pippin’s position is
evident in the move between a rejection of ‘any simplistic resort to the
imposition metaphor’ on p. 226 and his final diagnosis of Kant’s for-
malism on pp. 227–8.
38. Ibid., p. 142.
39. Ibid., p. 152.
41
40. Ibid., p. 152.

41. This sounds very similar to Henrich’s interpretation as we will see
below.
42. Pippin, KTF, pp. 209–10. See discussion earlier in this section,
p. 12.
43. See especially ibid., pp. 143–50.
44. Critique of Judgment, AA 179.
45. See discussion of the systematic function of reflective judgement in
Chapter 7, pp. 249–55.
46. Pippin, KTF, pp. 118–19 (my addition). See also p. 221.
47. Ibid., p. 118, n. 41. See also Pippin, ‘Kant on Empirical Concepts’,
p. 15: ‘It is the subjective nature of this demand or ‘need’ for unity
which, while it reveals how deeply connected Kant’s version of empiri-
cal knowledge is with his theory of reflective judgement, is of little help
with our problem here.’ In both cases Pippin gives exactly the same ref-
erence to Buchdahl’s Metaphysics and the Philosophy of Science,
pp. 504–6. Interestingly, in later work, Pippin introduces a quite dif-
ferent interpretation of reflective judgement as involving an orientation
to nature. See his ‘The Significance of Taste’, especially pp. 567–9. See
also ‘Avoiding German Idealism’, especially pp. 987–8 and pp. 990–1.
48. Pippin, KTF, pp. 227–8.
49. Ibid., p. 228.
50. Ibid., p. 230: ‘The broad problem at stake in all such critical enter-
prises is how to determine in general proper “account-giving” in phi-
losophy, and indeed in all knowledge or discourse. To make such a
determination implies some sort of comprehensive perspective on what
can or cannot be said, or known . . . I have suggested that, however
important and interesting many of Kant’s reflections on the problem
of form are, this comprehensiveness is not attained.’
51. See ibid., pp. 5–6, n. 6, on Hegel.
52. Ibid., p. 212: ‘If we are in fact “called on” to make assumptions about
experience, which assumptions can direct but whose terms transcend
such experience, then which assumptions would seem to be much more
a result of reflection on the actual, scientific questions asked in various
contexts, than on the formal nature of human reason. If in fact this
were true, then it would also follow that the more effective such “holis-
tic” assumptions were, especially in increasing actual empirical knowl-
edge, the more inclined we would be to treat such assumptions less as
regulative maxims, and more as correct explanations of nature.’
53. Ibid., p. 228.
54. See, for instance, Paul Guyer, ‘Kant on Apperception and A Priori
Synthesis’, p. 211.
55. Paul Guyer, Kant and the Claims of Knowledge ( KCK).
56. Guyer, ‘Kant on Apperception’; Pippin, KTF.
42
57. Guyer, KCK, p. 123.

58. Ibid., p. 55: ‘he does not suggest that the mind is such that it can
always ensure that experience is possible, but rather that it is so con-
stituted that experience will be possible only if the objects of experi-
ence, as a matter of fact, conform to the requisite conditions’.
59. Ibid., p. 123.
60. Ibid., p. 55 (quoted on p. 9, this chapter).
61. Ibid., p. 55.
62. Ibid., p. 363. Guyer refers to the discussion of the debate between
Trendelenburg and Kuno Fischer in Vaihinger, Kommentar (vol. 2,
pp. 290–326), and in Kemp Smith, Commentary, pp. 113–14.
63. I agree with Guyer’s diagnosis that, for Kant, the fit between the spatio-
temporal features of mind and of objects cannot be merely coinciden-
tal. However, this does not mean that form is imposed on nature, but
rather that form supplies a structure necessary for any experience. This
is only the beginning and not the conclusion of how order arises within
experience.
64. See Karl Ameriks’ excellent account in ‘Kant and Guyer on
Apperception’, in Archiv für Geschichte der Philosophie, vol. 65.
65. Guyer, ‘Kant on Apperception’, p. 206.
66. Ibid., p. 206.
67. Ibid., p. 206.
68. Ibid., p. 208.
69. Ibid., p. 208.
70. Ibid., p. 206, A 99–100: ‘synthesis of apprehension must also be exer-
cised a priori, that is, in respect of representations which are not empir-
ical. For without it we should never have a priori the representations
either of space or of time.’ Guyer notes in this article and at the outset
of his book that the translations he uses are normally his own, although
guided by Kemp Smith. I leave his translations unaltered in this section.
71. Ibid., p. 206. References are to A 99–100 and to A 110.
72. Ibid., p. 206. The reference is to A 113.
73. Ibid., p. 206. The reference is to A 101–2.
74. Ibid., p. 206. The reference is to A 117–18.
75. Ibid., p. 206. The reference is to Critique of Pure Reason ( CPR), B
134. The emphasis is Guyer’s.
76. This suggests that Guyer’s critique of Kant may be more indebted to
Peter Strawson’s The Bounds of Sense than he suggests. See the dis-
cussion of Strawson’s critique of Kant’s use of an idiom of the facul-
ties in the final section of this chapter. Guyer refers to Strawson at the
outset of his article, seeking to distinguish his own position from the
latter’s. See Guyer, ‘Kant on Apperception’, p. 205. In the final section,
Guyer claims that ‘“transcendental psychology” may be removed from
“what is living” in Kant’s philosophy’ (p. 211).
43

78. Ibid., p. 207. Reference also to CPR, A 130.
79. Ibid., p. 207. The reference is to CPR, A 125. Guyer’s translation is
slightly different from Kemp Smith’s. See also the discussion of A
‘Deduction’ in Chapter 4, pp. 121–32.
81. CPR, A 125.
82. Thus I agree with Guyer’s translation on this point.
83. CPR, A 125–6 (my translation).
84. Guyer, ‘Kant on Apperception’, p. 207. The reference is to CPR, A 125
(Guyer’s emphasis).
85. Ibid., p. 207. The reference is to A 130.
86. See discussion of the hierarchical nature of the full version of Kant’s
deduction of the categories in Chapter 6 below.
88. Guyer, KCK, p. 6.
89. See ibid., p. 418, for Guyer’s analysis of the structure of Kant’s ‘tran-
scendental deductions’ in contrast to ‘transcendental proofs’ under-
stood more widely.
90. Ibid., p. 438, n. 2.
91. Ibid., p. 131.
92. As we will see in the next section (pp. 25–35), Guyer’s account of the
identity of the self throughout experiences owes a debt to Henrich.
93. Guyer, KCK, p. 131 (my addition). Once again this is very close to
Henrich as we will see in the following section (pp. 25–35). See Guyer,
‘Review of Identität und Objectivität’, p. 153, where he refers to
pp. 16–53 of Henrich’s German edition.
94. CPR, A 127; cited in Guyer, KCK, p. 132.
95. CPR, A 122; see Guyer, KCK, p. 132.
96. Guyer, KCK, p. 132.
97. Ibid., p. 26.
98. In contrast Guyer asks why we shouldn’t rehabilitate ‘transcendental
realism’ regarding space and time. See ibid., p. 349.
99. In Guyer’s view this is to be found principally in some unpublished
remarks and in a letter written in 1792 to Beck and not in the CPR. It
thus stands in contrast to what Guyer interprets as Kant’s official posi-
tion.
100. Guyer, KCK, p. 415.
101. Ibid., p. 414. See discussion of Allison’s Kant’s Transcendental
Idealism in Chapter 2, pp. 61–9.
102. Ibid., p. 414.
103. In this section I will principally discuss Henrich’s Identität und
Objectivität. An English version, ‘Identity and Objectivity’, is to be
found in Henrich, The Unity of Reason.
44
104. See Guyer’s ‘Review of Identität und Objectivität’.

105. We will come back to what is meant by this particular kind of identity.
See Henrich, ‘Identity and Objectivity’. See especially pp. 186–7 on
numerical identity and Cartesian certainty (Identität und Objectivität:
pp. 86–7). See also Henrich, ‘The Identity of the Subject in the
Transcendental Deduction’, in Schaper and Vossenkuhl, Reading
Kant, especially pp. 254 and 261–6 on ‘numerical identity’.
106. Henry Allison, Kant’s Transcendental Idealism ( KTI) (1983),
p. 140. (This discussion is omitted in the 2004 edition.) In a note,
Allison says that Guyer criticises Henrich for his Cartesian reading of
Kant (see p. 353, n. 23). But it is rather that Guyer criticises the
Cartesian tendency in Kant and agrees with Henrich’s reading. See
Guyer, ‘Review’, p. 162. Speaking of Henrich’s claim that, for Kant,
we have a priori knowledge of numerical identity, Guyer states: ‘I have
no doubt that Kant does subscribe to this argument’. Admittedly
Guyer questions Henrich’s use of the term ‘Cartesian’. He suggests that
to call the certainty of apperception ‘Cartesian’ is misleading as
Descartes was referring to a singular starting point, whereas Henrich
seeks to establish the foundation of a plurality of certain forms of
judgement. Moreover, for Descartes the certainty of the Cogito does
not imply certainty about the continuing existence of a thinking sub-
stance or of the conservation of finite substances, for the latter requires
God. However, it is likely that Henrich uses the term ‘Cartesian’ in a
way similar to Husserl, that is to express an apodeictic starting place,
rather than to subscribe to Descartes’ wider metaphysics. See Husserl,
Cartesian Meditations. See Chapter 2, pp. 61–9 below for an extended
discussion of Allison.
107. Allison, KTI (1983), p. 144 (my emphasis).
108. Ibid., p. 144.
109. This is in part because (as we will see later), for Henrich, apperception
is not only real but also formal.
110. Guyer, ‘Review’, p. 166. See also p. 167, where he says that absolute
certainty entails ‘we can always force our future representations to
conform to our categorial rules’.
111. See discussion of logical derivation and real constitution later in this
section.
112. Guyer, ‘Review’, pp. 166–7.
113. Yet again Guyer refers to Kant’s account of the ‘affinity’ of nature as
his trump card. See ‘Review’, p. 166, n. 12. The reference is to CPR,
A 121–125. See my discussion of this in the previous section.
114. See discussion of Guyer’s article, ‘Kant on Apperception’, on
p. 18.
115. See in particular Guyer’s discussion of Henrich’s insistence on the
necessity of self-consciousness on pp. 163 and 164 of his ‘Review’.
45
116. Henrich, ‘Identity and Objectivity’, p. 198; references to the German

edition follow in brackets (p. 100).
117. Guyer does not discuss this specific argument in his ‘Review’.
118. Henrich, ‘Identity and Objectivity’, p. 201 (p. 103).
119. Ibid., p. 201 (p. 104).
120. CPR, A108. Cited by Henrich, ‘Identity and Objectivity’, pp. 199–200
(p. 102).
122. Ibid., p. 201 (p. 104).
123. CPR, A 108: ‘The original and necessary consciousness of the identity
of the self is thus at the same time a consciousness of an equally nec-
essary unity of the synthesis of all appearances according to concepts,
that is, according to rules, which not only make them necessarily
reproducible but also in so doing determine an object for their intu-
ition, that is, the concept of something wherein they are necessarily
interconnected’ (my emphasis).
124. See Chapter 6 on Kant’s discussion of special laws in the ‘Deduction’.
125. Henrich, ‘Identity and Objectivity’, p. 201 (p. 104). Note Henrich’s
mention of the capacity of the subject to initiate synthesis.
126. Ibid., p. 186 (pp. 85–6).
127. Ibid., p. 186 (p. 85).
128. Ibid., p. 181: ‘for the states of the representing subject are dependent
upon what is given to it in intuition’ (p. 79).
129. Ibid., p. 204 (p. 107).
130. Ibid., p. 203 (p. 106).
131. Ibid., p. 164 (pp. 58–9).
132. Guyer, ‘Review’, pp. 152–3. Guyer argues that it is Henrich’s
commitment to the certainty of self-consciousness that leads to
absolute necessity, but it is the conclusion of this entailment that really
worries him.
133. Ibid., p. 153. Guyer refers to pp. 52 and 109 of the German edition.
134. The first passage is from Henrich, ‘Identity and Objectivity’,
pp. 159–60 (p. 52). The second is from Ibid., p. 206 (p. 109). See
Guyer’s ‘Review’, p. 153.
135. Karl Ameriks, ‘Recent Work on Kant’s Theoretical Philosophy’, p. 16.
He refers to Ralph Walker, A Selective Bibliography on Kant, p. 195.
137. See ibid., p. 180 (p. 78): ‘Particular things can remain the same
throughout change in their states as long as several of their constitutive
properties remain unchanged or as long as the thing in question endures
through the continuous, but never total change of its states.’ In con-
trast, Leibniz’s strict numerical identity is committed to the view that
‘all states of substances are present in them from the beginning onward’
(p. 179 (p. 77)). This results in the theory of substances as monads.
46
138. Ibid., p. 187 (p. 86).

139. Ibid., p. 188 (p. 88).
140. Ibid., p. 188 (p. 88).
141. Ibid., p. 189 (p. 89).
142. Ibid., p. 189 (p. 90).
143. Ibid., p. 201 (pp. 103–4). See discussion on pp. 28–9.
144. Ibid., p. 207 (p. 111).
145. Ibid., pp. 207–8 (pp. 111–12).
146. Ibid., p. 208 (p. 112).
147. Henrich, ‘The Identity of the Subject in the Transcendental Deduction’,
Section 6: ‘Identity as a Formal Property of Self-Consciousness’, in
Schaper and Vossenkuhl, Reading Kant, pp. 266–70. See in particular
p. 268.
148. Ibid., p. 268.
149. Ibid., p. 268.
150. Ibid., p. 269. It should be noted that Henrich’s ‘fact’ is not the same as
Allison’s. See Allison, KTI, p. 140, cited on p. 25 above. Whereas
Henrich’s ‘fact’ would be a successful actualisation of form and would
not count as a priori, Allison’s is the transcendental presupposition of
a formal structure.
151. Allison, KTI (1983), p. 140. See discussion on pp. 25–6 above.
152. Chapter 6 is a working-through of this perspective.
153. I present a version of this argument in my ‘Kant’s Phenomenological
Reduction?’, where I argue that aesthetic judgement reveals the
primary temporalisation and spatialisation necessary for any judge-
ment (Hughes 2007).
154. ‘The Moral Image of the World’, in Henrich, Aesthetic Judgement and
the Moral Image of the World, p. 3. At the same point he also says ‘it
can be shown that the unity of self-consciousness could not even be
conceived unless that very unity functions as the point of departure for
constituting a world of objects’.
155. Strawson, The Bounds of Sense, p. 30.
156. Ibid., p. 16.
157. Ibid., p. 18.
158. Ibid., p. 44 (Strawson’s emphasis).
159. See Chapter 5 on the subjective side of the deduction, and Chapter 4,
pp. 116–19, on the status of faculty talk.
161. Ibid., p. 38.
162. Ibid., p. 22 (my emphasis).
163. Ibid., p. 22.
164. Ibid., p. 22.
165. Ibid., p. 32.
166. Ibid., p. 32.
47
167. Ibid., p. 32.

168. Ibid., p. 44.
169. For a related claim, see Henrich, ‘Kant’s Explanation of Aesthetic
Judgment’, in Aesthetic Judgement and The Moral Image of the World.
See especially pp. 32 and 54.
171. See John Locke, An Essay Concerning Human Understanding
‘Introduction’ pp. 46–7: ‘Our Business here is not to know all things,
but those which concern our Conduct’. See also p. 44.
172. Kant was undoubtedly metaphysical in this sense, as is any philosophi-
cal position that goes beyond bare naturalism. If any version of tran-
scendence is in question, metaphysics is at issue. But this is not to say
that the projection of a transcendent or dogmatic position is necessary
for metaphysics. If transcendence emerges within, or at least in relation
to, experience, metaphysics can be critical or dualist and avoid dogma-
tism.
173. See discussion on Pippin, pp. 10–17 ff. above.
48
Formalism and the Circle of Representation
The purpose of this chapter is to show how formalism is defended by

some of Kant’s most important recent supporters. Gerd Buchdahl’s
Metaphysics and the Philosophy of Science, first published in 1969,
set the scene for the riposte to the dominant Strawsonian critique of
Kant, published three years previously in 1966.1 Henry Allison
openly acknowledges his debt to Buchdahl, as does Pippin in his less
sympathetic interpretation of Kant. Béatrice Longuenesse’s direct ref-
erence to Buchdahl is restricted to one critical note, but his influence
can be indirectly traced through Allison.2
While in the accounts offered by Buchdahl, Allison and Longuenesse
we will find more positive accounts of formalism than offered by Kant’s
critics, I will nevertheless argue that we discover an emphasis on sub-
jective structure at the expense of a convincing account of the relation
between subjective form and the material given in experience. Both
Buchdahl and Longuenesse, for all the many strengths of their inter-
pretations, risk falling into what I call a circle of representation. Allison
has a more convincing account of affection, but still fails to provide a
sufficiently robust account of the material side of experience.
I Buchdahl’s Reductive Formalism

Buchdahl’s intricate and forceful interpretation of Kant has not often
been given due attention. While his extensive discussion of Kant in
Metaphysics and the Philosophy of Science is regularly cited, it is not
so often analysed or assessed. Nevertheless, Buchdahl’s interpretation
of Kant has been of great importance for Pippin, as we saw in the pre-
vious chapter, and also for Allison, whom I will discuss in the fol-
lowing section. Buchdahl’s reading diagnoses a hierarchy of formal
frameworks for experience in Kant’s epistemology. He argues that the
categories of the understanding provide a provisional template for the
derivation of methodological principles that establish how we should
judge in empirical cases.3 The categories do not, however, automati-
cally supply the content or even the form of empirical judgements.4
49
His later work, Kant and the Dynamics of Reason, develops this per-
spective and, in particular, the view that an approach inspired by
Husserl’s phenomenological reduction allows for a new assessment of
the status of Kant’s ontological commitments. I will concentrate
exclusively on establishing the formalist status of Buchdahl’s inter-
pretation and on showing how his version of formalism leads to a
problem in the account he gives of extra-mental reality.
Buchdahl insists that Kant’s critical project is not intended to
secure the certainty of particular descriptions of the world, but only
the formal conditions which first make possible the certainty of those
propositions:
We are no longer concerned with the question of the certainty of our
knowledge of such and such a set of statements, but only of the formal
conditions that have to be satisfied before questions of certainty can even
be asked.5
This is critical philosophy’s transcendental turn to the conditions of

the possibility of experience. Only once transcendental conditions
have been established will it be possible to assess the validity of con-
crete propositions about experience. Despite widespread recognition
that conditions of possibility are the trademark of the critical turn,
there remains, in Buchdahl’s opinion, much confusion between and
conflation of transcendental and empirical levels of analysis. One
leading example of this is the assumption that in the analysis of
causality in the Second Analogy, Kant is already referring to actual
causal connection, rather than investigating the ground of the possi-
bility of any causal connection.6 Kant’s focus, Buchdahl insists, is on
the transcendental form of experience, although he prefers to refer to
this as structure, and, in particular, as transcendental structure.7
In addition to the distinction between transcendental and empiri-
cal levels of experience, Buchdahl insists on further distinctions at the
transcendental or formal level. In Metaphysics and the Philosophy of
Science, Buchdahl distinguishes between the categorical and regula-
tive orders of experience.8 While the first coincides with the most
general order of experience or phenomena, the second aims at estab-
lishing an order among the range of particular natural laws. Thus
there are two orders of nature. The first arises from the understand-
ing and its generation of categories, while the second arises only
insofar as reason projects an orientating idea that makes possible
order in empirical nature. Buchdahl argues that the second order is
only loosely or analogously modelled on the first, which thus does not
50
provide the validating grounds of the former.9 While categorical prin-

ciples are objective in the sense that, without them, the experience of
objects would be impossible, regulative principles are only subjective
in that they arise from an intervention on the part of the subject.10
The order among empirical laws of nature is a construct that facili-
tates the investigation of that world, particularly by science.
In Kant and the Dynamics of Reason, Buchdahl distinguishes three
levels of formal structure. These coincide with what he calls general,
special and systems ontology. All three operate at the level of possi-
bility, that is, they establish the possibility of a certain level of expe-
riencing nature. General ontology investigates the possibility of
nature in general, while special ontology is concerned with material
(or physical) nature and systems ontology with ‘nature as an
“ordered” system of objects and of the empirical laws that govern
their behaviour’.11 The first coincides with the constitutive principles
and the third with the regulative principles. The second denotes the
territory of Kant’s Metaphysical Foundations of Natural Science and
‘investigates the possibility, and thus intelligibility, of the basic con-
cepts and laws of Newtonian science’.12 The identification of this dis-
tinctive ontology reinforces Buchdahl’s point that the Critique of Pure
Reason does not directly offer a foundation for Newtonian physics.13
In laying out the three levels at which nature can be ordered, I follow
Buchdahl’s order of presentation. However it may be useful to
remember that the hierarchical sequence from generality towards
empirical application would, rather, be from general via systems to
special ontology. Systems ontology provides the mediating principle
that allows for the transition from general ontology to the construc-
tion of a formal structure for Newtonian physics.
Primarily Buchdahl is concerned with the first and third levels of
ontology, that is, with ‘the real possibility of objective cognition or
experience in general’ and with ‘the problem of the validation of the
methodological maxims and ideas of natural science’.14 As in his
earlier work, the analysis of causality is a touchstone for the distinc-
tion between levels. General ontology establishes only that a concept
of causality must be presupposed. Any actual causal connection can
only count as contingent.15 Buchdahl goes so far as to say that the
transcendental category of causation ‘might be quite compatible
with the absence of a network of laws, or of any laws whatsoever’.16
Systems ontology alone is capable of establishing that experience
as a whole is causally connected, but it only operates regulatively
or subjectively, that is, from the point of view of the judging subject.
51
The methodological approach, which treats nature as a system of

empirical laws, allows for the emergence of a further level of analysis
of nature as material or physical. This is the level of special ontology,
which, while it does not directly coincide with Newtonian physics,
establishes the philosophical perspective from which the latter
operates.
Each level of possibility, or ontology, ultimately counts as formal.
In general ontology the forms of space and time and the categories
supply the framework of nature in general. In systems ontology this
same framework indirectly supplies a form for empirical laws. The
Metaphysical Foundations of Natural Science offers a further deter-
mination of what is ultimately the same form by establishing the basic
concepts and laws of Newtonian science, focusing principally on the
concept of force. Importantly, even at this level of analysis Kant is
concerned not with empirical experience and only with its form.
In analysing the formalist status of Buchdahl’s interpretation of
Kant, it is illuminating to examine the relation the former draws
between the transcendental and formal status of the frameworks
under discussion. The following passage from his earlier work
addresses what he later refers to as general ontology:
[T]he ‘critical’ approach will (as already mentioned) formulate the ground
of the possibility of an individual material thing as such in terms of tran-
scendental structure, i.e. of construction in space and time, under the guid-
ance of the categories: an essential requirement for the definition of the
‘reality of the data’ being that the formal framework should contain an a
posteriori element, i.e., be anchored in an instance of actual experience of
the completed object (qua ‘phenomenon’).17
A problem arises that will turn out to be crucial for Buchdahl’s for-
malist interpretation of Kant. Buchdahl claims that the formal frame-
work must ‘contain’ or ‘be anchored in’ an a posteriori element. How
could this be, if the qualifying characteristic of the transcendental
level of analysis is that it operates strictly at an a priori level? Does
this suggest that the formal framework and the transcendental struc-
ture are not equivalent? Moreover, is there not a contradiction
between saying on the one hand that the formal framework contains
the a posteriori while on the other hand it is anchored in the latter?
The metaphors seem to push in two directions, suggesting contrary
theses. We will see later that this tension arises from Buchdahl’s
commitment to there being a balance between the framework and the
contents of experience, that is, between the a priori and a posteriori.
52
I need to pre-empt the conclusion to my later discussion of this

problem and Buchdahl’s solution, because his fashion of resolving the
issue allows us to clarify the relation between transcendental struc-
ture and formal framework.
Buchdahl rearticulates what was to have been a balance in a one-
sided manner, for he claims that the a posteriori is finally grounded in
the a priori. Thus transcendental structure and formal framework are
identical here – despite the fact that the latter contains the a posteri-
ori – because, in the final analysis, the given arises from the sponta-
neous contribution of the subject.18 Clearly, though, this solution
severely compromises any suggestion that the formal framework is
anchored in or, in any meaningful sense, balanced by experience.
Later I will look at Buchdahl’s account of the status of the given in
more detail.
Buchdahl is hesitant as to whether the second level of framework,
i.e. that of the fundamental laws of the natural sciences, qualifies as
transcendental. On occasion he signals ambivalence as to this and
even as to its formal status:
Particularly in the case of EC [Special Ontology] it is not so much any
formal ‘validity’ that is in question, as whether the explication involved is
fruitful, anchored in the general structure of physical experience. In the
case of EC, the argument is only barely transcendental. . .19
Later he takes a negative position on the question saying that ‘in
Kant’s technical terminology of this it is not something “transcen-
dental” but an analysis in “special metaphysics” ’.20 We can conclude
that special metaphysics provides a framework, but not necessarily a
transcendental one and thus it may not be formal in the strong sense
where the two terms are equivalent.
The discussion preceding Buchdahl’s negative judgement on the
transcendental status of special metaphysics reveals that the other
two levels of framework do so qualify.21 This is also clear from the
following more moderate contrast:
[l]eaving it open whether Special Ontology is transcendental in quite the
same way in which General Ontology is construed; unlike the case of
Systems Ontology, where we found that there is a relatively clear-cut dis-
tinction between the employment of the methodological framework, as
yielding the phenomenology of a scientific theory, and its employment as
generating a Systems Ontology.22
This passage identifies the stakes in qualifying not only as formal but
also transcendental. What is required is that there be a clear distinction
53
between the framework or form and the particular descriptions of the

world founded on that framework. General and systems ontologies
achieve this and thus count as transcendental. In contrast, Special
Ontology lacks a sufficient gap between ontology or structure and phe-
nomenology, that is, the range of experiences falling under that struc-
ture.23 Buchdahl’s criterion for transcendental status would bear much
more examination, but we can sum it up quickly for our purposes by
saying that only if a clear distinction is achieved between the a priori
and the a posteriori can the former operate as the condition of the pos-
sibility of the latter.
In Buchdahl’s account, form is equivalent to determination and
matter is the determinable.24 Form and matter are two sides of the
experiential world of appearance qua forms of intuition and empiri-
cal content. As such, they also function as the poles of what he calls
the reduction of that appearance to its transcendental conditions.
This aspect of Buchdahl’s approach is most developed in Kant and the
Dynamics of Reason but is already latently present in Metaphysics
and the Philosophy of Science where he says that the notion of the
object is ‘“bracketed” by virtue of the transcendental condition’.25
The Husserlian idea is that the transcendental conditions of the pos-
sibility of experience, which Buchdahl, unlike Husserl, calls its ontol-
ogy, can be accessed by abstracting specific qualities from objects and
returning to their transcendental or formal conditions. This strategy
of reduction turns out, on Buchdahl’s reading, to be the crux of the
critical turn in philosophy. The reduced aspect of form is captured in
Husserlian terminology thus: ‘the noetic aspect of form is defined as
“determination” [Bestimmung], “relation”, or again, as “copula”.’26
Thus form is an element – and, we will find, threatens to become the
unique element – of the transcendental structure that emerges under
reduction. Buchdahl goes so far as to say: ‘the ground of the world
has become “transcendental form” ’.27
But Buchdahl is insistent that this strategy does not signal a sub-
jective turn. The logical or grammatical reduction of the phenomenal
world does not undermine the latter’s empirical reality. It is only its
ontological status that is put in question, in that the transcendental
method, as interpreted by Buchdahl, eschews reliance on an external
ontological source for the being of objects. What is left is the mere
‘fact’ that objects are given to us in experience. Buchdahl refers to
this by the German term Sachheit.28 Givenness, considered at the
transcendental level, i.e. at the level of the possibility of experience, is
nothing but the fact that there is something rather than nothing.
54
While there are causes within experience, there is no cause, no double

affect, of experience. Talk of the given at the transcendental, as
opposed to the uncritical, level is simply ‘Pickwickian’, that is, merely
a form of speaking that arises from the internal logic of a technical
discourse.29 In other words, we speak of and, perhaps, have to think
of reality standing at the limits of experience, yet this image is mis-
leading when we assume that there is something outside experience.30
The form of experience is reached by a reduction, but lived expe-
rience is re-attained through a construction that realises the form
within a concrete experience. Realisation constructs a particular
object through giving spatial and temporal form to matter. Thus the
two sides of the reduction are repeated or realised at the empirical
level. Buchdahl calls his method in general ‘Reduction-Realization
Process’, and usually refers to it as RRP.31 Only what he calls
‘Transcendental Reflection’ can reduce experience to its transcenden-
tal ground and then progressively construct, that is, realise that
ground at different levels of actuality.32 Matter is undifferentiated
prior to its formal construction. Buchdahl concludes that there is only
one object that is analysed or interpreted in many ways or at many
different levels. Only realisation differentiates the singular transcen-
dental object into a plurality of empirical objects.
As I mentioned earlier Buchdahl is committed to the view that
there is a balance between form and materiality.33 While this way of
describing experience is intended to show that Kant’s transcendental
project does not threaten the objective existence of objects, the
account given by Buchdahl in both Metaphysics and the Philosophy
of Science and in Kant and the Dynamics of Reason is unlikely to
convert the unconverted. Not only is the objective situated entirely
within the immanent, it is ultimately given a subjective ground. For
while Buchdahl distinguishes the subjective and objective trajectories
within constitutive judgement, it turns out that the given or objective
side ultimately comes from the subject:
It follows further that the character of the ‘given’ as ‘objective’, i.e., as
other-than-self, will likewise have to be construed as a contribution made
a priori. The character of the object as being ‘other-than-self’ will have to
be located, not in what is received, but in what is spontaneously con-
tributed by the cognitive subject in the context of judging. We might
describe this by saying that the ‘self’ is here regarded as the seat of the a
priori, and of what is ‘other-than-self’: objectivity being located within the
framework of possible experience, instead of residing in a realm that tran-
scends experience.34
55
The given as objective is grounded in the spontaneity of the subject.

If this is what balancing amounts to, it is very one-sided. While
Buchdahl could have argued that what is other-than-self always
stands in relation to the self, he goes a step further by saying the
former is located in the contribution made by the subject. Moreover,
he says this contribution is spontaneous, not receptive, and that there
is no role for the given. The suspicion is strengthened that his initial
insistence that there is no perspective beyond the world of the subject
results in the subjectivisation of the object.
However, might not Buchdahl simply be committed to the position
that the form of the given comes from the subject? This would be quite
uncontroversial for it is unquestionably Kant’s position that the form
of sensibility is subjective. If this is right, Buchdahl’s argument is only
that our ability to take up the given – and thus to refer to an object in
the first place – requires the a priori capacity of intuition insofar as
the latter introduces spatial and temporal form into the manifold of
apprehension. And this would mean that he is simply not concerned
with the extra-mental element of the given. But the problem is exactly
one of omission. Buchdahl fails to adequately address the question of
affection, or at least the side of it that, in passing, he refers to as ‘other-
than-self’.35 He concedes that what is passively received is one of the
component elements of the given, but quickly resolves the element
that is ‘other-than-self’ into a discussion of the a priori structure of
the given, which he says ultimately derives from the self.36 Buchdahl
is quite right to say that what is given is not simply passively received
because of the formal element Kant discovers in intuition. However,
his accounts of both receptivity and what is received by our power of
intuition are inadequate.37 Givenness is reduced to the form of matter
that can be realised or constructed as empirical objects.
In Kant and the Dynamics of Reason there is independent and sug-
gestive evidence that Buchdahl does not intend to adopt a subjectivist
position. He argues that the transcendental object [Buchdahl’s abbre-
viation is To], which is arrived at in the reduction, ‘ “determines” what
particular concrete empirical intuitions will come up a posteriori’.38
To, in one of the many varieties Buchdahl distinguishes, qualifies as a
‘boundary notion [Grenzbegriff] (Ton) which “limits sensibility”, and
thereby allows the “problematic” possibility of something “positive
beyond the circumference [Umfang] of sensibility” ’.39 Buchdahl is
most concerned to counter the view that affection amounts to the
causal affect of objects on us and his alternative is that To is the ground
of appearances. For this reason he rejects Graham Bird’s interpretation
56
of the transcendental object as a logical myth.40 Affection [Affizierung]

is best understood not as a causal relation, but as ‘ “being brought to”
– or “being in” a state of conscious awareness’.41 These passages reveal
that while the transcendental object does not causally influence us, no
more is it a mere thought experiment.
Admittedly Buchdahl says that Sachheit is ‘the merely conceptual
aspect of the holding of a certain state of affairs; in particular, to the
fact that such and such is the case’.42 But this does not mean that he
considers experience as a logical construction. Instead he wants to
emphasise the contingency of experience and the fact that it ‘involves
a ‘finding’ as well as an ‘originating’; that there is a particular state
of affairs rather than another is simply the case.43 This is not some-
thing that bears further reflection, but is the beginning point for all
reflection. The ‘fact’ of experience does not render the latter subjec-
tive, but merely establishes that its contingency is not open to further
analysis.
Buchdahl’s intent of avoiding attributing to knowledge a merely
subjective status is further evident in a comment he makes on Fichte:
It may be of interest to note that this reading echoes the approach taken
by Fichte already during Kant’s lifetime, except that we have sought to
replace the Fichtean somewhat misleading emphasis on the ‘self’, the Ich,
by the more neutral language of the process of reduction and realization.44
While he concedes his project is comparable with Fichtean idealism,
Buchdahl seeks to establish the distinctiveness of their approaches. So
what is it that is to be denied in Fichte’s approach? Buchdahl does not
accept that experience in general arises from self-affection. This is
particularly clear earlier in a harsher judgement on Fichte:
If this is not an ‘external cause’, neither should it on any account be
described as standing for the ‘affection of the self by the self – the disas-
trous Fichtean direction of post-Kantian idealism. The self does not ‘posit’
itself as an entity affecting itself; rather, it is our analysis of realization
which posits the self (sensibility) as affected or determined, a concept
intended to render nothing more than the aspect of ‘the given’ in the object
of that analysis.45
Buchdahl wants to replace Fichte’s subjective idealism with language
that establishes the logical or grammatical ground of experience in a
neutral place between subject and object. It remains to be seen
whether the turn to the logical subject and the ‘Reduction-Realization
Process’ accessed only by transcendental reflection is as neutral as
Buchdahl’s emphasis on balance would require.
57
In Metaphysics and the Philosophy of Science Buchdahl discusses

a ‘shift in the grammar of the Kantian “subjectivity” ’.46 There is ‘no
reason to locate concepts, just because they are logical entities, in a
realm of subjectivity’.47 Further the subjective is not to be equated
with the psychological or empirical.48 This may seem rather confus-
ing. Are concepts subjective or not? But as Buchdahl is about to make
clear, what is at issue here is what he calls a sharpening of Kant’s use
of subjectivity. What he means is that Kant’s range of reference is
always subjective, where this term corresponds with experience in
general in contrast to the thing-in-itself.49 We can therefore only make
distinctions between different degrees of subjectivity within that
broader horizon. In other words, Kant says that experience and the
categories that provide order are subjective so as to emphasise that
we are not concerned with a world beyond the bounds of possible
access. Our world is always just that, the world that is accessible to
us. Within our world, the subject has a greater or lesser role to play
depending on the level of analysis. For instance, the projective role of
reason involves a much more active and explicit role for the judging
subject than does the relatively uncreative role of the understanding.50
On the face of it, this is a convincing representation of Kant’s tran-
scendental or Copernican turn.
Buchdahl’s aim, then, is to reappraise the relation in which subjec-
tivity and objectivity stand to one another within the Kantian project.
This will be of the greatest importance for, as he says: ‘[t]his “bringing
together” of the “subjective” and the “objective”, of the “act” and
what results from the act, is the very epitome of the Kantian method.’51
However the suspicion remains that what results is a rather
subjective synthesis of subject and object, as the following passage
suggests:
This ‘sharpening’ of the Kantian meanings is graphically illustrated in
putting Kant’s main point by saying that the ‘objectivity’ of the phenom-
enal object is not ‘objectively given’ but wrought ‘subjectively’. This at
once makes it clear that the second ‘objectivity’ in this account no longer
has a normal sense.52
Admittedly, Buchdahl simply says that the objectivity of the phenom-

enal object and not the phenomenal object per se is ‘wrought subjec-
tively’. On the face of things this, once again, is a straightforward
interpretation of Kant: for an object, strictly speaking, arises from
the subsumption of an intuition under a concept of the understanding
and thus entails the operation of our subjective faculties. But the
58
problem with Buchdahl’s position is, as previously remarked, one of

omission. Buchdahl omits to mention that our intuition operates on
something given in experience. This suggests that the given too – not
just the forms of intuition – comes from the subject. The result is that
subjectivity contains objectivity or the a posteriori within it, rather
than what was at first presented as its equivalent, that the a priori is
anchored in the latter. When we recall that one page earlier, Buchdahl
has declared Pickwickian the status of the given, he is left with a very
unbalanced account of the relation between the form and the matter
of experience.
What has happened to dualism? Certainly for Kant the forms of
intuition are subjective, but they establish the possibility of our recep-
tivity to something given in experience and, in the case of spatial
form, something external. Buchdahl is, of course, right to insist that
there is no absolute externality qua thing-in-itself. But it seems that
the sharpening of subjectivity situates the limits and not just the form
of experience within the subject. Such an interpretation, while offered
in defence of Kant, opens the way for critiques such as Pippin’s.
We have seen that the account given by Buchdahl in his earlier
work of the balance within general ontology is, in his own terms,
rather Pickwickian. But we have also seen that, especially in Kant
and the Dynamics of Reason, Buchdahl seeks to establish distance
between his position and subjectivism. We might hope that we will
discover a more convincing account of objectivity in the contrast he
draws in the later work between general and systems ontologies.
There is no balance between the activity of reason and the empirical
world into which it introduces order as a system of empirical laws.
For this reason Buchdahl considers systems ontology to be subjective.
In contrast, there is a balance between the categories of the under-
standing and the matter that is taken up and formed as an object:
There is however a difference between the two cases. The categories of the
understanding (e.g. causality) are valid only if they apply in the context of
a sensory manifold; Kant calls this their ‘constitutive employment’. Only
in this context is the notion of an ‘object’ admissible; only thus can the
understanding ‘create’, so to speak, its object. For reason the case is dif-
ferent . . . nothing as such ‘corresponds’ to the theoretical frameworks of
science; the ‘unity’ which such theories ‘mirror’, is not ‘given’ but – as
Kant ceaselessly reiterates – only ‘projected’.53
The objectivity of general ontology arises from the fact that it is

applied to the sensory given. However, as the use of the word creates
59
suggests, the understanding is spontaneous and does not merely

mirror an already existing order. The latter claim is, of course,
correct, but does not entail that the understanding creates order. In
the end, Buchdahl’s distinction is difficult to pin down. While reason
‘projects’, understanding ‘injects’ order into nature at the most
general level of its constitution. ‘The understanding is said to proceed
“spontaneously”, so that the feature of “objectivity” which Kant
seeks to define as “possible” is secured by showing it to be “self-
wrought”.’54 Buchdahl insists that a balance between transcendental
structure and a posteriori given qualifies general ontology as objec-
tive. Yet on examination there seems to be at best a difference of
degree between an understanding that injects order and a reason that
projects it.55
The way in which this plays into the hands of interpretations such
as Pippin’s and indeed Guyer’s and Strawson’s is by now clear and
becomes even more explicit when Buchdahl declares:
As with the understanding, reason must not ‘borrow’ unity from what is
presented to it, but instead ‘impose it’. The explanation of this procedure
in both cases is the same: one can extract only as much with ‘apodeictic
certainty’ from ‘the world’ as is injected into it by the spontaneous proce-
dures of understanding and reason.56
In this passage ‘imposition’ can be substituted for both ‘injection’ and

‘projection’ and it would appear that the crucial distinction between
understanding and reason collapses, leading us to conclude that not
only reason, but also understanding, necessarily imposes form on
matter. Despite Buchdahl’s talk of balance between the a priori and
the a posteriori, the distinction between general ontology and systems
ontology is in danger of disappearing. While it is clear that Buchdahl
intends to secure a meaningful sense in which there is an objective
element within constitutive judgement, it is not clear that he can
achieve this within the perspective he develops.
Buchdahl recognises the problem only too well. In Metaphysics
and the Philosophy of Science, Buchdahl remarks that it is very diffi-
cult to discover what Kant believes is ‘brought into being’ and what
is ‘found’ or ‘given’.57 His response to this dilemma in Kant and the
Dynamics of Reason is to conclude that: ‘the kantian world is entirely
“enclosed” within its transcendental framework’.58 It would appear
that, in the end, not even a Fichtean Anstoß sets experience in
motion.
Reinforcing Buchdahl’s failure to adequately address the given as
60
given is his resistance to the philosophical relevance of applicability.

He denies that Kant is in any way motivated by the ‘empirical ques-
tion whether what we observe to be the case . . . is in fact the case’.59
He says this can only count as trivial in contrast with his project of
establishing the possibility of cognition of objects in general.60
Buchdahl is, of course, correct that Kant is not directly concerned
with empirical validity, but what he underestimates is that the valid-
ity of the formal structure of objects in general will only be established
once Kant has shown that the latter applies to empirical objects.
Admittedly, he only ever addresses this question at a general level, but
the formal framework must be shown to be the form of matter and
this cannot be achieved if the latter is merely viewed in a Pickwickian
fashion. Kant is committed to an extra-mental given, as I will argue
in the next chapter, and unless he can show how the latter is taken up
through the forms of experience, these will only count as logical
forms and not as forms of experience.61 The fact that matter is empir-
ical and that we are not affected by things-in-themselves, does not
rule out that what is given in experience is genuinely external to the
subject. Kant holds that we are affected by something given in expe-
rience and this is not merely a technical issue. Admittedly, the other-
than-self can only affect us insofar as we have a capacity to be
affected, but Buchdahl focuses on that capacity at the expense of what
affects us and, moreover, characterises the object as rooted in spon-
taneity rather than, equally, in receptivity. He thus denies that the
question of affection is a genuine philosophical question.
What we have found in our investigation of Buchdahl’s account is
that he establishes a powerful formalist interpretation of Kant’s epis-
temology. This version of formalism establishes the need for a reci-
procity or balance between form and matter. However we have also
discovered that, in the last instance, the objective side of the balance
achieved by his version of formalism falls back into a monism of form
or subjectivity. This has encouraged Pippin and others in their cri-
tiques of Kant’s supposed subjectivism. It remains to be seen whether
there is an alternative elaboration of formalism that does not end up
in the same cul-de-sac.
II Allison’s Methodological Formalism

Allison’s Kant’s Transcendental Idealism has transformed Kant
studies in the English speaking world and beyond by rediscovering
a Kant not covertly dependent on a hidden noumenal world of
61
things-in-themselves.62 Allison establishes the human focus of critical

philosophy and shows how the latter is still of great interest for
current debates on the ‘transcendental’ in general.
Allison shares much with Buchdahl’s formalist reading, insisting
that the a priori must be understood as the form of experience and
not as a self-standing apparatus. In Kant’s Transcendental Idealism,
Allison enters into neither analysis nor criticism of Buchdahl, simply
appealing to his authority on a number of occasions.63 A notable sim-
ilarity between their readings is the way in which they interpret the
relation between the ‘Transcendental Deduction’ and the sections that
follow on from it. For both, the full story of the legitimation of the
categories for experience is only to be found in the ‘Analytic’ as a
whole.
Allison’s particular take on Kant’s epistemology is to insist that the
transcendental turn uncovers the ‘epistemic conditions’ of experience.
These provide the fundamental framework without which experience
would not be possible. I will argue that although it may seem that
Allison engages in what Pippin calls an ‘independent analysis of rules
for knowing’, his commitments are more complex.64 Importantly,
Allison does not adopt Buchdahl’s talk of imposition, an interpreta-
tion he ascribes to the ‘standard picture’ for which Kant counts as a
phenomenalist.65
Allison favours an analysis of epistemic conditions rather than one
of the conditions of the possibility of experience because of the way
in which the former focuses attention on the conditions for knowl-
edge in contrast to logical, psychological and ontological condi-
tions.66 It is straightforward to establish that Allison’s epistemic
conditions should be construed as formal in status. In an early section
of Kant’s Transcendental Idealism entitled ‘Transcendental Idealism
as Formal Idealism’, Allison remarks that Kant notes in at least two
places that formal idealism is an appropriate alternative term for his
approach.67 Allison goes on to say that Kant would have done better
to have followed his own recommendation as this may have discour-
aged the phenomenalist standard picture that has been so prevalent
in interpretations of Kant. He continues: ‘Kant’s formal idealism is
“formal” in the sense that it is a theory about the nature and the scope
of the conditions under which objects can be experienced or known
by the human mind.’68
This short passage holds within it the most important elements of
the formalist status of Allison’s interpretation. The Copernican turn
is a formalist one to the epistemic conditions of experience and, as
62
such, necessarily turns our attention to the role of mind in experience.

However, Allison insists that the mind is to be understood logically
and not psychologically. At this stage of his argument the importance
of the relation between form and mind is clear, while its specific
nature is less so:
. . . behind Kant’s formal idealism, lies a principle that is implicit in the
Critique as a whole, but is nowhere made fully explicit: that whatever is
necessary for the representation or experience of something as an object,
that is, whatever is required for the recognition or picking out of what is
‘objective’ in our experience, must reflect the cognitive structure of the
mind (its manner of representing) rather than the nature of the object as
it is in itself.69
The question arises, what is meant by the claim that the objective
reflects the structure of the mind? Is the object other than the mind or
merely a reflection of it? Admittedly, as Allison says in this passage,
the object is not in-itself, but this does not eliminate the possibility
that the object is empirical and yet still external to the mind. The
question is: in what way can objectivity be dependent on subjectivity
without Kant’s account falling back into the trap of impositionalism?
Later, Allison emphasises the centrality of mental form for his epis-
temic or methodological reading thus:
Once again, then, the point is that the appeal to the formal, a priori con-
ditions of human experience and their characterization as ‘epistemic’ are
the defining features of Kant’s idealism. The position is idealistic because,
as we have seen, it grants to these conditions the function of defining the
meaning of ‘object’ or equivalently, of determining what can count as
‘objective’ for the human mind.70
Epistemic conditions are formal in status and establish the conditions
of objectivity. It is thus that Allison insists on the indexing of objec-
tivity to the mind, without falling into psychological introspection.
Objectivity does not so much mirror subjectivity as depend on the
latter for its conditions of possibility. Nevertheless, it is still not clear
in what sense objectivity requires that there is something outside the
subject. If it does not, then the spectre of impositionalism – with its
attendant inattention to the affective or aesthetic side of experience –
returns.
Reinforcing the equivalence between forms and epistemic condi-
tions, Allison tells us: ‘[c]orrelatively, “form” must be taken to mean
condition, while “matter” means that which is conditioned or deter-
mined by the form’.71 Were there any doubt, he sums up his account
63
of the sensible conditions of experience thus: ‘[i]n the last analysis

then, everything turns on Kant’s claim that the representation of space
functions as a form or condition of human experience.’72
The epistemic conditions comprise the forms of intuition and the
categories of the understanding. Distancing himself from orthodox
interpretations – and, importantly, anticipating Longuenesse’s
extended defence of the relevance of the ‘Metaphysical Deduction of
the Categories’ for the Critique as a whole – Allison argues that
this section counts as a genuine first step in Kant’s argument. The
‘Table of Judgements’ presents the ‘conditions of all human thought,
for all such thought is judgemental, and they are the conditions of
judgement’.73 Judgemental functions require certain concepts, but
the objective or empirical reality of these concepts requires the
‘Transcendental Deduction’.74
Having thus rehabilitated the beginning of the ‘Analytic’, Allison
continues to show how its full scope is necessary if Kant is to estab-
lish that the categories are the conditions not just of thought but also
of experience. Two stages of the ‘Deduction’, the ‘Schematism’ and
the ‘Principles’ chapters, are all necessary for this. While the
‘Schematism’ chapter establishes that a schema ‘furnishes a meaning
condition for a corresponding pure concept’, the ‘Principles’ chapter
establishes that the same schema provides ‘the conditions of the deter-
mination of appearances in time and thus of the possibility of experi-
ence’.75 The second Analogy and its transition from ‘the subjection of
the perceptions to the rule to the subjection of the perceived event’ is
the ‘culmination’ of the ‘Transcendental Analytic’.76 Allison follows
Buchdahl in favouring the weak interpretation of this portion of the
text, arguing that Kant’s intent is only to show that there is some
antecedent condition in any succession that counts as an event and
not the stronger claim that other such events will follow the same
pattern.77 Thus concludes Allison’s consistently formalist and hierar-
chical reading of Kant’s epistemology as presented in the ‘Analytic’ of
the first Critique.
We have already seen the problems Buchdahl falls into through his
insistence that the given has a merely Pickwickian status. Despite the
temptation to conclude that Allison’s focus on epistemic conditions
amounts to what Pippin calls ‘an exclusive, foundational concern
with our “mode of knowledge” ’, there are reasons for suspecting that
this is not Allison’s intention. 78 He contrasts his position to that of
the phenomenalist interpretation, which he calls the standard picture,
for which an appearance is equivalent to a mere representation.79
64
Empirical realism commits Kant, he argues, to the position that ‘our

experience is not limited to the private domain of our own repre-
sentations, but includes an encounter with “empirically real” spatio-
temporal objects’.80 Explaining the significance of the Copernican
Revolution’s insistence that objects must conform to our knowledge,
he comments: ‘But the mind-dependent objects of philosophers such
as Berkeley and Hume (ideas and impressions) can no more be said
to “conform to our knowledge” than can a humanly inaccessible
“object” such as Lockean real essence’.81
Allison’s point is that a merely ideal object would not conform to
our knowledge, simply because it would already be internal to the
latter. Epistemic conditions are the mental conditions that make the
representation of an object possible, and thus render the latter know-
able. Mind supplies the structure or form of objects, not their mater-
ial or substance. Epistemic conditions are, thus, not intended as an
independent alternative for, but rather make possible an intentional
reference to objects.
Nevertheless Allison’s relative inattention to the affective side of
Kant’s project – a topic that Kant also insufficiently elaborated – has
the result that the epistemic conditions seem to promise a neutral and
self-standing apparatus. Objectivity is explicated wholly with refer-
ence to epistemic conditions. What is ‘“objective” in our experience,
must reflect the cognitive structure of the mind (its manner of repre-
senting) rather than the nature of the object as it is in itself’.82 This is
unquestionably true, for Kant holds that an object that can be known
must conform to the structure arising from our minds. But although
it sometimes sounds as if this is not only the necessary but also the
sufficient condition of objectivity, I will show in the next chapter that
Kant’s position is rather more complex.
In the early stages of the original edition of Kant’s Transcendental
Idealism, there is no treatment of the material condition of objectiv-
ity. Allison’s view, quite simply, is that an object is whatever conforms
to the mind’s conditions (both sensible and intellectual) for the rep-
resentation of it as an object’.83 If matter has a necessary role to play
in objectivity, he gives us little in the way of a clue as to what this
would involve. However, in the final section of the first edition Allison
discusses the question of affection in relation to the thing-in-itself and
the transcendental object, while in the second edition he treats this
topic in the revised first part of the book. 84 In both editions
his account is based on a distinction between two levels at which
affection can be approached within the transcendental perspective.
65
Empirically formed matter is the result of affection: it therefore

cannot be also the latter’s cause or ground. To consider the ground of
matter is to view it as not yet empirically formed, even though we
cannot know or experience it as such.85 When we abstract from expe-
rience in this way, we are thinking of an indeterminate affect that
gives rise to sensation insofar as it is taken up through sensibility.
Allison concludes that it is justifiable to refer to these material condi-
tions of human cognition methodologically as the ground or cause for
the matter in experience.86 This establishes a distinction between
empirical matter, which is always already formed, and its ground that
counts as transcendental matter.87 We only think of the latter insofar
as we abstract from the empirical character of an object and consider
the empirical object as if it is unconditioned by the forms of space and
time or the categories of the understanding.
Allison argues that many of Kant’s discussions of the transcenden-
tal object can be understood as a treatment of the matter that is pre-
supposed in any empirical object. But instead of treating matter in a
transcendent fashion as a thing-in-itself, a transcendental approach
allows Kant to consider what is only ever experienced in relation to
epistemic conditions as if it were not constrained by the epistemic
conditions that are characteristic of anything represented by us. This
thought experiment, whereby matter is abstracted from empirical
experience, makes use of the notion of the transcendental object.88
This development reinforces other evidence already mentioned above
that Allison does not intend to merely provide an independent analy-
sis of rules for knowing.
A philosophical reconstruction of our affection by the object is
expressed by reference either to the object considered as in itself or as
the transcendental object.89 This means that affection can be recog-
nised as the ‘necessary (material) condition of the possibility of
experience, and in this sense as part of a “transcendental story” ’.90
Experience is only possible insofar as we are affected by something. It
is therefore a necessary condition of a transcendental reconstruction of
experience that the something in question is viewed as a ‘ “something
in general = x” ’ that is, as the transcendental object.91
This account of affection makes a crucial adjustment to the signif-
icance of Allison’s account of epistemic conditions. In addition to the
latter, which are mental in status, there is a necessary material condi-
tion. Matter must be given to us if experience is to be possible. While
matter can be thought in abstraction from experience and is not
reducible to epistemic conditions, it is nevertheless only known in rela-
66
tion to the latter. The material condition of experience always precedes

any experience or reconstruction we can make of it. Nevertheless this
does not render it extraneous to the transcendental system, but as
Allison insists, makes it ‘a necessary condition of a transcendental
account’.92
As we saw in the previous section, Buchdahl insists that Sachheit,
or the fact that something is given, is a condition of experience.
However he denies that this is of philosophical significance, saying
that matter is only ever empirical. Allison would, of course, agree if
this were taken to mean that we are only ever affected by appearing
objects and not, indirectly or otherwise, by things in themselves.
Nevertheless the advance in Allison’s account is that he is willing to
concede the need for a philosophical consideration of the limits of
experience. We must be able to transcendentally reflect on matter as
given on the basis of a ground not conditioned by our sensibility.
Recognition of this is not only philosophically possible, but necessary
to a transcendental account.
Nevertheless, Allison’s relative inattention to the role of affection,
especially in the early chapters of the first edition of Kant’s
Transcendental Idealism, has encouraged the view that the a priori
form of experience can be discussed without any reference to the
affective side of experience. Even when he eventually addresses the
material element in experience, it is treated as simply the other side of
the formal story. This is particularly marked in the new edition where
he describes the claim that matter has a transcendental ground as an
‘analytic claim’.93 Is there any sense in which form must anticipate
matter or stand in a dynamic relation to the latter? Allison does not
raise any of these questions. That Allison is committed to the mater-
ial side of the transcendental story has been established, but as the
material given is merely an analytic entailment of epistemic form, it
is questionable as to whether he has, in fact, escaped the charge that
he offers nothing but a reflection on ‘our mode of knowledge’. If he
is to show how the latter intentionally grasps an object, he would
have to provide more than a merely analytic account of the relation
between form and matter. This would entail focusing on the way in
which epistemic conditions take up the affective given.
My suggestion will be that a deeper examination of the subjective
faculties reveals how the subjective conditions of experience provide
access to extra-mental objects.94 Allison, like Buchdahl before him,
seeks to establish that the formal framework of experience is not
psychological. However his determination to avoid any trace of
67
subjectivism, leads to a failure to adequately explore the subjective

conditions that have now been established qua epistemic conditions
as the conditions of objectivity.
Allison rejects Strawson’s view that formalism is inextricable from
Kant’s use of the idiom of the faculties. A methodological considera-
tion of epistemic conditions is meant to avoid any suggestion of expe-
rience being ultimately grounded in a mental or psychological source.
Allison clearly shares Buchahl’s view that it would be preferable
to adopt a more neutral or logical vocabulary than Kant does.
Nevertheless, Allison does not reject the value of faculty talk outright
– his answer to the question of whether there is any justification for
undertaking an examination of the subjective side of cognition being
difficult to pin down.
In Idealism and Freedom, which gathers together articles origi-
nally published between 1986 and 1995, Allison defends Kant’s tran-
scendental psychology or the ‘subjective deduction’ as ‘part of the
project of justifying or establishing the objective validity of the cate-
gories’.95 Arguing against Patricia Kitcher’s empirically oriented
defence of ‘faculty talk’, Allison says that the subject of trans-
cendental psychology is neither phenomenal nor noumenal, being
rather, using an expression introduced by Gilbert Ryle, ‘systemati-
cally elusive’.96 This suggests that it may be defensible to appeal to
faculty talk, albeit indeterminately.
In another article where he discusses his teacher Aron Gurwitsch,
Allison highlights what he sees as the negative results of focusing on
the subjective apparatus of cognition. He argues just as strongly
against the phenomenological orientation of Gurwitsch, as against
the empirical materialism of Kitcher. His major disagreement with
the former is that despite Gurwitsch’s recognition of the need to read
the three stages of the A ‘Deduction’ methodologically – that is, as
outlining three conditions of experience – he still tends on occasion
to treat apprehension as if it were ‘a unique form of pre-conceptual
awareness that does not involve the categories’.97 This would
suggest a pre-conceptual order ‘ratified’ by the categories.98 Allison
strongly contests this, saying that through the introduction of the
rule of the category we are able to represent ‘a distinctive objective
order through these perceptions, not an order of them’.99
Gurwitsch’s focus on the aesthetic component within Kant’s episte-
mology has, he thinks, severed what must stay in necessary con-
junction, namely the intuitive and conceptual elements of cognition.
A focus on the more subjectively oriented A ‘Deduction’ results in
68
the undermining of the methodological approach Gurwitsch estab-

lished.
While in these two quite differently oriented critiques Allison
argues against taking faculty talk as substantive – that is, as referring
to empirical mental entities or pure acts – he does not rule out that
there is a valid way of talking about the faculties as formal. Indeed,
his response to Kemp Smith’s charge that Kant confuses two conflict-
ing notions of form suggests that he may be in favour of this.100 Allison
draws our attention to an ambiguity in Kant’s talk of the form of intu-
ition as implying both the ‘form of intuiting’ and the ‘form of the intu-
ited’ and defends Kant’s suggestion that space as the form of intuiting
outer objects is the subjective formal source of the manifold.101
Nevertheless Allison insists that the ‘Transcendental Aesthetic’ is prin-
cipally concerned with the form of the intuited, or the objective side
of the formal analysis of intuition.
It has become clear that Allison need not merely offer ‘an exclu-
sive, foundational concern with our “mode of knowledge” ’.102
However, just as his account of affection is underdeveloped as to the
relation in which the formal conditions of experience stand to the
material given, correspondingly he is relatively silent on how a sub-
jective representation takes up an affect. While Allison’s account does
not enclose us in a circle of representation, it also fails to explain how
representations could stand for something other than themselves.
Interestingly, it may be the relative inattention he pays to the subjec-
tive side of experience that results in the enduring suspicion that for-
malism somehow misses the extra-mental status of the object.
III Longuenesse’s functionalist formalism: the appropriation

of sense and imagination
Béatrice Longuenesse’s Kant and the Capacity to Judge directly
addresses the question of the relation between representations and
the objects of representations. 103 Her claim that the ‘Table of
Judgements’ in the Critique of Pure Reason offers a key for answer-
ing this question is an important and challenging addition to the
recent literature. I will show how her account, focusing on the formal
status of transcendental idealism, highlights what she calls the col-
laboration of the faculties necessary for knowledge. However, I will
also argue that, despite her insistence on the dualist character of
Kant’s project, the working through of her complex position finally
puts at risk the affective side of dualism insofar as she insists that the
69
same function of understanding is the source both of the forms of

understanding and of intuition.
In ‘The Divisions of the Transcendental Logic and the Leading
Thread’, Longuenesse rehearses the Leitmotiven of her earlier book.104
Her article gives a textual reading of the introductory passages of
the ‘Transcendental Analytic’ where Kant presents the ‘Table of
Judgements’ and claims that it is the ‘leading thread’ for the discovery of
a priori categories governing our experience.105 Longuenesse argues that
this often criticised claim, and the accompanying apparatus of logical
judgement, is defensible if we read these passages as an analysis of
the form of thought in general, a form that anticipates the categorial
table.
Longuenesse identifies the central role of form in transcendental ide-
alism. From the outset she talks not only of ‘forms of sensibility’ but
also of ‘forms of the understanding’.106 Whereas talk of intellectual
capacities was already widespread, what is striking in Kant’s project
is his transcending of the ‘quite standard empirical-psychological
meaning’ of capacity.107 Indeed she insists that ‘the originality of Kant’s
position lies in his attributing to each capacity a specific form, or
mode of combination and ordering of representational data’.108 Instead
of simply describing conventionally distinguished mental activities,
Kant moves to the transcendental level insofar as he attributes to
each capacity a specific way or mode of combining and ordering
the elements of experience. Whereas intuition allows us to combine
sensible data, understanding is the source of order among concepts.
As Longuenesse expresses the situation, it is the collaboration
between two kinds of forms that properly qualifies Kant’s logic as
transcendental.109
Kant’s formalism is thus linked to faculty talk, which has been re-
identified as transcendental rather than psychological. Longuenesse
points out that Kant is operating with a quite different understanding
of the relation between formal logic and psychology than would be
the case in the twentieth and twenty-first centuries, especially in the
analytical tradition:
Note that for Kant – contrary to the view of logic prevalent after Frege –the
opposition between logic and psychology does not rest on the fact that the
former has nothing to do with our mental activities. For Kant the opposi-
tion resides, rather, in the normative character of logic, as opposed to the
descriptive character of empirical psychology. But even the normative char-
acter of logic is, in the end, the expression of the normative capacity of our
minds . . .110
70
Faculty talk is normative in that it analyses how thought should

operate, rather than describing how we in fact think. Thus there is no
need to equate Kant’s position with faculty psychology. This is an
important corrective not only to Strawson, but also to Guyer and
Pippin, whom I have argued suggest indirectly that Kant’s faculty talk
contributes to his supposed impositionalism. Longuenesse here takes
up a theme that Allison and Buchdahl try to avoid.
Longuenesse remarks on the particular alliance between form and
general logic insofar as the latter ‘deals with the mere form of
thought’.111 The form of thought is set in opposition to the matter or
object of thought, that is, its content. Formal logic is concerned only
with the combination of concepts regardless of their content – in con-
trast to transcendental logic, which is concerned with the content of
thought, at least at its a priori level.112 Longuenesse’s account of the
status of content is persuasive. She insists that transcendental logic is
concerned with the way in which our mode of sensibility forms empir-
ically given data and not with the mechanics of physical affect.113
Longuenesse also says that both transcendental and general logic are
pure insofar as they ‘borrow nothing from psychology’.114 We can
conclude that both levels of logic are concerned only with the nor-
mative dimension of thought insofar as they deal with the a priori
form of thought.
Form implies ‘universality insofar as by virtue of this form alone
concepts can be combined’.115 Prima facie it seems difficult to link this
statement with aesthetic form. Admittedly Longuenesse’s account
echoes Kant’s view that aesthetic form is the condition of an intu-
ition’s unification under a concept. But the directness with which she
links form and conceptualisation risks losing sight of what
Longuenesse herself has recognised as a collaboration between two
forms.
The relative bias toward conceptual form is further evidenced by
the way in which she contrasts conceptual function to sensible affec-
tion, finally claiming that ‘form’ and ‘function’ are used interchange-
ably, distinguished only by the latter term capturing the process of
which the former is the result.116 She is clearly right in attributing to
Kant the view that function and conceptual form are equivalent
terms. The question is: is sensible form distinct from function? If form
and function are equivalent and function is distinct from affection,
then where is the place in her account for the latter? The suspicion
arises that not just affection as a mechanical or physical process, but
affectivity per se has been moved outside the bounds of her
71
philosophical horizon. Agreed, the form of sensibility is a priori and

thus is not given, being rather the form of the given. Its a priori status
reveals that it arises from the subject, but Kant makes a distinction
between the unifying forms of the understanding and the receptive
forms of sensibility. The forms of intuition make possible our capac-
ity for sensibility, although, admittedly, this is a peculiar ability to
take in or receive. The distinction between the two forms of experi-
ence is constitutive of the possibility of experience.
The link Longuenesse makes between function and form is
explained by her chief thesis of synthesis speciosa. Longuenesse
argues that the same function of understanding is the source for the
forms of sensibility and for the categorical forms. The initial analysis
of that general ability of understanding – Vermögen zu urteilen or the
capacity to judge – is laid out in the ‘Table of Judgements’.
Longuenesse could have claimed simply that the ‘Table of
Judgements’ anticipates the ‘Table of Categories’ by laying out a
sketch of the general form of thought necessary for any judgement.117
It would then have been possible to combine this with the claim that
judgement requires a source independent from thought, that is, our
capacity for taking up the sensible given. But Longuenesse wants to
say that even the latter is governed by the Vermögen zu urteilen. ‘His
goal is to argue that these same functions by means of which judg-
ments are formed (by analysis of the sensible given), first guide syn-
thesis of sensible manifolds, with a view to analysis.’118 Synthesis
relies on imagination or sensibility, but ‘in order to subsume a sensi-
ble manifold under concepts, one first needs to combine it in such a
way that it is recognizable under concepts’.119 This bringing to con-
cepts presupposes a unity of synthesis.
In Longuenesse’s Kant and the Capacity to Judge this leading idea
is identified as the thesis of synthesis speciosa. The categories play a
double role: first preparing the sensible given for determination by
introducing unity, and then determining what they have already
unified:
the categories have a role to play ‘at each end’ of the activity of judging:
as mere ‘logical functions of judgement’, they play a role in guiding reflec-
tion of the sensible given with a view to forming empirical concepts, and
for this they must first also guide synthesis in view of reflection. But as con-
cepts, they are applied only in judgements of experience.120
Elsewhere it emerges that the goal of applying the categories within

judgements of experience is what first gives the analysis of the logical
72
functions of judgement its role as ‘guiding thread’. This is a function-

alist argument. The application of the forms of thought as categories
within experience legitimates the presupposition of them at the outset
of Kant’s argument. Were this the whole story it would be persuasive
and count as a helpful expansion of Allison’s assessment of the role
played by the ‘Metaphysical Deduction’. However, as the passage
just cited reveals, Longuenesse’s position is more radical than this,
for she is committed to the view that the categories in their initial
guise as logical functions of judgement guide not only thought but also
sensibility.
Double accounting for the role of the categories is the key to
Longuenesse’s account. It is also the key to grasping how she risks
reducing dualism to a monism of a broadened understanding where
the original distinction between two forms is traced back to one form
or function, that of thought or judgement in general.121 The form of
intuition has been transformed from the basis of our affective access
to objects in the world into a subsidiary part of the internal appara-
tus of representation.
The standard reading is that conceptual synthesis alone achieves
the unity characteristic of determining judgement or knowledge.
Nevertheless, this unity is only achieved insofar as sensibility com-
bines the given within the forms of space and time. ‘Synthesis of
apprehension’ is a distinct, though not a discrete, necessary condition
of the unification of a sensible given under a concept that qualifies as
knowledge. This account does not entail a commitment to intuition
achieving a unity prior to that of the understanding. The holding
together achieved in intuition is a component in the process of unifi-
cation only finally completed in a concept.122 The standard account
also holds that the forms of intuition, in Allison’s sense of forms of
the intuited, are indeterminate until they are explicitly unified or
determined under a concept.123
Longuenesse disagrees with this position. Crucially, she holds that
the intuitions of the ‘Transcendental Aesthetic’ are already unified. 124
Kant, she claims, rarely speaks of form as the mere capacity to receive
representations, and, retrospectively, in Section 26, we discover that
in the ‘Transcendental Aesthetic’ intuitions already qualify as formal
intuitions, that is, they have content. What in the ‘Transcendental
Aesthetic’ are called forms of intuition are in fact formal intuitions and
arise from the pre-formation of sensibility by the understanding.125
The claim that forms of intuition arise from the Vermögen zu
urteilen derives from Longuenesse’s reading of Section 26 in the
73
second part of the B ‘Deduction’. The second step there, as she

reconstructs it, is that the forms of space and time ‘being themselves
unified intuitions, are under the transcendental unity of apperception,
which is the source of the categories’.126 Longuenesse goes on to
explain an ambiguity in Kant’s position in that he claims both that
space and time are determined by the understanding and that they are
prior to concepts.127 She resolves this apparent paradox in the fol-
lowing way. The forms of space and time are prior to any determinate
concept and to the categories in their second role as reflective univer-
sal representations, but they are not prior to the first exercise of
understanding as synthesis speciosa, that is of a self-affection of the
understanding on our sensibility. This is an ingenious way of explain-
ing what might otherwise appear to be a non sequitur at B160.
Despite its initial plausibility, I believe it purchases the consistency of
Kant’s position at too great a price.128
Longuenesse cites another passage in support of her reading:
The same function, which gives unity to the various representations in a
judgment also gives unity to the mere synthesis of representations in an
intuition . . . The same understanding . . . brings by means of the synthetic
unity of the manifold in intuition a transcendental content to its repre-
sentations.129
Longuenesse reads this passage as establishing that the same logical

function gives unity to concepts in a judgement and to the synthesis
of intuition.130 In a sense this is incontestable, while in another sense
it is highly contestable. Most readers of the Critique would agree that
insofar as a sensible given is determined as to its unity, synthesis by
the understanding is necessary. Many interpreters, although not all,
hold that this determined intuition is what Kant means when he talks
of a formal intuition as opposed to a form of intuition, where the
latter is a pure term and the former a hybrid one.131 But most would
not agree that the fact that the unity or unification of intuition can be
traced to the understanding means that the latter is the source of the
form of intuition. Instead, the usual interpretation would be that the
application of the unifying power of understanding to the indepen-
dent capacity for sensibility gives rise to knowledge. And there could
be no independent capacity for sensibility, should the form of the
latter ultimately resolve into an activity of understanding.
The problem as to the identity of the faculties reveals a substantive
problem about Longuenesse’s characterisation of Kant’s account of
experience. If the understanding – even in its broader sense as intellect
74
in general – is ultimately the source of the form of sensibility, then it

is difficult to maintain that the latter is affected by something other
than the mind. We would be very close to saying that the data of sen-
sibility has its source in the mind itself. Of course, Longuenesse
strictly says only that space and time as unified forms fall under the
understanding. However, she gives no serious role to the forms of
intuition prior to unification with the result that the logic of her
argument is that intuition does not so much cooperate with under-
standing, as count as an outpost of the latter. While Longuenesse is
right to stress that for Kant there is no absolute given, there is just as
much danger in suggesting that there is nothing other than construc-
tions of the mind.132 The issue here is not the mechanics of physical
affect, which Longuenesse is quite right to exclude from Kant’s tran-
scendental account, but rather the very possibility of an affective
element at the experiential or phenomenological level. For this we
need a faculty, namely intuition, as the ground of the possibility of
affection. And we still need to know how representations relate to
objects at this level.
Longuenesse does not altogether omit affection from her discus-
sion, but she does underplay its importance. She rightly insists that
Kant rearticulates ‘the manner in which things are given to us’.133
There is no brute given in Kant’s account of sensibility, but only a
given that is formed in space and time. Longuenesse characterises the
situation as one of a ‘transformation’ from outer affect to intuition.134
It would appear that outer affect is the physical cause of an experi-
ence, but in order for it to count as anything for us, it must be taken
up as a representation through the synthetic power of imagination.
And is this not Kant’s position? Longuenesse has argued that the form
of sensibility is an exercise of the understanding, not that the matter
of experience is. She is committed not to internal realism but to indi-
rect realism.135 The primacy afforded to the understanding is at the
level of representation and not of the represented.
But Longuenesse’s thesis of synthesis speciosa turns sensibility
towards the internal system of representation, giving no account of our
capacity for receptivity of the represented. This is particularly evident
in the contrast she draws between the position she attributes to Allison
and her own. While, for the former, what she calls the ‘“internalisation
of the object to representation” would simply mean internalisation to
the point of view of (human) representing subjects’, her own preferred
position is that ‘objects “as” objects-of-representation are appearances
“in us” ’. She concludes that ‘things do exist in themselves outside us,
75
but we perceive them only by means of states of consciousness’.136

Allison’s epistemically formalist reading has been replaced by a newly
tuned version of what he calls ‘the standard picture’. Objects, as
things-in-themselves, stand hidden behind representations. Under this
description, representations cannot count as epistemic conditions, that
is, as our modes of access to objects.
A further problem arises for Longuenesse’s interpretation of Kant’s
epistemology in relation to his account of aesthetic judgement.
Although addressing this issue may seem to take us onto quite differ-
ent terrain, Kant claims that aesthetic judgement reveals the condi-
tions of ‘cognition in general’, a claim which entails that a successful
interpretation of his epistemology must also be able to account for
aesthetic judgement, as discussed in the Critique of Judgement.137
Consideration of this issue allows us to reconsider the first stage of
Longuenesse’s argument, namely, the view that apprehension already
entails the determining power of understanding.
Longuenesse’s thesis of synthesis speciosa entails that under-
standing ‘appropriates’ imagination in contrast to the view that the
latter is the mediator between sensation and intellect.138 She speaks
of the imagination in this context and not of sensibility, for it is the
former in its guise as productive imagination that generates the
forms of intuition.139 She concedes that there is also an activity of
imagination that is not appropriated by the understanding, but goes
on to say that Kant is only marginally interested in this ‘merely sen-
sible/ associative/ fictional activity of imagination’.140 What Kant is
really interested in, she believes, is the rule-governed activity of
imagination at the behest of the understanding exercised in the
interests of a cognitive goal.
But in the Critique of Judgement, Kant focuses on a free exercise
of the imagination that is autonomous and most definitely not appro-
priated by the understanding. Aesthetic judgement arises as a
‘harmony of the cognitive powers’ in which the faculties of imagina-
tion and understanding combine with one another in the absence of
a rule.141 Moreover he argues that the ‘free lawfulness’ of the imagi-
nation, that is, the latter faculty exercised beyond the legislative
authority of the understanding, reveals the subjective conditions of all
judgement or ‘cognition in general’.142 Thus the unappropriated
imagination turns out to be far from marginal to Kant’s project and
to his cognitive project in particular. Later I will take up the relation
between appropriated and unappropriated imagination. For the
moment, I simply want to show that if Kant’s characterisation of
76
imagination as ‘free’ in aesthetic judgement is later shown to be defen-

sible, then Longuenesse’s account of synthesis speciosa is put in
further question.
The contribution of aesthetic judgement is marked by the fact that
it displays a free harmony between imagination and understanding
and that it does not give rise to cognition. This would not be possible
if imagination were already appropriated by understanding. If imagi-
nation is always already directed by the Vermögen zu urteilen, then it
is difficult to see how we could employ our imagination without being
oriented towards a cognitive goal. 143 Judgement (urteilen) is, for
Longuenesse, cognitive judgement. In later chapters of this book I will
challenge this equation.144 From her perspective, aesthetic apprehen-
sion can only be a failed or, at least, a not yet achieved cognition.
Allison also suggests that Longuenesse is wrong to characterise reflec-
tive judgement as ‘failing to reach determination under a concept’.145
Longuenesse’s account differs in many important respects from the
impositionalist one: namely, in her sensitivity to the relations between
the faculties, in her sympathy for formalism and her insistence on the
reciprocity between matter and form, and in her subtle reading of
understanding as involving different levels of operation – to name but
a few. Nevertheless, the internalisation of the relation between mind
and world effected by the thesis of synthesis speciosa indirectly brings
about the same consequence as impositionalism. The understanding
comes to be the source of order among all sensory data. In a sense she
goes even further than impositionalism insofar as she makes sensibil-
ity not just the recipient of cognitive activity, but a colonised outpost
of the latter.
Longuenesse insists she is using the term ‘understanding’ in a
broader sense than usual. The general capacity for judgement
analysed in the ‘Table of Judgements’ is not simply that of the deter-
mining understanding. It is, rather, that of intellect in general, cover-
ing determinant judgement (‘understanding’ in a narrow sense),
reason and judgement. This distinction, she thinks, will prevent her
from falling into the trap allegedly fallen into by Kant himself, that
is, of making the whole project of the first Critique subservient to
determinant judgement or understanding in the restrictive sense.146
Longuenesse hopes to open up the breadth of intellect and the inter-
connections between different branches of it. One example of her
success in this project is the way in which she has revealed the neces-
sity for reflection within the scope of determining judgement.147
Nevertheless the synthesis speciosa argument does not so much
77
achieve her intended goal as put it at risk. While not all judgements
are cognitive, they are all governed by the Vermögen zu urteilen, which
sets a broadly cognitive end for experience.148 Longuenesse’s goal of
identifying the full range of judgement could be attempted in another
way, by contrasting between a narrower and wider sense of reason,
where the latter includes the whole range of the three Critiques.149
Cognition would then be situated within the systematic framework of
reason without necessitating that all judgement is cognitive. This
would also allow for what I call a pluralist account of the faculties in
which they express distinct orientations open to us in experience while
only ever being experienced as cooperating within experience.
Conclusion
I hope to have established in this chapter that Buchdahl, Allison and
Longuenesse each offer a highly sophisticated defence of Kant’s formal-
ism. In their subtly different accounts, they rebut the charge that Kant
offers a merely subjective epistemology, resting on a questionable faculty
psychology. However, I have also argued that, due to their inattention to
the role played by affect within experience, they risk falling back into
impositionalism. Buchdahl and Longuenesse, in particular, are in danger
of suggesting that Kant is trapped in a circle of representation, while
Allison gives an insufficiently dynamic account of the relation in which
matter stands to form. In the next chapter I commence my own account
in which the affective or aesthetic side of experience is shown to be nec-
essary for formalist idealism. In later chapters and in contrast to
Buchdahl and Allison in particular, I will argue that it will only be pos-
sible to establish that formalism escapes the charge of subjectivism once
the subjective element in knowledge is given its proper due and revealed
as a necessary part of the project of establishing objectivity.
Notes
1. Gerd Buchdahl, Metaphysics and the Philosophy of Science: The
Classical Origins Descartes to Kant (= MPS).
2. See Béatrice Longuenesse, Kant and the Capacity to Judge (= KCJ),
p. 223, n. 21. It is also worth noting that Graham Bird published a
quite alternative interpretation to that of Strawson in 1962. See his
Kant’s Theory of Knowledge. Bird rejects the two worlds interpreta-
tion of Kant and should have been recognised as a viable alternative to
Strawson. While Allison cites Bird favourably, he does not enter into a
78
detailed exchange with an account he considers to be close to his own,

but from which he differs ‘on many issues’. See Allison, KTI, p. 333 n.
9; see also p. 334, n. 12. Buchdahl does not reference Bird in MPS and
neither does Longuenesse.
3. For illumination on how Buchdahl uses the term ‘methodological’, see
MPS, pp. 511–12.
4. My reconstruction of Kant’s broader epistemological argument in
Chapters 6 and 7 is deeply influenced by what I call Buchdahl’s ‘hier-
archical’ reading of Kant.
5. Buchdahl, MPS, p. 483, n. 1. Buchdahl is comparing Kant’s position
to that of his predecessors discussed in previous sections of MPS.
6. For Buchdahl’s account of the Second Analogy, see MPS, pp. 648–51.
For his account of the relation between transcendental and empirical
analyses of causality and his critique of competing positions, including
Strawson’s, see pp. 651–65.
7. Buchdahl, MPS; see pp. 489, 490 and 494.
8. For a discussion of the relation between these, see Buchdahl, MPS,
pp. 501–2.
9. Buchdahl, MPS, p. 501. For an extended discussion of the relation, see
pp. 651–65.
10. Buchdahl, MPS, pp. 592–3. See also his Kant and the Dynamics of
Reason (= KDR), pp. 183–4. It is Buchdahl’s view that regulative prin-
ciples are not needed for the empirical experience of an object, but only
for the lawful relations between empirical objects. Contrast my
account in Chapter 7, pp. 254–5.
11. Buchdahl, KDR, p. 56.
12. Ibid., p. 56.
13. Ibid., p. 170.
14. Ibid., p. 56.
15. Ibid., p. 205.
16. Ibid., p. 204.
17. Buchdahl, MPS, p. 494. Note again that, for Buchdahl, general ontol-
ogy establishes the conditions of possibility for an individual material
thing. I will argue in Chapter 7 (pp. 249–55) that full determination
of the possibility of a material empirical thing is not supplied by con-
stitutive principles alone.
18. Buchdahl, MPS, p. 629 (discussed below).
19. Buchdahl, KDR, p. 34 (my addition). EC is Buchdahl’s abbreviation
for Special Ontology. This stands for ‘Explicative Component’. See
KDR, p. 25, Fig. 1.5.
20. Ibid., p. 90.
21. Ibid., p. 89.
22. Ibid., p. 34. I explain the idea of a framework yielding a phenomenol-
ogy below. Buchdahl calls this the process of realisation.
79
23. Ibid., p. 34.

24. Ibid., p. 62.
25. Buchdahl, MPS, p. 638.
26. Buchdahl, KDR, p. 62. Buchdahl refers to CPR, A 266, B 322.
27. Buchdahl, MPS, p. 632, n. 4.
28. Buchdahl, KDR, p. 118. Kant uses the term at CPR, A 143, B 182, in
the ‘Schematism’ in a discussion of the schema of reality and, in par-
ticular, the material component in things, which he equates with things
in themselves, Sachheit and reality. Kemp Smith, who translates
Sachheit as ‘thinghood’, inserts a ‘not’, negating the sense of the orig-
inal as stated in the Akademische Ausgabe version of both editions.
Buchdahl remarks that this mention of Dinge an sich refers to the sense
of the latter phrase as Transcendental Object, or To. Compare
Allison’s interpretation of transcendental object as the material dimen-
sion in the object discussed in the next section pp. 65–7.
29. Buchdahl, MPS, p. 639. In the first chapter of Charles Dickens’ The
Pickwick Papers a row develops between Mr Pickwick and another
member of the Pickwick Club, a speculative debating society. Other
members are horrified that one member should accuse the other of
being a humbug, while the accuser refuses to withdraw this insult.
Nevertheless, calm is restored when Mr Pickwick’s opponent clarifies
that he used the term in its Pickwickian, not its common, sense.
Clearly, to label something Pickwickian is to say that its usage is
limited to a restricted technical discussion and is about as relevant to
our everyday concerns as Mr Pickwick’s seminal dissertation entitled
‘Speculations on the Source of the Hampstead Ponds, with some
Observations on the Theory of Tittlebats’ – or, as some might say, it is
‘merely academic’!
30. I give a quite different account of the philosophical importance of
affection in Chapter 3.
31. Buchdahl, KDR, pp. 53–103, especially p. 66.
32. Ibid., pp. 79 ff.
33. Buchdahl frequently mentions a question of balance. See, for instance,
MPS, pp. 608 and 618.
34. Buchdahl, MPS, p. 629 (Buchdahl’s emphasis).
35. Ibid., p. 629.
36. Ibid., p. 629.
37. See my insistence on the two-sided character of sensibility in Chapter
3, pp. 105–8.
39. Ibid., p. 44. He refers to CPR, A 255, B 311.
40. Ibid., p. 73. Buchdahl refers to Bird, Kant’s Theory of Knowledge,
pp. 68–9 and pp. 76–81. Compare Allison’s account of the transcen-
dental object discussed on pp. 65–7.
80

42. Ibid., p. 118 (Buchdahl’s emphasis).
43. Ibid., p. 119. See also p. 132 on the ‘misleading empirical model of
an external world acting on the subject’. There is nevertheless a ‘non-
subjective element’ in experience, namely that we find ourselves with
such-and-such sensations or representations.
44. Ibid., p. 149.
45. Ibid., p. 75. Just to be clear, I include this characterisation of Fichte
strictly to illuminate Buchdahl’s position. In no sense should this be
taken as an attempt to characterise the relation between subjectivity
and objectivity for Fichte.
47. Ibid., p. 639. Buchdahl thus distinguishes his position from Strawson’s
critique of faculty talk.
48. Ibid., p. 640. This distinction is extremely important for both Allison
and Longuenesse.
49. Ibid., p. 640.
50. Below I will suggest that Buchdahl does not render this distinction as
clearly as he intends.
51. Buchdahl, MPS, p. 593. In my own account, my aim is to demonstrate
the intertwinement of the subjective and objective trajectories of crit-
ical idealism.
52. Ibid., p. 640.
54. Ibid., p. 170.
55. Buchdahl, MPS, p. 640. See above. Moreover, I will argue in Chapter
7 that the system of nature has an objective trajectory insofar as it is
necessary for empirical knowledge of objects.
60. Ibid., p. 624.
61. I develop this argument in Chapter 6.
62. Henry Allison Kant’s Transcendental Idealism was first published in
1983. A revised and enlarged edition appeared in 2004. It has not been
possible to incorporate an assessment of the new edition in this work,
but I believe that the issues addressed here are not substantially
changed in the new edition. In other respects, however, the second
edition offers so many developments in Allison’s thinking that Pippin’s
remark cited on the dust jacket of the 2004 edition is just right, namely,
that ‘the result is almost a new book altogether’. It is tempting to think
that this is Allison’s B edition!
63. See Allison, KTI (1983), p. 97, on the indeterminacy of space (2004),
81
p. 116; (1983), p. 231 on the second Analogy; and p. 234 on the

importance of regulative principles.
64. See Chapter 1 for a discussion of Pippin.
65. Allison, KTI (1983), p. 4; (2004), pp. 38–42.
66. Ibid., p. 11; (2004) p. 11.
67. Ibid. (1983), p. 26. Allison refers to the appendix to the Prolegomena
Ak. IV, 375, and to CPR, B 519. See (2004), p. 35.
68. Allison, KTI, p. 26 (Allison’s emphasis); (2004), p. 35.
69. Ibid. (1983), p. 27; (2004), p. 37.
70. Ibid. (1983), p. 61. The Antinomies are given an extended treatment
in the new edition, see pp. 357–95.
71. Ibid. (1983), p. 106. Allison also recasts this argument radically in the
2004 edition, see pp. 122–8.
72. Ibid. (1983), p. 111. Another passage important for establishing the
formalist status of Allison’s interpretation is found on p. 106: ‘a form
of appearance is a feature of the appearance in virtue of which its ele-
ments are viewed as ordered or related to one another in experience’.
As noted above, this section is altered in the new edition. See (2004)
pp. 122–8.
73. Ibid. (1983), p. 122. For a revised account of ‘discursivity’ and the
conceptual side of dualism, see (2004) Chapter 4, pp. 77–96.
74. Ibid. (1983), p. 129; (2004), p. 156.
75. Ibid. (1983), p. 195; (2004), p. 225.
76. Ibid. (1983), pp. 226 and 216. Allison’s analysis of the Analogies is
greatly extended in the new edition, where he endeavours to show the
interconnectedness of all three principles of time determination. See
(2004) pp. 229 ff.
77. Ibid. (1983), pp. 230–1; (2004), pp. 255–6.
78. Pippin, KTF, p. 23. See Chapter 1, p. 41.
79. Allison, KTI (1983), pp. 5–6; (2004), pp. 4–6.
80. Ibid. (1983), p. 7. See (2004), p. 14 ‘our intuition and, more generally,
that of any finite cognizer, must be sensible, that is, receptive, resulting
from an affection of the mind by objects’.
81. Ibid. (1983), p. 30. For a revised version of the relation in which Kant
stands to Berkeley, see (2004), pp. 38–42.
82. Ibid. (1983), p. 27; (2004), p. 37.
83. Ibid. (1983), p. 30.
84. See (1983) Ch. 11, especially pp. 247–54. For the revised version in
the new edition, see (2004), Ch. 3 pp. 50–73, especially pp. 64–73.
85. Ibid. (1983), p. 250; see (2004), p. 67 ‘the object remains completely
indeterminate’.
86. Ibid. (1983), pp. 250–2; (2004), pp. 71–3.
87. Ibid. (1983), p. 253. Allison prefers to refer to empirical matter and its
transcendental ground, rather than transcendental matter. However, in
82
both editions he notes that Kant employs the latter expression on occa-
sion. See, for instance (2004), p. 72.
88. Allison’s account of the transcendental significance of the transcen-
dental object can be compared to Gerold Prauss’s interpretation of phi-
losophy’s role in relation to the natural sciences. For both Allison and
Prauss, transcendental reflection allows us to grasp what cannot be
understood at the empirical level. Prauss describes transcendental phi-
losophy as the ‘non-empirical science of the empirical’. See Prauss,
Kant und das Problem der Dinge an sich, pp. 205–27 and in particu-
lar pp. 212–13. Allison, however, distinguishes his position from
Prauss’s, which he sees as rendering unintelligible the very concept of
a non-empirical affection. See Allison, KTI (1983), pp. 250 and 364,
n. 26; (2004), pp. 68 and 461.
89. Allison, KTI (1983), p. 249. For a revised discussion of these, see
(2004) pp. 51–64.
90. Ibid. (1983), p. 249; (2004), p. 67.
91. Ibid. (1983), p. 250. In the new edition, the ground of matter is aligned
with things in themselves, see (2004), p. 72. This, however, is quite
consistent with Allison’s account of the relation in which the tran-
scendental object stands to things in themselves.
92. Ibid. (1983), p. 250 (Allison’s emphasis); (2004), p. 67.
93. Ibid. (2004), p. 72. See also his claim on p. 73 that all talk about
things as they are in themselves, noumena and the transcendental
object, that is, the ground of matter, counts as technical. This sounds
very like Buchdahl’s claim that the given is strictly ‘Pickwickian’ in
status.
94. See Chapter 5 and subsequent chapters.
95. ‘On naturalizing Kant’s transcendental psychology’, in Allison,
Idealism and Freedom, p. 58.
96. Ibid., pp. 65–6.
97. ‘Gurwitsch’s interpretation of Kant’, ibid., p. 74. This article was first
published in Kant-Studien in 1992.
98. Ibid., see pp. 72 and 77–8.
99. Ibid., p. 78.
100. Allison, KTI, p. 97.
101. Ibid., p. 97.
102. Pippin, KTF, p. 23. See Chapter 1, p. 13.
103. Longuenesse, Kant and the Capacity to Judge (= KCJ), p. 18.
104. Longuenesse, ‘The Divisions of the Transcendental Logic and the
Leading Thread (= ‘Leading Thread’).
105. CPR, A 66, B 91, and A 67, B 92. Kemp Smith translates Leitfaden
simply as ‘clue’.
106. Longuenesse, ‘Leading Thread’, p. 131.
107. Ibid., p. 131.
83
108. Ibid., p. 133. In this and subsequent passages, the emphasis is

Longuenesse’s.
109. Ibid., p 133.
110. Ibid., p. 135.
111. Ibid., p. 134. Her reference is to A 54, B 78.
112. Ibid., p. 136.
113. Ibid., p. 136.
114. Ibid., p. 135.
115. Ibid., p. 136.
116. Ibid., pp. 139 and 143.
117. This is Allison’s position, as we saw in the previous section.
119. Ibid., p. 152.
120. Longuenesse, KCJ, pp. 199–200, n. 3. See also p. 196.
121. Understanding is ‘broadened’ because it now includes imagination and
is even the source of the forms of sensibility. See Longuenesse, KCJ,
p. 63, where understanding is identified as ‘the unity of apperception
in relation to the synthesis of imagination’.
122. See my discussion of the conditions of unification of objects as pre-
sented in the A edition of the ‘Deduction’ in Chapter 4.
123. See, for instance, Allison, KTI, p. 97.
124. Longuenesse, KCJ, p. 221. Longuenesse compares her position to
Heidegger’s. Indeed Heidegger already identifies the problem of the
source of synthesis speciosa (see Heidegger’s Kant and the Problem of
Metaphysics). But his solution is to deny that Kant subsumes imagi-
nation under understanding, preferring to construe the two faculties as
standing in ‘the rich totality of a complex activity’. See especially
pp. 67–8 (German edition p. 61).
125. Longuenesse, KCJ, p. 223. See also her interesting claim that the forms
of intuition are really potential forms (p. 221). A very similar idea is
to be found in Heidegger’s Kant and the Problem of Metaphysics,
p. 149 (German edition p. 137) where he cites Kant’s statement that
‘space and time are the pre-formative forms [Formen der Vorbildung]
in pure intuition’. He references Erdmann, Reflexionen II, 408, Kant’s
Posthumous Works in Manuscript Form, Vol. V, No. 5934.
126. Longuenesse, KCJ, p. 214 (Longuenesse’s emphasis).
127. She is referring to Kant’s famous footnote at B 160–161.
128. I give an alternative reading of the crucial footnote at B 160–161 in
Chapter 4.
129. Longuenesse, ‘Leading Thread’, p. 153. CPR, A79, B 105
(Longuenesse’s emphasis).
131. See Allison, KTI, p. 97.
132. Longuenesse, ‘Leading Thread’, p. 154: ‘so no object for a concept is
84
simply “given” ’. See, in this regard, the attack on the ‘myth of the
given’ in John McDowell’s Mind and World.
133. Longuenesse, KCJ, p. 213 (Longuenesse’s emphasis).
134. Ibid., see pp. 213 and 220.
135. Ibid., p. 20, n. 9.
136. Ibid., p. 20, n. 9.
137. For instance, Critique of Judgment, AA 218. See discussion in Chapter
5.
138. Longuenesse, KCJ, p. 207.
139. Ibid., p. 206.
140. Ibid., p. 208.
141. CJ, AA 218.
142. CJ, AA 240, is a development of the account at AA 218.
143. My account is in direct opposition to the interpretation offered by
Avner Baz in ‘Kant’s Principle of Purposiveness and the Missing point
of (Aesthetic) Judgements’, who relies on Longuenesse, among others,
for his reading of Kant.
144. See, in particular, Chapters 5 and 7.
145. Allison, Kant’s Theory of Taste, pp. 353–4, n. 2. The reference is to
Longuenesse, KCJ, p. 164, n. 47.
146. See Longuenesse, KCJ, pp. 203–4, for a very interesting, acute critique
of Heidegger’s reading of Kant. See also my discussion of Heidegger in
Chapter 4.
147. See, for instance, Longuenesse, KCJ, p. 112, n. 17.
148. This is what encourages Baz’s reception of her interpretation. See Baz,
‘Kant’s Principle of Purposiveness’, p. 14.
149. See discussion of the tripartite supersensible in the penultimate section
of Chapter 8, pp. 299–302.
85
Formal Idealism and the Aesthetic Condition

of Experience
In the previous chapter I have identified a problem in the accounts

given by some of Kant’s principal defenders of the relation between
representations and objects. While Buchdahl, Allison and Longuenesse
provide a robust defence of Kant’s formalism, each of their accounts
fail to give an adequate account of the relation in which form stands
to matter. I will now offer an alternative account of the relation
between representation and material given.
In the first section of this chapter, I discuss how Kant’s characteri-
sation of his epistemological project as amounting to formal, as
opposed to material, idealism reveals his commitment to an extra-
mental world of objects. Only the form of experience is initiated by
our minds.
In the second section, I turn to Kant’s ‘Copernican revolution’,
which has often been taken to be a corner stone of his commitment
to the view that mind imposes order on the world. I suggest that it
need not be read in this way, because of a limitation on the range of
its scope and also because it comprises two stages, the first of which
concerns our receptivity to an affect given in experience.1
Next I examine some of the passages in the ‘Transcendental
Aesthetic’ that most invite an impositionalist reading and suggest an
alternative perspective according to which representation is our mode
of access to an extra-mental given. Having established this in princi-
ple in the third section, in the fourth I argue that the account of affec-
tivity such a perspective would require is in fact provided in the
opening paragraphs of the ‘Transcendental Aesthetic’.
In the fifth section I give a resumé of the complex identity of recep-
tivity that has emerged. Sensibility is dual-faced: its mental component
being the formal condition of our capacity to be affected by something
given in experience. This is the first and crucial step in taking seriously
the subjective dimension of Kant’s epistemology, while not falling into
the impositionalist readings examined in Chapter 1 and insufficiently
86
Formal Idealism and the Aesthetic Condition of Experience
parried by those interpretations discussed in Chapter 2. Nevertheless,

Kant did not provide an adequately forceful account of the affective
side of sensibility, often leaving the impression that the a priori forms
of intuition were not so much modes of receptivity, as that they
actively shape matter. This would, however, be a curious result as sen-
sibility is also frequently characterised by Kant as ‘passive’. In fact, the
truth lies somewhere between these extremes in the receptive capacity
to be open to something other than ourselves.2
I Transcendental Idealism as Formal Idealism: The

Refutation of Material Idealism
In a footnote Kant makes the following remark about transcendental
idealism: ‘I have also, elsewhere, sometimes entitled it formal idealism,
to distinguish it from material idealism, that is, from the usual type of
idealism which doubts or denies the existence of outer things them-
selves.’3 The final phrase in this passage may appear to suggest that
Kant here commits himself to the existence of things-in-themselves
(Dinge an sich), that is, things regarded without any reference to
human experience of them. However there is no mention of Dinge an
sich, but only of außere Dinge selbst. The ontological status of these
outer things is left open by the latter phrase and does not necessarily
commit Kant to a two-worlds ontology of appearances and things-in-
themselves.4 Kemp Smith omits the subsequent sentence: ‘In many
cases it seems advisable to use this [i.e., ‘formal idealism’] rather than
the latter expression [i.e., ‘transcendental idealism’], in order to avoid
all misinterpretation.’5 At times even Kant seems to lose sight of this
important advice. My intention is to rearticulate the formal status of
transcendental idealism and show how we can defend the latter
against the charges of impositionalism and subjectivism.
Kant’s point in this crucial note is that, in contrast to his own posi-
tion, material idealism ‘doubts or denies’ that there are independently
existing objects. Who are Kant’s opponents here? Certainly he has
within his sights Berkeley, who denied the external existence of
objects and insisted that the latter are only ideas in the mind.6 Formal
idealism, in contrast, does not deny the extra-mental existence of
objects. But no more is it committed to the extra-experiential exist-
ence of things-in-themselves. It must be that ‘outer things themselves’
in the passage under consideration refers to objects independent of
the mind, but nevertheless experienced by us. These ‘außere Dinge
selbst’ are accessible to the mind, but are not projected by it. Formal
87
idealism preserves a crucial role for the mind as objects are only iden-
tified as such insofar as they have a certain structure or form, which
is mind-dependent. Formal idealism is, however, contrasted from the
material variety insofar as the former does not reduce all experience
to the activity of mind.
In the ‘Refutation of Idealism’, ‘material idealism’ refers not only
to Berkeley but also to Descartes.7 Kant contrasts the dogmatic ide-
alism of Berkeley to the problematic idealism of Descartes. The latter
species of idealism is sceptical, or at least uncommitted as to the inde-
pendent existence of extra-mental objects. It ‘doubts’ though does not
‘deny’ the existence of outer things. All that is certain is the Cogito.
It is only with the additional premise of a beneficent deity that our
certainty in the existence of external objects can be reasserted. In con-
trast, for Kant, any certainty that we have in our activity of thinking
must be based not only on a concept but also on an intuition. In order
to experience myself as thinking, I must do so over time. And, to
experience myself as temporal, is to mark out my experience of self
against a spatial experience of external objects. Thus, Kant concludes,
we simply cannot maintain a certainty with regard to the self and be
sceptical about the existence of external objects. We cannot have one
without the other and it is clear that, for Kant, the only reasonable
conclusion is that our common sense faith both in the existence of the
thinking subject and in external objects is reliable.8 Formal idealism
is thus distinguished both from the Berkeleian and Cartesian versions
of material idealism insofar as it starts from a commitment to the exis-
tence of extra-mental objects. Something must be given to the mind
if experience is to be possible.9
However we have seen in Chapter 1 that had Kant consistently
talked of formal instead of transcendental idealism, his critics would
have been no more convinced. It is the very formal status of his
method that – for Pippin, Guyer and Strawson – results in the con-
clusion that he is committed to subjectivism or the imposition of
mental structures on the external world. In contrast to what they see
as a reduction of empirical experience to an order projected on that
world by the form-giving mind, we could say in our terms, not theirs,
that they insist on the aesthetic or the given.10 For each of them, Kant
fails to account for what is outside the mind and slips back into a too-
easy model of ‘impositionalism’.
Meanwhile the authors we examined in Chapter 2 admit the
importance of mind in Kant’s method, while insisting that what is in
question here is the logical and not the psychological activity of mind.
88
Mind gives structure to objects, but it does not generate them. Thus
far, the approaches taken by Buchdahl, Allison and Longuenesse tally
with Kant’s distinction between formal and material idealism, as I
have read it. As I have argued, however, none of these three com-
mentators focuses sufficiently on the relation in which form stands to
the affective side of experience. This is certainly not to say that they
qualify as material idealists in either of Kant’s senses. They neither
deny nor doubt the external existence of things. Rather, they start
from the presupposition that the aesthetic dimension of experience is
unproblematic, philosophically speaking.
II Why an impositionalist interpretation of the Copernican

Revolution is not necessary
It cannot be denied, however, that Kant’s presentation of his method
at times invites the impositionalist reading developed by his critics.
Famously, in the account of the Copernican revolution in philosophy,
he says that there must be a reversal in the relation between knowl-
edge and objects.11 In this passage Kant consistently uses Gegenstand
rather than Object, suggesting strongly that he is concerned with the
possible experience of extra-mental objects and not merely with con-
structions of thought. Indeed, no revolution would be necessary if he
merely wanted to establish the possibility of access to the latter. If we
are to attain anything worthy of the title of knowledge, then the latter
cannot simply arise from the causal effect of objects on our minds.
The status of the empiricist causal account of knowledge became
explicit in Hume’s claim that causality is nothing other than a cus-
tomary association of certain events we call causes with certain others
we call effects.12 If this model were applied to the case of knowledge
in general, there would be no necessary connection between object
and representation and, as a result, our claims to knowledge would
not qualify for the title. The link between representation and object
that is sought cannot be derived from experience, which as Hume
rightly concluded, gives us nothing but associative patterns. Kant’s
alternative is to suggest that the only way of securing that our claims
to knowledge do, in fact, grasp a necessary connection between object
and representation is if the mind – or, more strictly, our cognitive
capacity – introduces an a priori element.
The reversal Kant achieves is not, however, that now the mind –
or the a priori form arising from it – causes objects. Causal relations
are to be understood strictly within the range of experience. Asking
89
the question as to the cause of experience is to pose an illicit, tran-

scendent and ultimately vacuous question. A priori elements, we
will see, provide the structure or form of experience and not the
cause of the latter. Nevertheless, does this not mean that the mind
imposes its own order on objects? We will see in our discussion of
the structure of the ‘Analytic’ in Chapter 6 that this is a question
that returns to haunt Kant at many levels.13 For now I intend to con-
centrate on evidence internal to Kant’s initial account of the
Copernican revolution in order to show that at this early stage of
his account, we need not leap to conclude that his position entails
impositionalism.
Having begun with the claim that objects must conform to knowl-
edge, Kant proceeds to develop his insight at two levels. In effect, he
separates out two conditions of knowledge, saying first that objects
must conform to the faculty of intuitions:
A similar experiment can be tried in metaphysics [i.e. similar to that of
Copernicus’ reversal of the attribution of movement], as regards the intu-
ition of objects [Gegenstand]. If intuition must conform to the constitu-
tion of the objects, I do not see how we could know anything of the latter
a priori; but if the object (as object of the senses) must conform to the con-
stitution of our faculty of intuition, I have no difficulty in conceiving such
a possibility.14
This is the aesthetic condition of knowledge. Kant asserts that if we
are to understand how we achieve knowledge of objects of the senses,
then we should not construe those objects as the causes of our knowl-
edge. It must rather be the case that the objects conform to our power
of intuition. In the ‘Transcendental Aesthetic’ we will discover what
Kant has in mind. If we are to come to attain knowledge of sensory
objects, there must be something in us that allows the given in sensa-
tion to be taken up, that is, we must be capable of being receptive to
that something.
The second condition is that objects must conform to concepts:
Since I cannot rest in these intuitions if they are to become known, but
must relate them as representations to something as their object, and
determine this latter through them, either I must assume that the concepts,
by means of which I obtain this determination, conform to the object, or
else assume that the objects, or what is the same thing, that the experience
in which alone, as given objects, they can be known, conform to the con-
cepts. In the former case, I am again in the same perplexity as to how I can
know anything a priori in regard to the objects. In the latter case the
outlook is more hopeful.15
90
Knowledge, although not perhaps some lower level of consciousness,

requires that we have not only an intuition, but also a concept. In the
‘Transcendental Deduction’ we learn that the forms of intuition are
only the initial conditions of the combination of the sensible given
which, in turn, facilitates its being unified by the concepts of the under-
standing.16 In the light of this, it is surprising that the second stage does
not say that intuitions must conform to concepts. Intuitions give sen-
sible content for concepts, while objects are the composite of intu-
itions and concepts. Yet in this passage Kant uses not only ‘object’ but
also ‘experience’ as referring to the sensory given that has not yet been
determined or unified by the understanding. Later I will suggest that
Kant uses these terms in both a technical and a non-technical sense.
Kant’s general point is that the extra-mental given in experience
can only be taken up into consciousness and thus qualify as known
by us insofar as we supply an a priori element that comes from mind
alone. There are two elements of form, both aesthetic and conceptual,
and they conjointly allow the unification of the sensory object under
the rules of understanding so as to give rise to knowledge. Insofar as
an object can be experienced – which for Kant is equivalent to its
being knowable – then it must ultimately be subsumable under certain
fundamental concepts that function as rules:
For experience is itself a species of knowledge which involves under-
standing; and understanding has rules which I must presuppose as being
in me prior to objects being given to me, and therefore as being a priori.
They find expression in a priori concepts to which all objects of experi-
ence necessarily conform, and with which they must agree.17
This is the locus classicus of the impositionalist interpretation of

Kant. Knowledge arises neither from our intuitions conforming to the
constitution of objects, nor from our concepts conforming to objects.
It can only arise insofar as objects conform to our concepts. The case
seems to be clear. The mind introduces order into objects and it does
so without any contribution from those objects. It seems difficult to
avoid the conclusion that the mind imposes order on objects.
Before considering that the impositionalist reading may simply be
too quick, we should consider what Kant means by the term ‘object’
in the passage just cited. I have already noted that he uses the term
indeterminately in anticipation of a later, more rigorous usage. But
what in general does Kant intend by the term in this context? It would
seem reasonable to assume that he is referring to experiential or
empirical objects. We can establish this by a via negativa, for the force
91
of these passages is that objects fail to ground knowledge because

they are empirical. A further illumination of the empirical status of
objects comes later in the Preface in the course of a discussion of the
limitation of knowledge in the interests of our practical or moral ori-
entation. Kant makes the point that if our knowledge is of appear-
ances and not of things-in-themselves, then a space is left open for the
thought of the latter within the practical domain.18 We can never have
knowledge of these speculative or moral entities, but we can have
practical faith in them.19 By contrast, the objects of knowledge are
appearances and thus empirical.
In insisting that the objects that we know are ‘appearances’, Kant
is saying that they must stand in relation to mind. They are objects
that appear to us. Now insofar as they are marked by the representa-
tional relation in which they stand to mind, they share the formal
structures characteristic of mind. Only insofar as they are formed by
the structures introduced by our mental powers are they capable of
counting as appearances. But to say as much is not to say that they
are mere projections on the part of mind. They are syntheses of the
material given, which arise from the unifying capacity of mind.
Synthesis must still be effected on a given that precedes mental activ-
ity and as such synthesis is not a mere projection but rather arises
from the intervention of mind within the material world. The form of
objects, admittedly, arises from the mind alone, but only as conditions
for synthesising the given. As I will argue in the next chapter, this syn-
thesis arises as a cooperation of a plurality of different mental orien-
tations in response to an extra-mental given.
We now must consider a limitation on the Copernican Revolution,
before moving on to insist on its full scope. The limitation is that Kant
is here concerned only with the a priori element in knowledge. The
‘objects’ – be they things in themselves or appearances – would be inca-
pable of supplying this dimension of knowledge. Things-in-themselves
are alien to the cognitive process, while appearances count as the
‘undetermined object [Gegenstand] of an empirical intuition’.20
Neither is capable of supplying the element that makes possible the
determining function characteristic of knowledge. Only an a priori
element that arises from the mind makes possible determination or uni-
fication. But this a priori element is merely the form or structure of
knowledge, that is, it provides only the initial conditions of knowledge
and is not the latter’s sole source. Admittedly, Kant’s presentation of
the Copernican revolution seems to suggest a less limited reading where
knowledge arises from the mind alone. However, seen in conjunction
92
with his characterisation of transcendental idealism as ‘formalist’, a

differently nuanced interpretation becomes possible. The a priori forms
supplied by the mind are the initiating conditions with respect to which
any claim to knowledge must arise, but the a priori is operative – and
indeed valid – only insofar as it is set in relation to what is given in
experience. The empirical or a posteriori object arises from the combin-
ation of a priori and the given. The task of achieving empirical knowl-
edge will require much more than the merely formal conditions, both
aesthetic and conceptual.21
Recognising this limitation goes hand in hand with establishing the
full scope of the Copernican revolution. If the structures introduced
by mind are to bring about order in the extra-mental world, mind
must be capable not only of unifying the given through concepts, but
also of being receptive to what is given in experience. This is the
first and aesthetic condition of knowledge that I identified above.
Receptivity also has a formal structure, namely, the forms of intu-
ition. But the forms of intuition allow us to take up something extra-
mental. This hidden side of sensibility has often been underestimated,
resulting in the view that the forms of intuition impose order rather
than make possible our reception of or response to objects. Indeed,
Kant’s statement that objects must conform to intuition seems to
suggest just this. However, Kant’s account of sensibility is more
complex, as I will argue later.22 For the moment I will simply claim
that the first stage of the Copernican revolution would not be distinct
from the second, if the forms of intuition were not receptive rather
than spontaneous. And we will see that the defining characteristic of
intuition is receptivity, that is, our capacity for taking in something
given to us via our senses. If receptivity is a necessary component of
knowledge, then it cannot simply be the case that mind imposes order
on the world. Admittedly the concept brings unity into intuition, but
both steps are necessary for knowledge to arise. This brings to mind
another classical reference point in the Critique of Pure Reason:
‘Thoughts without content are empty, intuitions without concepts are
blind.’23
Once we recognise the limitation of the Copernican Revolution
to the a priori and its full range of reference, expressed in two stages,
both aesthetic and conceptual, it is no longer necessary to read
the passage as advocating impositionalism. Only the form of experi-
ential objects comes from the mind and, moreover, aesthetic form is
a capacity for receptivity to something given, not a projection on
the world. Admittedly, it is possible to read the aesthetic side of
93
transcendental idealism as merely passive and not receptive in any

meaningful sense, but I believe this should be resisted.24 The remain-
der of this work is devoted to bringing out how the formal structure
of transcendental idealism makes possible access to the material but
empirical given.
There are at least three elements necessary for the cognition of an
empirical object.25 First, something must be given to us in intuition;
second, we must be receptive to what is given through the forms of
intuition, space and time; and, third, an a priori concept is required
for determination of the given. In his account of the Copernican revo-
lution in philosophy, Kant has often been taken to be concerned only
with the third element. And even though I have emphasised the neces-
sity of recognising the second element in the first stage of the
Copernican revolution, it must be admitted that at this stage of his
discussion the first element is not positively addressed. This has the
result that the complex structure of sensibility is not made fully
evident, and, until it is, the distinctiveness of the aesthetic and con-
ceptual dimensions of experience cannot be adequately grasped. Later
in this chapter we will see that Kant pays more attention to the first
element of knowledge at the outset of the ‘Trancendental Aesthetic’.
It will then become more apparent that the three elements of knowl-
edge are part of one complex story.
The limitation of the Copernican Revolution to the a priori and its
composition in two stages, both aesthetic and conceptual, has impor-
tant implications for what Kant intends by ‘knowledge’. Insofar as
knowledge has an aesthetic dimension, it is temporal and arises with
regard to spatial objects. The conceptual dimension emerges as a task
of determination of the determinable, that is, of appearances, which
are identified as the undetermined.26 This means that while the
Critique remains an examination of the a priori, it must also give
some account of the way in which the latter provides the form of the
a posteriori. Only the form of experience, or a priori knowledge,
arises from the mind alone and even it must anticipate application to
empirical instances. Empirical knowledge has an even more complex
genesis.27
Were the impositionalist interpretation correct, the remainder of
the Critique would merely be an exposition of the apparatus of mental
projection. Contrary to this, I will argue that the Critique furnishes a
lengthy argument for the complex operation necessary for the possi-
bility of any concrete knowledge in the world. This is the reason why
Kant has to develop his argument at a series of different levels.28
94
Through the progressive development of his argument, Kant gradually

approaches a point at which the a priori structures – which are,
indeed, initiated by the mind – come to grasp the genuinely extra-
mental. This process entails setting the initiating conditions, both con-
ceptual and aesthetic, in relation to the affective condition of
experience. The aesthetic forms are particularly crucial because they
provide the point of access to an affective given. Without them the
conceptual forms would have nothing to operate on.
One common way of presenting the situation is to say that Kant
aims to get beyond the polarities of the rationalist and empiricist
worldviews and their concentration on thought or sensation, respec-
tively. Kant’s alternative path is to establish the relation between
thinking and sensing. In the next two chapters I will show how this
relation arises as the activity of synthesis. Synthesis is the capacity to
take up the sensory given through the introduction of formal struc-
tures. The Copernican revolution first uncovers this initiating condi-
tion of experience, while also revealing that mental activity has a dual
structure, being both receptive and unifying. This puts in question the
thesis that the mind projects an order onto an indifferent world.
The alternative is that knowledge arises as a process of negotiation
between mind and world.
My account will not suggest that the world independently displays
structural features that just happen to fit with the mind’s own order.
This would be to revert to Trendelenburg’s overlooked alternative
rehabilitated by Guyer.29 This would, in turn, destroy any claim that
the mind provided an a priori structure for experience, because the
conformity between mind and world would simply be contingent and
not necessary. To say that there is a necessary connection between
subjective representations and objects is to say that mind actually gets
hold of what is given to it in experience. We do not reach the right
conclusion by coincidence. The necessity characteristic of knowledge
arises neither from a contingent nor a pre-established harmony
between mind and world, but only from our ability to determine
things through the use of our cognitive powers, that is, our faculties
of understanding and intuition. The alternative I offer – both to the
view that the mind imposes order on the world and to the position
that there is a mere coincidence between mental order and order in
the world – is that there is an ongoing dynamic between mind and
world initiated at a formal level by the mind. But this will only be
established by the full extent of Kant’s argument, of which the
Copernican Revolution is merely the beginning.
95
III An alternative to the circle of representation:

representation as the mode of access to objects
I have suggested that the mind initiates the possibility of knowledge
insofar as it provides forms for experience. I have highlighted that the
conceptual determination of objects is only possible insofar as we are
capable of receptivity to something given in experience. This is the
aesthetic dimension of Kant’s formal idealism. These two insights into
Kant’s position have been extracted from his statement of the
Copernican Revolution in philosophy. In this section I will argue that
we can recognise the centrality, and indeed the necessity, of represen-
tation without falling into a circle of representation.
My aim is not to try to return to the affect (or even the ‘double
affect’) of a thing in itself. It is rather to reinstate the force of the affect
of the appearing thing in Kant’s account. I have argued that while the
‘outer things themselves’ that Kant mentions in his contrast between
formal and material idealism are not things in themselves, they are
extra-mental entities.30 Their defining characteristic is that they are
accessible to us, and in this sense are ‘for us’ or for the mind. Their
status is not privatively subjective, but relational. The question is:
how are we to account for the affect of a thing that is related to mind
and yet extra-mental?
Throughout the ‘Transcendental Aesthetic’, Kant repeatedly insists
that we encounter objects as representations for our minds. Crucial
to this perspective is the claim that both time and space are modes of
representation – that is, they are nothing if we abstract from ‘that
mode of representation [Vorstellungsart] which is peculiar to us’.31
But I will argue that these claims are consistent with the position that
the forms of intuition are modes of access to extra-mental things.
Kant appears to go one step further when he says ‘what we call
outer objects [Gegenstände] are nothing but mere representations
[Vorstellungen] of our sensibility, the form of which is space . . .’32
This could sound as if what we call outer objects are merely subjec-
tive mental representations. Even this claim, however, is compatible
with the view that we are affected by something extra-mental.
Representations belong to sensibility insofar as the latter makes them
possible, but representations arise from the affect on us of something
other than our representations. This something other than ourselves,
the source of the affect, can only affect us insofar as our conscious-
ness represents things given in experience. Were this not the case, the
object would count as a thing-in-itself.
96
It might seem that Kant goes yet further when he says that form
represents only what is ‘posited’ by the mind and thus is ‘nothing but
the mode [Art] in which the mind is affected through its own activity
(namely, through this positing of its representation), and so is affected
by itself’.33 But in this case Kant is referring strictly to the formal con-
ditions of experience, which do indeed arise from the mind. He need
not be taken as suggesting that what is taken up within form – that
is, representations in general or objects – are posited by the mind.
Only form, which is our mode or style [Art] of being affected counts
as self-affection. Admittedly, Kant’s account is misleading. It would
have been better had he said that form is the dimension of our recep-
tivity that comes from us. Nevertheless, this passage does not lead to
material idealism.
In yet another passage, Kant says that ‘all our intuition is nothing
but the representation of appearance’.34 This, too, can be read in
accordance with his formalist perspective. Representation can now
be recognised as a necessary condition of our access to appearing
objects, yet we really do access something extra-mental when we
have a representation. In other words something does appear to us as
a representation.
So far I have parried what appears to be an invitation to an impos-
itionalist reading in a number of passages in the ‘Transcendental
Aesthetic’. Now I turn to another passage where Kant appears to take
the opposite tack, going so far as to suggest he may have reverted to
transcendental realism. Late in the ‘Transcendental Aesthetic’, Kant
writes ‘Our mode of intuition is dependent upon the existence of the
object [das Dasein des Objekts], and is therefore possible only if the
subject’s faculty of representation is affected by that object’.35 Affect,
it would appear, arises from the object, which sets the faculty of rep-
resentation into action. This could count as a restoration of the causal
account of affection against which the Copernican revolution was
directed. Would this mean that things in themselves are the basis for
our representations of things as appearances? It is thus not surprising
that the interpreters discussed in the previous chapter emphasise the
priority of representation over any talk of access to objects and
restrict philosophical investigation of objects within the scope of
what I have called the circle of representation.
But Kant avoids transcendental realism just as much as he avoids
empirical idealism. The last cited passage merely states that the
appearing thing affects us insofar as it is accessible to us in experience
and that the form of intuition is only exercised in conjunction with
97
such an affect. Now does this not lead to the danger raised by Allison
and mentioned in the previous chapter, namely, that the empirical
object, which is the result of affection, is conceived as the latter’s
cause? We have seen that the problem can be avoided, as Allison sug-
gests, once we recognise that the matter in the empirical object can be
considered in philosophical reflection in abstraction from the epis-
temic conditions that are necessary for experience.36 We can thus
maintain that it is the empirical object that affects us and that strictly
speaking there is nothing other than empirical matter.37
An empirical object comprises a material and a formal compo-
nent. The formal component establishes the possibility of introduc-
ing order into experience and arises from ourselves, thus counting as
a priori. The material component is the given element in the empiri-
cal object. We only have access to this given insofar as we have capac-
ities – both sensible and conceptual – for introducing order into
matter. Matter requires form. However form is the form of matter.
In explaining the material given we do not need to proliferate levels
of matter, although we do need to distinguish different ways in which
we consider it.
When we consider an object as empirically formed, we view it as
the result of a process whereby matter has been given a formal struc-
ture. The empirical object viewed in this way does not reveal the
initial affect that first gave rise to that formative activity, because we
are concerned with the object as formed matter. However, we can also
view the same empirical object in respect of its constitutive elements,
that is, as to its matter and its form, even though within experience
these are never encountered discretely. From the vantage point of
philosophical reflection, I consider the object as material, that is, I
focus on the matter in the object. It is true that this matter has only
affected me insofar as I am the bearer of formal capacities that give
shape or figure to the material given, however there must also be
matter that is formed and this matter is nowhere other than in the
empirical object. The thought experiment I am engaged on is one in
which I imagine what always occurs together as prised apart. The
gain from doing so is that I am able to focus on the matter in the
appearing object and do not need to conjecture a further transcen-
dental matter beyond experience.
The transcendental move involves reflecting in an alternative
fashion on the experience with which we are already engaged. Kant’s
transcendental turn is necessarily committed to a material given, and,
at the same time, to the position that the latter can be viewed not only
98
empirically but also from the standpoint of transcendental reflection

that reveals the unformed element within experience.
But so far, my account treats matter simply as accessible from the
perspective of transcendental reflection. I have not addressed how
empirical matter affects the subject. Empirical matter in the appear-
ance has an affect on us, even though this only happens insofar as
matter arises within a form, that is, as a representation. While
Allison’s account suggests that matter is simply the obverse of form,
a distinction between matter understood as result and as the event of
affection will highlight the process through which matter is taken up
by form and thus gives rise to an affect. While only formed entities
can be known, transcendental philosophy must face the question of
the limits of experience and it does so insofar as it addresses the point
at which matter becomes accessible to us. The transcendental turn
entails that this event of affection does not happen outside of experi-
ence, but rather at its limits. The event of the emergence of a thing for
consciousness is usually simply assumed by us, even insofar as we are
philosophers.38
Kant’s occupation of what could seem like opposing positions, at
times insisting on the priority of mind, at others on the priority of the
object, can now be explained. His general project is one of express-
ing the relation in which the formal structure initiated by mind stands
to given objects in the world. His commitment to the relation between
form and world is what leads him sometimes to emphasise one term
and sometimes the other.
Admittedly, the affect things have on us occurs at the limit of our
powers of representation and hence of reflection. Nevertheless, this
does not render the question of affect philosophically redundant. If
Kant is to avoid material idealism he must show how it is that we can
be in contact with real things and not just mental projections. How
are we capable of being affected by something other than our own
subjective impressions and yet only in the medium of our representa-
tions? He must show how our faculty of representation takes up
something given in experience and does not simply generate the latter.
What affects our power of representation will necessarily stand at the
limit of our experience, but this is not to say that the object stands
outside it in the way that a thing in itself would.
Clearly the authors I examined in Chapter 2 do not think that the
faculty of representation generates affect. However, by inadequately
addressing the philosophical relevance of the question of affect,
they invite the rejoinder of interpreters such as Pippin, Guyer and
99
M710 HUGHES TEXT M/UP.qxd 29/3/07 8:12 AM Page 100 Gary Gary's G4:Users:Gary:Public:Gary's
Strawson for whom Kant’s account gives no convincing story as to

how the mind encounters something extra-mental and for whom
instead the mind imposes order on a neutral given.
The stakes are high, because I do not want to revert to the position
that experience arises out of an unknown something, ‘I know not
what’. How are we to understand the claim that appearances are both
merely for our power of representation and yet, at the same time, arise
out of an affect on that power of representation? The model I will
introduce in the next section is one in which there is a necessary rela-
tion between mind’s formative or representational power and the
matter given in experience, that is the empirical matter, that could not
even be recognised as such were there not a representational form in
which to take it up. Correspondingly, mental formative activity
would be incapable of initiating experience were it not prompted or
affected by the material in appearance.
IV A reading of the initial paragraphs of the ‘Transcendental

Aesthetic’
In the previous section I established a possible reading of passages in
the ‘Transcendental Aesthetic’ where Kant appears most guilty of
adopting an impositionalist position. But I now need to establish that
Kant actually presents affection as a condition of representation. In
this section I will show that Kant provides a model of affectivity in
the initial paragraphs of the ‘Transcendental Aesthetic’ before he
moves to his principal topic, the status and structure of the forms of
intuition, which, as we have seen are modes of representation of
appearances.
Intuition is the means through which knowledge stands in imme-
diate relation to objects.39 While an object, strictly understood, is the
result of the determination of an intuition under a concept, Kant ini-
tially uses ‘object’ in a less technical sense to refer to something extra-
mental given to our senses. The object in this context is the
underdetermined given in intuition. I have already used this distinc-
tion to clarify the Copernican revolution. I will now show that my
suggestion is supported by the initial account Kant gives in the
‘Transcendental Aesthetic’. The relation to things through intuition is
only possible insofar as the object is given to us.40 This can only occur
insofar as ‘the mind is affected’.41 The mind can only be affected
insofar as we have a capacity of receptivity. Sensibility is our capac-
ity ‘for receiving [bekommen] representations through the manner in
100
which we are affected by objects’.42 We can now draw together these

claims. Sensibility is our capacity for being affected by objects and,
for us, being affected entails that we have representations of objects.
Sensibility is the capacity to take up affects as representations. This is
confirmed when Kant goes on to say that sensibility ‘yields us intu-
itions’.43 But affection is the necessary condition of there being rep-
resentations in the first place. Intuitions are representations and they
arise from sensibility, our capacity to be affected.
We can so far conclude from the first paragraph of the
‘Transcendental Aesthetic’ that our ability to be affected by objects is,
at the same time, the initial condition of our representational capac-
ity. Insofar as we are, through sensibility, capable of being affected by
objects, we are capable of having intuitions of them, that is, of relat-
ing to them immediately. Intuitions are Vorstellungen, that is, they are
modes of knowledge that allow an object, literally, to be placed before
us (vorstellen). In having an intuition we stand in relation to things
that are accessible to us. These things are appearing objects or appear-
ances.
In the second paragraph, Kant characterises an appearance as ‘the
undetermined object of an empirical intuition’.44 We can now under-
stand the implication of this characterisation. An appearance is an
object that affects us through our capacity for sensibility. Our only
access to the object is in a representation. At this stage of Kant’s argu-
ment, the appearance is undetermined simply because it affects us
only as an intuition and has not yet been taken up by understanding,
the capacity for determination of the given so as to achieve knowl-
edge. The initial, non-technical sense of object refers us to our sensory
awareness of what would, if it were determined by a concept, count
as an object in the strict sense.
In the first sentence of the ‘Aesthetic’ Kant says that all thought is
finally directed to intuition, our ability to stand in immediate relation
to objects.45 This reveals that it is not only the case, as McDowell
says, that concepts ‘go all the way down’, but also that intuition goes
all the way up.46 And it would appear that not all empirical intuitions
are taken up in thought. In the ‘Transcendental Deduction’, Kant says
that objects may appear to us without ‘their necessarily relating to
functions of the understanding’.47
Now although this passage could be taken to suggest that we can
have intuitions without their standing in any relation to concepts, it is
not necessary to come to this conclusion. All Kant need say is, first,
that intuition is a necessary and irreducible condition of an object in
101
the full sense of the term. Second, sensibility provides us with objects
in a non-technical sense, that is, with appearances in intuition. Objects
in this weaker sense count as representations, yet ones that are not
determined by the faculty of understanding. However, Kant’s claim
that intuitions are not necessarily connected to the functions of the
understanding is misleading. He should have said that, insofar as our
intuition is not intellectual and is necessarily combined with a capac-
ity for conceptualisation, intuitions must necessarily relate to the pos-
sibility of determination under a concept. What is not necessary is that
intuitions are in fact determined by a concept. This maintains his com-
mitment to the dual necessity of intuitions and concepts for experi-
ence, without concluding that all intuitions give rise to cognition.
Leaving open the possibility of indeterminate intuitions is important
and not just insofar as it leaves a gap for underdetermined awareness
in everyday life, without which Kant’s account of experience would be
implausible. It also leaves a place for aesthetic judgements in which
intuitions relate to concepts, but only indeterminately.48 It seems likely
that the ambivalent nature of Kant’s account of the relation in which
intuitions stand to concepts in the first Critique arises from his not yet
having established the distinction between determinate and reflective
judgement.49
At the most primitive level, an appearance is the representation of
an object’s affect on us. It is empirical because the object is only ever
that thing that affects us within the range of experience. This thing
can only be accessed through our modes of knowledge, in this case
the forms of intuition that give rise to representations. But represen-
tations would not arise if the object did not affect our sensibility. The
relation between object (in the non-technical sense) and representa-
tion is a two-sided one. Without representation there would be no
affect by the thing, without the affect by the thing there would be no
representation. This account of the role of representation highlights
the necessity of affect and thus does not fall back into the circle of
representation. The conclusion is that an appearance is an object (in
the looser sense) given in intuition and only accessible to us through
our powers of representation. If that appearance were determined by
a concept it would count as knowledge of an object (in the technical
sense).
In the second paragraph of the ‘Aesthetic’ Kant says that sensation
is the effect (Wirkung) of an object upon the faculty of representation
insofar as we are affected (affiziert) by it.50 The question is not here
of the effect of some unknown cause, but rather of the affect that
102
makes possible an empirical intuition.51 In the third paragraph,

Kant develops his point by making a distinction within appearance
between matter and form. Matter in an appearance is that which cor-
responds to sensation.52 Sensation is the empirical result of our capac-
ity to be affected by an object. The form of appearance, in contrast,
is that which allows the manifold, that is, the given, to be ordered in
certain relations.53 The ordering of sensation cannot arise from the
given itself, just because the latter is empirical. Order must arise from
the mind, as only as a priori can its necessity be secured. Any appear-
ing thing arises out of the conjunction of the material given, which is
empirical and thus a posteriori, with the formally a priori.
Matter is thus established as necessarily empirical. It is easy to con-
clude that this entails that the a posteriori is simply the after-effect of
a priori form. But, as we saw in the previous section, there are two
levels at which the matter necessary for intuition can be analysed.
Matter experienced as one of the component parts of an appearance
is viewed as the result of a process of formation. We only experience
matter insofar as it is taken up by representation, that is, in relation to
form. However this result arises out of two irreducible elements.
Although we cannot experience matter in abstraction from form, we
can think it. I suggested in the previous section that we can express the
affect of matter as an event, in contrast to its result in sensation.
Thinking matter as event allows us to consider the a posteriori in a dif-
ferent way. The matter of an appearance counts as a posteriori now in
the sense that it is what is necessarily given as an affect. The a poste-
riori cannot come from the mind and must be given. This reveals an
alternative nuance of the a posteriori, where the latter is not simply
subsidiary to the a priori, but radically distinct from it. The radical
sense of the a posteriori is that of the necessity of the givenness of
matter within the empirical. Matter viewed in this way is the neces-
sary complement of the a priori.54 Thus while sensation viewed as the
material aspect of appearance counts as an effect or result, we can also
think about affect as a necessary condition of experience or event.
While we only ever experience the result or trace of this event – that
is, some combination of form with matter – there would be no mate-
rial given, were there not an initial event of affection.55
Matter is that in the appearing thing that sets our representational
capacities to work. Neither the affective condition of representation
nor its a priori form is temporally prior, so it is not that the object first
exists and then has an affect on us. Rather, to view something as an
object is to consider it as standing in an affective relation to us, that
103
is, as accessible to our representational powers. The affect and the

representation arise in one and the same event as a relation between
mind and world. This is the characteristic status of an appearance in
contrast to a thing-in-itself that would not stand in any affective rela-
tion to the representational mind.
To say that the object only exists as such insofar as it affects us and
is capable of being taken up by our powers of representation may
appear to fall back into material idealism, albeit of a new form.
However, Kant is not committed to the position that we produce
objects, but simply that we only attend to those that our subjective
faculties are capable of taking up. This does not entail the existence
of another category of objects we cannot access.56 Rather, we relate
to objects as our environment; they are part of the world we also
inhabit and it is only by treating them as such that we are capable of
encountering and, on occasion, knowing them.
The appearance, as we have seen, is a composite of form and
matter, but Kant now goes on to focus exclusively on the formal side
of intuition. This central emphasis of the ‘Transcendental Aesthetic’
has encouraged the impression that his account of mind is wholly
constitutive and not receptive. Form counts as pure insofar as to
achieve it we abstract all the features that arise from sensation, that
is the material side of the appearance. These include impenetrability,
hardness and colour.57 Additionally, we must abstract the features
that arise from the understanding – for instance, substance, force and
divisibility.58 What is left is characteristic only of pure intuition, that
is, the aesthetic form of appearance, namely, extension and figure.
Pure intuition, Kant says, ‘is to be found in the mind a priori as a
mere form of sensibility’.59 This formal residue of every appearance
is what Kant puts down to self-affection, as we saw in the previous
section.60
Thus form cannot be derived from sensation; nevertheless, we
have seen that there would be no appearance in the first place, were
we not affected through sensibility by the object. Form is not
imposed on the matter of sensation, but is the framework within
which the latter is represented. An appearance arises from two
sources, one extra-mental and the other mental. The appearance
differs from the thing in itself and from the merely phenomenal illu-
sion in that it is relational, that is, it arises from a relation between
object and the mind. We have already seen in a provisional way in
our discussion of the Copernican revolution that mind initiates the
forms not only of intuition but also of understanding.61 Later we will
104
see that the combination of a plurality of modes of knowledge or fac-

ulties is necessary for Kant’s account of experience.
In the ‘Transcendental Aesthetic’, Kant chiefly investigates the
principles of a priori sensibility, that is, the aesthetic form of appear-
ances, which he presently announces are only two, namely space and
time.62 But the first few paragraphs of the Critique have revealed that
these mental forms are only one side of the story of intuition and that
appearances arise only insofar as we are affected, in a non-technical
sense, by an empirical object. The forms of intuition are representa-
tional forms through which we take up what empirically affects us.
V The complex identity of receptivity

In the first paragraph of the ‘Transcendental Aesthetic’ Kant identi-
fies sensibility as ‘the capacity (receptivity) for receiving representa-
tions through the way in which we are affected by objects’.63 I have
construed this as the view that we have representations only insofar
as we are affected through our capacity of sensibility. This affect gives
rise to a sensation. But, as we have also seen, this is only one side of
the story of receptivity. Our intuitions – that is, our capacity to stand
in an immediate relation to objects – also require an a priori element,
that is, a form that makes possible the ordering of the manifold of
matter. Receptivity arises out of an affect, but the latter can only
count as such – that is, affect us – if it can also take on a certain order
and this requires a pure intuition. The forms of intuition make possi-
ble that we take up something and hold it together in an empirical
form.64 The aesthetic side of our experience is (philosophically)
revealed as a point of contact between objects and mind.65 A full
account of the structure of intuition must include not only its formal
conditions, but also the way in which form makes possible access to
a material given. Without this crucial move concealed in the initial
paragraphs of the first Critique, Kant’s formal idealism would have
fallen back into material idealism.
The form of space makes possible that there are empirical spaces,
that is, that these can be taken up by us as representations. In the first
and third propositions of the ‘Metaphysical Exposition’ of space,
Kant rules out that space counts as an empirical or general concept,
respectively.66 Having entertained and rejected one possible account
for the status of space, in the second proposition he claims that space
is the pure intuition underlying all our outer intuitions.67 Space, as
pure form of intuition, counts as the condition of the possibility of
105
appearances.68 Thus space and time as pure intuitions are the formal
conditions that make possible that we have empirical intuitions, that
is, that we can be affected by an object through sensibility so as to
give rise to a sensation. But this is to say that pure intuitions are neces-
sary but insufficient conditions of empirical spaces and times, not that
they generate the latter.
The necessity of form does not undermine the status of intuition as
given. Rather, it reveals the complex structure of our receptivity nec-
essary for taking up any affect. Form is simply the ordering process
through which we take up the given. Intuition arises out of a mater-
ial and a formal source; first, our sensibility as the capacity to be
affected; and second, the pure intuitions of space and time. These are
the two sides of our capacity of sensibility or receptivity, revealing
how affection and the possibility of representation are inseparable
from one another.
Receptivity or sensibility is thus, on further investigation, twofold.
It comprises, first, the capacity for being affected, which gives rise to
the material element in intuitions, and, second, pure intuitions or
forms. These are two sides of one coin. We could not be affected were
we not capable of taking up the given within a form or order. In the
next chapter we will examine how experience or knowledge arises out
of a combination of understanding and intuition. We have now learnt
that even the intuitive component of experience is not a simple event,
being directed both toward the subject and toward an indeterminate
object.
It is because of this two-sided character of receptivity that Kant is
not an impositionalist. It is also because of this that transcendental
idealism qualifies as transcendental, that is, as a philosophical reflec-
tion on the conditions of experience. The conditions are ‘of’ experi-
ence with the genitive force of belonging to experience, in contrast to
the suggestion that they are an independent set of rules applied to
experience. The dually oriented structure of receptivity is what allows
our minds to get a grip on the world. Admittedly, Kant’s failure to
fully develop his account of the affective side of experience and, at the
same time, adequately explain the complex structure of receptivity
are responsible for many of the misunderstandings that have arisen.
While Kant and his interpreters have tended to characterise sensi-
bility as passive in contrast to the activity of the understanding, we
can now see that this is misleading. Sensibility is a capacity, albeit a
curious one insofar as it is the ability to receive what is given in experi-
ence. It is not active because it is the mode in which we are affected
106
by things in the world. But no more is it merely passive, first and most
obviously because it requires a formative ability. The second of these
features has led many readers to conclude that sensibility does not
really count as receptive. However, sensibility is a capacity to open
ourselves to things in the world. This entails that we anticipate or
prepare for the event of the material given, not that we create or
impose the conditions of that event. Intuitive form is nothing other
than the anticipation of the event of the affect of the material given.69
Kant’s account of sensibility shares some of the force of the phenom-
enologist’s notion of passive synthesis.70 The latter is a pre-reflective
capacity to make sense of things in the world.71 Kantian sensibility is
not pre-reflective, however, and is always combined with some level
of reflection. Nevertheless there is a parallel insofar as both passive
synthesis and receptivity inhabit a middle ground between activity
and passivity.72
At this stage, I simply wish to establish the point of departure for
the reading of the relation between form and matter that I will
develop in later chapters, especially in Chapter 6 where I examine the
overall structure of the objective deduction. In what sense has the
approach adumbrated here escaped the circle of representation? I
fully agree with Buchdahl, Allison and Longuenesse that any object
we experience is a representation. But I also have some sympathy with
the insistence by Pippin, Guyer and Strawson that our experience
must be of empirical objects and not just of representations. My solu-
tion has been to argue that representations are our mode of access to
things. It is not that we experience representations instead of things.
It is rather that we experience things as representations, not through
the latter but in them.
The interpretation I have developed in this chapter allows me to
suggest provisionally a way in which we might construe the relation-
ship between empirical realism and transcendental idealism. The
forms of experience are our modes of representation that give us access
to an empirically real world. Transcendental idealism secures the
empirical reality of the world and does not threaten it. No more is
empirical realism a cheap substitute for transcendental realism, which,
being incapable of proving its premises, cannot secure what it tanta-
lisingly promises, namely, knowledge. It is not so much that Kant
offers empirical realism because this is all that can be derived from
transcendental idealism, as that transcendental idealism is the strategy
he adopts to safeguard the empirical realism to which he is ultimately
committed. Empirical realism is the only realism compatible with the
107
relation in which we stand to an extra-mental world, the form of

which arises from our minds.
Notes
1. An affect is a sensory event, viewed as something that happens or is
given to a receptive subject.
2. Or, in the case of temporal sensibility, to be aware of ourselves as other,
that is, over time.
3. Added in the second edition of Critique of Pure Reason, at B 519.
4. In Chapter 6 we will see that Kant uses the expression ‘appearances
themselves’, Erscheinungen selbst, at A 190, B 235, to express that
appearances are capable of bearing objective validity.
5. B 519 (my additions). The German is: ‘In manchen Fällen scheint es
ratsam zu sein, sich lieber dieser als der obgenannten Ausdrücke zu
bedienen, um alle Mißdeutung zu verhüten’.
6. In the mind of God, though, not merely in our minds. See Berkeley ‘A
Treatise Concerning the Principles of Human Knowledge’ in
Philosophical Works ed. Ayers, pp. 116–17.
7. B 274. For a discussion of the systematic position of the ‘Refutation’
and its contribution to a possible spatial schematism, see Chapter 6,
pp. 234–5.
8. Though both are defensible only within the context of experience and
not as purely theoretical or conceptual theses. They require intuition
and not just concepts.
9. My analysis of the early paragraphs of the ‘Transcendental Aesthetic’
later in this chapter will show that the position of the Refutation cannot
be seen as a divergence from Kant’s main project.
10. Although not on the absolute given, as all three concede the need for
reflective mediation of the given.
11. Preface to second edition of CPR, B xvi. Strictly, the reversal applies not
only to objectively valid cognition or knowledge, but all cognition or
Erkenntnisse.
12. See Hume, A Treatise of Human Nature, p. 11, on the role of fancy in
relations of cause and effect. See also p. 86 on the role of belief or assent.
13. See, for instance, my discussion in Chapter 6, pp. 238–41 of the princi-
ple of hypothetical necessity, which is a precursor for the reflective prin-
ciple of the systematicity of empirical nature.
14. B xvi/xvii (my additions).
15. B xvii.
16. See discussion of the first edition ‘Transcendental Deduction’ in Chapter
4, pp. 122–7.
17. B xvii.
18. B xxi.
108
19. See also B xxv/xxvi. In Chapter 2 we saw that there is a problematic role
for things-in-themselves even within the epistemic domain insofar as
thought of them facilitates a focus on undetermined matter.
20. A 20, B 34 (my addition). See discussion in following section.
21. See Chapters 6 and particularly Chapter 7, pp. 249–55.
22. See final section of this chapter, pp. 105–8.
23. A 51, B 75.
24. See final section of this chapter, pp. 105–8.
25. Later I will argue that these are not all that are necessary. See Chapter
7, pp. 249–55.
26. A 20, B 34. See my discussion of the initial paragraphs of the
‘Transcendental Aesthetic’ below. It will emerge that appearances may
also count as the ‘not yet fully determined’ and not merely the ‘indeter-
minate’.
27. In the Preface to first edition of CPR, A xii, Kant characterises the task
of the Critique thus: ‘I do not mean by this a critique of books and
systems, but of the faculty of reason in general, in respect of all knowl-
edge after which it may strive independently of all experience’. Clearly
the last phrase refers to a priori knowledge. My argument will be that
a priori knowledge must nevertheless anticipate the possibility of empir-
ical knowledge, the full account of which requires more than the argu-
ments of the Critique.
28. See Chapter 6 for a hierarchical reading of the ‘Transcendental
Analytic’.
29. See Chapter 1, p. 18, for a discussion of Guyer.
30. B 519.
31. A 35, B 51. See also A 37, B 54.
32. A 30, B 45.
33. B 67/8.
34. A 42, B 59.
35. B 72.
36. See Chapter 2, pp. 65–7, on Allison’s reading of the transcendental sig-
nificance of affection.
37. Gerold Prauss’s interpretation is in the same vein. He argues that affec-
tion is empirical; however, empirical affection cannot be grasped at the
level of empirical experience nor by the empirical sciences. Only tran-
scendental reflection can explain the empirical affect on a non-empiri-
cal subject in a ‘non-empirical science of the empirical’. See Prauss, Kant
und das Problem der Dinge an sich, especially pp. 213–27. My own for-
malist account shares with Prauss the conviction that the subjective
structure of the empirical is not empirical, but transcendental.
38. See Chapter 8, p. 304, where I argue that aesthetic apprehension makes
possible a glimpse of the event as event.
39. A 19, B 33.
109
40. A 19, B 33.

41. A 19, B 33.
42. A19, B 33 (my translation).
43. A 19, B 33.
44. A 20, B 34.
45. A 19, B 33.
46. McDowell, Mind and World. See, for instance, pp. 13 and 23, which
seem to emphasise only the first of these relations.
47. A 89, B 122; my translation of ‘ohne daß sie sich notwendig auf
Funktionen des Verstandes beziehen müssen’. He reinforces this at A
89/90, B 122–3.
48. See my discussion in Chapter 4, pp. 137–47, of B 160 where Kant
appears to rule out the possibility of undetermined intuitions. See also
Chapter 5, pp. 175–6, on the relation between aesthetic judgement and
cognition in general. Sarah Gibbons holds a similar view. See Kant’s
Theory of Imagination, p. 18, note 6, where she distinguishes between
the possibility and the actuality of conceptualisation.
49. This distinction is first introduced in the Critique of Judgment at AA
179. I first mentioned this in my discussion of Pippin in Chapter 1,
p. 15.
50. A 19/20, B 34: ‘the effect . . . so far as we are affected by it, is sensa-
tion’.
51. A 20, B 34.
52. A 20, B 34.
53. A 20, B 34.
54. In later chapters I will argue that a priori form anticipates instantiation
in the a posteriori. See Chapter 6, pp. 221–9, and 7, pp. 249–55.
55. But although matter is always experienced as in some sense formed, it
is not necessarily experienced as determined by form. See the discussion
of aesthetic form in Chapter 8. It is because in an aesthetic judgement
form and matter stand in relation and yet are not unified in a cognition
that the relation between form and matter is displayed.
56. At least not epistemically.
57. A 20/1, B 35.
58. A 21, B 35.
59. A 21, B35 (my translation).
60. B 67/8.
61. See discussion of forms of understanding in Chapter 4, pp. 113–16.
62. A21–2, B 35–6.
63. A 19, B 33 (my translation).
64. See discussion of synthesis of apprehension in intuition in Chapter 4,
pp. 122–4.
65. Later I will argue that aesthetic judgement provides a phenomenolog-
ical revelation of the aesthetic or sensible dimension of knowledge.
110
66. I will not discuss here the remainder of the content of the 3rd proposi-
tion, which taken in conjunction with the 4th establishes that space as
a whole is the horizon of all particular spaces. See my ‘Kant’s
Phenomenological Reduction?’ for the claim that the forms of intuition
are best understood as horizons of sensory experience.
67. A 24, B 38. I will leave discussion of the ‘Transcendental Exposition’
until Chapter 4.
68. A 24, B 39.
69. See Chapter 6, pp. 221–9, where I bring out that the principles of under-
standing are anticipatory of a material given in space and time.
70. See Edmund Husserl, Analyses concerning passive and active synthesis,
especially pp. 508–12, Hu 408–11.
71. Volume 11 of Husserl’s collected writings is devoted to this theme. A
more accessible version is to be found in Merleau-Ponty’s discussion of
‘general synthesis’ in the Phenomenology of Perception. See, for
instance, p. 428 (French edition, p. 489). See also Heidegger’s discus-
sion of Sein lassen or ‘letting-be’ in Being and Time; for instance, H354.
72. This ground that is not a ground can be grasped through the trope of
the middle voice, a notion that has been fruitfully explored by John
Llewelyn. See, for instance, his ‘Heidegger’s Kant and the Middle Voice’,
and the ‘Preface’ to The HypoCritical Imagination.
111
The Deep Structure of Synthesis
In the previous chapter I examined the receptive condition of knowl-

edge and its source in sensibility. This faculty is able to take up the
object given in experience because of its introduction of intuitive
forms into the manifold or empirical matter. I have claimed that there
are three conditions for knowledge. First, there must be something
given to us; second, we must be receptive to that given; and third, we
must be capable of unifying the given under a concept. In the last
chapter I discussed the first and second of these conditions. I now turn
to the third, the condition of conceptual determination.
Knowledge as opposed to intuition is only possible if we can unify
and thus identify what Kant provisionally called the object in the
initial paragraphs of the Critique of Pure Reason. In the first section
of this chapter I discuss the other side of Kant’s dualism and argue
that it is best understood as reflective form. I also argue that, on closer
examination, dualism requires a plural iteration of the operations of
the mind as the synthesis of affectivity and conceptualisation requires
a third term, imagination. The combination of these different ele-
ments of experience counts as synthesis.
In the second section, I discuss the faculty talk that I have employed
in my initial characterisation of synthesis. I suggest that the faculties
need be seen neither as psychological constructs nor as curious entities.
Kant’s faculty talk allows for reflection on a complex model of mind in
which only the combination of distinct orientations gives rise to the
structure or form of experience. Within his epistemology Kant princi-
pally distinguishes sensibility from understanding. At a more general
level of analysis of human existence, he distinguishes the cognitive, the
moral and the aesthetic. At each level of analysis Kant portrays the
human subject as operating with a plurality of different orientations.
Faculty talk is a way of capturing that plural constitution.
In the third section of this chapter I turn to Kant’s account of deter-
mining judgement in the two editions of the ‘Transcendental
Deduction’. Kant attempts to validate the claim that the concepts of
the understanding are applicable to all empirical objects. I argue that
112
both editions reveal a commitment to a plurality of faculties working

in conjunction with one another and that the imagination plays a vital
role in making this possible. However, in both editions Kant’s account
reveals the hegemony of the understanding and not a free cooperation
among a plurality of faculties.
In the final section of this chapter I argue that the synthetic activ-
ity arising from a combination of a plurality of faculties is examined
directly, not merely presupposed, in Kant’s account of aesthetic judge-
ment. The latter counts as synthetic insofar as its characteristic
‘harmony of the faculties’ reveals the synthesis in process necessary
for any judgement.
I Reflective form and the plural conditions of synthesis

If the ‘object’ given in experience is to be more to us than mere intu-
ition, then we must be capable of taking it up in such a way that we
can identify or determine its appearance in space and time. In order
to do this, we need not only intuitions but also concepts. Just as intu-
itions are traced back to the human capacity for sensibility, concepts
are ultimately traced back to the human capacity for understanding.
If we did not have the latter, we would be incapable of any level of
experience beyond that of sensation. Appearances can only count as
objects for us in the strict sense, that is, as identifiable particulars,
insofar as we are able to distinguish them as falling under concepts.
What was termed an object in the ‘Transcendental Aesthetic’ in fact
fulfilled only one of the conditions for the latter, namely, that there is
something given to us, not merely constructed by our minds. Kant’s
considered account of objects is that the proper use of the term is epis-
temic and refers to the result of the synthesis of a concept with an
intuition. Our experience of objects, thus, is dependent on the exer-
cise not only of the faculty of sensibility, but also on that of the under-
standing.
In the previous chapter a close link emerged between form and
mind. It is we who introduce form into experience. A priori forms are
of two types: first, the forms of intuition – namely space and time –
and, second, the forms of understanding, that is, the categories.
Sensibility has been identified as the source of the forms within which
any empirical intuition arises. Now I will examine understanding as
the source of the forms that make conceptualisation possible. The
form of experience arises from the combination of these two elements
of experience.
113
Dualism expresses a distinction within subjectivity between our

receptivity to the outside world and our capacity for conceptualisa-
tion that makes possible unification of what is given to us in intuition.
The most famous statement of this position is: ‘Thoughts without
content are empty, intuitions without concepts are blind’.1 Our dual
makeup means that we are reflective beings, for our thinking only
arises in relation to a world not of our own making. This means
that our peculiar mode of thinking is that of synthesising, that is, we
make sense of this world through the combination of receptivity and
understanding.
Dualism thus requires a relation between concepts and intuitions
and this entails a third term, the imagination, which makes possible
their synthesis. This is to say that our mental activity is plurally and
not merely dually iterated. Later in this chapter I will show that in the
heart of Kant’s epistemology the operation of a third term is essential.
While the plural constitution of human experience only becomes
explicit in the Critique of Judgement, at all stages of his account the
faculty of imagination is the mediating faculty that makes unification
possible. This is true even though it is also the case that the lineage of
imagination is often put in question, as it is sometimes allied with
intuition and often with understanding. Although at a surface level
dualism suggests a fragmented psyche, at a deeper level it must be
recognised as entailing the cooperation of a plurality of mental ori-
entations that constitute the subjective condition of experience.
At first glance dualism names the complexity within a subject who
is both receptive and reflective. But as a receptive subject necessarily
stands in relation to a world of things that are given, dualism also
reveals the relation in which we stand to a world external to our
minds. A receptive subject is not the author of its experience and thus
dualism reveals the human condition of finitude. The result, however,
is not an impasse between subject and object when dualism is inter-
preted as entailing a pluralist relation to an extra-mental world. The
third or relational term makes possible not only unity within a
subject, but also access to an external world insofar as we are capable
of taking up the given in a synthesis and achieving cognition of it.
This is in contrast to the two-term relation of the impositionalist
model, which were it true of Kant’s epistemology would mean that he
was committed to the view that the understanding projected its forms
upon an external world. Impositionalism gives a reductionist inter-
pretation of dualism, according to which dualism results in the dom-
inance of mind. But dualism is neither committed to a simplistic
114
relation among the faculties nor between mind and world, and for
good reason. The initial triumphalism of mind would fall back into a
subjectivism based on understanding alone where instead of encoun-
tering the world, we would be faced only with our own reflection.
Contrastively, if the philosopher rests his or her account on the faculty
of intuition as the sole effective source of knowledge, this version of
monism would establish only that we are capable of having impres-
sions of the world, not that we can know objects.
In the previous chapter I developed the idea that mind initiates
form. By this I mean that the initial possibility of ordering the world
comes from the mind, but experience only becomes possible insofar
as mental forms are set in relation to possible experience. This
requires that the forms be rearticulated so that they are capable of
‘anticipating’ empirical experience, as I will argue later.2 The limita-
tion on the role of form was revealed by my discussion of Kant’s dis-
tinction between formal and material idealism. Our mental activity is
principally that of ordering and making sense and we do so by gen-
erating structures that allow us to take up the contents of the world
we find ourselves within.
Kant characterises the conceptual side of experience as a capacity
for ‘spontaneity’. Understanding is capable of spontaneously or
actively introducing a priori forms of unification, that is, categories
into experience. These forms are not to be found in the given nor can
they be supplied by the power of receptivity, sensibility. Unification
arises from the human capacity for thought and, in particular, from
our capacity for understanding. However, it is important to recognise
the limitation Kant places on our ability to generate the conceptual
form of experience. The latter can only give rise to the form of experi-
ence insofar as it is combined with the aesthetic form arising from our
capacity for receptivity. For this reason I believe it is important to
temper the characterisation of our power of conceptualisation as
spontaneous, justified by its status as the source of priori forms, with
a recognition of its reflective status.3 This is particularly important
when considering the application of a priori forms. The forms intro-
duced spontaneously only gain validity in that they take up content
that they themselves cannot deliver.
We do not have an intuitive intellect that would directly grasp the
inner essence of things. Our consciousness is necessarily a synthesising
one. We work at making sense of the world and have no immediate or
final access to the truth of things as they are in themselves. The latter
would require an operation of an intuitive understanding independent
115
of any other faculties and for finite beings such as we are this is simply
not an option. We lack the royal road to truth and must always strug-
gle for an ideal, which, necessarily, is absent. But this does not make us
indifferent to truth. We do not simply accept that what appears to be
the case, is the case. Instead we seek out ever more layers of appear-
ance, tending towards an infinite and unrealisable idea of total revela-
tion. This goal must, however, function strictly as a regulative idea of
reason if we are not to fall back into a new form of dogmatism that
would arise from assuming that we can achieve absolute truth. To do
so, would be to ignore the necessarily reflective and synthesising nature
of our consciousness.4 While the position just adumbrated may appear
to go beyond Kant’s commitment to the certainty of knowledge,
his claim for apodeictic certainty refers only to the synthetic a priori
judgements that provide the minimal framework, not the content, of
experience.5
II The form of experience and faculty talk

We can make sense of Kant’s faculty talk in the following way. Our
consciousness is plurally oriented and comprises a number of differ-
ent capacities that must be distinguished and coordinated, that is, syn-
thesised within experience. The responsive, yet spontaneous human
mind necessarily stands in a complex relation with the world. It is
open to the world and yet it is not wholly determined by it. The thesis
of dualism is expressed by a faculty theory that draws out the various
different orientations we take within experience. At a more general
level of analysis than the epistemological one within which we are
currently operating, Kant is committed to the view that human
experience cannot be analysed using one singular explanatory model,
for instance, a mechanical one of natural entities, a teleological one
of natural purposes, or a moral one. Instead his account of human
experience requires recognition of a combination of different orien-
tations. Human beings are neither wholly rational as angels perhaps
would be, nor wholly sensory as Kant holds animals are. Rather, we
are the kind of beings who combine rationality with sensible experi-
ence. We do so through the mediation of a capacity arising from a
third faculty, imagination.6
Many commentators, understandably, have cavilled at this appar-
atus and the psychological foundationalism it seems to entail.7 In my
view, the faculty theory is neither psychological nor does it provide
an autonomous foundation for experience, but rather is an attempt,
116
sometimes clumsily handled, to account for the way in which a

complex mental life makes experience possible. In order to avoid
unnecessary reification, we should say that synthesis is the activity of
the faculties cooperating with one another and that the faculties are
nothing other than the elements or poles of that cooperation. The fac-
ulties need not be taken as predating, or as substrates of, our experi-
ence. They are not even in principle independent of experience, for
they bear validity only insofar as they are at work within particular
experiences of the world. However, this does not mean that they
are simply superstructural or fictional. Kant’s talk about faculties
expresses the dynamic activity of mind and provides a way of address-
ing the latter’s necessary moments. In order to do this he is forced to
talk as if there were poles or constituent elements in this activity,
while this is not strictly true. It is necessary to identify specifiable ten-
dencies within our mental lives. We orientate ourselves to the world
in a number of distinguishable ways and an adequate philosophy of
mind must account for these differences. We know what it is like to
attempt to achieve knowledge about a state of affairs, or to respond
to it morally or imaginatively. Human experience would be unimag-
inable without these necessary possibilities, which although inter-
twined with one another are also distinguishable. Faculty talk makes
possible a philosophical reflection on the necessary elements of our
mental activity, allowing certain primary orientations to be individu-
ated, while allowing them to be treated as necessary structures of pos-
sibility, not as substantial mental entities. This is in contrast both to
the deflationary account that would treat faculties as merely possible,
yet cannot explain why they are not just some capacities among
others, and, on the other hand, the reificationary account that can
only establish the individual characteristics of the faculties at the cost
of positing them as substantial.
The fear of psychologism has led many philosophers to excise the
necessary role for mental activity in Kant’s epistemology, leaving such
concerns to the psychologists. Yet Kant believes that objective knowl-
edge arises from subjective sources. An advantage of Kant’s use of
faculty talk in his epistemological argument is that it allows him to
focus on the differentiation and interaction within mind, without
cutting off mental activity from the world. I agree with what I will
argue is his position, namely, that establishing a relation between sub-
jectivity and objectivity is necessary if knowledge is to be shown to be
possible. I also agree that the subjective process by which knowledge
emerges should be recognised as depending on the ability to combine
117
a plurality of different orientations or perspectives. Whether it is

necessary to characterise the subjective side of this relation as com-
prising faculties is debatable, but that Kant was committed to such a
model is undeniable.
For Kant the particular faculties of understanding, sensibility and
reason are the sources of knowledge, perception and speculation,
both moral and scientific.8 However it is only through the interaction
of the faculties that anything like thought or experience emerges.
Knowledge, for instance, has a primary connection to the under-
standing, but cannot arise unless the latter takes up the sensible given
to which we are receptive in intuition and indeed, as I will argue later,
Kant holds that knowledge also requires that we exercise reflective
judgement.9 The need for cooperation among our faculties is first
introduced as a theme in the Critique of Judgement under the title of
‘the harmony of the faculties’ although there is plenty of evidence,
which I will endeavour to uncover, that the pluralist model was
already operative as a presupposition in the first Critique.10
All of this underlines the pervasiveness of reflection within human
experience in the Kantian account. Reflection is the thinking activity
of a consciousness that is also receptive. When Kant talks of the com-
bination of sensibility and conceptualisation within experience, he is
not, of course, intending that there are two sorts of experience that
are then forged together. Instead his point is that all human experi-
ence necessarily displays both a sensory and a conceptual dimension.
This is what McDowell means by saying that spontaneity goes all the
way down.11 There is no sensible impression that is wholly detached
from the possibility, at least, of reflection on it. This is true even when
the power of expression breaks down. Faced with the highest peak in
Europe and the expanses of snow surrounding it, I may say ‘there are
simply no words for this!’ I may search for an expression that at least
points to my lack of a description, saying perhaps, ‘this is sublime’.
But I cannot escape the power of reflection. My sense of awe is ine-
liminably connected to the inadequacy of my power of words and
draws on a contrast to the latter. Correspondingly, any exercise of my
power of reflection or spontaneity stands in some relation, however
distant, to sensible experience. Receptivity goes all the way up, just as
spontaneity goes all the way down.12 Kant expresses this complex
structure using the model of a plurality of faculties.
While worries about the reintroduction of foundationalism – now
psychological – should not be underestimated, the importance of
faculty talk is in providing Kant with a way of expressing the complex
118
identity of the form of experience. Orientations, irreducible to one

another, cooperatively give rise to a structure without which experi-
ence would not be possible. Kant captures this in talking about dis-
tinct but related forms of experience that jointly give rise to the
transcendental form of experience. It is thus that faculty talk allows
him to develop a complex philosophy of mind. Mind is not recon-
structed as a singular entity, facing an undifferentiated world, but,
rather, a complex and internally relating activity facing a diverse
world – the manifold of sensibility. The unification that the mind
effects is necessary for the unifications in our experience of the world,
for it is principally our capacity of synthesis that makes any unity in
the world possible. The model or notion of mind and world that arises
from all of this is one of a constant dynamic process and yet one in
which unity or sense, though no final unity or sense, can and must
arise. Experience and knowledge are projects or tasks which comprise
the ongoing unification or determination of the indeterminate, but
determinable, which we take up through the combination of sensibil-
ity and understanding.
There is, however, a danger that the account of the pluralism of the
faculties I have sketched and will defend in the chapters that follow
appears both blandly moderate and implausible as a representation
of Kant’s position. What is the advantage of a cooperation of a plu-
rality of the faculties in contrast to the dominance of one singular
mental capacity? And, is not Kant, in the end, of the view that moral
reason is the dominant force in our experience of the world – a posi-
tion his own account of orientation in thinking reinforces?13 In con-
trast to this prima facie persuasive view of Kant’s commitments, I will
defend an interpretation of the relation of the faculties, not simply as
a cooperation of indifference, but rather a dynamic interaction that
requires the distinctiveness of the elements in the relation. While one
faculty is dominant in each of the legislations covered by the three
Critiques, the effectiveness of the principle governing each requires a
plurality of mental orientations. Even moral law cannot operate by
independent Diktat. In Kant’s philosophy of history we find that any
hope we may have for the emergence of a kingdom of ends on earth,
rests on our ability to organise our relations with other moral agents
according to political strategies, which are, necessarily, amoral.14
Kant’s system continually relies on a cooperative model of mental
activity, though not one of mere equivalence. This only becomes
explicit for his readers and for Kant himself in his articulation of the
‘harmony of the faculties’ in the Critique of Judgement.
119
III How the legitimation of the categories requires the

cooperation of a plurality of faculties
The categories are ‘concepts of an object in general, by means of
which the intuition of an object is regarded as determined in respect
of one of the logical functions of judgement’.15 It is thus clear that the
conceptual side of Kant’s account of knowledge necessarily stands in
relation to the affective or aesthetic side that I was concerned with in
the previous chapter. The relation in which concepts stand to intu-
itions is distinctive of transcendental logic, in contrast to general logic
which abstracts from all content.16 The ‘Table of Judgements’ belongs
to general logic, but nevertheless provides the most general shape of
thought characteristic also of transcendental logic. Judgements are
functions of unity in our representations.17 Transcendental logic, in
contrast, will establish the functions of unity or synthesis among
objects.18 Thus Kant identifies the most general functions of judge-
ment and uses these as his touchstone for discovering the most general
ways of thinking of objects, namely, the categories.19
Kant warns that it is very tempting to use the categories ‘by them-
selves and even beyond the limits of experience, which can alone yield
the matter (objects) to which those pure concepts of understanding
can be applied’.20 By doing so, the understanding makes a material
and not a properly formal use of its principles.21 Kant’s point is that
within transcendental logic, a logic that is concerned with the a priori
conditions of empirical experience, the understanding can only be
used in conjunction with sensibility, that is, in application to matter
or objects that are as yet undetermined. This is what he aims to estab-
lish in the ‘Transcendental Deduction of the Categories’; any other
use of the categories would count as dialectical.
The ‘Transcendental Deduction’ is the centrepiece of the logical or
conceptual side of Kant’s account of knowledge. In it he seeks to
prove the validity of the categories of the understanding. This entails
showing that the categories or subjective conditions of thought
‘furnish conditions of the possibility of all knowledge of objects’.22
He remarks that the forms of intuition do not need a deduction. This
is because, although they are a priori, they must necessarily relate to
objects and therefore bear objective validity.23 This supports my
insistence that sensibility is two-sided.24 Establishing the objective
validity of the categories requires a specific argument because, Kant
says, appearances can arise ‘independently of functions of under-
standing’.25 Kant’s point is that everything we experience must arise
120
within the forms of space and time and therefore counts as appear-
ance. But he also makes clear that not all appearances are subject to
the unifying power of understanding. It sounds as if Kant’s commit-
ment to the dual structure of experience has collapsed. On other
occasions he appears to suggest quite the opposite, namely, that all
intuitions must be unified by the understanding.26 This, too, would
threaten dualism. Kant should have said that while not all appear-
ances are determined by the understanding, they must stand in some
relation to our reflective power of understanding and thus are deter-
minable, if not determined.27
I will not provide here a continuous reading of the two editions of
the ‘Deduction’. There are many excellent accounts available and it
would not be possible to rehearse Kant’s arguments in full or assess
them.28 Instead I will examine the ways in which both editions reveal
the pluralist relation among the faculties I have been adumbrating.
Nevertheless, because Kant is ambivalent about the role played by the
imagination and this leads to much disagreement among his inter-
preters, I will need to go into some detail even on this issue. I will show
how an insufficiently articulated account of the relation between the
faculties contributes to Kant’s prevarication as to whether or not
knowledge exhausts the field of possible experience.
(i) The A ‘Deduction’ and The Identity of Productive

Imagination
Following an introductory section retained in the B edition, the pre-
sentation of the ‘Transcendental Deduction of the Categories’ in the
A edition is divided into two further parts. In Section 2 Kant analy-
ses what he presents as three syntheses, while in Section 3 the same
syntheses are treated as three aspects of one systematic account.29
It may seem most easy to support a non-impositionalist reading
starting from the A edition ‘Deduction’. This is the position of
Heidegger, for whom the A edition has the potential to become an
ontological account insofar as the imagination affords access to
things and not just representations.30 However, I will argue that both
the A and B ‘Deductions’ provide an ambivalent account of the rela-
tionship between imagination and understanding. In both editions the
relations between the faculties and the central mediating role of the
imagination have to be teased out, while ultimately the understand-
ing is dominant. In short, it is not, as Heidegger claims, the case that
the A ‘Deduction’ reveals the imagination as the root of sensibility
121
and understanding, while the B ‘Deduction’ makes the understanding

the final source of knowledge. Both versions require that imagination
is the necessary mediator between the faculties, while at times com-
promising the identity of imagination in favour of understanding.
In Section 2 of the A edition Kant presents his position through an
analysis of three syntheses. Commentators have often suspected that
this version of his argument falls into the trap of offering a faculty psy-
chology, as he identifies each of the syntheses with an operation of the
mind. It should be remembered, however, that Kant’s analysis of
mental activity, which is to be found in both editions, is normative, as
Longuenesse points out, in that it operates at the level of the structure
of the mind and not its contents.31 Kant’s account in Section 2, in par-
ticular, has also been criticised for giving the impression that intuition
is a prior operation, detached from that of the understanding. I will
show, however, that in Section 2 Kant avoids this trap as well, although
I will suggest that he leaves room for a level of integrity within intu-
ition that falls short of the full unity required for knowledge.
The synthesis of apprehension in intuition establishes coherence
within the sensory given in that the temporal manifold of intuition is
‘run through, and held together’.32 If an intuition is a representation
for consciousness, then the impressions out of which it is made up
must be unified.33 This occurs over time. Kant is not concerned here
with re-identification over time, or even with our capacity for main-
taining attention to an object over an extended period of time. It is
rather that he is insisting that our consciousness of something in the
present requires that we are able to hold together temporal moments
as the content of our consciousness.
Kant says that without this synthesis the representations of space
and time would not be possible.34 If it later transpires that under-
standing already has a role to play in the synthesis of apprehension,
this might seem to count as persuasive evidence for Longuenesse’s
synthesis speciosa thesis.35 However, I will show that there is scope
for resisting the conclusion that the forms of intuition are a covert
operation of the understanding.
There are grounds for arguing that the synthesis of apprehension
entails more than the forms of intuition tout court. Whereas the man-
ifold of intuition is given in the forms of space and time, that mani-
fold is only ‘held together’ through a synthesis in apprehension. If this
distinction goes through, then we can argue that even if the under-
standing plays a role within the synthesis of sense, the former is
excluded from the framework conditions of the latter. The forms of
122
intuition turn out to be independent of the understanding, even if the

synthesis in apprehension is not.
Potential additional evidence for a distinction between the forms of
intuition and the manifold appearing within those forms arises in
Kant’s occasional mention of a synopsis of sense that coincides with
the manifold in intuition. He says on one occasion that a synthesis
must correspond to this synopsis.36 This sounds as if the manifold in
intuition coheres (in some sense that has not yet been identified) prior
to a synthesis of apprehension. The suspicion is reinforced by Kant’s
immediately going on to distinguish between receptivity and spon-
taneity as the sources of knowledge, suggesting that the first corres-
ponds to synopsis, while only the second counts as a synthesis.37 This
may correspond to a distinction Kant makes in the second edition of
the ‘Deduction’ where he says that the forms of intuition give only a
manifold, whereas formal intuition gives unity of representation.38 We
could, then, conclude that the manifold in intuition arises as a syn-
opsis, whereas a synthesis is required if there is to be unity of that
manifold. Thus the idea would be that the forms of intuition, inde-
pendently of understanding, introduce within the manifold a coher-
ence of sorts that does not yet count as unity. However, this suggestion
must remain speculative and stands in tension with an earlier passage
in the A edition where Kant appears to identify the synopsis of sense
with the first of the three syntheses necessary for knowledge.39 While
it is possible that Kant meant to distinguish a way in which the man-
ifold holds together prior to synthesis, the evidence is at best incon-
clusive and I will not investigate the potential nature of such a synopsis
any further here.
Even if the evidence for a distinction between synopsis and synthe-
sis is not persuasive, it is still possible to argue that in his account of
the synthesis of apprehension Kant is concerned with space and time
as experienced, that is, strictly, with representations in space and time
and not with the pure forms of experience. While the first paragraph
reiterates the general finding of the ‘Transcendental Aesthetic’ –
namely, that all appearances belong to inner sense – the second para-
graph turns to the conditions of the unification of the sensory given.40
Thus when Kant says that the representations of space and time arise
from the synthesis of apprehension, he is explaining the first stage or
condition of the process of unification of the given in space and time
and is no longer addressing the initial aesthetic condition of experi-
ence, namely, that the given must arise within the singular context of
space and time as the forms of intuition. So even if this first synthesis
123
makes possible representations in space and time, it does not make

possible the forms of intuition.41 The latter remain distinctive condi-
tions of the unification of experience. This reading is reinforced by
Kant’s account of apprehension in Section 26 of the B ‘Deduction’, as
I will argue later in this chapter.42
But there is a further and independent way in which Longuenesse’s
interpretation can be answered. If I can show that the synthesis of
apprehension in intuition is not grounded in the understanding, while
still standing in a necessary relation to the latter, then we need not
conclude that the synthesis of apprehension results from a covert
exercise of understanding.
The account of the synthesis of reproduction in imagination is par-
ticularly resistant to clarification. Kant moves between two levels of
argument. At the outset of his discussion the second synthesis appears
to be merely supplementary to apprehension. If we are to re-identify
an object over time, then we must be able to presuppose that objects
are subject to a law of association. His examples reveal that he is
thinking about the re-identification of empirical objects or events, for
instance cinnabar, the human form and the weather conditions on the
longest day of the year. Association, be it causal or merely reproduc-
tive as is under discussion here, is not simply the result of the cus-
tomary association in our minds, but must also must be found in
objects. Kant says that this is the case because objects, as appearances,
are subject to a rule of association.43 Only in the second Analogy of
Experience will Kant present his considered account of how a rule
grounds causal association in objects. Kant’s response to Hume in the
A ‘Deduction’ seems to be, at best, indirect. He is not principally con-
cerned with causal association, but with the re-identification of one
image over time.
But now Kant moves to a different level of analysis, for he says:
But if I were always to drop out of thought the preceding representations
(the first parts of the line, the antecedent parts of the time period, or the
units in the order represented), and did not reproduce them while advanc-
ing to those that follow, a complete representation would never be
obtained . . .44
At first he was concerned with the problem of re-identification over

time, but now he turns to the retention of consciousness of a phe-
nomenon within the present. ‘Reproducibility of appearances’ now
refers to a curious sort of reproduction, for at a deeper level Kant
considers how we hold onto an object of consciousness within the
124
extended present.45 Kant’s examples reveal that he has turned his

attention to the internal structure of our time-consciousness of a par-
ticular phenomenon, rather than the problem of its re-identification
over time. He says that we could not draw a line in thought, think of
the time within one twenty-four-hour period or even represent a par-
ticular number if we could not retain earlier impressions in attending
to such phenomena.
It is this shift in levels that explains Kant’s comment that ‘the syn-
thesis of apprehension is thus inseparably bound up with the synthesis
of reproduction’.46 Examination of our ability to retain consciousness
of something within the extended present reveals that we can only run
through and hold together the manifold in intuition in that we are
capable of reproducing or retaining a content of consciousness as the
same thing. While our intuition allows us to take in something given,
the retention of that given over time requires the faculty of imagina-
tion. Imagination is our capacity for representing objects even when
they are not present to consciousness. In this case the object is present
to consciousness and the imagination holds together our fleeting
impressions (Eindrücke). The synthesis of apprehension is thus nec-
essarily connected with the synthesis of reproduction, without which
there would be no unity in intuition. If the second synthesis were
simply concerned with re-identification over gaps in time and not
retention within an extended present, the activity of imagination
would be merely supplementary and not necessarily connected with
the operation of intuition. An impression can only qualify as an image
insofar as it is retained over time; this requires we have not only a
capacity of intuition, but also imagination.47
If a complete representation and ‘even the purest and most elemen-
tary representations of space and time’ require an operation of the
imagination, then must we conclude that the first and second synthe-
ses are not distinct from one another?48 It is not necessary, however,
that we reduce apprehension and imagination to one function, in
expectation perhaps of their subsequent reduction to a higher power
of understanding. Kant’s point need only be that having a complete
representation (and perhaps even an incomplete or indeterminate one)
requires a complex operation of more than one mental capacity. He is
not therefore committed to the view that imagination is, on consider-
ation, capable of replacing our capacity for apprehension. Such a con-
clusion would in any case be at odds with his rejection of material
idealism, for it would entail that imagination supplied the affective
material of experience and did not have to rely on external, though
125
empirical, matter. All we need conclude is that the synthesis of repro-

duction extends the range of the synthesis of apprehension, by making
us capable of sustaining the affective given in consciousness beyond a
merely fleeting impression.
What is the unity to which Kant refers in his account of the syn-
thesis of apprehension? So far it seems only to entail that the mani-
fold can be run through and held together. This could count as a
rather low level of integrity. However, Kant has now claimed that a
complete representation requires an operation of the imagination.
Does this development in his account mean that sense achieves no
holding together whatsoever or only that a higher degree of integrity
is required if we are to focus on an impression as an image? The text
at least seems to leave open the possibility that the latter could be the
case, for Kant does not yet claim that objects that are not known are
nothing to us. We will see that Kant closes down this possibility in
later versions of his argument, arguing that even apprehension
requires unification under a concept.49
In my consideration of the synthesis of recognition in a concept, I
will focus exclusively on the relation in which it stands to the previous
syntheses. Kant now announces that a third synthesis is necessary if the
reproductive activity of imagination is to give rise to knowledge: ‘If we
were not conscious that what we think is the same as what we thought
a moment before, all reproduction in the series of representations
would be useless.’50 Reproduction requires consciousness of unity if it
is to be of any use and we can only suppose that the use in question is
epistemic. Unity – Kant slips in without anything much in the way of
explanation – entails that we have a concept. He immediately provides
the missing link by defining a concept as ‘unitary consciousness’.51 A
concept simply is a combination of the manifold in time, which as he
has just argued entails that it is reproduced or retained.
We can now return to Kant’s mention of unity in intuition in his
account of the first synthesis. Only under a concept is a fully-fledged
unity of intuition achieved, although this rests on the first necessary
condition of unity, namely that the manifold is run through and held
together in intuition. We cannot be sure if Kant intended to leave open
the possibility that some intuitions may not be determined by con-
cepts.52 But we can be certain that knowledge as a unity of our intu-
itions over time requires not only understanding, but also the
contributions of intuition and imagination. We can thus address the
question: are the three faculties best captured as reducible to one
complex operation of the Vermogen zu urteilen or as three distinct
126
orientations working in conjunction? In Section 2 there seems to be

little reason for coming to Longuenesse’s radical conclusion. The
forms of intuition are presupposed by, rather than being the primary
topic of, the synthesis of apprehension; and while understanding
works in conjunction with imagination to give rise to a form or image,
this is best understood as a complex operation linking our receptive
and reflective capacities.
Kant begins Section 3 with a complex commitment. Initially he
announces that the syntheses that were examined in distinction from
each other in the previous section will now be treated in systematic
interconnection.53 He then adds a statement that appears to supply
strong evidence for the distinctiveness of the faculties: ‘There are three
subjective sources of knowledge upon which rests the possibility of
experience in general and of knowledge of its objects – sense, imagin-
ation, and apperception.’54 While there is only one integrated synthe-
sis, it arises from three distinctive sources. A cooperation of a
plurality of faculties is thus necessary
The cooperative status of the faculties is reinforced later in the
section when Kant says that understanding is the unity of appercep-
tion, exercised in relation to the synthesis of imagination. He also says
that the categories of the understanding relate to objects only by
means of both intuition and the faculty he now presents as operating
on the latter, namely, imagination.55 The faculties must work in con-
junction with one another if experience is to arise. Interpreting the
first of these claims as entailing that imagination is a covert operation
of the understanding simply does not capture the intricacy of the
overall picture, where Kant sometimes traces imagination back to
intuition and sometimes to understanding. Only a relational model of
the faculties can account for this complexity.
Kant now introduces the notion of a productive synthesis of imag-
ination that, in contrast to the reproductive imagination, is a priori.56
Productive synthesis is identified as the active faculty for synthesis or
combination and is immediately directed to perceptions or apprehen-
sions.57 Not only is imagination identified as one of the three funda-
mental sources of experience, but Kant also suggests that it may be
first among equals: ‘Thus the principle of the necessary unity of pure
(productive) synthesis of imagination, prior to apperception, is the
ground of the possibility of all knowledge, especially of experience.’58
The supposed priority of the imagination supports Heidegger’s
reading that the imagination is the hidden root that joins intuition
and apperception.59
127
However Kant’s distinction between reproductive and productive

imagination eventually puts in jeopardy the supposed autonomy of
the latter. Imagination only counts as transcendental and productive
if it stands in relation to the original unity of apperception; only as
such is imagination the pure form of all knowledge.60 Reproductive
imagination alone cannot achieve the unity or affinity of appearances
that arises from the principle of the unity of apperception.61 Kant goes
on to say that affinity ‘is a necessary consequence of a synthesis in
imagination which is grounded a priori on rules’.62 It is thus imagin-
ation exercised in conjunction with apperception and, indeed, acting
according to the rules of the latter that counts as productive.63 While
this account does not necessarily threaten the claim that apperception
has to cooperate with imagination, it raises considerable doubts
about the view that the latter is in any sense prior to intuition and
apperception.
Having initially established the productive role of imagination,
Kant announces he will show the necessary connection between the
categories of the understanding and appearances ‘starting from
below’, that is, starting from empirical appearances.64 As was already
established in the ‘Transcendental Aesthetic’, what is first given to us
counts as appearance. When we are conscious of the given in intu-
ition, this qualifies as perception. Kant adds in parentheses that con-
sciousness is a necessary condition of knowledge, a claim that is
clearly in line with his dualist perspective for which knowledge arises
from a combination of an aesthetic given with reflective conscious-
ness. However the precise statement Kant offers is that if appearances
do not stand in ‘relation to a consciousness that is at least possible,
appearance could never be for us an object of knowledge, and so
would be nothing to us’.65 The latter phrase seems to rule out a pos-
sibility he previously asserted, namely, that there could be an aware-
ness or intuition of things, without our achieving knowledge of them.
It could be argued that here he restricts the range of consciousness to
coincide only with the level of awareness that qualifies as knowledge,
leaving open that there could be another level of awareness that falls
short of the unity that corresponds to apperception. Nevertheless the
suggestion that an appearance would be nothing for us if it were not
an object of knowledge tells against this. I believe there is sufficient
ambiguity in Kant’s account to support the view that he has not ade-
quately clarified the levels at which objects function in his account.
At this stage the problem has not clearly emerged and, consequently,
is impossible to dissipate.
128
The problem of the exclusion of not-yet determined objects is more

explicit at the outset of Section 3 where he says: ‘Intuitions are
nothing to us, and do not in the least concern us if they cannot be
taken up into consciousness, in which they may participate either
directly or indirectly. In this way alone is any knowledge possible.’66
However, while, appearing to rule out unsynthesised intuitions, this
earlier passage leaves a loophole by introducing a distinction between
direct and indirect consciousness of things. This leaves open the pos-
sibility that not all intuitions give rise to the direct or determinate
awareness necessary for knowledge, while all representations in prin-
ciple or indirectly stand in relation to the categories and so could be
the objects of knowledge. This can explain Kant’s subsequent claim
that representations must ‘at least be capable’ of being connected in
a unitary consciousness.67 While we are only in principle capable of
connecting all our representations within a unified system of con-
sciousness, those that are so connected will qualify as knowledge. But
Kant is not attentive to the limitations of knowledge in the central
argument of the A ‘Deduction’, where, as we have seen, all appear-
ances count as perceptions and thus are subject to apperception.68
Whereas I argued that, in Section 2, Kant left open the possibility
that there could be a low level of integrity among representations
without their necessarily attaining the unity afforded by apperception,
Kant now seems to rule this out by presenting the relation between the
three elements of synthesis as one of entailment. Appearances contain
a manifold that must be combined. This combination ‘such as they
[appearances] cannot have in sense itself’ arises from the imagina-
tion.69 Thus, what was first presented as the synthesis of apprehension
in intuition is now announced as the work of an ‘active faculty’, that
is, imagination in its productive guise.70 Kant goes on to declare that
the representation of the form of an image requires that the imagina-
tion is able to reproduce the combination it has achieved.71 It should
be noted that the form of an image referred to here clearly denotes the
form of an empirical appearance and not the pure forms of space and
time. Therefore there is no implication, in this statement at least, that
the forms of intuition arise from the productive imagination. Even in
Section 2 the holding together of impressions over an extended
moment in time required an operation of the imagination. Knowledge
only arises insofar as the form or image conforms to a rule, the empir-
ical rule of reproduction.72 This subjective rule only allows us to grasp
objects in that it, in turn, has an objective ground in the affinity of
appearances, which requires an operation of apperception.73 Thus any
129
perception whatsoever requires the application of the categories to

intuitions.
The three syntheses appear to be firmly bound to one another,
leaving no room for a distinction between different levels of conscious-
ness or unity. But, in fact, there is a potential gap between the second
and third stages of this new argument. Kant could have said that only
intuitive representations giving rise to knowledge are unified by under-
standing. He could then have distinguished between determined intu-
itions – for which he could have retained the title ‘perceptions’ – and
underdetermined intuitions, which require only intuition and imagin-
ation, while falling short of full determination under a concept. If Kant
had not started out by remarking, almost in passing, that an object
would be nothing for us at all were it not to give rise to knowledge, his
main argument need not have come to such a radical conclusion.
Despite Kant’s withdrawal of a certain level of autonomy from
both apprehension and imagination, we need not necessarily con-
clude that this implies that imagination and intuition are covert
activities of the understanding even in strictly cognitive activity.
Admittedly, the cooperation among the faculties is not one of full
reciprocity, for in the cognitive case, the understanding is dominant
in its role of introducing unity into the manifold of intuition.
However, as I have shown, it can only achieve this through the coop-
eration of intuition and imagination, thus the model is one of hege-
mony rather than of imposition. The understanding must work
through the intercession of the faculties of intuition and imagination.
And, thus, while the account of the productive synthesis of imagina-
tion in Section 3 of the A ‘Deduction’ identifies the third faculty as
exercising rules that ultimately are derived from the understanding,
we need not conclude that apperception simply imposes order on
sensory appearances. In Chapter 1, I argued in response to Paul Guyer
that the centrality of apperception in Kant’s accounts of affinity and
necessity did not necessarily entail impositionalism, because form ini-
tiates, but is not the sufficient condition of, order. Now it has emerged
that form arises not as a unitary activity on the part of one faculty,
but as a cooperation of a plurality of faculties. Apperception achieves
affinity and necessity only as a complex operation, one dimension of
which is receptivity to empirical objects given to sensibility.
While I have argued that imagination is not primary, no more is it
merely subsidiary to understanding, being instead the mediating
faculty par excellence. A passage, often remarked on, tells against
concluding that imagination is merely an instrument of apperception:
130
A pure imagination, which conditions all a priori knowledge, is thus one

of the fundamental faculties of the human soul. By its means we bring the
manifold of intuition on the one side, into connection with the condition
of the necessary unity of pure apperception on the other. The two
extremes, namely sensibility and understanding, must stand in necessary
connection with one another through the mediation of this transcenden-
tal function of the imagination, because otherwise the former, though
indeed yielding appearances, would supply no objects of empirical knowl-
edge, and consequently no experience.74
We can make sense of the apparent tension in which this stands to

other passages, where Kant appears to reduce imagination to an exer-
cise of the understanding only insofar as we recognise the relational
status of the faculties. In this passage Kant uses faculty talk to insist
on the dualist perspective, that is, there are at least two fundamental
dimensions of experience, namely sensibility and understanding. His
insistence that these must be set in relation to one another necessitates
the introduction of a third term, the imagination, which also counts
as one of the fundamental faculties of the human soul. This latter
claim is prima facie evidence for the claim that there must be a
cooperation among a plurality of faculties.75
Imagination is not only the third faculty but is also the relational
faculty par excellence in that it is our capacity for mediating between
different orientations. If there were not something given to the mind
in sensible form, then there would be no knowledge; just as if there
were no mental capacity for unification, there would be no knowl-
edge. The extra-mental given and mental formative power are both
necessary for knowledge and are in no way reducible to one another.
The two poles of dualism are not simply opposed, or indeed two
aspects of one faculty, but instead must be related and for this a third
term is necessary. The impositionalist account cannot accommodate
the intricate examination of the determination of the empirical by
these a priori forms. It is bound to find the elaboration of the roles of
the faculties an idle wheel. To grasp the complexity of Kant’s position,
we need a relational and dynamic account of formalism.
Kant’s account, while recognising the distinctiveness of receptivity
and understanding, at times undermines this by the emphasis he puts
on the role of spontaneity within synthesis. Had he emphasised the
reflective status of consciousness, commentators might have been less
inclined to conclude that he is committed to impositionalism. While
spontaneity – that is, the mind’s capacity to initiate order at the
formal level of experience – is a necessary feature of his account, Kant
131
sometimes sounds as if he is saying that the mind introduces order as

a pure act. When this is combined with his tendency in the first
Critique to leave insufficient space for non-cognitive awareness of the
world, Kant can appear to have a very top-down account of experi-
ence. However, I have shown that even in the face of this, the A
‘Deduction’ reveals his commitment to the need for a cooperation of
a plurality of faculties. This provides at least the beginnings of a more
comprehensive account of experience, where mind must be receptive
if it is to achieve knowledge and non-cognitive apprehension has a
place.
(ii) The B ‘Deduction’: Figurative Synthesis and a Renewed

Attention to the Mode of Intuition
We have already seen how the imagination facilitates the relation of
understanding and intuition in the A ‘Deduction’. I intend to show
that the B ‘Deduction’ also reveals the mediating power of imagina-
tion through the new notion of figurative synthesis, which Kant con-
trasts to intellectual synthesis. On closer examination, we will see that
it is not so much that there are two different syntheses as that there
are two levels of analysis of what is necessary if the categories are to
be synthesised with intuitions or objects.76 Synthesis is analysed first
as intellectual, and second as figurative. I will argue that the contrast
between the former’s two-term and the latter’s three-term relation
establishes the plural conditions of the epistemic task. There has been
much debate about the B ‘Deduction’, which has generated perhaps
more literature than any other section of the Critique. My limited task
here will be that of assessing how pluralist or reductionist an account
Kant gives of the relation between the faculties.
I agree with Allison and Henrich that there are two parts to the
‘Deduction’ as presented in the B Edition and that these two parts cul-
minate in Sections 21 and 26, respectively. Allison’s interpretation,
with which I am in general agreement, hangs on a terminological dis-
tinction between Objekt and Gegenstand where the former stands for
a logical object and the latter for an object of experience. As is now
well established, Allison’s interpretation is that the first part of the
‘Deduction’ is concerned with establishing only that the categories
apply to logical objects, while the second establishes that they apply
to objects of experience. The general claim that the ‘Deduction’ grad-
ually works from a more logical position towards a more experien-
tial one is, I believe, convincing. My reading differs from Allison’s
132
insofar as I draw out the subjective side of the story. The two stages
reveal not only two different analyses of the object, but also two
stages of analysis of the subjective conditions of knowledge. My point
is not to replace an objective account with a subjective one, but rather
to show how the two sides are mutually implied by each other. Kant’s
question, as I read it, is: how can we philosophically reconstruct the
activity of the subjective modes of knowledge or faculties so as to
establish how they can give rise to objective knowledge? Or, how can
a subject know an object?
In agreement with Longuenesse and Allison, I take the first part
of the B ‘Deduction’ to be a development of the ‘Metaphysical
Deduction of the Categories’ from the ‘Table of Judgements’.77 What
Kant calls intellectual synthesis recapitulates the forms of thought,
which in general logic function in abstraction from application to
objects.78 Those same forms of thought are now transformed into cat-
egories that make possible the unity in our sensible intuitions. Mere
forms of thought become categories when they are set in relation to
the content of experience, namely, intuitions. Thus, one more element
is added to the logical monism of the ‘Metaphysical Deduction’,
where thought operates in isolation.79 The first step towards a plu-
ralism of faculties has been taken.
Anticipating a more complex account, Kant reveals on several
occasions that the emphasis in this part of the text is one-sidedly ori-
ented towards the understanding in its guise as the ‘original unity of
apperception’. He reminds us that there is another side to the story.
For instance, in Section 17 he states:
Insofar as the manifold representations of intuitions are given to us, they
are subject to the former of these two principles [the forms of space and
time]; insofar as they must allow of being combined in one consciousness,
they are subject to the latter [original synthetic unity of apperception].80
Thus Kant is well aware that the presentation of his argument results
in emphasising one side of dualism over the other, temporarily at
least.
This same passage poses a question I raised in my discussion of the
A ‘Deduction’. In what sense is anything given to us necessarily
unified or combined by apperception? If this were the case then it
would be impossible for us to have sensations without also concep-
tually identifying the latter. As we have seen, Kant states in both edi-
tions that not all appearances are conceptually determined, but we
have also seen that he is inconsistent on this issue. The situation is
133
exacerbated by the fact that, whereas the imagination acts as inter-

mediary in the A edition, here concepts operate directly on intuitions.
This would encourage the impositionalist reading in that concepts
would immediately apply to sensations. There would be no gap
between mind and world, no task for mind to undertake in the world
other than one already guaranteed success. This danger is diminished
in the second part of the B ‘Deduction’ by the introduction of figura-
tive synthesis, while the formalist interpretation allows us to read the
passage under consideration as saying that it is only if the manifold is
taken up in such a way that it is unified so as to give rise to knowl-
edge that it is subject to apperception. This would leave open the pos-
sibility that there may be different modes of experience such as the
aesthetic where determinate unity is not achieved, nor indeed sought.
In the conclusion of the first part of the ‘Deduction’, Kant remarks
that he has abstracted from the mode in which the manifold is given,
although he has not abstracted from the fact that the manifold is given
‘prior to the synthesis of understanding and independently of it’.81
This last phrase reinforces the argument that synthesis arises out of a
cooperation of distinct but connected faculties. As we have seen, even
in the account of intellectual synthesis, concepts stand in relation to
intuition. But we can now make sense of the distinction between the
two versions of synthesis, as he goes on to say: ‘How this takes place,
remains here undetermined’.82 Kant’s point is that in the first part of
the deduction he has simply taken the relation between understand-
ing (in its guise as apperception) and intuition for granted. He merely
states that intuitions must be subsumed under concepts if knowledge
is to arise. Now he is going to identify the process through which the
relation between the two faculties is possible and this will require
focusing on the combination of the distinctive form or mode of intu-
ition with the forms of understanding or categories.
While in the first part of the B ‘Deduction’ Kant progresses to a
two-term model of knowledge comprising apperception on one side
and sensibility on the other, this account is inadequate on two inter-
connected grounds. The polarised account presented as intellectual
synthesis is incapable of showing how categories can be applied
within experience and, at the same time, of showing how two seem-
ingly opposed mental faculties can be coordinated in the interests of
knowledge. Indeed these two failures converge, for it is only if Kant
can show the possibility of the cooperation of the faculties, that he
can vindicate his claim that we have access to an empirical given. The
relation between the subjective conditions of experience must be
134
examined if the objective validity of the categories is to be established.

This is not because subjective form is imposed on external matter, but
rather because one of those faculties, sensibility, is nothing other than
our capacity to take in the empirical given, as we saw in the previous
chapter. The other subjective condition, the understanding, must be
able to unify the given if knowledge is to be achieved. This entails a
cooperation of the faculties.
In the two-term account of mind, the only possible scenario, given
the Copernican revolution’s insistence on the a priori element in
knowledge, is that apperception imposes form on the sensible given.
But this would mean that what counts as knowledge would be no
more than the projection of our subjective impressions and not a
grasp of an extra-mental world. It is because this would not be a sat-
isfactory conclusion to his epistemic project that Kant goes on to
develop the argument of the deduction so as to focus on a relation of
form and matter and not simply an imposition of one on the other.
In the second part of the B ‘Deduction’, the relation between
understanding and intuition is brought into focus, for knowledge
arises not only from the forms of thought, which Kant now retro-
spectively suggests were the topic of the first part of the ‘Deduction’.83
The unity of apperception in itself knows nothing ‘but merely com-
bines and arranges the material [der Stoff] of knowledge, that is, the
intuition, which must be given to it by the object’.84 The intuition that
must be given is not merely the formal condition of intuition, that is,
the pure intuitions of space and time. This would give rise only to
mathematical concepts that do not themselves count as knowledge.
Knowledge requires an empirical intuition, with its material content
or sensation.85 Apperception thus provides merely the form for
arranging a material given that must be given. Only thus do the cat-
egories achieve objective reality.86
If pure concepts are to be combined with empirical intuitions so as
to give rise to knowledge, synthesis must be more than merely intellec-
tual.87 Figurative synthesis replaces a two-term relation with a three-
term one. Sensibility and understanding are now joined by imagination,
which is not simply a third party but the very ground of the possibility
of mediation between the other two mental orientations. The tran-
scendental imagination makes possible the application of concepts to
empirical intuitions that is necessary if knowledge is to arise.
Nevertheless, the status of the imagination as third faculty is diffi-
cult to express. As mediator, it is hybrid in status. Kant claims both
that imagination belongs to sensibility and that it is an action of
135
understanding on sensibility.88 Just because the latter claim comes

later in the account of imagination’s origin, it might seem that it
trumps the former statement of allegiance to sensibility. Kant suggests
that figurative synthesis, which he also calls the transcendental syn-
thesis of imagination, is comparable to the role of the productive
imagination, first mentioned in the A ‘Deduction’.89 As we saw, the
latter is closely allied with the understanding. Leaving aside the
lineage of the imagination for the moment, what is critical about both
accounts is that imagination emerges as the capacity to generate the
form of an image.90 This explains the relevance of Kant’s comment
that imagination is the ‘faculty of representing in intuition an object
that is not itself present’.91 The form is the representation of an
object that can be attended to even when its material correlate is not
presented in intuition. Imagination is thus the linking term between
the mere impressions of intuition and the concepts of understanding.
Grasping the form of an object is the first step towards knowledge
and this is achieved by the imagination.92
We have two alternatives in response to the account of the rela-
tionship in which imagination stands to the other faculties. Either
we tidy up Kant’s account and conclude that only some of his
statements express his considered view. We may then conclude that
the imagination is not really a third faculty but only an outpost of the
understanding. This would lead to problems later on insofar as
aesthetic judgement rests on a use of the imagination that is free
from the rules of understanding. Moreover in the strictly epistemic
context it condemns Kant’s epistemology to the pyrrhic victory of
impositionalism.93 Alternatively, we can see Kant’s various character-
isations of the relations in which the different faculties stand to one
another as revealing a struggle which he has not yet the means of
resolving. We can then recognise that imagination is a problematic
third term, which, at times, is an ally of sensibility and, at others, of
understanding. It is this very status as mediator par excellence
that leads to the problem of identifying it in a definitive way. 94
Kant will not find a methodological position that is sufficiently
subtle for expressing its problematic status before the third
critique.
What is undeniable is that from Section 21 through to Section 24,
Kant insists that apperception is the framework within which the
material given must be ordered. This requires what he calls figurative
synthesis, which mediates between the faculties of sensibility and
understanding. While the status of the imagination is difficult to pin
136
down, it is clear that its function is to make possible the application

of the formal conditions of experience to empirical matter.
(iii) The Re-identification of the Synthesis of Apprehension

in Section 26
The introduction of figurative synthesis in Section 24 is evidence for
Kant’s continuing commitment to the view that knowledge arises from
a plurality of faculties. However, the culmination of his argument in
Section 26 once again invites a more reductionist interpretation. This
is one of the most difficult passages and particularly challenging for
my reading. What we can call broadly the proof of the ‘Deduction’ can
be reconstructed in four steps with a prefatory definition:
Def: Synthesis of apprehension is ‘that combination of the manifold in an
empirical intuition, whereby perception, that is, empirical consciousness
of the intuition (as appearance) is possible’.95
1. The synthesis of apprehension must always conform to the forms
of sensible intuition.
2. Space and time are not just forms, but are also intuitions contain-
ing a manifold. As this is the case, their representation requires
‘the determination of the unity of this manifold’.96 At this point
Kant refers us to the ‘Transcendental Aesthetic’ and adds a now
famous footnote.97
3. Thus, unity of the synthesis of the manifold and therefore also
combination ‘to which everything that is to be represented as
determined in space or in time must conform’ is the a priori con-
dition of the synthesis of apprehension.98 He adds that this condi-
tion of unity is given ‘not indeed in, but with these intuitions’.
4. This unity is in an original consciousness in accordance with the
categories ‘insofar as the combination is applied to our sensible
intuition’.99
Thus Kant is able to conclude:
All synthesis, therefore, even that which renders perception possible, is
subject to the categories; and since experience is knowledge by means of
connected perceptions, the categories are conditions of the possibility of
experience, and are therefore valid a priori for all objects of experience.100
This proof offers a recasting of the argument of the A ‘Deduction’,
addressing the problem of how the given in intuition is determined in
time. The main difference is that there is no direct correlate for the
137
synthesis of reproduction, or, indeed, for the productive synthesis of

imagination. In the conclusion to the B ‘Deduction’, Kant fails even
to integrate the account of figurative synthesis he has just introduced
in Section 24.
The Definition makes clear, were there ever any doubt on the issue,
that Kant’s aim is to establish the applicability of the categories at the
empirical level. The first step establishes that all objects of empirical
experience appear within the horizon of space–time or the forms of
sensible intuition. This may count as a correlate for the synopsis of
sense that was mentioned in the A edition. However Kant does not
suggest that the forms of intuition provide a distinguishable way in
which the manifold has a degree of coherence short of conceptual
unity. What he does claim is that the forms of intuition are a first con-
dition of the determination of empirical experience. He states that the
synthesis of apprehension must always conform to these forms,
because otherwise synthesis would be impossible. While this does not
rule out the possibility that the forms of space and time may later be
trumped by the categories of understanding, at first sight this is strong
evidence for resisting the view that the forms of intuition can be
traced back to an operation of the understanding, however covert.
In the second step Kant distinguishes between the forms of intu-
ition and experienced or empirical intuitions thus reinforcing the dis-
tinctiveness of both. 101 Space and time are either the conditions of
experience, or they are experienced. Insofar as they are experienced
they are not merely formal: they have content. Empirical intuitions
mark out identifiable segments of the infinity of space–time. Whereas
we could never intuit the whole of space and time, we can have intu-
itions of individual spatio-temporal objects. But if intuitions are to be
identifiable as having content, then they must have some degree of
integrity. This entails that they are unified.
Now while empirical intuitions, which I discussed in Chapter 3,
must display some unity, it is not clear that Kant should have gone so
far as to claim that they require ‘the determination of this unity’ as he
now does.102 We have seen that Kant denied this on two occasions at
the outset of the ‘Deduction’ and in both editions.103 I have also sug-
gested that the version of the deduction presented in Section 2 of the
A ‘Deduction’ leaves open the possibility that there could be two
levels of unity, where the first requires only that intuition conform to
the general form of the categories. However, in the second step of the
conclusion to the B ‘Deduction’, Kant implies that all apprehension is
determinately unified. While it is clear that knowledge requires deter-
138
mination, Kant is in danger of suggesting that all sensory apprehen-

sion is cognitive.
We can now turn to the especially difficult footnote at B 160 where
Kant expands Step 2 of his account of the possibility of empirical
knowledge. I have already established that in this second stage of his
argument, Kant is concerned with intuitions as experienced, rather
than merely with the forms of intuition that were the subject of
the first stage. In the footnote he reintroduces geometric intuitions,
because he is intent on revising the account he gave in the
‘Transcendental Aesthetic’, where he provided only a very provisional
account of experienced intuitions. The exception was in his discus-
sion of geometric forms. The ‘Transcendental Exposition’ provided a
defence of the interpretation of space and time as forms of intuition
established in the ‘Metaphysical Exposition’. Kant argued that such
an interpretation underwrites the apodeictic status of geometry.
Retrospectively, we can see that geometric forms, albeit pure or
formal, are experienced intuitions. They are distinct from the forms
of intuition discussed in the Metaphysical Exposition, which are best
understood as the framework within which experience arises and
never directly experienced as such.104 In line with the development
between Step 1 and Step 2, Kant now establishes that the geometric
shapes under consideration in the ‘Aesthetic’ are formal intuitions
and not merely forms of intuition. He must hope that the reintroduc-
tion of this example will clarify his general point about experienced
intuitions and in particular empirical ones, the conditions of which
this proof is intended to clarify.
In the ‘Transcendental Aesthetic’, Kant takes the unity of empiri-
cal intuitions for granted, treating the latter as if they arose from the
power of intuition alone. Intuitions are treated as if they were identi-
fiable and capable of being experienced and yet no account is given
of the conditions of their unification. They are examined only with
respect to their affective side and not as to the possibility of taking up
affect in consciousness. Kant does not explain the reason for this at
any length in the B ‘Deduction’. One reason we can suggest is that his
primary concern in the ‘Transcendental Aesthetic’ was with the forms
of intuition, not with empirical intuitions directly. His intention at the
outset of the Critique was to establish that space and time must count
as pure intuitions, derived neither from experience nor from concepts.
I believe this is why he says that he did not supply an account of the
unity of intuitions in the ‘Transcendental Aesthetic’, where he was
intent only on establishing that intuitions arise from a source that
139
‘precede[s] any concept’.105 The Aesthetic establishes an element of

experienced intuitions, but not their full structure.
Having conceded this omission, he goes on to say that in fact their
unity ‘presupposes a synthesis which does not belong to the senses but
through which all concepts of space and time first become possi-
ble’.106 Longuenesse has taken this statement as crucial evidence for
her claim that the forms of space and time are a first operation of the
understanding on sensibility.107 The first question we must address is
this: is the synthesis in question the synthesis of apprehension in intu-
ition, or the synthesis of imagination, which in the A ‘Deduction’
turned out to be necessarily bound up with the combination of intu-
ition? Here Kant mentions only intuition, but he is concerned with
intuitions as represented and this can only be achieved through the
figurative or productive power of imagination. In the footnote at B
160–1, Kant elides the distinction between intuition and imagination,
presumably on the grounds that the former cannot count as a repre-
sentation unless its content is taken up by the faculty of imagination
so as to give rise to a figure or form. This would then explain why
Kant says that the unity in intuition entails a synthesis that does not
belong to the senses. The synthesis in question is imaginative and not
necessarily intellectual.
Now we know from the first edition that if knowledge is to arise,
imagination must stand in a necessary relation to understanding. But
Kant did not need to conclude that the synthesis of apprehension is
already an operation of the understanding. Unity in intuition can
require more than the senses, without that entailing that the under-
standing alone is the source of unity. Admittedly, the full story of the
unification of intuitions as representations will require the under-
standing, but there is no need to make the imagination an outpost of
intellect. A cooperative model of the relation between the faculties
will account for the complexity of Kant’s account in the footnote. It
is true, though, that Kant’s omission of any mention of imagination
at this stage of his argument encourages the conclusion that under-
standing operates immediately on intuition.
When Kant goes on to say that ‘all concepts of space and time first
become possible’ through this synthesis, Longuenesse concludes that
this means the forms of intuition are only possible through the syn-
thesis of apprehension, which is now identified as an operation of
understanding.108 The possibility that the forms of intuition are out-
posts of the understanding is the second issue we need to address with
regard to the phrase under consideration. But this runs against Kant’s
140
recently repeated distinction between forms of intuition and formal

intuitions. The concepts of space and time in question are determina-
tions of the infinite whole of space and time. Determinations of space
and time, as experienced intuitions, are perceptions and entail unifi-
cation. They require a synthesis beyond that of sensibility, but the
forms of space and time do not. We may suspect that Kant is some-
times insufficiently careful and refers to space and time as concepts
when he should have referred to them as intuitions or, at least, used
a neutral term that left their status open. However, in this passage he
has just used the term concept to refer to an operation of the under-
standing, so it seems likely that he sustains that usage, as he always
should. I, therefore, conclude that the phrase under consideration
tells us that all representations in space and time, not the forms of
space and time, become possible through a synthesis requiring not
only intuition, but also imagination.109
The final sentence of the note begins with the claim that space and
time are ‘first given as intuitions’ insofar as understanding determines
sensibility.110 Now for Longuenesse’s reading this mention of the
understanding merely renders explicit what was assumed all along,
namely, the operation of the understanding on intuition. How,
according to my alternative reading, can I explain the introduction of
the understanding at this stage of Kant’s account? If representations
in space and time are to give rise to knowledge, the establishment of
which possibility is the goal of the overall argument, then they must
be subject not only to a synthesis of imagination, but also must be
determined by a concept of the understanding. Only thus will space
and time qualify not only as the forms of intuition, but also as given
to us as determinate intuitions. As we have seen, at this stage of his
argument Kant does not leave a space for undetermined intuitions. So
his initial phrasing of Step 2, with its early mention of determination,
is echoed by the footnote’s moving rather abruptly from the opera-
tion of the imagination in making possible the form of an object to
the determination of that form by the understanding. While this raises
a problem about the relation between apprehension in general and
knowledge, the distinctiveness of the forms of intuition is not at risk.
In my view Kant simply moves too quickly at this stage of his argu-
ment. All he needs say, as he did in Section 2 of the A edition, is that
the synthesis of apprehension must be bound up with a function of
the understanding, through the mediating role of the synthesis of
reproduction in imagination. However he suppresses the mediate
stage and threatens to transform an indirect relation in which
141
understanding relates to apprehension into a direct one. By failing to

account for the role of imagination, which he had established as the
centrepiece of figurative synthesis in Section 24, he collapses the com-
plexity of his position into a flattened account that invites an imposi-
tionalist interpretation.
My interpretation so far requires no straining of Kant’s sense,
although I do criticise the speed with which he moves from under-
standing to sensibility and his requirement that all apprehension must
be unified. In the second clause of the last sentence of the note my
reading is faced with a final sticking point. Kant now says – as if it fol-
lowed naturally from his claim that the synthesis of apprehension is
the means by which understanding determines sensibility – the unity
in question ‘belongs to space and time, and not to the concept of the
understanding’.111 But why should this be if, as I am arguing, he is
talking not about forms of intuition, but rather about formal intu-
itions, which are necessarily hybrid in origin? Longuenesse is in a posi-
tion to account for this peculiar move in Kant’s argument through her
distinction between an implicit and an explicit operation of the under-
standing on intuition. The first is only preparatory for determination
and thus could count as belonging to intuition.
In answering this question we should remember that the second
step of the proof to which this note is appended is concerned with
unity in intuition. So far Kant has argued that this requires not only
intuition, but also a further synthesis, which I have argued should
be identified as that of imagination. Full determination of intuition
so as to give rise to knowledge would require the additional input
of understanding, in line with Kant’s final goal of showing that even
empirical intuition is governed by the categories. But he has not yet
got so far in his argument. The introduction of the unifying role of
understanding does not occur until Step 4, although it is prepared
for in Step 3. In Step 2 there had been no mention of the under-
standing, prior to the final sentence of the note, where we must
suspect that Kant jumps ahead of himself. My suspicion is that
having prematurely introduced what is as yet only a goal, that is, the
claim that the understanding finally gives rise to the determination
of representations, he steps back to the matter in hand. His argu-
ment at this stage corresponds to the relation between the first and
second syntheses of the A ‘Deduction’, namely to the way in which
intuition is ‘run through and held together’ in intuition in conjunc-
tion with imagination.
Kant remains committed to dualism and thus protects the distinc-
142
tion between the sensory and conceptual conditions of experience.

The synthesis of apprehension in intuition falls under the first of
these conditions. And although, as he later says, it must conform to
the synthesis of apperception, this is not to say that the two condi-
tions of knowledge are identical.112 If Kant had not suppressed and
rather had explored the mediating role of imagination, pivoted
between the sensible and intellectual conditions of experience, he
could have made his point without inviting the conclusion that he is
reducing intuition to an operation of understanding. Admittedly,
later in the main text he says that the imagination and understand-
ing amount to ‘one and the same spontaneity’, but this need not be
read as denying their distinctive roles and only as insisting on the
necessity of the cooperation of the faculties.113 Thus I take the final
clause of the last sentence in the note at B 160 and its claim that
unity ‘belongs to space and time, and not to the concept of the
understanding’ to express a continuing commitment to the plural
sources of experience, however clumsily presented. Even in the B
‘Deduction’, Kant resists reducing the synthesis of apprehension to a
mere operation of the understanding.
Another factor that may well have influenced Kant’s conclusion to
the footnote is the leading example he has used to elaborate the status
of representations in space and time. Geometric figures, while they do
not coincide with the framework of space are also unlike empirical
intuitions in that they express only a pure synthesis of space.
Determinate empirical intuitions are subsumed under a concept, but
what is the role of a concept for a geometric figure? The concept of a
triangle does not subsume or explain the figure: it names something
that can only be apprehended in intuition or imagination. Clearly, in
the case of geometric figures the unity of the formal intuition does
indeed belong to space and time and not to a concept of the under-
standing.
Kant now resumes his argument in the main text. Step Three
states that the unity of the manifold is the condition of the synthesis
of apprehension. He explicitly equates unity with combination,
whereas in the A edition a possibility remained that combination
arose from intuition and imagination and unity only from a concept.
However, even in the B ‘Deduction’ the third step of Kant’s account
leaves open the possibility that he is claiming only a conditional
necessity, that is, unity is necessary only for those representations
that qualify as determinate and amount to knowledge. But by the
same stroke, the potential distinction between combination and
143
unity is annulled by Kant’s conclusion that unity is necessary for the

synthesis of apprehension and not simply for synthesis under a
concept. And if there is no way in which intuitions hold together
short of displaying a determinate unity, then we must conclude that
all our apprehension amounts to cognition. If Kant wanted to main-
tain the possibility that there could be intuitions that are not (yet)
determined, he would have to leave open the possibility that there
were representations that did not yet display the full level of unity
arising from subsumption under a category. In the footnote, as we
have seen, Kant at best fails to articulate this possibility and is in
grave danger of suppressing it entirely.
We need not, however, conclude that this is his definitive intent, for
at the conclusion of the third step he remarks that the unity of appre-
hension is given ‘not indeed in, but with these intuitions’.114 This is
evidence for Kant’s continuing commitment to the distinctiveness of
the aesthetic conditions of experience. What is less clear is whether
the phrase once more opens up the possibility that some apprehen-
sion may not be cognitive. My suggested extension of his argument is
as follows. Unity operates on intuition, but the former is not consti-
tutive of the latter. Intuitions are distinct not just because they arise
from a specifically aesthetic formal ground, the forms of intuition, but
because even empirical intuitions are only unified insofar as the
understanding operates on apprehension, which though necessarily
bound up with the understanding, is not simply identical with the
latter. And if unity arises from a distinctive faculty, then it is at least
in principle possible that we could have intuitions that were not
unified under a concept. While this possibility is left open by the text
under consideration – and we have seen that on other occasions Kant
explicitly espouses it – at this stage he makes no restriction on the
range of knowledge.
The fourth step of Kant’s proof states that unity in appearances coin-
cides with unity in apperception.115 Understanding now, by right,
enters the argument explicitly for the first time. Although Step Two
rather precipitately referred to determinate unity, the need for unity of
apprehension was only established in Step Three. It is not until Step
Four that unity is traced to the understanding. But even now, the under-
standing cannot operate in isolation. The unity of consciousness would
not be achieved, if understanding were not directed to an object, that
is, applied to our sensible intuition.116 Were this not the case, we know
from earlier arguments of the ‘Deduction’, apperception would merely
give rise to forms of judgement and not to categories. The subjective
144
forms of the understanding only attain knowledge when they grasp

appearing objects. This requires that the unity achieved by under-
standing is applied to sensible intuition and thus unity arises from a
complex operation of the faculties.
Kant completes the proof of the ‘Deduction’ by concluding that the
unity of empirical objects can be nothing other than the unity of the
combination of the manifold arising from the categories of the under-
standing.117 As I have argued, this is a foreshortened version of what
is a complex account. While the logic of his argument certainly
requires that the synthesis of apprehension must stand in some rela-
tion to the understanding, he need not have concluded that the
combination of empirical intuition is simply an operation of the
understanding. And Kant could have allowed that some empirical
intuitions have a level of integrity, while failing to be wholly unified
under a concept. Such underdetermined intuitions would still stand
in a potential relation to the understanding.118 These would include
sensory affects that we could, but do not yet, take up in reflection. But
they would also include judgements of taste where both intuitions and
concepts are in play, yet the first is not determined under the second.
Kant need only claim that knowledge requires the full level of unifi-
cation by the understanding. In collapsing the distinctiveness of levels
between the syntheses of apprehension and recognition, while sup-
pressing the mediating role of imagination, he is at risk of making his
technical sense of experience as knowledge coincide with the only
possible way in which we engage with the world. This would result
in cognitive reductionism and invite an impositionalist interpretation
of his epistemology. If he is to avoid this, he must separate out the
component parts of knowledge and, by the same stroke, show how
experience short of knowledge is possible. We will see that he
attempts the first of these in the development of his argument that
follows on from the ‘Deduction’, and indirectly addresses the second
issue in the Critique of Judgement. It is only in the third critique that
the aesthetic – that is, sensory – dimension of Kant’s epistemology is
finally safe-guarded against the reductionist tendencies that we have
seen emerge in the B ‘Deduction’.
Kant’s account of apprehension has shifted since Section 2 of the
A edition where the synthesis of apprehension was presented as a
distinctive synthesis from that of recognition in a concept.119
Nevertheless even there we saw that unity in apprehension finally
arises not just from intuition, but also from the cooperation of
imagination and understanding.120 This extended account is not
145
inconsistent in substance with Section 3 of the A edition where appre-

hension is still associated with intuition, although it is explicitly iden-
tified as an action of the imagination on intuition.121 Nor is there any
necessary shift of position in the final stages of the B ‘Deduction’,
where apprehension is presented as standing in conformity with the
synthesis of recognition in concepts.122 But Kant is at risk of collaps-
ing the necessary relation between two elements of synthesis into an
identity. This is because his omission of any account of imagination
leaves the impression that the unity arising from the understanding
operates immediately within the synthesis of apprehension. By the
same omission, he covers over the possibility that there could be a
non-cognitive apprehension. In the Critique of Judgement another
possibility emerges, for aesthetic judgement counts as an indetermi-
nate and thus non-cognitive apprehension of a sensory object. This
alternative account of apprehension arises from a free operation of
the imagination, the cognitively oriented exercise of which has been
suppressed in Section 26.
At the same time as Kant has difficulty establishing the limits of
knowledge in the ‘Deductions’, he struggles to express the combina-
tion of faculties that is necessary if knowledge is to arise. As I
have tried to show, his relative inattention to the subjective side of his
story contributes to the difficulties that have emerged in his account
of knowledge. At times it sounds as if he is reducing imagination to
an operation of the understanding. But even in Section 26 of the B
edition, more complex formulations show this not to be the case:
‘Now it is imagination that connects the manifold of sensible intu-
ition; and imagination is dependent for the unity of its intellectual
synthesis upon the understanding, and for the manifoldness of its
apprehension upon sensibility.’123 What this reveals is that imagina-
tion is a mediator. It is, indeed, identical to understanding when it
unifies intuition: ‘It is one and the same spontaneity, which in the one
case, under the title of imagination, and in the other case under the
title of understanding, brings combination into the manifold of intu-
ition.’124 But it also belongs to sensibility in that the imagination
makes available intuitions for concepts.125 Kant’s struggle to finally
identify the role of imagination is due to his lack, as yet, of a suffi-
ciently sophisticated methodological approach for capturing the plu-
rality of the faculties and the process of mediation among those
faculties. His account of experience is at risk of being overly
epistemic, not because he relies on a faculty theory, but
because his account of the relation in which the faculties stand to
146
each other and to possible experience is as yet insufficiently devel-

oped.
(iv) Heidegger’s Account of Imagination as the Root of

Sensibility and Understanding
Heidegger’s Kant and the Problem of Metaphysics, while flawed, is
one of the most insightful works into the lasting significance of Kant’s
Critique of Pure Reason.126 His interpretation inspires readers to look
beyond the surface of Kant’s account and search out what remains
unstated, but is made evident in the letter of the text.127
For Heidegger, Kant’s project is not one of epistemology, but is
rather an attempt to provide a fundamental ontology in the style of
his own Being and Time. Such an ontology is not a science of beings
or entities, but rather of the ‘Being’ that is prior to all experienced or
ontic beings.128 Kant, according to this account, is not simply con-
cerned with establishing the epistemic conditions of experience, but
with showing how human beings are capable of transcending experi-
enced beings towards their ontological ground, which he calls Being.
In drawing out this potential in Kant’s position, Heidegger focuses on
the role played by the imagination.
Heidegger’s great insight is in drawing attention to the intermedi-
ary role of the imagination between sensibility and understanding.
His greater insight is to insist that imagination is not simply part of
the apparatus of faculty theory, but is the point of access to human
beings’ finitude and their transcendence. Imagination sets us in rela-
tion to something other than ourselves and makes possible our tran-
scendence of mere experience; at the same time, it reveals that we are
finite beings. Transcendence and finitude are ineliminably linked for
human existence or Dasein.
For Heidegger, the transcendental imagination is linked to our
receptivity or sensibility and, at a deeper level, is the source of our
sensible capacity of intuition. As I have brought out in my reading of
the A edition ‘Deduction’, receptivity is bound up with the capacity
for retention that arises from our capacity for imagination. But, as we
saw, imagination not only cooperates with intuition but also with
understanding. Heidegger holds that imagination is the linking term
between the two extremes, which by virtue of its mediation can be
understood as inseparable from one another.129
Unlike most contemporary interpreters of Kant, Heidegger prefers
the A edition ‘Deduction’ because of the importance it gives to the
147
role of imagination. He does not accept the usual contrast between a

‘psychological’ and a ‘logical’ account:
It should be noted, in truth, that the laying of the foundation is no more
‘psychological’ in the first edition than it is ‘logical’ in the second. On the
contrary, both are transcendental, i.e., necessarily ‘objective’ as well as
‘subjective’.130
Heidegger rightly puts in question the assumption that the emphasis
on the understanding in the B edition qualifies that account as more
‘logical’ than the earlier account.
But it is this emphasis on the understanding that, in his view, puts
at risk the ontological orientation of Kant’s project. The understand-
ing can only determine objects insofar as it stands in a (complex) iden-
tity with imagination.131 This position, established in the third Section
of the A edition is forfeited, in Heidegger’s view, in 1787: ‘In the
second edition of the Critique of Pure Reason the transcendental
imagination, as it was described in the vigorous [leidenschaftlichen]
language of the first edition, is thrust aside and transformed – to the
benefit of the understanding.’132 The imagination still has a role to
play, but Heidegger, in a move that prepares for Longuenesse’s
reading, considers that Kant’s account of the synthesis speciosa entails
a co-opting of the imagination by the understanding:
he shows by this expression [synthesis speciosa] that the transcendental
imagination has lost its former autonomy. It receives this name only
because in it the understanding is referred to sensibility and without this
reference would be synthesis intellectualis.133
Heidegger says that Kant strikes out two passages presenting imagi-
nation as the third fundamental faculty.134 The imagination is even re-
identified as a ‘function of the understanding’ in a handwritten note
found on the margin of Kant’s copy of the Critique, whereas the pub-
lished text in both editions reads that the former is a ‘fundamental
function of the soul’.135 While I have shown how the culmination of
the B ‘Deduction’ in Section 26 suppresses discussion of the imagina-
tion, I have also argued that Kant’s account of figurative synthesis
establishes the mediating role of the imagination. I have further argued
that Section 3 of the A edition does not protect the position of the
imagination in the way Heidegger suggests, but my disagreement with
his account goes deeper than this and is more difficult to tease out.
Heidegger’s concern for the ‘autonomy’ of the imagination may be
misplaced and reveals more about his own concern for origins than it
does about Kant’s faculty theory. Once we recognise that the role of
148
the imagination is principally one of mediation, as I have argued in

previous sections, it is unsurprising that its autonomy or identity will
frequently be in question. As the middle term between the principal
faculties of intuition and understanding, it is both necessary and liable
to inclusion in one or other pole.136 Heidegger is inclined to make the
middle term into the origin. This is especially evident when he
announces that the imagination is the ‘condition of the possibility of
all the faculties’.137
For Heidegger, imagination is not equidistant between intuition
and understanding and leans more towards the first than the
second.138 He insists that intuition is the primary term for knowledge,
for it is what allows us to take up an affect.139 If imagination is to hold
the primordial position he claims for it, it is hardly surprising that the
mediating faculty is more frequently shown in its systematic relation
to intuition than to that of understanding. Heidegger is so committed
to his view of imagination as a further elaboration of intuition that
he goes so far as to say that transcendental imagination is time. Time
is the medium within which all intuition affects us and is thus the
horizon within which we encounter beings. Heidegger particularly
focuses on Kant’s characterisation of time as ‘self-affection of the self’
and concludes that, ultimately, time and not space is the source of
affection.140 ‘Time as pure intuition is in one the formative act of intu-
iting and what is intuited therein’.141
In taking this position Heidegger remains true to the letter of
Kant’s Schematism chapter in which the necessity of the temporalisa-
tion, not the spatialisation, of the categories is analysed. However
there are grounds for thinking that Kant’s account is insufficient, even
within his own terms. I have argued in the previous chapter that the
initial paragraphs of the ‘Transcendental Aesthetic’ reveal the event
of the material affect of objects. Later I will argue that while his
project of establishing the subjective conditions of objective objects
must start with a temporal schematism, it cannot stop there.142 In
order to show that mental forms grasp the material given he needs to
show how the categories are combined with the spatial form of intu-
ition as our mode of access to outer things. And while Heidegger goes
on to say that space is ultimately grounded in time, thus establishing
that human existence is spatial, this is not good enough for Kant’s
purposes. Kant must show that the different, although intertwined,
trajectories of time and space establish the condition of the possibil-
ity of the subject’s encounter with an external object. This is exactly
what he tries to establish in the ‘Refutation of Idealism’.143
149
My present concern is with the way in which Heidegger presents the

relation between the faculties. He sets up the perspective from which
we can see the imagination as the relation between two extremes.144
But he falls into treating the third faculty not, as he says, as ‘only a
mediating, intermediate faculty’, but rather as ‘the primordially unify-
ing element’.145 The transcendental imagination is the ‘root’ that joins
the two extremes, but it now appears that the joining in question is not
so much a process, but rather a source or origin. 146 In his interpreta-
tion of the Schematism we can find a further aspect of his ontologis-
ing of the faculties while he seeks, legitimately in my view, to establish
that they serve as points of access to things in the world. Whereas Kant
says that the schema is a rule for the generation of images, Heidegger
interprets it as a pure image.147 Even though he adds that it is the pos-
sibility of a pure image that is in question, his account encourages the
view that the schemata are pure entities rather than processes.148
There can be no doubt that the unity Heidegger is seeking to estab-
lish is not static, for it is nothing other than the flowing of time
towards the future and from the past towards the present. The imag-
ination is only origin as temporal synthesis. But the tension between
seeing imagination as mediator and as origin leads to Heidegger’s
continually complicating its identity and the relations in which it
stands to the other faculties. The complexity of his account is not
wholly justified by the intricacy of the relation he is trying to express.
Unnecessary complexity leading to obfuscation arises from his temp-
tation to establish the imagination as more than merely intermediary.
Heidegger’s complex presentation goes hand-in-hand with an over-
simplification of the role of imagination. He seeks to make imagin-
ation first among equals, when the situation is rather that the
imagination is the mediator par excellence, as I have argued in previ-
ous sections. In slipping into his preferred lexicon of ‘primordiality’
and ‘origins’, Heidegger does not go far enough in his examination of
the relational status of the imagination. If he is to capture Kant’s
concern with how we are affected by something given in experience,
he must show how our capacity for affection allows us not only to
self-affect, but also to be affected by something other than our-
selves.149 Heidegger believes he can achieve this through an analysis
of the temporality of human existence, or Dasein. However as
Heidegger’s consideration of the primacy of temporality over spatial-
ity already reveals, an account of self-affection simply cannot estab-
lish the possibility that Kant is aiming to legitimate, namely, our being
affected by objects external to ourselves. This would require not only
150
time, but also space.150 Revealingly, when Heidegger cites Kant’s

comment that ‘space and time must affect the concept’, he goes on to
talk exclusively of time.151
The great potential of Heidegger’s interpretation of Kant has been
taken further in the work of the contemporary philosopher John
Llewelyn, who has revealed the role of imagination as an ongoing
process, unashamed of its merely intermediary status.152 Heidegger’s
own shortcomings are, I believe, related to his failure to address the
role played by the imagination in the Critique of Judgement where
Kant examines the faculty of judgement as such.153 Kant’s account of
reflective judgement is of an operation of mediation between intuition
and understanding. In the next section, I will examine how this
account sheds light on Kant’s earlier analysis of cognitive judgement.
IV Synthesis as the activity of judgement

As I have already suggested earlier in this chapter, I believe that there
are resources in the Critique of Judgement for resolving a problem
that has emerged in Kant’s epistemology, namely his uneasy identifi-
cation of the role of imagination. I have suggested that the imagina-
tion is best understood as mediating between the other faculties. This
perspective, while already presupposed in the first Critique, is at the
same time compromised by judgement’s being linked not only to sen-
sibility, but also to understanding. In this final section I will argue that
aesthetic judgement reveals the synthetic process that is necessary for
any judgement. In the course of my analysis of determining judge-
ment, I have shown how Kant displays a further ambivalence about
the limits of knowledge. At times, he suggests that all apprehension
amounts to knowledge, while at others he states clearly that not all
intuitions are unified under the understanding. A deeper analysis of
the relation between the faculties, afforded by the Critique of
Judgement, will allow us to show the importance of his retaining the
possibility that there are unsynthesised intuitions. In Chapter 8 I will
argue that the Critique of Judgement also allows us to resolve this
problem.
Aesthetic judgement is not cognitive simply because it does not aim
at a determination of the object. It counts strictly as reflective; first,
insofar as it aims only at an ideal unification in experience that is never
actually achieved; and, second, in that we are thrown back on our
own subjective activity. Nevertheless, in the course of his analysis of
aesthetic judgement, Kant has much to say that illuminates his
151
epistemological stance. Most importantly, Kant suggests that aesthetic

judgements are based on a harmony of the faculties that is necessary
for ‘cognition in general’ (Erkenntnis überhaupt).154 Although the
latter claim suggests a direct and problematic lineage between cogni-
tive determinative and aesthetic reflective forms of judgement, I will
argue that they stand in an indirect, yet systematic relation.
Aesthetic judgements are reflective in that they throw us back on the
subjective activity that makes them possible, namely, the harmony of
the faculties. My suggestion will be that they also indirectly reveal the
less harmonious (but, on inspection, cooperative) activity necessary for
any cognition whatsoever. This activity is the subjective process that
makes synthetic conclusion possible. Kant’s claim that aesthetic judge-
ments qualify as synthetic has been roundly criticised, as it seems to
suggest that aesthetic judgements fall under the rubric of cognitive syn-
thetic judgements. I will argue that aesthetic judgements are synthetic
in the sense that they display the synthetic process necessary for any
cognitive judgement whatsoever. Aesthetic judgements are syntheses in
action, with the important caveat that they do not aim at a cognitive
conclusion. The cognitive activity of judgement – which we have just
been examining in the ‘Deduction’ – is a means to an end, whereas in
the aesthetic case synthetic activity is an end in itself.
(i) A Re-reading of Kant’s Account of Synthesis in the

Critique of Judgement
The clearest statement of the synthetic status of reflective judgements
comes in the first Book of the third Critique:
We can readily see that judgments of taste are synthetic; for they go
beyond the concept of the object, and even beyond the intuition of the
object, and add as a predicate to this intuition something that is not even
cognition: namely [a] feeling of pleasure (or displeasure). And yet, that
these judgments are, or want to be considered, a priori judgments as
regards the demand that everyone assent, a demand they make despite the
fact that their predicate (of one’s own pleasure [as] connected with the pre-
sentation) is empirical, is also already implicit in the expressions used to
make that claim. Hence this problem of the critique of judgment is part of
the general problem of transcendental philosophy: How are synthetic
judgments possible a priori?155
We have already seen in the previous section that figurative synthesis
entails going beyond the mere concept of an object and seeking the
intuition for that concept. A concept alone would not achieve knowl-
152
edge. Kant now says that judgements of taste share this general
feature of synthesis, in that they, too, go beyond the mere concept of
an object. But in this case the concept does not determine an intuition,
so what does Kant mean? Moreover, for Kant, aesthetic appreciation
does not start from a concept, but rather from a sensory response to
something given in experience. One way in which we can make sense
of his first claim is to say that an aesthetic judgement goes beyond the
concept insofar as it abstracts from the conceptual determination of
an object and focuses instead on the feeling we have in response to it.
In finding something aesthetic, we may be well aware of what it is,
but in taking pleasure in it, we abstract from considerations of iden-
tification and concentrate on the pleasurable effect it has on us.
What is really crucial for the status of aesthetic judgements as syn-
thetic is that they go beyond the intuition of the object, in that we
have a feeling of the reflective harmony of our faculties that is insep-
arable from the intuition, qualifying the latter as aesthetic. An aes-
thetic judgement starts from an empirical intuition of an object, but
in this case the intuition is not determined by a concept. We aim
towards a concept that would explain the given intuition, but all that
is available is an indeterminate idea of beauty. It is thus not so much
that these judgements go beyond the concept as that they stop short
of determination by the latter.156
Synthesis in general links a concept or an intuition with something
further. In cognitive judgement, the concept is linked with an intu-
ition or vice versa, but in aesthetic judgement there is a gap between
the intuition and a possible explanatory concept. The concept of
beauty that we attribute to this object is simply too general to deter-
mine the empirical object with which we are faced. Yet the gap
between concept and intuition is not a void, for we experience an aes-
thetic feeling that serves as a link between the particular thing under
our gaze and the indeterminate predicate ‘beautiful’. In an aesthetic
judgement we become aware of the movement or activity between
the two poles of a given object and a subjective reflection that would
in other circumstances lead to conceptual determination. We are
aware of the synthetic process that creates a link across the gap iden-
tified by dualism. We are able to remain at the level of the process
because of the focus afforded by a feeling of pleasure. The feeling
allows us to remain at the imaginative level of attention that in cog-
nition was resolved into a unitary consciousness. In the aesthetic
case, we stop short of unity or determination and direct our atten-
tion to the feeling. This has the result that we can become aware of
153
the mediating role of imagination that was necessary for, but con-
cealed in, cognition.
While aesthetic synthesis does not subsume the intuition under a
concept as in the cognitive case, the intuition bears a universal value
from its being associated with a very particular kind of feeling. The
peculiarity of the feeling characteristic of these judgements is that it
is not merely private, despite the fact that the feeling is mine and arises
in response to an empirical object. Aesthetic feeling raises a claim for
universal validity.157 When I say that a particular object is beautiful,
I implicitly appeal to the agreement of all other judging subjects.
Kant concludes the passage under consideration with the claim that
the third critique contributes to the general problem of transcendental
philosophy, which he identifies with the question: How are synthetic
judgments possible a priori? This is because aesthetic judgements
belong to the class of judgements that are under consideration in tran-
scendental philosophy. This species of judgements is identified as syn-
thetic, insofar as they go beyond the intuition and the concept. They
also qualify as a priori in that they are universally valid for everyone.
But now we are faced with two questions. Is Kant right in claim-
ing that aesthetic judgements count as synthetic within the logic of his
own argument? And, if so, does the a priori status of aesthetic judge-
ments simply add a new species of synthetic a priori judgements, or
does it illuminate the account of synthesis as a whole?
My answer, which will occupy the remainder of this book, involves
showing that aesthetic judgements are synthetic in that they reveal the
process of synthesis that is at work in all judgements. I will now
examine two passages that will provide us with a clue to the identity
of synthesis in process and the relation in which it stands to determi-
nate synthesis.158
In the published Introduction to the Critique of Judgment, Kant
makes a connection between aesthetic judgement and judgment
(or ‘cognition’) in general. In an aesthetic judgement, which counts
as empirical and singular, there is a purposive harmony or fit
between the object and the mental activity arising in response to
that object:
For the basis of this pleasure is found in the universal, though subjective,
conditions of reflective judgments, namely, the purposive harmony of an
object (whether a product of nature or of art) with the mutual relation of
the cognitive powers (imagination and understanding) that are required
for every empirical cognition.159
154
Pluhar’s translation illuminates something that is understated in the

original German text. In using the expression ‘the mutual relation of
the cognitive powers’ Pluhar brings out how any empirical judgement
requires a cooperation of the faculties.160 An aesthetic reflective
judgement is distinguished by the presentation of the object being
purposive or conducive to the mental activity that arises in response
to it. There is a harmony between the object and our response to it.
Later Kant will say that the mental response itself counts as a
harmony of the faculties.161 These two harmonies are indeed corol-
laries of one another. Our faculties spontaneously cooperate with one
another because the object is propitious for our judgement. However,
in this passage Kant also claims that the purposive harmony is – in
some way yet to be established – related to a necessary condition of
all empirical cognition.
What is at stake in Kant’s claim that an aesthetic object stands in
a purposive harmony with the mutual relation of the faculties neces-
sary for all empirical cognitions? He surely cannot be saying that a
purposive harmony between object and mind is the necessary condi-
tion of knowledge?
The key to making sense of this passage is to understand that it is
only the mutual relation between the faculties and not the purposive
harmony between the object and the faculties that is characteristic of
empirical judgements in general. Empirical judgements necessarily
display some mutual relation of the faculties in that the cognitive syn-
thesis of concept and intuition requires a cooperation between under-
standing and intuition. We can conclude that the play or harmony of
the faculties that arises when we encounter an object we judge to be
beautiful, is a heightened version of the pervasive coordination of the
faculties necessary for any judgement, that is, any experience at all.
In this harmony imagination stands in for intuition insofar as the
former provides the conditions for the combination of the latter and
thus for its formal status.162 In the harmony of the faculties imagina-
tion and understanding harmonise freely with one another. Thus the
mutual relation necessary for cognition becomes apparent only in an
aesthetic judgement where a specific cognitive goal is not in view.
Unfortunately for the persuasiveness of a series of Kant’s arguments,
he does not deploy a clear distinction between a mutual relation and
a harmony of the faculties and this lack becomes more acute in the
body of the Critique of Aesthetic Judgement.163
The second passage that will serve as a clue to the relation in which
synthesis in process stands to determinate synthesis comes in the
155
‘General Comment’, at the end of the ‘Analytic of the Beautiful’. Kant

remarks that aesthetic judgements display free lawfulness: ‘the object
may offer it [the imagination] just the sort of form in the combina-
tion of its manifold as the imagination, if it were left to itself [and]
free, would design in harmony with the understanding’s lawfulness in
general. . .’164
In a determining judgement, synthesis is governed by a rule or law
introduced by the understanding. An aesthetic judgement is prompted
by an object that has a form that encourages the imagination to
combine the manifold in a way that is conducive to the unifying activ-
ity of understanding, that is its unification of the manifold under a
law. But in this case there is no law, so the imagination’s combination
counts as a free lawfulness. Kant concedes that such a notion appears
contradictory.165 However, he goes on to resolve this apparent
paradox by arguing that in an aesthetic judgement the imagination is
capable of following the understanding without being compelled by
it. The combination that would usually prepare for unification under
a concept in this case stops short of its resolution. This counts as free
lawfulness.
We can explain this with reference to the reading I have given of the
‘Transcendental Deduction’. In a determinate cognitive judgement, the
faculties cooperate when the understanding gives the rule to intuition
through the intermediary of the imagination. In a reflective aesthetic
judgement, the faculties cooperate without any rule being applied,
however the relation in which the faculties stand to one another is so
harmonious that it is as if a rule had been applied. Aesthetic judgement
thus mimics cognition’s lawfulness, but does so without a law operat-
ing. This is why the aesthetic harmony of the faculties is capable of
revealing the subjective structure of synthesis, that is, its synthetic
activity. While determining judgements aim at syntheses as results, aes-
thetic judgements reveal the process of synthesis that otherwise would
only be available to us in a philosophical reconstruction such as that
offered by the Critique of Judgement. For the moment, I have not
explained how aesthetic judgements could exhibit this general struc-
ture of cognition; I will only consider this question in the final chapter.
(ii) Makkreel’s Rejection of the Synthetic Status of

Aesthetic judgement
It has often been concluded that the account of synthetic a priori
judgements is restricted to the first Critique and its concern with
156
objective knowledge. Rudolf Makkreel provides one of the most

systematic rejections of the claim that aesthetic judgements count as
synthetic.166
First, Makkreel argues that the role of the imagination in the figu-
rative synthesis of the first Critique is that of the handmaiden of the
understanding.167 He concedes that the imagination brings some of
its own role to bear in the determination of sensibility, insofar as it
provides the schemata necessary to figurative synthesis.168 However,
despite this concession, his reading of the ‘Deduction’ of the first
Critique is impositionalist and this is the fundamental reason why he
denies that aesthetic imagination is synthetic.169 Given the restricted
and subservient role he apportions to the imagination within figura-
tive synthesis, it is perhaps unsurprising that he insists on such a deep
cleft between cognitive experience and what he prefers to call ‘aes-
thetic consciousness’.170 If synthesis is always determinative and is
also central to knowledge, then Makkreel would be right to conclude
that otherwise aesthetic apprehension would be reduced to a subset
of cognition.171 However, I have shown that we can at least make
sense of the idea of a synthesis that is not determinative.
Makkreel’s reading of the B ‘Deduction’s’ account of figurative
synthesis runs along the familiar lines of the impositionalist interpre-
tation. However, he pauses at the following claim, coming, as it does,
so late in the development of Kant’s account:
Synthesis in general, as we shall hereafter see, is the mere result of the
power of imagination, a blind but indispensable function of the soul,
without which we should have no knowledge whatsoever, but of which we
are scarcely ever conscious.172
He comments that this was an ‘oversight on Kant’s part’.173 In a

textual correction never incorporated into a published edition, Kant
amends the passage thus:
‘Synthesis in general, as we shall hereafter see, is the mere result of the
power of the imagination, a function of the understanding.’174
This unpublished remark seems at first sight to cancel the distinctive-

ness of the faculties, resulting in there being no need for a coopera-
tion between understanding and imagination.
But should we take this comment as the decisive move in a struggle
to express the relation between the faculties, evident throughout the
published text? Heidegger, while also noting this correction, remarks
that Kant nevertheless left unaltered the subsequent Schematism
157
chapter, in which the imagination is presented as the source of tran-

scendental synthesis.175 In any case, discovering another statement in
favour of the primacy of the understanding does not solve the
problem, but rather reopens it. Makkreel’s assumes this unpublished
remark represents Kant’s final view on the matter and is thus conclu-
sive. This is questionable as an interpretive strategy. We cannot know
this was Kant’s last thought on the matter and, even if it was, the last
opinion an author expresses is not always most representative of his
or her position as a whole. A further problem arises from the fact that
this is not the only occasion on which Kant identifies synthesis with
the imagination, making it unlikely that this particular passage counts
as a mere oversight.176 Moreover, even if we did take the amendment
as decisive, we would still be faced with the question of how we should
interpret it. Is the imagination a function of the understanding in the
sense that it is merely an exercise of the latter, or is it that imagination
sometimes functions in cooperation with sensibility and sometimes
with understanding, as other passages suggest?
Makkreel has further reasons for denying that aesthetic judge-
ments qualify as synthetic. He remarks that the term synthesis is never
used in the accounts of aesthetic apprehension and comprehension.177
He dismisses the idea that this might be due to a wish to avoid self-
repetition, saying that Kant never hesitates to repeat himself on other
occasions. This is, however, hardly a conclusive claim. As we have
seen above, Kant not only announces that aesthetic judgements are
synthetic, but also makes it clear that he holds them to be central to
the general transcendental project of establishing the possibility of
synthetic judgements. Whether he would have been willing to repeat
himself or not is difficult to ascertain. Sometimes Kant is repetitive,
while at other times he fails to state the obvious: there is no a priori
rule for deciding which is the case here. All we can be sure of is that
Kant declares that aesthetic judgements are synthetic.
Makkreel’s most interesting suggestion is that reflective judge-
ments are synthetic ‘only in form’. Aesthetic judgements:
are not synthetic in the objective sense applicable to cognitive judgements,
in which we add to the concept of an object a concept of one of its attrib-
utes. Instead of claiming something about the objective properties of an
object, the judgment of taste discloses something about our own subjec-
tive state of mind in apprehending the form of an object.178
I agree with this and conclude that it is quite appropriate to characterise

aesthetic judgements as synthetic in that they display the subjective
158
conditions of synthesis in general. Makkreel is prevented from coming

to this conclusion because he generally treats aesthetic judgements as if
they were concerned only with the subject and not with the relation
between subject and object. The subjective bias of Makkreel’s inter-
pretation is picked up by Ameriks, unsurprisingly given his own objec-
tivist reading of Kant’s aesthetics.179
Interestingly, Makkreel does not wholly rule out that there is some
relation to an object in an aesthetic judgement. Towards the end of
the second section of his book, he retrospectively claims that objects
were at issue and corrects his previous subjectivist bias by saying that
aesthetic, like teleological, judgements refer to objects but are never-
theless not directly cognitive.180 From this perspective, it would have
been possible for Makkreel to reach the conclusion that aesthetic
judgements are not only formally or subjectively synthetic, but that
they reveal the relation between subject and object. It would then
have been a short step to the position that I will defend, namely that
aesthetic judgements are capable of contributing to the general
project of transcendental philosophy, namely, the project of showing
how synthetic a priori judgements are possible. But this is exactly
what Makkreel denies.181
I agree that aesthetic judgements are not synthetic in the same sense
as is cognition. The question is: do the former reveal the deeper struc-
ture of the synthesis that is the subjective condition of cognition? This
is what I will try to establish in the chapters that follow. Makkreel
prefers to insist that aesthetic judgements are post-categorial, that is,
that they expand rather than reveal the conditions of the possibility
of cognition.182 For this reason he is not inclined to concede that aes-
thetic judgements reveal the deeper structure of synthesis.
In conclusion of this discussion, I wish to draw attention to a par-
ticularly suggestive connection made by Makkreel. In discussion of
the way in which the concept of life binds together the extremely
unwieldy text of the Critique of Judgement, he suggests that this
concept refers to the power to move, and that this, at its deepest level,
is grounded in the life of the mind.183 Despite my disagreement with
the role he apportions to synthesis, his suggestion could have initiated
a very illuminating way of addressing what I am calling synthesis in
process. Revealingly, Makkreel says that considering life as the power
to move ‘allows us to interpret the aesthetic feeling of life as a tran-
scendental point of unity for both the active power of the under-
standing and the receptivity of sense’.184 This, he believes, suggests a
way of ‘mitigating Kant’s dualism of understanding and sense’.185
159
While I agree with the way in which Makkreel develops Kant’s sug-
gestive, but elliptical comment on life, I have argued in this chapter
that dualism is always more dynamic and more pluralist than his
interpretation would suggest. At a deeper level of analysis, under-
standing and sense are to be seen, not as opposed, but as standing in
necessary relation to one another. In order that they can relate, the
intermediary capacity for imagination is required. Thus dualism is
unveiled as entailing a pluralist model of mind and a dynamic rela-
tion between the faculties. Aesthetic judgements of beauty, in partic-
ular, facilitate an indirect access to the cooperation of the faculties as
a ‘feeling of life’ of the mind.186
Conclusion
In this chapter I hope to have established that, for Kant, synthesis
relies on a cooperation of the subjective faculties or synthesis in
process. This is the deep structure of any synthetic achievement. A
priori synthesis, as analysed in the ‘Transcendental Deductions’
requires the activity of a plurality of orientations, even though the role
apportioned to imagination is ambiguous. I have also argued that aes-
thetic judgement properly counts as synthetic and contributes to the
general project of establishing the possibility of a priori synthetic
judgements, just as Kant claims. This is because the play of the facul-
ties is nothing other than an instance of the cooperation of the facul-
ties necessary for any judgement, free from any cognitive conclusion.
In the next chapter I will argue that the deep structure of synthesis
is best understood as the subjective side of the deduction of the
categories.
Notes
1. CPR, A 51, B 75, at the outset of the ‘Transcendental Logic’.
2. See Chapter 6 on the anticipatory character of knowledge.
3. Recognition of the reflective status of the understanding leaves open
the possibility for comprehending the role of concepts in relation
to aesthetic judgement. See Chapters 7, pp. 260–9, and 8, pp. 280–90.
Longuenesse has very helpfully highlighted the role of reflection
throughout Kant’s account of experience (see Chapter 2, p. 77).
4. For a discussion of the ‘task’ of knowledge, see Chapter 7, pp. 252–3.
5. See my account of empirical knowledge in Chapter 7, pp. 249–55.
6. This is especially evident in aesthetic judgement, which, Kant believes,
is peculiar to human beings. See CJ, AA 210.
160
7. Notable exceptions include Patricia Kitcher’s Kant’s Transcendental

Psychology and Onora O’Neill’s ‘Transcendental Synthesis and
Developmental Psychology’.
8. Later we will see the importance of the further faculties of reflective
judgement and imagination. Both of these are ‘linking’ or relational
faculties without which experience would not be possible.
9. I argue this in Chapter 7, pp. 249–55.
10. CJ, AA 218.
11. See mention of McDowell’s Mind and World in Chapter 3, p. 101.
12. See Chapter 3, p. 101, where I have already made both claims.
13. See ‘What is Orientation in Thinking?’, AA, VIII, pp. 131–47. This
was first published in 1786 and thus predates the Critique of
Judgement’s development of the idea of a ‘harmony of the faculties’.
14. I come back to this issue in a discussion of the ‘Dialectic of Taste’ in
Chapter 8.
15. B 128.
16. A 55, B 80.
17. A 69, B 93/4.
18. A 77, B 102–3.
19. This is the highly criticised ‘Metaphysical Deduction’ recently
defended by Béatrice Longuenesse. See Kant and the Capacity to
Judge, pp. 73–80.
20. A 63, B 87–8.
21. A 63, B 87–8.
22. A 89/90, B 122.
23. A 89, B 121–2.
24. As I argued in Chapter 3, pp. 105–8.
25. A 89–90, B 122.
26. See discussion of A 116 below. See also discussion of B 160–1 below.
27. On this question, see Chapter 3, p. 102.
28. See Allison, KTI, pp. 133–5; and Henrich, ‘The Proof-Structure’.
29. A 115.
30. See discussion below.
31. See discussion of Longuenesse in Chapter 2, pp. 70–1. Patricia Kitcher,
in Kant’s Transcendental Psychology, suggests that ‘normative and
factual claims commingle in transcendental philosophy’ (p. 21).
32. A 99.
33. A 99; ‘Impressions’ translates Eindrucke.
34. A 99.
35. See discussion of Longuenesse in Chapter 2, pp. 72–8.
36. A 97.
37. A 97.
38. B 160 note.
39. A 94.
161
40. A 98–9.
41. A related problem arises towards the end of the account of the syn-
thesis of reproduction in imagination where Kant says that the ‘purest
and most elementary representations of space and time’ require the
operation of imagination (A 102). I believe that this should be under-
stood in parallel with the problematic footnote at B 160–1 where, as I
read it, Kant says that pure geometric forms, not the forms of space
and time, require a synthesis of imagination. See discussion below,
pp. 139–43.
42. See discussion below, p. 138.
43. A 100.
44. A 102.
45. This is what phenomenologists sometimes call the ‘living present’.
46. A 102.
47. Kant remarks that it is because reproduction is necessarily bound up
with apprehension, one of the transcendental acts of the mind, that
reproductive imagination also qualifies as transcendental (A 102). He
withdraws this status in Section 3. See below, p. 127.
48. A 102.
49. See discussions of Section 3 of the A edition and of B 160–1 in the
second edition below, pp. 127–30 and pp. 145–6.
50. A 103.
51. A 103.
52. See discussion in Chapter 3, p. 102.
53. A 115.
54. A 115.
55. A 119.
56. A 118. This is in contrast to his earlier claim that the reproductive syn-
thesis of imagination counts as transcendental. See A 102 .
57. A 120.
58. A 118 (my emphasis).
59. See discussion of Heidegger below, pp. 147–51.
60. A 118.
61. A 121–2.
62. A 123.
63. A 123.
64. A 119/120.
65. A 120.
66. A 116.
67. A 116; ‘wenigstens müssen verknüpft werden können . . .’
68. A 120.
69. A 120 (my addition). Compare the very similar claim in the footnote
at B 160–1 discussed below, p. 137.
70. A 120.
162
71. A 121.
72. A 121.
73. A 121–3.
74. A 124. See, also, discussion of an almost identical claim at A 78, B 123,
in my discussion of Makkreel in the final section of this chapter,
pp. 157–8.
75. See, however, the next section for a discussion of Makkreel’s assess-
ment of an emendation to the B edition where Kant seems to withdraw
imagination’s status as a fundamental faculty. I argue, however, that
this withdrawal cannot be taken as decisive.
76. Sarah Gibbons also thinks the distinction is of level, not of type of syn-
thesis. See Kant’s Theory of Imagination, p. 40.
77. See discussions of Allison and Longuenesse in Chapter 2, p. 64 and p.
70.
78. B 151. Kant also calls this combination through the understanding in
contrast to figurative synthesis, which he associates with the imagina-
tion. This distinction suggests that the understanding must cooperate
with imagination if knowledge is to arise.
79. This, however, can only count as an abstraction from a more concrete
level of experience, if we take seriously Kant’s claim that thought
always refers to intuition (A 19, B 33). See discussion in Chapter 3,
p. 101.
80. B 136/7.
81. B 144–5.
82. B 145.
83. B 150.
84. B 145.
85. B 147.
86. B 150.
87. B 151.
88. B 151–2.
89. B 152.
90. A 120. The link between imagination and form is also assumed in
Kant’s account of figurative synthesis in Section 24 of the B edition.
91. B 151.
92. This is also why the form of the object is so important in aesthetic
judgement’s non-cognitive exhibition of the general possibility of cog-
nition. See Chapter 8, pp. 280–90.
93. See Chapter 2, pp. 72–8 on Longuenesse for both these issues.
94. See my ‘Taste as Productive Mimesis’.
95. B 160.
96. B 160.
97. B 160–1. I will return to the footnote later, restricting my reconstruc-
tion for now to the main text.
163
98. B 161.
99. B 161.
100. B 161.
101. This distinction is repeated on several occasions in Section 26. See, for
instance, the beginning of the note at B 160. See also his example of
the apprehension of a house at B 162. See, again, B 164.
102. B 160.
103. A 89–91, B 122–3.
104. In my ‘Kant’s Phenomenological Reduction?’, I argue that the forms of
space and time are analogous in status to a phenomenological
‘horizon’ (Hughes 2007).
105. B 161.
106. B 161.
107. See discussion of this point in Chapter 2, pp. 72–8.
108. B 161.
109. I argued that the distinction between forms of representation and the
representations in space and time was also in force in the synthesis of
apprehension in the A ‘Deduction’.
110. B 161 (Kemp Smith’s emphasis).
111. B 161.
112. B 162 note.
113. B 162 note. See below, pp. 151–6.
114. B 161.
115. B 161.
116. B 161.
117. B 161.
118. As I argued in Chapter 3, pp. 101–2. Later I discuss how aesthetic
judgements relate to cognition in general and yet are not cognitive in
status.
119. A 98–100.
120. A 100–3.
121. A 120.
122. B 162 note.
123. B 164.
124. B 162 note.
125. B 151. In the last section we saw that Kant says imagination
belongs to sensibility just as he says that it is an action of the under-
standing.
126. Heidegger, Kant and the Problem of Metaphysics (= KPM); Heidegger,
Kant und das Problem der Metaphysik. For an incisively critical, while
sympathetic account of Heidegger’s reading of Kant, see John
Llewelyn, ‘Heidegger’s Kant and the Middle Voice’.
127. Heidegger, KPM, p. 206; references to German edition in brackets
(p. 195).
164
128. At its deepest level, Being is temporal and human beings are beings for
whom time is at issue. To consider things ontologically and not merely
ontically is to consider them in their temporal givenness and to under-
stand our own finitude in respect to that givenness.
129. It is tempting to think that Merleau-Ponty’s notion of a chiasm – that
is, a relation between two intertwined and ultimately inseparable
beings – would have been helpful for the complex set of relations
Heidegger tries to express. See Merleau-Ponty, The Visible and the
Invisible, Ch. 4 ‘The Intertwining – The Chiasm’.
130. Heidegger, KPM, p. 175 (p. 164).
131. See, for instance, Heidegger, KPM, p. 252 (p. 236), where Heidegger
says that time, which has now been identified with imagination, ‘is
essentially one with pure apperception’.
132. Ibid., p. 167 (p. 155). Leidenschaftlichen could also be translated as
‘passionate’.
133. Ibid., p. 170 (p. 159).
134. Ibid., p. 167 (p. 156). He refers to CPR, A 94 and A 115. In both
passages Kant says there are three subjective sources of experience.
It must be admitted that the suppression of these passages is signi-
ficant.
135. Heidegger, KPM, p. 168 (p. 156). See A 78, B 103. Heidegger gives the
reference as Nachträge XLI. Makkreel also comments on this alter-
ation as I discuss below. He refers to AA XXIII, 45.
136. I discuss a related question about the ‘heautonomy’ of aesthetic judge-
ment in my ‘Taste as Productive Mimesis’ (Hughes 2006).
137. Heidegger, KPM, p. 154 (p. 148).
138. Ibid., pp. 135, 136, 137 (pp. 123, 124, 125).
139. Ibid., p. 28 (p. 21). ‘Cognition is primarily intuition’. Heidegger’s
point is that intuition stands in immediate relation to objects.
140. See discussion of self-affection in Chapter 3, p. 97.
141. Heidegger, KPM, p. 180 (p. 169).
142. See Chapter 6, pp. 229–37.
143. Again, see Chapter 6, pp. 234–5.
144. See, for instance, Heidegger, KPM, p. 86 (p. 78).
145. Ibid., p. 202 (my emphasis) (p. 190).
146. Ibid., p. 152 (p. 140).
147. Ibid., pp. 108–9 (p. 100). Pippin’s interpretation is rather similar to
this. See Chapter 1, p. 14.
148. Ibid., p. 109 (p. 100).
149. See my discussion of affection in Chapter 3, pp. 105–8.
150. Heidegger sees this problem and intends to address it in On Time and
Being. As Llewelyn remarks, there are already many spatialising ele-
ments in Being and Time; see his ‘Heidegger’s Kant and the Middle
Voice’, p. 115.
165
151. Heidegger, KPM, p. 193; German edition, p. 182; where he refers to

CPR, A 77, B 102.
152. See not only Llewelyn’s ‘Heidegger’s Kant and the Middle Voice’, but
also his The HypoCritical Imagination, especially the Prologue; and
The Middle Voice of Ecological Conscience: a chiasmic reading of
responsibility in the neighbourhood of Levinas, Heidegger and others,
especially the Preface and Ch. 10, ‘Something like the middle voice’.
153. Heidegger remarks that he cannot discuss the Critique of Judgement;
see KPM, p. 167; German edition, pp. 155–6.
154. CJ, AA 218. Although Kant’s claim clearly covers all cognition and not
merely that which gives rise to objective knowledge, our concern here
will be entirely with the latter.
155. CJ, AA 288–9, Section 36, ‘On the Problem of a Deduction of
Judgments of Taste’.
156. My concern in this discussion is to provisionally identify the primary
structure of aesthetic judgements and to establish the relation in which
they stand to cognition. Judgements of taste also go beyond the
concept in the sense that they say more than any particular conceptual
determination could express. But this over-determination of sense
arises from the lack of determinacy characteristic of any aesthetic
judgement. It is because an aesthetic judgement rests only on the activ-
ity of judgement in general that it is open to a range of different
determinations. Makkreel suggests that aesthetic judgement is post-
determinative, rather than pre-determinative; see the discussion in the
next section. Although I cannot argue the case here, my view is that
aesthetic judgements can be post-determinative just because they are,
primarily, pre-determinative.
157. CJ, AA 211–12.
158. For a related discussion, see Howard Caygill, Art of Judgement,
pp. 318–20, on formative activity and judgement.
159. CJ, AA 191. The following is the crucial passage in the original
version: ‘mit dem Verhältnis der Erkenntnisvermogen unter sich, die
zu jedem empirischen Erkenntnis erfordert wird (der Einbildungskraft
und des Verstandes)’.
160. Makkreel also makes a helpful contrast between the ‘accord’ necessary
for any cognition and the ‘agreement’ characteristic only of aesthetic
judgements. See Makkreel, Imagination and Interpretation in Kant,
p. 62. However, I disagree with his claim that it is only in the aesthetic
case that the faculties must adapt to one another.
161. CJ, AA 217/18.
162. See discussion of the ‘Transcendental Deduction’, especially in the A
edition, above.
163. I will discuss the problems this gives rise to in Sections 9 and 21, in
particular. See Chapter 5, pp. 174–6 and p. 188.
166
164. CJ, AA 240/1. Additions to passages from the Critique of Judgement

in square brackets here and elsewhere are Pluhar’s, unless otherwise
signalled.
165. AA 241.
166. See Makkreel, Imagination and Interpretation in Kant (= IIK),
pp. 47–51. See also his ‘Response to Guenter Zoeller’.
167. Makkreel, IIK, p. 30.
168. Ibid., p. 30, where he says that imagination introduces its own forma-
tive power. Consistently, Makkreel suggests that the schemata are nec-
essary for the determination of a concept. See, for instance, p. 124 and
129. Sometimes he additionally talks as if schemata were already at
issue in the account of figurative synthesis in the B ‘Deduction’. See
IIK, p. 31. This position bears some similarity to Longuenesse’s sug-
gestion that the schemata come before the categories. She means, of
course, that the schemata are prior to the explicit determination of
intuition by concepts. See Chapter 2, pp. 72–8.
169. See especially IIK, pp. 42 and 46, where he identifies figurative syn-
thesis with ‘definite objective cognition’.
170. See IIK, pp. 4, 45, 47–9, 52, 67, 73, 92, 106, 118. See also, ‘Response
to Guenter Zoeller’ p. 278.
171. Though my reconstruction of the Deductions shows that he would not
be right to conclude that Kant’s position is simply impositionalist.
172. CPR, A 78, B 103.
174. Ibid., p. 29. He refers to XXIII, 45.
175. Heidegger, KPM, p. 168 (p. 156).
176. See, for instance, CPR, A 124; also in the A ‘Deduction’ and discussed
earlier in this chapter.
178. Ibid., p. 48. See also ‘Response’, p. 277.
179. See K. Ameriks, ‘Rudolf A Makkreel, Imagination and Interpretation
in Kant’ (= ‘Review’), pp. 232 and 233, referring to Makkreel, IIK,
pp. 53 and 63 respectively. For Ameriks’ own position, see ‘Kant
and the Objectivity of Taste’, and ‘How to Save Kant’s Deduction of
Taste’.
181. As does Allison, KTI, p. 172.
182. See, for instance, Makkreel, IIK, p. 47. See also p. 57. The contrast
between the pre- and the post-categorial is used by Ameriks, ‘Review’,
p. 230, referring to IIK, p. 45. Ameriks’ remarks on a pre-categorial
reading Makkreel attributes to Guyer and to an objection raised by
Meerboote, namely that such a position leads to the conclusion that
all objects are beautiful. (I discuss this problem in Chapter 8,
pp. 284–90.) See Makkreel, IIK, p. 49; Guyer, Kant and the Claims of
167
Taste, pp. 86 ff. Meerboote, ‘Reflection on Beauty’, in Essays in Kant’s

Aesthetics, ed. Ted Cohen and Paul Guyer, p. 79. Ameriks then argues
for an alternative and, he believes, more successful argument for the
pre-categorial significance of taste. The general conditions of cognition
are made use of by, but are not sufficient conditions of, taste.
See Ameriks, ‘Review’, pp. 233–4; Ameriks, ‘How to Save Kant’s
Deduction of Taste’, pp. 295–302. I agree with Ameriks that a pre- and
a post-categorical account of aesthetic judgement need not conflict.
However my account of the pre-cognitive role of aesthetic judgement
is quite different from his. I will argue in the next chapter that it is the
relation in which we stand to the object, rather than the object per se,
that is central for aesthetic judgement.
183. See especially Makkreel, IIK, pp. 105–7. See also Douglas Burnham,
Kant’s Philosophies of Judgement, pp. 149 ff., see especially p. 172.
Aesthetic judgement is schematising without a concept and this is pos-
sible because a free sensible lawfulness can be felt. Later Burnham goes
on to investigate the way in which feeling reveals the structure of life.
185. Ibid., p. 106.
186. CJ, AA 204. My turn of phrase here allies Kant’s phrase with Hannah
Arendt’s expression ‘the life of the mind’; see Hannah Arendt, The Life
of the Mind.
168
The Completion of the Subjective Deduction in

the Deductions of the Critique of Judgement
In the previous chapter I suggested that aesthetic judgement reveals

the synthesis in process necessary to all judgements. In this chapter I
argue that synthesis in process is best understood as the subjective
side of the deduction, often referred to as ‘the subjective deduction’.
In the first section (pp. 170–6) I discuss Kant’s distinction between
subjective and objective deductions and insist that these are two sides
of the deduction, rather than two separate deductions. The subjective
side of the deduction is the cooperation of the faculties, or synthesis,
necessary for any judgement. However at this stage of his presenta-
tion, Kant is hesitant, although not entirely negative about the signif-
icance of the faculties for his epistemology. I suggest that the positive
presentation of the subjective deduction is provided in the Critique
of Judgement. I take as a clue Kant’s statement that all objectively
valid judgements are also subjectively valid. I then show how aes-
thetic judgements – bearers of subjective universality – are not solely
a class of judgements set apart, but are internally linked to the struc-
ture of judgements already analysed in the first Critique.
In the second section (pp. 177–89) I discuss the claim in Section 21
of the third Critique that aesthetic judgement contributes to a non-
sceptical epistemology by uncovering a principle of common sense. I
assess the degree to which Kant is successful in avoiding the conclu-
sion that cognition is dependent on aesthetic judgement. In contrast
to the interpretation of Henry Allison, I suggest that this section
counts as a first attempt at providing a deduction of aesthetic judge-
ment insofar as it offers a completion of the subjective deduction.
Allison denies that Section 21 is concerned with aesthetic judgement
and I discuss his interpretation in the third section of this chapter
(pp. 189–93). Although Kant does not as yet provide a convincing
deduction, I cannot accept Allison’s view that Section 21 has merely
epistemic resonance. I agree, however, that cognition entails a feeling
and that sensus communis has a cognitive application.
169
In the fourth section of this chapter (pp. 193–201), I turn to the

structure of the official deduction. I argue that the distinctive contri-
bution of the latter is its focus on the ‘power of judgement as such’
and show how the latter is only revealed directly in judgements of
taste. However, I also suggest that the failure of this version of the
deduction is that there is no explanation of how aesthetic judgement
relates to the subjective conditions of cognition.
I The subjective and objective deductions

On first sight, it may seem that a central passage in the first Critique
runs totally against the emphasis I have put on the cooperation of the
faculties and the account of synthesis I have built upon it. In the
Preface to the A edition, Kant distinguishes the subjective and objec-
tive deductions. The subjective deduction is the account of the role of
the faculties in cognition and, as we will see, is given at best an ancil-
lary role. But we will see that it is not so much that there are two
deductions – one subjective and one objective – but that there are two
sides to cognition. On one side there is the determination of the object
by the understanding, while on the other is the cooperation of the fac-
ulties necessary for that determination. Recognising this is the first
step to showing how the aesthetic play of the faculties reveals the
necessary condition for knowledge of objects in the world.
The ‘subjective deduction’ is often taken to refer either to the
A ‘Deduction’ in general or more precisely to its second section,
where Kant outlines the three syntheses necessary for knowledge.1
Admittedly, Kant makes the distinction between the subjective and
objective deductions only in the Preface to the A edition, but this dis-
tinction reveals a general structure of his method. I will now examine
the problematic status of the ‘subjective deduction’ in its first and only
explicit appearance in the text and show how Kant holds out a ten-
tative promise of its having a more positive role. While this alterna-
tive role can be discerned implicitly throughout the arguments of the
first Critique, it is not until the third Critique that its full potential is
realised.
In the Preface to the first edition of the Critique of Pure Reason,
Kant says that the subjective deduction is of great importance, but not
essential for the main argument which is supplied by the objective
deduction. Kant says that there are two sides to the enquiry ‘for
exploring the faculty which we entitle understanding, and for deter-
mining the rules and limits of its employment’.2 He refers, almost in
170
Subjective Deduction in the Critique of Judgement
passing, to these two sides as the objective deduction and the subjec-
tive deduction, saying of them:
The one refers to the objects of pure understanding, and is intended to
expound and render intelligible the objective validity of its a priori con-
cepts. It is therefore essential to my purposes. The other seeks to investi-
gate the pure understanding itself, its possibility and the cognitive faculties
upon which it rests; and so deals with it in its subjective aspect.3
The objective deduction establishes that the categories are valid in
respect of the objects that provide the content of the judgements we
make using our power for unification, the understanding. The sub-
jective deduction reflects back on that capacity. Kant elaborates that
such a reflection would concern the faculties upon which the under-
standing rests. Why does he not simply refer to the capacity of under-
standing alone? It is because, as we saw in the last chapter, the power
of understanding is only able to achieve unification of objects insofar
as it cooperates with the faculties of intuition and imagination.
For his immediate purposes, Kant announces the relative lack of
importance of the subjective side of the deduction:
Although this latter exposition is of great importance for my chief
purpose, it does not form an essential part of it. For the chief question is
always simply this: – what and how much can the understanding and
reason know apart from all experience? not: – how is the faculty of
thought itself possible?4
Kant refers us forward to the ‘Transition to the Deduction’ at A 92–3
for a discussion of the objective deduction. This passage is repeated
in the B edition and serves as an introduction to both editions of the
‘Deduction’ in providing a short account of their shared purpose.5
The argument is almost identical to the account of the Copernican
revolution presented in the Preface to the B edition and discussed
in Chapter 3, pp. 89–95.6 Just as in the latter passage, in the
‘Transition’ Kant says that either the object makes the representation
possible or vice versa and opts for the latter option, but now he goes
on to clarify something that was not stated in the account of the
Copernican revolution, namely, that this does not entail that the rep-
resentation produces the existence of the object and only that it deter-
mines the latter.7 We can conclude that insofar as the representation
does not ‘produce’ the object, Kant distinguishes his formal idealism
from material idealism. As I also argued in Chapter 3, pp. 100–5,
representations take up the affect given in sensibility and thus we are
not condemned to the circle of representation.
171
In the ‘Transition’, Kant anticipates the ‘Deduction’ concluding

that:
The objective validity of the categories as a priori concepts rests, therefore,
on the fact that, so far as the form of thought is concerned, through them
alone does experience become possible. They relate of necessity and a
priori to objects of experience, for the reason that only by means of them
can any object whatsoever of experience be thought.8
The objective deduction qualifies as objective insofar as it establishes

the application of the categories to all objects of experience. But
are the subjective and objective sides so distinct from one another?
What is to be proven by the objective deduction is that a subjective
faculty, the understanding, is capable of introducing order into the
objective world. Experience becomes possible through a form of
thought. This shows that the objective deduction is dependent on the
operation of a subjective faculty, the understanding. Moreover, the
cooperation of faculties already observed in the initial statement of
the Copernican revolution is still required in the ‘Transition’, for Kant
insists that there must be two conditions under which the knowledge
of an object is possible. There must not only be a concept, but also an
intuition.9
In the previous chapter, I discussed the way in which faculty talk
is used to draw out the character of figurative synthesis in the B
‘Deduction’. The latter form of synthesis prepares for the application
of categories to empirical objects and its distinctiveness lies in being
presented as a three-term relation among the faculties, in contrast to
the two-term intellectual synthesis. The faculty talk is not restricted
to Section Two of the A ‘Deduction’, although it is undoubtedly high-
lighted there. Throughout the ‘Analytic’ of the first Critique, and in
both editions, Kant develops his account of the rules and limits of the
understanding – that is, of the objective deduction – using a vocabu-
lary of the cooperation between the faculties of intuition and under-
standing. I have given a case study of this in my interpretation of the
B ‘Deduction’ and close examination of the ‘Schematism’ and of the
‘Principles’ will reveal that Kant continually presents his argument in
terms of a relation between understanding and intuition.10
Often commentators regard Kant’s faculty talk as, at best, a stylis-
tic irrelevance and, at worst, a serious mistake, resting on a covert
metaphysical agenda. Kant himself remarks that the subjective side of
his undertaking is of great importance and, against the grain of his
initial characterisation of the situation, I will argue that it can be
172
established as a necessary condition of cognition and thus of the

objective deduction. It is tantalising that even in the initial statement
in the A Preface where Kant characterises the subjective deduction as
not essential, he suggests that he intends to give a further defence of
its status. He first remarks that it may seem ‘hypothetical in charac-
ter’ insofar as it seems to involve ‘the search for the cause of a given
effect’.11
This would involve trying to establish how the faculty of thought
itself is possible, in contrast to restricting investigations to the ques-
tion ‘what and how much can the understanding and reason know
apart from all experience?’12 Thus the danger is that the subjective
deduction appears to go beyond the limitations of the transcendental
analysis of knowledge in search of a cause for thought. This would be
to seek a transcendent explanation of experience, whereas Kant
insists that it is only valid to seek out the transcendental structures of
experience. He is worried that the subjective deduction risks seeking
a genetic explanation for thinking, which would involve deriving the
latter from a metaphysical or psychological foundation.
Kant goes on to say that he will show elsewhere that the subjec-
tive deduction is not hypothetical in character.13 I take this as
showing he believes that it is possible to investigate the subjective
conditions of knowledge on the basis of a properly transcendental
method. But it is unclear how – or, indeed, where – he intended to
return to this question. My suggestion is that, whatever his inten-
tions at the time of writing the Preface to the first Critique, the argu-
ment of the third Critique, with its establishment of the synthetic
activity of the faculties as a pre-condition for the possibility of any
cognition, fulfils the promise made earlier. The faculty talk is no
longer at risk of straying into a search for the transcendent origin of
thought and is given a transcendental role insofar as the activity of
thinking is revealed as providing the subjective, though universal
and necessary, form of experience. As such, the subjective deduction
is no longer simply an inessential, though important vehicle for the
expression of the possibility of knowledge and has become part of
the internal structure of the ‘chief purpose’, that is, of the objective
deduction. At this stage the subjective and objective trajectories
truly become ‘two sides’ of one argument. If we are to determine
‘how much understanding and reason can know apart from all
experience’ we cannot ignore the activity of thinking, which on
deeper investigation is revealed as a cooperation of the faculties
necessary for any judgement.
173
If we now turn to Section 8 of the third Critique, we find a clue to

the link between objective and subjective validity:
Now a judgment that is universally valid objectively is always subjectively
so too, i.e., if the judgment is valid for everything contained under a given
concept, then it is also valid for everyone who presents an object by means
of this concept. But if a judgment has subjective – i.e., aesthetic – univer-
sal validity, which does not rest on a concept, we cannot infer that it also
has logical universal validity, because such judgments do not deal with the
object [itself] at all. That is precisely why the aesthetic universality we
attribute to a judgment must be of a special kind; for although it does not
connect the predicate of beauty with the concept of the object, considered
in its entire logical sphere, yet it extends that predicate over the entire
sphere of judging persons.14
All objective cognitive judgements are valid objectively insofar as
they determine the object by means of a concept. They state, for
instance, that such and such a thing is an x. They thus achieve a deter-
mining or explanatory synthesis vis à vis the object. But in doing so,
they also bear subjective universality because they must be valid for
all judging persons. Grasping the object simply means to represent
reality in a way that is valid for everyone. Thus the objective deduc-
tion necessarily has a subjective dimension in commanding the agree-
ment of all judging subjects.
In contrast, aesthetic judgements do not determine the object.15
They are, however, subjectively valid and thus valid for all judging
persons. In Section 9 Kant develops his account of subjective validity,
arguing that aesthetic judgements display the ‘merely subjective deter-
mining basis’ of all cognitions.16 They are characterised by a ‘relation
between the presentational powers insofar as they refer a given pre-
sentation to cognition in general’.17 He calls this a ‘free play’ in which
the cognitive powers harmonise.18 Kant concludes that this relation
must hold for everyone and thus counts as universally communicable
in that it is the subjective condition of cognition.19 Thus the subjec-
tive validity of aesthetic judgements arises from their being based on
a relation between the faculties necessary for any judgement whatso-
ever. The subjective validity that was established as necessary for any
objective judgement in Section 8 is revealed in the following section
as the subjective cooperation of the faculties on which aesthetic judge-
ments are exclusively based.
In Section 9, Kant’s account of the relationship between aesthetic
judgement and cognition reveals a problem that haunts his account on
many occasions. How can it be that aesthetic judgement is distinct
174
from cognition and yet displays the necessary subjective conditions of

cognition in general? If cognition were founded on the specifically har-
monious play of the faculties, then it would count as aesthetic. We can
assume that Kant does not want to argue this. From the outset of the
first Critique he is committed to the distinctiveness of cognitive and
aesthetic principles, while he comes latterly to recognise their system-
atic connection. However, the establishment of that systematic con-
nection causes him many problems. For the moment he is at risk of
suggesting that the harmony of the faculties characteristic of aesthetic
judgement is necessary for judgements giving rise to knowledge.
There are two possible amendments that could solve the local
problem in Section 9. Either ‘cognition in general’ is taken to refer to
something distinct from the general case of cognition, or, the harmony
of the faculties is exemplary of, but not identical to, the subjective
grounds of cognition.
In the first case ‘cognition in general’ would be taken to refer to the
general possibility of cognition displayed in aesthetic judgements,
despite the fact that they are not cognitive. This general possibility
would stop short of actual cognition, establishing only its form.
Aesthetic judgements would display the form of cognition, that is, its
subjective conditions, but not its actuality. In an aesthetic judgement
the faculties are prompted to cooperate in a way that is particularly
conducive for cognition, but no knowledge is achieved because our
mental activity is so free that it stops short of the determination that
is the objective condition of cognition. It is as if we were engaged in
a cognitive exercise, but our exploration is so free and open-ended
that cognition will not be achieved.
The other option is to retain the natural understanding of ‘cogni-
tion in general’ as referring to all cognition and fine-tune Kant’s argu-
ment so as to say that aesthetic judgement is characterised by a
harmonious relation of the faculties, which is a heightened expression
of the mutual relation of the faculties necessary for any cognition.20
As in the previous interpretative option, aesthetic judgement displays
the form of cognition, but is not included within the range of the
latter.21
The two interpretations converge insofar as they establish that aes-
thetic and cognitive judgements are systematically connected, but dis-
tinct from one another. This is surely what Kant intends, although he
has not yet established the apparatus to express his position. We will
see that in later discussions he makes progress in articulating his intri-
cate position, while never quite pinning it down. For the purposes of
175
our current discussion both amendments allow the same conclusion,

namely, that aesthetic judgement is subjectively valid insofar as it
reveals the subjective activity necessary for any cognition whatsoever.
We saw in the last chapter that aesthetic judgement is characterised
by a double harmony insofar as the thing we deem beautiful is pur-
posive for our mental activity, while the latter is characterised by a
harmonious play. While this does not amount to a determinate grasp
of reality, as in the cognitive case, we nevertheless feel that it is appro-
priate to attribute such a harmony to all other judging subjects. The
subjective validity of the aesthetic judgement is based on the latter’s
being grounded in the subjective activity displayed by all judging sub-
jects, which makes possible our response to an object in any judge-
ment. Thus the subjective side of the deduction – that is, the
cooperation of a plurality of the faculties – and the relation in which
that subjective side of judgement stands to the objective side, is
brought to the fore in aesthetic judgement. It is this link to the general
form of cognition that establishes the universal validity of what
would otherwise be merely subjective.
Aesthetic judgements reveal the subjective side of the deduction, as
first outlined in the Preface to the A edition of the first Critique.
Moreover, they do so in such a way as to show that the subjective
deduction is not to be understood hypothetically or genetically, but
rather as part of the formal structure of validity in general, a validity
that always has two sides, that of the subjective, but universal struc-
ture of judgement and that of the latter’s application to an intentional
object. Put in this way, we can see that the subjective side of the
deduction is an essential component of establishing objectivity. Thus
we can justify the suggestion I made in the previous chapter that aes-
thetic judgements contribute to the general transcendental project of
establishing the possibility of a priori synthetic judgements, for they
reveal the synthetic activity of the faculties necessary for any cogni-
tion. Although there is a systematic link between aesthetic and cog-
nitive judgement insofar as they both rely on a subjective process of
synthesis and both can claim subjective validity, they are distinct. In
the absence of objective validity, subjective validity functions not just
as one side of a larger story, but as a constitutive principle in its own
right. It is for this reason that it becomes possible to inspect the sub-
jective side of the transcendental project in a way that was not possi-
ble when the principal concern was with the determination of
knowledge per se. However Kant only achieves this distinction in the
official version of the deduction, as we will see.
176
II The anti-sceptical force of aesthetic judgement

Kant does not, however, believe that he has provided the definitive
account of the relation between aesthetic and cognitive judgement in
Section 9 of the Critique of Judgement. In the fourth Moment of the
‘Analytic of the Beautiful’, he provides an argument for the position
he has already sketched in the second Moment. In Section 21, Kant
claims that aesthetic judgement is based on a principle that is neces-
sary for any non-sceptical logic or epistemology. While this may
appear to be a remarkable claim, it was in fact entailed in nuce in the
earlier position we have just examined. What is clear is that he cannot
mean that the harmony of the faculties characteristic of aesthetic
judgement is a necessary condition of determining cognitive synthe-
sis. This would indeed qualify aesthetic judgement as necessary for
resisting scepticism, but would at a stroke render otiose the distinc-
tion between determining and reflective judgement.22 In the light of
my reading of the ‘subjective deduction’, we can now make sense of
his claim, explaining how judgements that are strictly subjective in
their universality nevertheless have a role to play in what must be a
cognitive argument. We will now discover the helpfulness of the clue
discovered in Section 8, namely, that there are necessary subjective
conditions for every objective judgement.
The link Kant makes between aesthetic judgements and anti-
sceptical arguments is significant for a theme to which we will return.
It seems that it would be too simple to take at face value the many
occasions on which Kant says that aesthetic judgements have nothing
to do with the object. In the passage from Section 8 cited in the last
section, he says first that ‘such judgements do not deal with the object
at all’. However, he then goes on to say that they do ‘not connect the
predicate of beauty with the concept of the object, considered in its
entire logical sphere’.23 This second, more careful statement leaves
open the possibility that aesthetic judgements relate to the object,
while not determining it under a concept. If aesthetic judgements con-
tribute to anti-sceptical arguments, we can say, at the very least, that
they stand in some relation to objects insofar as they reveal the neces-
sary, though insufficient, subjective conditions for the possibility of a
successful objective judgement.
The general goal of the fourth moment of the ‘Analytic of Taste’ is
to establish that the liking we feel for beautiful objects is not merely
contingent, but necessary. In Sections 18 and 20, Kant argues that if
a pleasure is necessarily associated with judgements of taste, then they
177
must be determined by a subjective principle that grounds their uni-

versal validity.24 This is the principle of common sense presupposed
in all aesthetic judgement. Common sense is not to be confused
with outer sense, as it is ‘the effect arising from the free play of
our cognitive powers’ (die Wirkung aus dem freien Spiel unserer
Erkenntniskräfte).25 We might be tempted to conclude from this that
common sense is the empirical effect of a priori activity. But Kant
cannot mean this, for common sense is the principle on which aes-
thetic judgements are based. As such, it is a transcendental principle
of experience and is distinct from the other principles we have exam-
ined so far, strictly in that it functions as a feeling.
For a clarification of Kant’s elaboration of common sense we need
to look to Section 40 where Kant characterises it as ‘an effect that
mere reflection has on the mind’.26 Common sense is a feeling that
arises from the activity of the mind and acts as the principle govern-
ing aesthetic judgements, that is, it supplies the standard for the latter.
Insofar as aesthetic judgements presuppose common sense, they rest
on our ability to respond to objects with a free play of the mind that
gives rise to a feeling of pleasure. In responding aesthetically to things
we judge them on the basis of a free play of the faculties. The feeling
of common sense is this play of the faculties affecting our minds so as
we become conscious of it, at least implicitly. In Section 20, common
sense is linked with aesthetic judgement, but later it emerges that it
has a wider application. Common sense is, thus, an ability to coordi-
nate a plurality of distinct mental operations, but, for Kant, this
description is not restricted to the mental activity of a solitary judging
subject. Judgements based on common sense are necessarily commu-
nicable to other judging subjects, as we will see in what follows. Kant
thus establishes the inter-subjectivity of mind in the course of his
analysis of judgements of beauty.
The title of Section 21 asks ‘[w]hether we have a basis for presup-
posing a common sense’.27 Given the preparation for this question in
the discussion of aesthetic common sense in Section 20, it would be
very unnatural if Kant now restricted his discussion, as Allison thinks
he does, solely to cognitive judgements.28 Admittedly Kant fails to
make an adequate distinction between cognitive and non-cognitive
judgements, as we will see. This, however, results not in a wholly epis-
temic proof as Allison suggests, but in an epistemic bias in the way
Kant connects aesthetic to cognitive judgement. As I read this section,
the problem Kant now poses is this: how can the aesthetic principle
introduced in the previous section count as subjectively necessary for
178
experience? He attempts to answer this affirmatively by showing that

common sense is a necessary subjective presupposition of ‘cognition
in general’.
The argument is most naturally divided into seven steps.29 After a
promising start where he refers to judgements in general, he restricts the
scope of his argument to cognitive judgements. This restriction oper-
ates as the effective premise of the rest of his argument and results in
a progression from cognition, via aesthetic judgements, to common
sense, culminating in the claim that common sense counts as a presup-
position of cognition. As Henry Allison has recently put in question the
idea that this section is concerned with aesthetic judgement at all, I must
reconstruct its stages in some detail. My aim is to show how Kant’s
account provides an extended, although flawed, argument for the claim
he made in Section 9, namely, that aesthetic judgements display the sub-
jectively valid structure necessary for all cognitive judgements.
The first step refers to judgements in general, not only determining
ones or cognitions. Nevertheless, insofar as Kant focuses on the
problem of grasping objects, he seems to be concerned principally
with judgements giving rise to knowledge of objects.
1. Cognitions (Erkenntnisse), judgements and the conviction accom-
panying them must be communicable. Otherwise they would
amount to no more than a subjective play of the faculties and there
would be no guarantee that they grasp their objects, just as scep-
ticism suggests.30
Kant’s claim is that if cognitions are to count as objectively valid –
and thus give rise to knowledge – then their truth must be communi-
cable, that is, they must be acknowledgeable by all judging subjects.
Something that is known is not simply the opinion of one or more
subjects. Kant thus begins his argument by restating the connection
between what, in Section 8, he called objective and subjective valid-
ity. Objective validity requires subjective validity, that is, the agree-
ment of all judging subjects.
While admitting the plausibility of Kant’s claim for the inter-sub-
jective communicability of knowledge, Guyer fails to find a successful
defence for this thesis at any stage in the critical opus.31 Kant’s posi-
tion in the ‘Transcendental Deduction’ of the Critique of Pure Reason
is one of methodological solipsism in Guyer’s view and he sees this as
standing in tension with the development of Kant’s argument in
Section 21.32 His charge that Kant fails to provide a convincing argu-
ment for the inter-subjective status of cognition is a fair one, for if any
179
is provided, it is at best indirect. Nevertheless, while it is true that the

‘Deductions’ – and even more relevantly the ‘Analogies’ – distinguish
between subjective and objective order without any explicit reference
to inter-subjectivity, Kant surely held as unquestionable that his analy-
sis was of mind in general and not of an independent subjective
psyche.33 And if it is necessary for any mind to apply the same epis-
temic rules, then, in principle at least, a space is left for communication
between minds at the transcendental level. Admittedly, this potential
is insufficiently developed in Kant’s epistemological account of the
status of mind and we must conclude that the premise of Kant’s argu-
ment in Section 21 is not subjected to proof because it counts as an
unquestioned and undefended axiom. But, while not proving the
validity of this fundamental presupposition, the account of communi-
cability in the third Critique reveals Kant’s commitment to it, not only
in aesthetic judgements, but for cognition in general.
The second stage of Kant’s argument deepens the analysis of the
subjective side of cognition. He moves from the subjective conviction
characteristic of any successful claim to knowledge, to the subjective
grounds that make that conviction possible. Notably he now speaks
only of cognitions and not of judgements in general.
2. If claims to knowledge are communicable, then the attunement or
proportion of the faculties, as the subjective condition of those
claims, must also be communicable.34
The reference to the proportion or relation of the faculties shows that
this is the real locus of the subjective side of the deduction. A valid
claim to knowledge and its attendant conviction are communicable
because they are based on a cooperation among the faculties and this,
too, must be shown to be universally valid. As Guyer remarks, this
does not simply repeat the subjective aspect of the first step, but
extends the latter, introducing the ‘“subjective condition” of knowl-
edge’.35 But knowledge can only be based on subjective capacities that
are shared by all judging subjects and thus cannot be merely psycho-
logical as Guyer concludes.36 Cognitions must display a universal
validity, complementary to that of the objective side of the deduction
of cognitive claims. The subjective condition of cognition that Kant
now calls the attunement or proportion of the faculties is what, in the
previous chapter, I identified as the cooperation or mutual relation of
the faculties necessary for any cognition.37
The next two sentences of Kant’s argument correspond with two
distinct steps. First, he claims:
180
3. This attunement actually takes place when an object induces the

imagination to combine the sensory manifold and the imagination
in turn induces the understanding to provide unity for this mani-
fold in concepts.38
Kant now makes the crucial claim that the subjective conditions nec-
essary for objective validity do, in fact, hold. Our faculties actually
cooperate in the way that allows them to grasp an object on condi-
tion of a dual operation. Kant’s account here recalls the two stages of
the Copernican revolution.39 It is also very similar to the account of
synthesis given in the A ‘Deduction’.40 On the one hand, the object
affects us in such a way that intuition in collaboration with imagina-
tion combines the sensible given. On the other hand, understanding
unifies what has been ‘run through and held together’ by the imagi-
nation.41 The third step provides an anatomy of the proportion or
cooperation among the faculties that the second step claimed was nec-
essary for any cognitive judgement. There is no reason to suspect that
Kant is, as yet, speaking of the special proportion characteristic of
aesthetic judgement.
This account of the attunement or proportion of the faculties
counts as the subjective side of the deduction. The determination of
the object only occurs insofar as there is a cooperation of the cogni-
tive faculties. Thus, the subjective side of the deduction is not merely
hypothetical but necessary for cognition. The objective deduction is
only concluded when its subjective side is taken into consideration.
However, we have not yet heard the full story.
4. There is however variation in the proportion of this attunement,

‘depending on what difference there is among the objects that are
given’.42
This is the crucial distinction that makes possible the introduction of
aesthetic judgements in the fifth step and for this reason I identify it
as a distinct step, in contrast to Allison.43 Not all cooperation among
the faculties is identical. But how can there be a difference of pro-
portion among cases in which the understanding unifies the mani-
fold? We can concede that the content of judgements about objects
vary according to their referent, but does it make any sense to say that
the formal subjective conditions of cognition vary? Surely any cogn-
itive judgement will display the same proportion or relation of the
faculties, because otherwise they would not count as subjectively
universal?
181
But Kant is not trying to establish a diversity of subjective condi-

tions among cognitive judgements, the formal conditions of which are
consistent. He is, rather, preparing the ground for establishing the
distinctiveness of a different set of judgements – aesthetic judgements
– while maintaining their systematic connection to cognitive judge-
ments. Aesthetic judgements, as a class, display a different attune-
ment that will be described in the fifth step.
Admittedly, the difficulty of establishing that this is Kant’s intent
is compounded by the fact that the third step appears to refer only
to cognitive judgements characterised by unification. This encour-
ages the conclusion that Kant is now making a distinction within the
range of determining judgements. As the special attunement he is
about to identify is characteristic of aesthetic judgements, this would
mean that the latter are a subset of judgements displaying unity. But
aesthetic judgements qualify as non-cognitive insofar as they are not
determined, that is, unified. They remain strictly reflective. What
Kant should have said was that all judgements entail that the mani-
fold is taken up through a combination of imagination and under-
standing, but there are, nevertheless, some that do not attain unity.
In these judgements there is a qualitatively different relation among
the faculties that nevertheless reveals the general subjective condition
of cognition. He would then be in a position to say that the peculiar
subjective proportion of the non-cognitive aesthetic judgement is
capable of revealing the mutual relation necessary for any cognitive
judgement.
The problem about the range of the third step can be traced back
to the second step, and even to the elaboration of the first step. While
Kant starts his putative deduction of common sense from the per-
spective of the range of judgements in general, he quickly suggests
that he is concerned only with cognitive judgements. A problem arises
when he needs to distinguish the species of aesthetic judgements from
cognitions in the fifth step. Kant should have maintained the position
that his argument concerns judgements in general, within which he
could have distinguished determining and aesthetic judgements. This
would have resulted in the following amended argument:
[Step 1] All judgements entail universality of communicability and

this is true even of those that bear objective validity.
[Step 2] Universality of communicability entails that the subjective
conditions of judgements must be based on a cooperation of the
faculties, characteristic of all judging subjects.
182
[Step 3a] This attunement of the faculties does in fact take place when
the manifold of sensation is unified in a cognitive judgement.
[Step 3b] But a subjectively universal relation between the faculties is
also displayed by aesthetic judgements, which appear to follow a
law or rule, while not giving rise to determinate unity.44
[Step 4] The particular relation between the faculties in any judge-
ment depends on variation in the objects under consideration.
This would allow him to proceed to the next stage of his argument,
namely,
[Step 5] Only aesthetic judgements display the particular proportion
of the faculties that qualifies as a harmony.
I suspect that Kant thought he could move directly from cognitions
to aesthetic judgements because he was committed to the view that
the latter rest on the subjective structure of ‘cognition in general’.
In the course of my discussion of Step 4, I have touched on an
important move that once again puts the impositionalist reading of
Kant under pressure. Kant distinguishes not only between different
forms of judgement, but also says that the difference in attunement
between judgements rests on a difference among the objects. Different
objects give rise to different subjective affects and thus have a role to
play in the content of cognition. But this does not mean that the object
qualifies as a pure given. The object can only give rise to an affect
insofar as it sets the subjective faculties in motion. What we finally
know is the object as it has been taken up through our powers of rep-
resentation. There is no epistemic thing-in-itself, but only an object
that stands in relation to a judging subject. Nevertheless, formal ide-
alism requires that the material existence of the object supplies the
content for cognitive judgements. Form takes up content and is not
simply an imposition on the latter.
I have also argued that the variation in proportion in question is
between the forms of cognitive and aesthetic judgement. If this dis-
tinction arises from a difference among objects, then it strongly sug-
gests that, for Kant, there is something in the object that gives rise to
aesthetic pleasure and that only some objects will give rise to judge-
ments of taste.45
Kant now moves to discuss the special case of aesthetic judge-
ments:
5. But there must be one attunement ‘in which this inner relation is
most conducive to the (mutual) quickening of the two mental
183
powers with a view to cognition (of given objects) in general: and

the only way this attunement can be determined is by feeling
(rather than by concepts)’.46
Following the adjustment I have suggested, namely that Kant should

have consistently referred to judgements in general and not just to
those giving rise to cognition, we can now make sense of the distinc-
tion expressed in this step. Judgements determined by a feeling of
common sense are aesthetic and are characterised by a cooperation of
the faculties that is highly congenial to cognition. If this reconstruc-
tion is right, then it might strike us as paradoxical that no knowledge
arises. Surely, if the activity of the faculties is most conducive to
cognition, cognition will result! The answer lies in Kant’s claim that
the attunement is conducive to ‘cognition in general’ (Erkenntnis
überhaupt). This is the same phrase that was used in the problem-
atic account from Section 9 we have already considered in the
previous section and for which I have suggested two possible inter-
pretations. The crux of both options is that aesthetic judgement dis-
plays the formal conditions of cognition, while falling short of actual
knowledge.
Guyer, as I have already noted, concludes that aesthetic judgement
can only have a psychological status due to a problem he finds in
Kant’s account of proportion. How can it be that all cognitions
display a proportion or attunement between the faculties and yet aes-
thetic judgements display a different attunement?47 If all judgements
display the same attunement of the faculties, this would render impos-
sible a distinction between cognition and taste, resulting in the con-
clusion that all things are beautiful. On the other hand, Guyer argues,
if the difference between proportions is stressed, then it would mean
that aesthetic judgements are not linked to the transcendental ground
of cognition, and rather count as merely psychological variations on
our capacity to judge.48
But the key to answering Guyer lies in challenging his reading of
attunement as implying the unification of the manifold. Following the
adjustment to Step 3 that I have suggested, and making use of the
terms I introduced in the previous chapter, the attunement of the fac-
ulties is the activity of synthesis that results only in some cases in uni-
fication of the manifold. If we distinguish between synthesis as
activity and synthesis as result, we can make sense of the distinction
between different proportions of judgement without breaking the link
between aesthetics and cognition. In an aesthetic judgement we are
184
aware of the sort of synthetic activity that would, under other cir-
cumstances, give rise to a determination or unification of the mani-
fold. However, on this occasion a determinate unification does not
arise because of certain qualities of the object and the way in which
we respond to it.
Guyer assumes that either aesthetic judgement can display an
optimal attunement, in which case in both forms of judgement the
proportion must always be the same, or taste fulfils no such role in
relation to cognition.49 But why should this be so? This would mean
that if cognition and aesthetic judgement are linked, then they must
have an identical subjective formal structure. For Kant, the formal
structure of a judgement is the subject of a transcendental investiga-
tion. It was necessary to write three Critiques and not just one,
because the formal structures of cognition, morality and aesthetics
are not the same as one another. Nevertheless, the three main orien-
tations in human life are connected, not only empirically insofar as
we connect all three in living our everyday lives, but also because
there are systematic connections between the distinct forms by which
we introduce order into our lives. Aesthetic judgement’s form is dis-
tinguished insofar as aesthetic judgement displays a different attune-
ment of the faculties from the relation required for cognition.
Despite my conviction that the text can be reconstructed so as to
make sense, I have argued that it does not work as it stands. I am par-
ticularly uneasy with the stipulative style in which Step 5 is expressed.
Kant says that there ‘must’ be an attunement most conducive to cog-
nition. Why, we might ask, must there be such an attunement? There
is no reason in the world why there must be, unless we believe, for
instance, that there is a metaphysical order of things that dictates that
beauty and knowledge are, perhaps at some higher level, one. Kant’s
presentation in Step 5 should have followed that of Step 3 where he
makes a claim about the actuality and not merely the possibility, of
cognitions. He should have said that there are, in fact, experiences –
namely aesthetic ones – in which we are aware of a particularly con-
ducive relation among the faculties. This is a fact of experience from
which transcendental analysis begins.
At the end of this stage of his argument, Kant says that, in these
special cases, the relation of the faculties is determined by a feeling.
Why must the attunement be determined at all, especially given that it
counts as free? The answer for this comes in the first sentence of Step
6, which builds on Step 2’s axiom that the subjective condition of cog-
nition – the proportion of the faculties – must bear a universal validity
185
comparable to objective validity and, indeed, necessary for the latter.

Both the objective and subjective conditions of knowledge must be
communicable. Any judgement insofar as it is universally valid must be
governed or determined by a principle. If this were not the case, my
judgement would simply be for me and not for all judging subjects, as
all objective judgements and all subjectively valid judgements must be
(see Steps 1 and 2). Insofar as they bear subjective universal validity,
aesthetic judgements are determined by a feeling. Therefore, Kant now
extends his account of the universal validity of the subjective condi-
tions of knowledge and insists that feeling too must be communicable.
The next move in Kant’s account may appear to contain two steps:
6. Additionally, this attunement and the feeling arising from it must
be universally communicable. The communicability of a feeling
presupposes common sense.50
However I have already shown how the first sentence follows from
Step 2. Thus, there is only one genuinely new piece of information.51
The second sentence, as stated above, could mean one of two things.
Either the claim is that if feelings are acknowledgeable by all, this can
only be because there is a shared common sense on which those feel-
ings are based. Or it could mean that common sense is the only feeling
that bears universal communicability. In the first case feeling would
be based on a further principle, the status of which would have to be
established. But Kant has already said in Step 5 that attunement is
determined by a feeling. We must therefore conclude that the subjec-
tive principle that determines aesthetic judgements is a feeling.52
Aesthetic judgements presuppose common sense, so is the latter
aesthetic? It certainly is tempting to conclude so. We have already
seen that, in Section 20, common sense was identified as ‘the effect
arising from the free play of our cognitive powers’, a distinctively aes-
thetic characterisation.53 Section 21, as I have reconstructed it, has
progressed from a limitation of the argument to a consideration of
cognitive judgements to the special case of aesthetic judgements. It is
only at this stage that Kant introduces common sense as the ground
of the latter, so it seems natural to conclude that common sense, too,
is aesthetic. Now if the feeling characteristic of aesthetic judgement
were based on a further principle of common sense, it would be
arguable that the latter need not be aesthetic. However, I have already
argued that common sense is a feeling. Does this mean that the
reading of Section 21 developed here must conclude that common
sense is aesthetic in status?
186
Kant does not see fit to clarify his position and we are left with the
now familiar problem of an insufficiently articulated account of the
relation between cognitive and aesthetic judgements. As a result,
some commentators have concluded that here he makes cognition
dependent on an aesthetic ground, whereas, going in the opposite
interpretative direction, Allison has argued that Section 21 is not con-
cerned with the relation between aesthetics and knowledge at all. In
contrast to both these options, I believe that Kant is trying to estab-
lish that aesthetic judgements reveal a broader characteristic of ‘cog-
nition in general’, that is, the subjective conditions of cognition that
establish the subjective validity of both cognitive judgements and of
aesthetic judgements. All judgements are based on common sense,
that is, the ability to coordinate a plurality of faculties exercised by a
community of judging subjects. This is the form of cognition, as I sug-
gested in my discussion above of the two possible connotations of
‘cognition in general’.
If Kant is to make his deduction of taste as presented in Section 21
persuasive, he requires, in addition to the adjustments I have sug-
gested above, a distinction between two levels at which common
sense operates. Such a distinction can be discovered in the official
‘Deduction of Taste’, although even there interpretative work is
required in order to draw it out.54 Only in aesthetic judgement does
the subjective condition of all cognition count as a principle in its own
right, that is, as a special principle of reflection.55 We could say that
this is a subjective principle of cognition, but it is preferable, I think,
to say that the aesthetic principle reveals the subjective side of all cog-
nition. The principle of common sense reveals the capacity for
common sense displayed in all our judgements. Thus all judgements
are based on a capacity for cooperation by the faculties that is written
large and exemplified in what is later given the Latin tag of sensus
communis, that is, the principle of aesthetic judgement. The recon-
struction of how aesthetic judgement exhibits the form or possibility
of cognition will be explored in the final chapter of this book.
Kant concludes his argument, claiming that common sense is nec-
essary for epistemologies that reject scepticism. As I have suggested,
his view should be that the principle of, although not the general
ability for, common sense is aesthetic. This would mean that aesthetic
judgements have a part to play in the project of establishing the legit-
imacy of claims to knowledge. They do so, however, only in an indi-
rect fashion, the nature of which will only be clarified at a later stage
of my account.
187
7. We do have a basis (and not a merely psychological one) for

assuming such a common sense as the necessary condition of the
universal communicability of our cognition. This must be presup-
posed by any logic and any non-sceptical account of knowledge.56
Kant believes he has achieved the task he set himself in Section 20.
Judgements of taste successfully appeal to the necessary agreement
of all judging subjects insofar as they are based on common sense.
And, as I have rearticulated his argument, Kant has shown that it is
valid to presuppose common sense as the principle for aesthetic
judgements because the latter express the subjective condition of all
judgements.
Thus construed, Kant’s claims that common sense is required for
any account of knowledge committed to providing an alternative to
scepticism is not implausible. It is merely the culmination of the steps
that have come before. Common sense arises as a particular attune-
ment of the faculties in aesthetic judgement, but, despite its parti-
cularity, it bears a general significance insofar as it is the relation
between the faculties ‘most conducive’ to ‘cognition in general’.
Common sense thus reveals the cooperation or mutual relation of the
faculties that is necessary if the subjective conditions of claims to
knowledge are to be validated. And we have seen that this is also nec-
essary if the objective conditions are to be legitimated. This is exactly
what any non-sceptical account of knowledge and, in particular,
Kant’s transcendental project sets out to achieve. A successful
account of knowledge will need to show that the faculties cooperate
in such a way as to be capable of taking up something extra-mental
given in experience. Aesthetic judgement displays a common sense
that counts as a heightened example of the mutual relation necessary
for any cognition.
The problem remains, however, that Kant has not yet established
the status of common sense. Steps 5 and 6 seem to suggest it is aes-
thetic, while Step 7 leads us to believe common sense is cognitive. I
have suggested a solution that would resolve this tension, but in
Section 21 at least Kant provides no such distinction between
common sense as a general cognitive capacity and sensus communis
as a principle. Moreover, the narrowing of range from judgements in
general in Step 1, to cognitive judgements in Step 2, makes the later
transition to taste extremely awkward. Kant has not yet arrived at an
account that is capable of expressing both the systematic connection
and distinctiveness of cognition and taste.
188
What is, however, clear from Section 21 is that the objective valid-
ity of cognitive judgements is dependent on the establishment of their
subjective validity. The subjective activity of the faculties is necessary
for objective validity.
III Allison’s Epistemic Reading of Section 21

Allison concludes that while it is normally assumed that Section 21
supplies a first attempt at the deduction, it should be read strictly
epistemically and is concerned only with sensus communis as a pre-
condition for knowledge and thus does not attempt to establish that
there is an aesthetic presupposition of all cognition. 57 The problems
we have found in the presentation of Kant’s argument in Section 21
explain why Allison opts for a wholly epistemic reading. Moreover,
it is only much later that Kant presents what he explicitly calls the
‘Deduction of Taste’. Nevertheless, the cost is that what should have
been the core of the fourth Moment of taste becomes a diversion.
And if Kant has not been concerned with showing that judgements
of taste are based on a principle of common sense, we would have to
conclude that the fourth Moment has not even attempted to estab-
lish that judgements of taste exhibit subjective necessity or make a
valid appeal for universal agreement. This is because the appeal to
universal assent concerning taste, which Kant believes it is the object
of the fourth Moment to establish, arises from aesthetic judgements
being based on the subjective conditions of cognition in general. If
Section 21 is not concerned with taste, then Kant has not given us
any reason to believe that the latter makes any such appeal to uni-
versal agreement.
My account diverges most radically from Allison’s at the step I
have identified as Step 5, where Kant introduces the idea of an
‘optimal attunement’.58 My reconstruction of this step is very close to
Allison’s version, the first and relevant part of which reads:
Nevertheless, there must be one optimal attunement, that is, one in which
the inner relation is most conducive to the mutual quickening of the cog-
nitive faculties with a view to cognition in general . . .59
Allison argues that if, as is normally the case, this is taken to refer to
aesthetic judgement, then it would ‘provide a transcendental ground-
ing for taste by linking it directly to the conditions of cognition’.60
Allison says that there are at least two problems with such an argu-
ment in addition to the fact that it ‘does not answer the skeptic’.61
189
Allison now discusses these two problems, which he considers to

be major ones. Firstly, he raises the question: why should the common
sense that must be presupposed as a condition of the communicabil-
ity of cognition have anything to do with the one that supposedly
must be presupposed as a condition of taste? This, he says, is partic-
ularly implausible given Kant’s critique of Baumgarten’s conflation of
cognition and beauty at the outset of the first Critique.62 The second
problem is, he says, the converse of the first. If we read Section 21 as
a deduction of the principle of taste, ‘then it follows that the aesthetic
common sense or taste must itself be presupposed as a condition of
cognition’.63 This is impossible, Allison continues:
if one keeps in mind that the common sense at issue in the case of taste is
the effect of the free play of the cognitive faculties. There is simply no way
in which a feeling resulting from the noncognitive condition of free play
could serve as a condition of cognition.64
Allison concludes that to read Section 21 aesthetically renders its con-

clusion not merely unconvincing, but incoherent.
Allison is quite right to ask if there must be a connection between
the communicability of cognition and that of taste. However, the way
in which he pursues this problem is not very persuasive. Surely it is
clear that Kant shifted position from the stance he took in his
comment on Baumgarten early in the first Critique, which, in any case,
he revised significantly in the second edition. The original version of
the footnote argued that a critique of taste is empirical and psycho-
logical and thus does not qualify for the title ‘aesthetic’, which should
be restricted to the science of sensibility.65 In 1787, Kant drew back
from his former position, saying only that taste is not based on deter-
minate laws and suggesting that a critique of taste could perhaps count
as ‘aesthetic’. The very writing of the Critique of Judgement is testa-
ment to his abandoning the view that the aesthetic, in the sense of
taste, has merely empirical status, though he retains his commitment
to the view that the principles of taste are not determinate. And while
he still insists that cognition and beauty should not be conflated, he is
no longer of the opinion that beauty lies outside the range of tran-
scendental investigation. Consequently, the content of Section 21 is
unlikely to be illuminated by a remark made before the author realised
the possibility, or indeed the necessity, of writing the book within
which its arguments appear. But it is even more odd that Allison sees
the footnote in the first Critique as relevant for this particular discus-
sion, for he later supports the view that aesthetic judgement reveals
190
the subjective conditions of cognition.66 Agreed, Kant owes an explan-

ation of how the principle of common sense can apply to both aes-
thetic and cognitive judgement, but there can be no doubt that he has
moved far beyond the position he espoused in the footnote about
Baumgarten.
The only real problem that Allison raises is that if Kant suggests
that common sense is aesthetic in status, insurmountable difficulties
would result for his claim that it is also the basis for cognition. I have
already suggested a solution to this problem in the previous section,
namely, that common sense reveals the subjective side of cognition
but only qualifies as a principle in the aesthetic case. This amendment
to the text of Section 21 would allow Kant to avoid making an aes-
thetic principle the ground of cognition. While this entails going
beyond Kant’s own argument, it does I think, show that there is no
need to diverge from Kant’s trajectory so radically as does Allison in
his reading of Section 21. While Allison may be more faithful to the
integrity of the section than I have been, he renders its place within
the wider text highly problematic. This removes a great deal of the
plausibility from his epistemic reading, which I will now examine in
its own right.
The sticking point, as Allison realises, for his alternative epistemic
reading is how one is to construe the ‘optimal attunement that can be
determined only by feeling’ without any reference to the feeling
proper to taste.67 He suggests that we can understand this by refer-
ence to the peculiar talent for judgement ‘which can be practised only
and cannot be taught’.68 He suggests that it follows from Kant’s char-
acterisation of judgement ‘that the subsumability of an intuition
under a concept must be immediately seen, that is, “felt” ’.69 Allison
goes on to suggest that this feeling for the attunement of the faculties
is nothing other than the feeling we have for our capacity to judge. If
this feeling and its object are to qualify as universally communicable
in the manner necessary for a non-sceptical position, then, Allison
argues, Kant is right to claim that we must presuppose an epistemic
and non-psychological common sense.70
Allison concedes a weak relevance of Section 21 for the broader
aesthetic argument of the fourth Moment in that, construed epistem-
ically, it has the additional merit of showing that the sensus commu-
nis of taste, from the discussion of which Kant has digressed in
Section 21, is at least not incoherent in principle.71 If I have shown
that the aesthetic reading need not fall into the trap Allison outlines,
there is no reason why my account cannot be compatible with many
191
aspects of his general interpretative approach. In particular, the link

Allison makes between the feeling of common sense and the capacity
for judgement is an interesting and important one.72 I agree that for
Kant judging in general requires some feeling of the fit between the
capacities for intuition and understanding.73
In contrast to Allison, I see no reason to conclude that there is more
than one feeling of judgement or sensus communis, while agreeing
that feeling has a role not only in aesthetic judgement, but also, tacitly
at least, in cognition. Common sense is at work in cognition, but as
its subjective condition and not as a principle. The subsumption
effected by judgement is possible because of the latter faculty’s capac-
ity for coordinating a plurality of faculties. This cooperation of the
faculties is normally unremarked on by us, but in aesthetic judgement
the usually invisible subjective side of cognition becomes available to
us in abstraction from its normal entailments. In this regard, the aes-
thetic principle of judgement uncovers the wit of judgement, insofar
as the latter is a capacity for connection among the faculties.74 At the
point at which common sense becomes a principle determining (aes-
thetic) judgement, it also becomes available for our reflection.
It is not so much that there are two distinct feelings, but rather
two different relations in which we stand to the feeling that arises as
‘an effect mere reflection has on the mind’ in the aesthetic case.75
Cognition necessarily entails a balancing or harmonising of under-
standing and intuition through the intermediary role of the imagina-
tion. To balance the faculties in this way is to be aware, although only
implicitly as our attention lies elsewhere, of an activity, a potential
tension and a possible balance within cognition. I agree that it makes
sense to say that this is a sort of feeling. But it is so subliminal, so
unremarked as to be almost invisible to us when we are concerned to
arrive at some end of cognition.76
In the aesthetic case we are not dealing with a wholly different sort
of feeling, but rather can become aware of the feeling of the activity
of judgement in a way that cognition simply does not allow.77 The
cooperation or activity of mediation necessary for cognition is now a
focus in its own right. In taste the subjective conditions of judgement
are held up for inspection in a way that transfigures their operation
within cognition. Taste is only achieved once these subjective condi-
tions are reflected on in their own right and this gives rise to an
explicit feeling or an effect.
But this also reveals that in cognition there is an aesthetic dimen-
sion, namely, something must be given in intuition if it is to be taken
192
up by the understanding. This process requires a cooperation of the

faculties, which only becomes available to us in experience at a
second level at which Kant uses the term aesthetic, that is, within aes-
thetic judgements. These reveal the subjective side of the deduction
and at the same time the necessary sensory dimension of our knowl-
edge. These two senses of aesthetic are at play in the title of this book.
IV The arguments of the ‘Deduction’

If I am right and Kant already offers an argument for the universal
communicability of taste in Section 21, then why does the ‘Deduction’
of the principle of taste only come much later in the text? Allison is
in a position to answer this, because he reads the earlier account as
concerned only with epistemic common sense.
In order to make sense of the structure of Kant’s account of aes-
thetic judgement, Allison re-positions a distinction deployed by Kant
in the Critique of Pure Reason. Kant distinguishes between the inves-
tigation of matters of fact, and the establishment of the legitimacy of
those facts, by referring to two distinct questions, namely, quid facti?
and quid iuris?78 Transcendental philosophy is concerned only with
the latter question, insofar as it seeks to legitimate the claims to
knowledge that we do, in fact, make. Allison suggests that this dis-
tinction can be used to illuminate the relationship between the
‘Analytic of the Beautiful’ and the ‘Deduction’ finally supplied in
Section 38. While the four Moments analyse the characteristics of
a pure judgement of taste, the ‘Deduction’, he argues, addresses
‘whether a judgment that meets the conditions of purity can make a
rightful demand on the agreement of others’.79 In other words the
‘Analytic’ establishes what a pure judgement of taste would be like,
whereas the ‘Deduction’ establishes its possibility. This is a helpful
heuristic device for reading the extremely complex structure of the
third Critique but is rather neater than a reading of the text supports.
Moreover, a distinction Kant uses to mark the limits of the transcen-
dental project is resituated within the transcendental analysis of
judgements of beauty. Allison concedes that the fourth Moment,
centred on Section 21, adds nothing to the content of the judgement,
although he argues that it has a bearing on the content insofar as
it aims to unite the three earlier elements in the idea of a sensus
communis.80
Establishing the role played by the fourth Moment leads Allison to
argue against Guyer, for whom the second Moment’s claim for the
193
subjective universality of judgements of taste already counts as a first

attempt at a deduction, while the fourth is yet another attempt.81
Allison replies that we should distinguish the second Moment’s claim
for the non-private status of aesthetic judgements from the fourth
Moment’s concern with the claim we make on others to agree with
the latter.82 While Allison is right to draw out the distinctiveness of
the second and fourth Moments, I share Guyer’s sense that the offi-
cial ‘Deduction of Taste’ in Section 38 emerges out of a series of
attempts that are not wholly distinct from one another. The text
would suggest that it was not of the utmost importance for Kant that
one stage of his argument be wholly distinct from its preparation at
a previous stage. For Kant, the fact that the fourth Moment concludes
the identification of the character of taste does not rule out that it
might additionally function as an attempted deduction of the validity
of pure judgements of taste.
But if, as I have suggested, Section 21 is a first attempt at a deduc-
tion of taste, then why does Kant feel the need to refine his account
further in Section 38? While Kant may not argue in a series of well-
defined and distinct steps, no more does he merely repeat himself
without aiming at moving his position forward. How does the offi-
cial deduction deepen the account of the subjective conditions of ‘cog-
nition in general’ that was first raised in Section 9 and returned to in
Section 21? The new element in Kant’s account is that of the power
of judgement as such. While Kant previously talked of the conditions
for judgement in general, he now takes as his theme the power that
makes judgements possible. It is the link between this power and taste
that will finally provide the latter’s legitimation.
The first stage in Kant’s final attempt at establishing the legitimacy
of judgements of taste begins in Section 35. The title of this section
establishes the stakes: ‘The Principle of Taste is the Subjective
Principle of the Power of Judgment as Such’.83 The judgement of taste
has as its basis ‘only the subjective formal condition of a judgment as
such’, that is, ‘the very ability to judge. i.e., the power of judgement’.84
This statement leaves open the possibility that taste is founded on a
principle of judgement that is not aesthetic, as Allison suggests. Kant
goes on to say:
When we use this power of judgment in regard to a presentation by which
an object is given, then it requires that there be a harmony between two
presentational powers, imagination (for the intuition and the combination
of its manifold) and understanding (for the concept that is the presenta-
tion [Vorstellung] of the unity of this combination).85
194
We might suspect that here we re-encounter the ambiguity that

haunted Kant’s earlier attempts at providing a legitimation of taste. For
while Kant seems to be talking specifically about aesthetic judgements
insofar as he refers to a harmony between the faculties, he could be
talking about judgements in general. If he were, he would once again
be guilty of introducing a harmony of the faculties when he should have
spoken only of the subjective conditions of non-aesthetic cognition.86
However, the preceding and succeeding discussion is exclusively con-
cerned with taste and Kant’s intent is to establish the principle that
governs the latter. The mention of harmony in this passage is associ-
ated only with aesthetic judgements, as it always should be. But if this
is right, then the passage also suggests that the power of judgement
exclusively give rise to aesthetic judgements, whereas does it not also
give rise to a range of other species of judgements?
The solution to this problem lies in the insight that a judgement of
taste is characterised by the power of judgement being exercised in
and for itself, in contrast to the cognitive case where judgement is the
means to a cognitive conclusion. Kant immediately goes on to argue
that, as a judgement of taste is not based on a concept of an object –
that is, does not subsume an intuition under a concept – then it ‘can
consist only in the subsumption of the very imagination under the
condition for the understanding to proceed in general from intuition
to concepts’.87 He calls this ‘schematising without a concept’ and says
that this counts as a subsumption of the ‘power of intuitions or exhi-
bitions (the imagination) under the power of concepts (the under-
standing)’.88 This, he concludes, is what he means by a harmony of
the faculties.89 The exercise of the power of judgement in its own right
is peculiarly allied to judgements of taste insofar as they have as their
ground the power of judgement per se.
Kant clarifies his position further in Section 36 insofar as he says
that in a judgement of taste the pure power of judgement is ‘subjec-
tively, object to itself as well as law to itself’.90 We can understand this
claim in the following way. In judgements of taste we have a rare
opportunity to reflect on judgement itself, or more strictly, on the sub-
jective conditions of its operation, just because the cooperation of the
faculties becomes visible. If this is right, then it is not puzzling that it
is at this point that Kant claims that judgements of taste are part of
the general problem of transcendental philosophy. However, his argu-
ment is much more modest than it need be. He merely claims that as
aesthetic judgements are synthetic then they belong to the general
project of establishing the possibility of synthetic judgements a
195
priori.91 He could have gone much further and argued that aesthetic
judgements do not just belong to the class of synthetic a priori judge-
ments, but that they also have a special status among the latter insofar
as they offer a reflection on the very activity of judgement itself.92 This
is surely where his account of the relation between judgements of taste
and the activity of judgement as a power leads. Judgements of taste
would thus qualify as transcendental judgements par excellence.
Section 37 appears to specify further what is to be deduced. In
Section 21 the principle of common sense was identified with a feeling
of a certain proportion among the faculties. Kant now says that it is
not the pleasure, but the universal validity of this pleasure that counts
as a universal rule for the power of judgement.93 This sounds as
though Kant has withdrawn the claim that the principle of taste is a
feeling, in favour of a rule prior to feeling. Yet, elsewhere he consis-
tently characterises aesthetic judgements as based on a feeling: only
this marks them out as aesthetic. In his discussion Kant shifts between
two levels of talking about liking.94 At the empirical level we experi-
ence a ‘liking’ (Wohlgefallen) for a particular object. This liking is
based on a feeling of pleasure (Gefühl der Lust) in the free play of the
faculties.
It is tempting to conclude that when Kant says that not pleasure,
but the universal validity of pleasure, is a universal rule for the power
of judgement, he is excluding my liking for a particular empirical
object and not the feeling of reflective activity or proportion among
my faculties that is the ground of that liking. The feeling of a propor-
tion among my faculties, in contrast to my particular liking for some-
thing, could then still count as a rule for aesthetic judgement. But the
problem with this solution is that Kant consistently distinguishes
between Wohlgefallen, which arises at the empirical level and Lust,
which arises only from the reflective cooperation of the faculties.95
I can only conclude that while it is indeed the feeling, and not some
prior rule, that is universal, the former is universal only insofar as it
operates as a rule. Kant’s intention in Section 37 is to stress the pecu-
liar status of the feeling that, as common sense, is the ground of the
possibility of judgements of taste. Feeling operates as a rule because
it expresses the subjective conditions of judgement and not because it
expresses a preference for a particular object.
So, in Section 36 when Kant mentions both ‘a feeling of pleasure
[Lust] (or displeasure) and a liking [Wohlgefallen] that accompanies
the objects’ presentation’, he must mean that I like this object, but I
do not expect others to like it in exactly the same way as I do. 96 My
196
liking is unavoidably empirical and thus is, in part, determined by my

own empirical history and projects. What I expect of others is that the
object I appreciate will encourage a freedom of mind that I discover
in myself as a necessary concomitant of my empirical liking, insofar
as the latter qualifies as aesthetic. It is this feeling of aesthetic freedom
that I expect others to share. Thus what is to be deduced in Section
38 remains the principle of common sense, which is a feeling of a
reflective proportion in response to a given empirical object.
The ‘Deduction’ proper offered in Section 38 pulls together the
strands of the account developed by Kant in the previous three sec-
tions. In a judgement of taste, the liking we have for an object arises
from ‘our mere judging of the form of the object’. 97 As I will argue in
the final chapter, this must be the spatio-temporal form of the object,
which in some – but only some – cases, gives rise to aesthetic plea-
sure.98 The liking (Wohlgefallen) for the object is for its subjective
purposiveness with the power of judgement.99 This is the ‘mere
judging prior to any concept’ identified as the power of judgement in
Section 35.100 Although Kant does not as yet explicitly say so in
Section 38, we already know from Section 35 that the judgement of
taste rests on the subjective conditions of judgement, which are, as a
cooperation of the faculties, the power of judgement as such. And
now we see that these subjective conditions are not wholly detached
from the object that would, in the cognitive case, be determined by
them. This will allow Kant to establish the grounds for a coherent
account of the relation in which taste stands to cognition, while main-
taining the distinctiveness of their respective status. Whereas in a cog-
nitive judgement the understanding would determine the form of an
object under a concept, in an aesthetic judgement the power of judge-
ment operates in response to that same form, without determining it.
Kant’s next step explicitly reintroduces the idea that the power of
judgement is directed only to the subjective conditions of any employ-
ment of judgement. On the basis of this characteristic he concludes
that taste is based on a power that can be presupposed of all people.
Were a judgement of taste determined by a particular content of sense
or by a concept of the understanding, it would not be based on a
general condition of all judging subjects. This subjective condition is
required for ‘possible cognition as such’.101 Thus, as we already dis-
covered in Section 35, the exercise of the power of judgement is the
cooperation of the faculties that is necessary for all judgements. In the
judgement of taste, however, this cooperation is experienced as avail-
able for our reflection and counts as a free play or harmony.
197
Kant concludes that the harmony between the presentation or the

form of the object and the subjective conditions of cognition is valid
for everyone.102 Judgements of taste are proven to bear the subjective
universality first claimed for them in Section 8 insofar as they are
based on the subjective coordination of the faculties that is necessary
for any cognition whatsoever.103 Kant goes on to say that this is equiv-
alent to the claim that when we judge an object of sense and are aware
through a feeling of pleasure of its subjective purposiveness for our
judgement, we are justified in requiring that the pleasure we feel is
shared by all.104 The pleasure that has been deduced as universally
valid is that taken in the harmony between the form of the object and
the activity of our subjective faculties, not our empirical liking for a
particular object. Strictly speaking I can only call on others to display
the form of aesthetic judgements, that is, the ability to judge on the
basis of a free play of the faculties.105 Thus the pleasure arising
from the exercise of the power of judgement in a judgement of taste
is proven to be universally valid insofar as it rests on the subjective
conditions of ‘cognition in general’.
My main disagreement with Allison concerns the status of the prin-
ciple of taste deduced in Section 38.106 Allison suggests that taste is
grounded on a principle that is ultimately not aesthetic, that is, the
principle of the subjective activity of judgement, whereas I hold that
the aesthetic principle of judgement is the only pure expression of the
subjective activity of the faculty of judgement.
Allison’s point is made especially clear at an earlier point when he
is concerned with the relationship between the systematicity of empir-
ical nature and aesthetic judgement:
the true relationship between formal or logical purposiveness and taste is
not that the former is itself the principle of the latter; it is rather that the
principle licensing the former (the conditions of a reflective use of judg-
ment) is identical to the principle underlying the latter.107
The principle of judgement is the ground both for the logical purpo-
siveness of nature and for taste. Allison is concerned to avoid a double
danger in Kant’s account. On some occasions it sounds as if an aes-
thetic principle of taste is dependent on empirical systematicity.108 On
others it looks as if taste provides a principle for cognition.109
Allison’s alternative epistemic reading of Section 21 was intended to
rebut the second horn of this dilemma, that is, of making cognition
in general dependent on taste. Allison’s solution is to argue that in the
deduction proper, Kant’s position is that taste is based on a principle
198
that expresses the subjective condition of judgement. This principle is

the ground for taste, but it would appear the principle itself is not aes-
thetic. This is Allison’s strategy for showing how taste can be based
on the subjective activity of judgement and thus displays the subjec-
tive conditions of cognition, while avoiding the conclusion that Kant
suggests that taste is the foundation for cognition.
My aesthetic reading starts from the fact that Kant at no stage
commits himself to there being a principle of judgement distinct from
that of taste. I agree that the principle of taste displays the subjective
conditions of judgement or ‘cognition in general’. However, I read Kant
as holding that taste rests on the subjective conditions of judgement,
which, only in the aesthetic case are taken in isolation and count as a
principle. There is no further non-aesthetic principle of judgement
underlying taste. The principle of taste simply is the expression of the
activity of judgement, which only operates as a principle in the aesthetic
domain. Elsewhere those same subjective conditions operate as the sub-
jective side of cognition, but never as a principle in their own right.
My strategy of interpretation has a further advantage. I will be able
to make sense of a connection that Allison dismisses out of hand, that
is, Kant’s claims that aesthetic judgement qualifies as an exposition of
the reflective principle of the purposiveness of empirical nature.
Allison’s insistence that taste is based on a non-aesthetic principle
entails that the former cannot make any direct contribution to Kant’s
account of the order of empirical nature.110
The primary difficulty with my account is that it could be thought
to imply that an aesthetic principle is necessary for cognition. If the
‘Deduction’ deduces an aesthetic principle of taste, then Kant has only
established that the latter entails the subjective conditions of cogni-
tion. But he has not explained how aesthetic judgements entail these
conditions in a distinctive way from cognitive judgements. We have
seen that he has prepared for such an account by focusing on the
power of judgement and by introducing, although not developing, the
suggestion that aesthetic judgement allows for a reflection on that
power. But as his argument stands, Kant has not yet established the
full systematic relation between cognition and aesthetic judgement.
My additional argument will be that aesthetic judgements are exem-
plary of the subjective activity of judgement necessary for cognition
in general. Kant gives the clue to this solution of his problem only in
a provisional way in the third Critique.
My solution has an additional benefit in that, insofar as we estab-
lish a direct connection between aesthetic judgement and empirical
199
cognition, we can restore the materiality of the aesthetic object. Allison

is inclined to insist on aesthetic judgement’s reliance on the subjective
conditions of cognition to the exclusion of an adequate account of the
aesthetic object. Allison insists that the judgement of taste is concerned
with the ‘object qua represented’.111 This is surely right, for what comes
to the fore in aesthetic apprehension is the presentation of the object.
However, the negative part of Allison’s comment risks reinforcing a
prevailing and rather simplistic view that if a judgement of taste is not
objective, then it has nothing to do with the object. He says that the
aesthetic judgement is not directed at ‘the inherent nature of such an
object, not even considered as phenomenon’.112 Our liking for some-
thing beautiful is most definitely not directed at a thing in itself, nor
indeed at qualities that are viewed as inherent, in the sense that they
are taken apart from the relation to the subject. However to say that
taste is not concerned with the object as phenomenon surely suggests
that Allison agrees with those interpreters who insist that Kant brought
about a subjective turn in his aesthetics.
Moreover, the contrast Allison proposes places him in danger of
obfuscating the relation in which representations stand to objects. An
object is always a represented object and a representation is always
intentionally directed to an object. This is true in Kant’s epistemol-
ogy, and in his aesthetics. To be concerned with the form or repre-
sentation of the object is not to be unconcerned with the object, but
rather to turn our attention to certain primordial features of its
objecthood. I have argued that in an aesthetic judgement it is the rela-
tion between subject and object that is potentially brought to our
attention. We saw in Chapter 2 (pp. 64–8) that Allison failed to give
a sufficiently direct and robust account of the relation in which rep-
resented objects stand to representations and that this leads to an
underestimation of the role of affect in knowledge. Allison’s account
of aesthetic judgement is similarly biased towards the subjective side
of experience, although he concedes a role for objects, just as he did
for matter in the epistemic case. The problem, as in the epistemic case,
is that he does not sufficiently examine the relation between subject
and object that is brought out in Kant’s examination of aesthetic
judgement. In the final chapter, I will argue that taste allows for a
reflection on the appearing of the appearance.
Perhaps suspecting that, even at this late stage, he has not found the
best way of establishing the universal validity of taste, Kant recasts the
main argument in a footnote at the end of Section 38, saying that two
conditions are necessary if we are to be justified in claiming universal
200
assent for an aesthetic judgement. 113 First, in aesthetic judgement the

faculties of all judging subjects must stand in the ‘same relation’ as
they do in cognition. This, he says, must be the case because other-
wise people could communicate neither about their presentations
(Vorstellungen) nor about cognition. Second, only this relation – and
hence ‘the formal condition of the power of judgement’ – must deter-
mine aesthetic judgements. While mistakes could be made as to the
formality of our judgement, this is a problem only of application.114
This restatement confirms that the relation of the faculties is the
formal condition of all judgement, necessary for any cognition or
judgement whatsoever. It still, however, does not resolve the problem
of how aesthetic judgement rests on the relation necessary for ‘cogni-
tion in general’ without counting as a variety of the latter.
Conclusion
In this chapter I have shown how the cooperation of the faculties,
identified as synthesis in process in the previous chapter, counts as the
subjective side of the deduction. In particular, I have argued that in
Section 21 of the third Critique, Kant seeks to establish that the syn-
thetic process displayed in aesthetic judgement is necessary for the
legitimation of claims to objective knowledge. This attempt fails,
however, due to an insufficiently finely-tuned argument and, not least,
the lack of a distinction between two levels at which sensus commu-
nis operates. The official ‘Deduction of Taste’ is much more success-
ful in establishing the necessary subjective conditions of cognition in
general, but even it fails to show the nature of the relation in which
the latter stand to aesthetic judgement. In Chapter 7 I will argue that
aesthetic judgement is best understood as standing in an exemplary
relation to cognition. I will also show how we can make sense of
Kant’s apparently ill-advised claim that the ‘Analytic of Taste’ offers
an exposition and a deduction of the principle of the purposiveness
of nature. First, however, I turn to examine the objective side of the
deduction in more detail.
Notes
1. See Makreel, IIK, for instance, p. 9 and also pp. 80–1; Ameriks,
‘Review’, p. 229. Kemp Smith suggests the same location for the sub-
jective deduction in Commentary, p. 236. However, he also suggests
that the subjective deduction has a wider relevance, especially in the B
201
edition ‘Deduction’. See pp. 234–45. See also Sarah Gibbons, Kant’s
Theory of Imagination, who argues for the general significance of the
subjective deduction in revealing the subjective conditions of judgement
and the relation in which this stands to aesthetic judgement. See, for
instance, pp. 52, 79 and 83. Patricia Kitcher’s Kant’s Transcendental
Psychology offers a sustained argument for the relevance of what she
calls the ‘dark side’ of the Critique and points to the ambivalence of
Kant’s retreat from the subjective side of the deduction (p. 13).
2. CPR, A xvi.
3. A xvi/xvii.
4. A xvii.
5. B 124–6.
6. Preface to B Edition, B xvi/xvii.
7. Compare B 519, discussed in Chapter 3, pp. 87–9.
8. A 93, B 126.
9. A 92/3, B125.
10. See Chapters 4, pp. 132–47, and 6, pp. 213–17, respectively.
11. A xvii.
12. A xvii.
13. A xvii.
14. CJ, AA 215.
15. Indeed Kant says that they have nothing to do with the object at all.
This is, however, much too strong a statement of his position. On
p. 177 I discuss Kant’s revision of his statement.
16. CJ, AA 217.
17. AA 217.
18. AA 217–18.
19. AA 218.
20. See the related discussion in Chapter 4, pp. 154–5 where I noted that
in the Introduction Kant avoided the suggestion that cognition is based
on an aesthetic condition. As Pluhar’s translation brings out, the
mutual relation of the faculties is necessary for all empirical cognition,
but distinct from the harmony of the faculties characteristic of aes-
thetic judgements.
21. Later I argue that this counts as a ‘contrapuntal exemplarity’. See
Chapter 8, pp. 296–9.
22. CJ, AA 179. See Chapter 1, p. 15.
23. CJ, AA 215.
24. AA 236 and 238. In Section 21, Kant intends to give a ‘proof’ for the
validity of this principle. If this is right, it complicates the picture sug-
gested by Allison when he says that the analysis of the four moments
of taste contribute only to the question of quid facti. See pp. 193–4.
25. AA 238 (Pluhar’s translation).
26. AA 295.
202
27. AA 238.
28. See pp. 189–93.
29. Allison and Guyer also divide the argument into seven steps but carve
it up slightly differently from each other and from the reconstruction
offered here. See Guyer, Kant and the Claims of Taste (= KCT)
pp. 279–97; and Allison, Kant’s Theory of Taste (= KTT), pp. 150–1.
30. CJ, AA 238.
31. Guyer, KCT, pp. 288–94.
32. Ibid., pp. 288–9.
33. Kant’s real error is that his commitment to the community of judging
subjects is too strong. He lacks any serious account of the interaction
and dissonance among subjects. I therefore agree with Gardner’s
comment that Kant is not committed to inter-subjectivity in a strong
sense where empirical reality would be constituted by inter-subjective
interaction. Gardner, Kant and the Critique of Pure Reason, p. 280.
However, inter-subjectivity becomes less of an automatic presupposi-
tion and more of a task in the account of aesthetic judgement where
we ‘require’ agreement from others. AA 214.
34. CJ, AA 238.
35. Guyer, KCT, pp. 284–5.
36. Ibid., pp. 296–7. His conclusion arises from his analysis of Kant’s
account of ‘proportion’ to which we will return.
37. CJ, AA 191.
38. AA 238.
39. See Chapter 3, pp. 90–1, and also the discussion of the ‘Transition to
the Deduction’ in the previous section of this chapter, pp. 171–2.
40. See previous chapter, pp. 122–7.
41. CPR, A 99.
42. CJ, AA 238.
43. See Allison, KTT, p. 150.
44. These two sub-stages should properly count as two distinct steps, but
I retain the existing numeration for reasons of comparison.
45. See Chapter 8, pp. 284–90, for a discussion of the charge that Kant’s
aesthetics results in the view that all objects are beautiful.
46. CJ, AA 238–9; The German for ‘mutual relation’ here is einer durch
die andere. Pluhar uses the same English phrase at AA 191 to translate
mit dem Verhältnis der Erkenntnisvermogen unter sich. This stage of
my reconstruction correlates with Allison’s Step 5 and is the crucial one
for his and any other reconstruction of Section 21, as he recognises.
See discussion in next section, pp. 189–93.
47. Guyer, KCT, pp. 294–7.
48. Ibid., p. 297.
49. Ibid., p. 295.
50. CJ, AA 239.
203
51. This is in contrast to Allison who treats the two sentences, as stated
here, as two steps. Allison, KTT, p. 51.
52. Or, perhaps, the form of a feeling. Any particular aesthetic pleasure
shares the transcendental form of a play of the faculties. Its content
would be specific to the particular experience, which may be expressed
as a ‘liking’. See discussion below, pp. 196–7.
53. CJ, AA 238; cited above, p. 178.
54. See discussion of the ‘Deduction of Taste’, and in particular of Section
35, below, pp. 194–5.
55. CJ, AA 212′, where Kant says that determining judgement ‘requires no
special principle by which to reflect’.
56. CJ, AA 239.
57. As examples of the usual reading, he refers principally to Guyer’s KCT
and Anthony Savile’s Aesthetic Reconstructions, both of which are
critical of Kant’s attempt to link aesthetics to cognition via sensus com-
munis. Allison, KTT, pp. 145 and 153. However, on p. 145 he also
refers to Ameriks, ‘How to Save Kant’s Deduction of Taste’,
pp. 295–302.
58. In Allison’s reconstruction, this is Step 4. See Allison, KTT, pp. 151
and 152.
59. Ibid., p. 151.
60. Ibid., p. 152.
61. Ibid., p. 152.
62. Ibid., pp. 152–3. Allison is referring to CPR, A 21, B 35–6.
63. Allison, KTT, p. 153.
64. Ibid., p. 153; Allison’s emphasis.
65. CPR, A 21.
66. See discussion of his interpretation of the ‘Deduction of Taste’ in the
next section, pp. 198–9.
68. Ibid., p. 154. Allison refers to CPR, A 133, B 172.
70. Ibid., p. 155.
71. Ibid., p. 155.
72. The most sustained account of a link between feeling and judgement
is offered by Jean-François Lyotard in his Lessons on the Analytic of
the Sublime. See, for instance, p. 9 (French edition p. 22): ‘In the ana-
lytic of taste, sensation no longer has any cognitive finality; it no longer
gives any information about an object but only about the “subject”
itself.’ Sensation translates la sensation, which in French could have
the connotation either of sensory impression or of feeling. Thus ‘sen-
sation’ here can stand for Gefühl in German. See also Makreel, IIK,
pp. 103–7, for a discussion of ‘a feeling of the life of the mind’. See
G. Zoeller, ‘Makkreel on Imagination and Interpretation in Kant’,
204
pp. 266–75; see p. 272. See also Makkreel, ‘Response to Guenter

Zoeller’, pp. 276–80, p. 279.
73. See also an earlier discussion of the relation between judgement and
feeling as the faculty for discrimination in Allison, KTT, p. 70.
Allison’s references are to CJ, AA 177–8 and AA 207′–8′.
74. See A. Bauemler, Das Irrationalitätsproblem. See particularly
pp. 141–66 on the relation between imagination and wit, our capac-
ity for connection.
75. CJ, AA 295.
76. This is linked to the general level of purposiveness discussed in Chapter
7, pp. 255–7.
77. Sometimes, however, we glimpse the activity of reflection at the limits
of our thought, especially whenever we are in a particularly reflective
mood. This may count as a feeling of the measure or balance of think-
ing. Aesthetic judgement gives us a less uncanny, more reliable,
although also more indirect access to the same phenomenon. See
Chapter 8, pp. 297–8, for a discussion of a reflective feeling of the
process of judgement.
78. CPR, A 84, B 116.
80. Ibid., p. 144. See also p. 157.
81. Guyer, KCT, pp. 161–4.
83. CJ, AA 286.
84. AA 287.
85. AA 287.
86. He was previously guilty of this conflation in Sections 9 and 21.
87. AA 287.
88. AA 287. In Chapter 7, I will argue that aesthetic judgements display a
general level of purposiveness between mind and nature that expresses
the possibility of synthesis. I believe that Kant’s comments in Section
35 can be made sense of in terms of ‘general purposiveness’.
89. AA 287. He also reintroduces the notion of free lawfulness from the
General Note at this point. See AA 240–1.
90. AA 288. Compare discussion of free lawfulness in Chapter 4,
pp. 155–6.
91. AA 289.
92. I claimed this in Chapter 4, p. 154.
93. AA 289.
94. Kant uses the term Wohlgefallen in a broad sense in Section 5, AA 209
ff., to encompass three sorts of liking: appetitive, moral and aesthetic.
He then narrows down his usage, reserving the same term as ‘free
liking’ strictly for aesthetic judgement alone. AA 210.
95. See AA 288 in Section 36.
205
96. AA 288.
97. AA 289.
98. Contrast Allison who denies that the form in question is spatio-
temporal. Allison, KTT, p. 175.
99. My account of Section 38 so far relates to the first sentence of both
Pluhar’s translation and of the German text.
100. CJ, AA 287.
101. AA 290. The reconstruction offered in this paragraph corresponds to
the second sentence of Pluhar’s translation and stops at the colon in
the second sentence of the original German version.
102. See Chapter 4, p. 155, on the two harmonies characteristic of aesthetic
judgement.
103. This is an account of the third sentence of Section 38 and of the con-
clusion of the second sentence in German.
104. This is an account of the final sentence of Section 38, both in Pluhar’s
translation and in the original.
105. Allison argues that the ‘Deduction’ concerns only a ‘pure’ judgement
of taste and not actual judgements. Allison, KTT, pp. 177–9.
106. For Allison’s reconstruction of the ‘Deduction’, see ibid., pp. 175–6.
107. Ibid., p. 64.
108. This is the case in the Introductions to the third Critique. See Chapter
7, p. 248.
109. See discussion of Section 21 above pp. 186–8.
110. See Chapter 7, pp. 260–8.
112. Ibid., p. 119.
113. CJ, AA 290.
114. Kant shows himself remarkably sanguine on this point also at AA 216.
206
A Priori Knowledge as the Anticipation of

a Material Given and the Need for
a Spatial Schematism
In this chapter I return to the objective side of Kant’s epistemological

project. In Chapter 4 I discussed the ‘Transcendental Deduction’,
which, it is often thought, provides the whole of the objective deduc-
tion. My aim is to show that the legitimation of the categories requires
not only the whole of the ‘Analytic’, but also recognition of the aes-
thetic dimension of his epistemological argument. It is unavoidable
that in our investigation of the relation between Kant’s accounts of cog-
nitive and aesthetic judgement, we are forced to proceed in a zigzag.
The reciprocal relation between cognition and aesthetic I am in the
course of defending simply cannot be traced out in a linear fashion.
As many commentators have recognised, Kant’s hopes for the con-
clusiveness of the ‘Transcendental Deduction’ turned out to be rather
premature. In both the A Preface and in the ‘Transition to the
Deduction’ Kant claimed that he would establish the objective valid-
ity of the categories insofar as they relate to objects of experience a
priori and necessarily.1 Meanwhile, in the B edition ‘Deduction’, Kant
claims that the categories apply to all perception, to the possibility of
experience and therefore to all objects of experience.2 In later sections
of the ‘Analytic’, he reveals that further articulation of the categories
is necessary if they are to qualify as the form of experience for all
empirical objects.
My account of the structure of Kant’s extended legitimation of the
categories owes much to Buchdahl’s reading in which every element
of the ‘Analytic’ has a part to play.3 However, my reading brings out
elements essential to the relation between Kant’s epistemology and
aesthetics, not to be found in Buchdahl or in the accounts given by
others who also resist the temptation to assume that the full deduc-
tion of the categories is supplied in the ‘Transcendental Deduction’.4
My aim in this chapter is to give a clear and systematic view of the
207
hierarchical stages through which Kant advances his argument in the

later stages of the ‘Analytic’ in the light of my insistence on the affec-
tive dimension of Kant’s epistemology. I will discuss how the prin-
ciples of the understanding are not merely an application of the
categories, but rather a further and aesthetically charged articulation
that is necessary for the legitimation of the latter. At the same time, I
will establish that within the ‘Principles’ chapter there is a develop-
ment from formal conditions toward a material given. Moreover, I
will argue that the systematicity of the principles is a necessary con-
dition of their legitimacy.5 Drawing on the account of ‘synthesis in
process’ I established in Chapter 4, I will show how the arguments of
the ‘Principles’ count as analyses of figurative synthesis and argue that
Kant’s account reveals that the subjective side of the deduction is a
necessary condition of objective determination. Most importantly,
throughout this chapter a priori knowledge emerges as a project of
determination, that is, as an anticipation of application to a material
given. In particular, and, relatively late in the development of his
thought, Kant comes to realise the centrality of space for establishing
the distinction between formal and material idealism. I argue that it
makes sense to speak here of a ‘spatial schematism’.
In the first section, I assess the progress achieved so far in provid-
ing a deduction of the categories in the ‘Transcendental Deduction’,
before turning to examine the extent to which the ‘Schematism’
chapter offers new input. It is more difficult to ascertain the role of
the principles, which are often taken to offer merely an application of
the categories. In the third section, I argue that the ‘Principles’ chapter
offers a further aesthetic development of the pure concepts of under-
standing. They extend the transcendental account in order to estab-
lish that the categories apply to what Kant now refers to as actual or
possible experience.6 It would thus appear that he believes this was
not achieved either in the ‘Deduction’ or in the ‘Schematism’. In the
fourth section, I argue that the development within the system of
principles from those that are merely a priori to those that are regu-
lative counts as a further articulation of the structure of the applica-
bility of the categories to existing objects. The significance of this
development turns out to be that objects are now considered as entail-
ing a material given in space and time. However, even now we will
find that Kant further hones his account of the categories in order to
show that they qualify as the form of experience and so bear objec-
tive validity. The ‘Transcendental Deduction’ was thus far from being
the end of the story.
208
A Priori Knowledge, Material Given and Spatial Schematism
In the fifth section, I examine the possibility that Kant may have
offered the beginnings of a spatial schematism, complementary to the
official temporal schematism in the ‘Principles’ chapter. This is a
development of Kant’s thesis in the ‘Metaphysical Exposition’ of
space in the ‘Transcendental Aesthetic’.
In the final section, I examine Kant’s final attempt to establish the
objective validity of the categories in the ‘Analytic’. He introduces a
new articulation of causal necessity and three associated ‘proposi-
tions’. I suggest that this late addition is evidence for a continuing
worry that he has not yet proven what he claimed much earlier in the
‘Transcendental Deduction’, in particular, because the given is only
now considered as material. In the next chapter we will see that the
same concern gives rise in the Critique of Judgement not only to a
further principle, but to a distinctive reflective level of analysis of
experience.
I The task of the ‘Transcendental Deduction’

In this section I build on the hierarchical reading established by
Buchdahl, emphasising the issue arising from the ‘Deduction’ as to
whether or not the transcendental form of knowledge is sufficient for
establishing the possibility of empirical experience. In particular, I
uncover a problem in Kant’s account as to whether the ‘Deduction’
already contains the conditions of possibility for singular empirical
objects. The discussion here sets the scene for my claim that a priori
knowledge is anticipatory of a material given.
In the ‘Transition to the Deduction’ Kant states that the aim of the
‘Deduction’ is to show that not only the forms of intuition, but also
the categories of the understanding count as a priori conditions of the
possibility of experience.7 More precisely, he says that pure concepts
are the conditions under which anything can be thought as an ‘object
in general’.8 As it stands, this claim would establish only intellectual
synthesis and set a minimal goal for the ‘Deduction’. But a few lines
later, Kant claims: ‘through them [the categories] alone does experi-
ence become possible’.9 The ‘Deduction’ aims to prove not only that
the categories are necessary for thinking objects, but also for experi-
encing them. Even this stronger claim does not entail that the cate-
gories generate experience and only that they count as its necessary
conditions insofar as they provide its reflective form. Kant says that
it is only insofar as categories supply the form of thinking experience
that they make the latter possible.10
209
Kant has already made clear that the categories must be combined
with an aesthetic component, namely the forms of intuition.11 So we
know that although the categories must provide not only the form of
thinking, but also the form of thinking the experience of objects, the
form of actual experience cannot be established entirely at the level
of thought. But the question is: can the possibility of actual experi-
ence be established at the transcendental level of analysis? And even
if this is possible, is the apparatus of the ‘Deduction’ sufficient for
doing so? Significantly, at the outset of the A ‘Deduction’ Kant says
that if an a priori concept did not relate to experience, it would
qualify only as a logical form and not as a concept ‘through which
something is thought’.12 Is the experience in question empirical expe-
rience and, if so, in what sense? It is clear that, for Kant, experience
is empirical and for this reason it may be thought that the question is
meaningless. But the issue is whether or not Kant’s account of experi-
ence establishes its empirical status from the outset. While Kant was
always committed to the empirical status of experience, his philo-
sophical reconstruction of it began at a very abstract or general level.
He only gradually worked towards providing a legitimation of
experience qua empirical. The evidence for this is that it is only fairly
late in his account that the givenness distinctive of the empirical
comes to play a central role in his account of the applicability of
the categories. Undoubtedly, empirical givenness is presupposed from
the outset of his account, as I argued in Chapter 3, pp. 100–5, but in
the initial accounts of the categories this aesthetic dimension of his
epistemology is underplayed in favour of an emphasis on the purely
formal elements of knowledge. For this reason it is necessary to stress
that the goal of Kant’s account is to legitimate the application of the
categories within empirical experience.
Kant immediately goes on to say that although the categories, as a
priori, have no empirical content, they must relate to experience.13 The
contrast between form and content in this statement already strongly
suggests that the experience in question is empirical. Kant concludes
that the categories only have ‘objective reality’ if they are the condi-
tions of a possible experience.14 So although Kant abstracts from every-
thing empirical in the a priori analysis of the categories, the deduction
will have to show how empirical objects, and not just a mere object in
general, finally fall under the conditions of the categories. Or he will at
least have to show that an object in general expresses the form of
empirical objects. It is not clear at this stage that Kant recognises how
much more articulation of his theory this may require.
210
It certainly appears on occasion that Kant believes that the

‘Transcendental Deduction’ supplies the full conditions of conceptual
determination for all empirical objects.15 When he says that cate-
gories ‘prescribe laws a priori to appearances, and therefore to
nature’, it sounds as if intuitions are not only universally but also
wholly determined by the understanding.16 But Kant suggests this is
not the case towards the end of the ‘Transcendental Deduction’, when
he says: ‘Particular laws, because they concern empirically deter-
mined appearances, cannot be wholly derived [from the categories],
although they nevertheless all stand under them.’17 In saying this,
Kant concedes that there is a gap between transcendental structure
and empirical experience, although he does not yet recognise that
there may be a problem in bridging that gap.18 The categories do not
complete the task of determination at the empirical level, for there
must be empirical determination in addition to the a priori determin-
ation arising from the categories.
This limitation on the ‘Deduction’ suggests two possibilities. First,
it could be the case that the full formal structure of any singular
empirical object is provided at the a priori level, leaving only the spe-
cific relations between objects to be determined at an empirical level.
Second, it could be argued that the full determination of any singular
empirical object must take account of the special laws that govern the
relations in which it stands to other objects. This view would draw
on the fact that an object of experience is distinguished by its stand-
ing in relation to other objects, as established by its falling under the
category of relation and by the argument of the ‘Analogies of
Experience’.19
While Kant does not even raise this question, he now announces
that the full a priori specification of experience is not yet complete:
How they [the categories] make experience possible, and what are the
principles of the possibility of experience that they supply in their appli-
cation to appearances, will be shown more fully in the following chapter
on the transcendental employment of the faculty of judgment.20
The contents of both the ‘Schematism’ and of the ‘Principles’ are part
of the account of how experience is possible: they thus still operate at
the transcendental level. All that the ‘Transcendental Deduction’ has
done is lay out the structure of experience at the pure a priori level. This
is what Kant means when he concludes the B ‘Deduction’ by saying that
he has established the principles of the possibility of experience insofar
as they determine appearances in space and time in general.21
211
If the categories do not fully determine empirical experience, then

it is at least in principle possible that we could have relatively indeter-
minate experiences. Intuitions alone do not qualify as experiences and
anything of which we are aware stands in some relationship to the cat-
egories and thus bears a minimum determination. But not all intu-
itions are fully determined by the categories, which underdetermine
empirical experience.22 This leaves open the possibility that there
could be a gap between the transcendental order supplied by mind and
the world of given objects. We can draw three conclusions from this.
First, the mind does not simply impose order on the world and instead
has to take further considerations into account in its empirical oper-
ation. Second, a gap is left for underdetermined intuitions, in particu-
lar those that are the starting point of aesthetic judgements. Third, the
picture that emerges is one of knowledge as not simply a fait accom-
pli and entailing a project. We have to go out into the world and
examine particular empirical affects if we are to achieve empirical
knowledge. We have a safety net insofar as we can rely on the a priori
framework of the categories and of space and time; however, these
formal conditions will not supply us with the full empirical determi-
nation of objects. Indeed, as I have already suggested, they do not even
supply us with the full transcendental determination of empirical
experience. Knowledge emerges out of an exploration of the world
rather than as a projection of mental order onto a featureless world.
I have shown how Kant concedes that the a priori categories
cannot supply the full account of the determination of empirical
knowledge, which also requires an empirical determination. I have
also shown that he promises that the ‘Schematism’ and ‘Principles’
will extend the transcendental story insofar as they give a greater
elaboration of how the categories make experience possible. A full
account of the way in which the form of experience initiated by the
mind is finally applicable within empirical experience will entail an
examination of these later parts of the ‘Analytic’. If the aim of the
‘Transcendental Deduction’ is to supply the proof that the categories
apply to all empirical intuitions, then it already looks as if, at the end
of the B ‘Deduction’, Kant realises that there is still work to be done.
In the remaining sections of this chapter, I will reconstruct the further
stages of the transcendental determination of empirical objects
through an analysis of the ‘Schematism’ and ‘Principles’, while in the
next chapter I will examine the conclusion of his account of the deter-
mination of empirical experience in the two Introductions to the
Critique of Judgement.
212
II The third thing: a depth analysis of figurative synthesis

In the reading of the ‘Schematism’ that follows, I bring out the rela-
tional status of the schema, allowing Kant to address how figurative
synthesis, already adumbrated in the ‘Deduction’, is first possible. I
suggest that the best way of grasping the possibility of figuration is as
‘synthesis in process’, first introduced in Chapter 4, and that this
recognition helps explain the ambiguous status of the schema.
In the ‘Analytic of Principles’, which comprises both the
‘Schematism’ and the ‘System of all Principles of Pure Understanding’,
Kant begins by saying that he will now provide a ‘canon’ for judge-
ment, that is, instructions for the application of the categories to
appearances.23 It sounds as if the schemata and principles will merely
serve as an instrumental implementation of a proof that is already
complete. However, examination of the significance of a canon for
Kant’s method reveals a more complex story.
Kant uses the term ‘canon’ on various occasions to express a limi-
tation on the critical project, and in particular its restriction to the
empirical domain.24 It is only, however, late in the Critique that he
supplies a definition: ‘I understand by a canon the sum-total of the a
priori principles of the correct employment of certain faculties of
knowledge.’25 If the ‘Analytic of Principles’, supplies a canon, then
this could mean it is simply a working through of the rules already
established in the ‘Transcendental Deduction’. But this does not seem
to be Kant’s intent, for he goes on to say that the ‘Transcendental
Analytic’ qualifies as the canon of the pure understanding.26 This sug-
gests that the ‘Transcendental Deduction’ alone does not supply the
principles of experience, although the ‘Analytic’ as a whole does do
so. And if the categories of the understanding are not established as
the principles of experience in the ‘Deduction’, then it would appear
that there is still work for the ‘Analytic of Principles’ in showing that
the categories apply to all possible experience.
If Kant were to have claimed that the categories simply are the
principles of experience, then he would be in danger of falling into an
impositionalist position, where the form of thought coincides with
the form of objects. But this is not Kant’s position. At the beginning
of the ‘Analytic of Principles’, he says that concepts would be mere
logical forms and not pure concepts of the understanding if they
did not ‘formulate by means of universal but sufficient marks
[Kennzeichen] the conditions under which objects can be given in
harmony with these concepts’.27 The distinction between logical
213
forms and categories strongly echoes the initial paragraphs of the A

‘Deduction’ and it would appear that, in Kant’s view, his initial goal
is not yet achieved.28 In the account of figurative synthesis in the B
‘Deduction’, Kant argued that the categories apply to objects only
insofar as the imagination mediates between concept and intuition.
Now he announces that a further stage is necessary. What are these
marks that will establish the conditions of applicability of the cate-
gories? Kant answers this question by immediately promising that in
his next chapter – that is, in the ‘Schematism’ – he will discuss the sen-
sible condition under which the categories can be employed.29 We
must conclude that the marks that are necessary to distinguish a mere
logical form from a principle of the understanding are the schemata.
At the outset of the ‘Schematism’ chapter, Kant introduces the
rather odd idea that the representation of the object must be homo-
geneous with its concept. He then presents as equivalent the reverse,
namely that ‘the concept must contain something which is repre-
sented in the object that is to be subsumed under it’.30 While the first
formulation makes it sound as if the object must conform to the
concept, the second suggests that the concept must conform to the
object. As it stands this could lead us to conclude that Kant does not
see that the two claims are contraries; neither does he appear to be
aware of the tension in which the second formulation stands with the
Copernican revolution. However, I believe that the combination of
these two positions arises from his attempt to express the relation in
which the concept stands to the object or underdetermined intuition.
The concept is nothing other than the condition for the possibility of
the unification of something, that is, the given in intuition. On the
other hand, the intuition can only be known insofar as it is capable
of being unified under a concept. The process of determination entails
at least three conditions, as we saw in Chapter 3, p. 94. Something
has to be given in intuition, our sensibility has to be capable of taking
up that given as an affect, and our understanding must be capable of
unifying the affect. If any of these three conditions failed, knowledge
would not arise.
But, surely, the need to unify an intuition under a concept was
already satisfactorily dealt with in the ‘Deduction’; at least in the second
part of the B edition, where Kant developed his account of figurative
synthesis? Imagination was the third thing that made possible the neces-
sary unification of intuition and concept and showed how the categories
were not only a form of thought, but also the form of experience. So is
something genuinely new added in the ‘Schematism’ chapter?
214
Whereas in the ‘Deduction’ Kant established the necessary role of

the transcendental synthesis of imagination, he now makes the add-
itional claim that the imagination must produce a third thing, or tran-
scendental schema, that makes possible the application of concepts to
intuitions or objects.31 This addition will allow for the resolution of
the problem, which he clearly thinks is not yet answered, namely
‘How, then, is the subsumption of intuitions under pure concepts, the
application, of a category to appearances, possible?’32 Kant remarks
that this question makes a transcendental doctrine of judgement
necessary, that is, the ‘Deduction’ must be supplemented by the
‘Schematism’ if the applicability of the categories is to be established.
But if it is not yet established that the categories apply to all objects,
then the ‘Deduction’ did not yet validate pure concepts as supplying
the form of experience even at an a priori level. Although the
‘Deduction’ revealed that figurative synthesis was necessary, the
‘Schematism’ is required if we are to understand how it is possible.33
The schema counts as a mediating representation insofar as it is
both intellectual and sensible.34 This dual status allows it to establish
the relation between pure concept and appearance, that is, the appli-
cation of a category to an empirical intuition. The schema must also
be pure, that is, it must have no empirical content. The third thing
qualifies as a ‘universal but sufficient mark’ insofar as it is capable of
applying the form of thought to all intuitions.35 As all of this must
occur at the a priori level, the schema makes possible the synthesis of
the pure form of the understanding with the forms of intuition.
Nevertheless, Kant’s aim is to show how a category is capable of syn-
thesising an empirical intuition or appearance, through an operation
on the pure form of intuition. He consistently states that the task of
the ‘Schematism’ is to show how an appearance can be rendered
homogeneous with a pure concept of the understanding.36
The mediating representation in question is temporal. Time is the
form of all intuitions. Yet the determination of time depends on the
unification afforded by the category. The schema can now be identi-
fied as the transcendental determination of time that allows the appli-
cation of the categories to appearances.37 This is the mark that makes
possible the application of categories. We will presently find that the
determination of time, in this context, is not a determinate temporal
intuition, but rather the temporal elaboration of a category so that it
can take up an intuition, which is necessarily temporally formed. In
the ‘Schematism’ the process of time has been introduced into the cat-
egories; a temporal conclusion has not been achieved.
215
Towards the end of the ‘Schematism’, Kant says that ‘[t]he

schemata are the true and sole conditions under which these concepts
obtain relation to objects and so possess significance [Bedeutung]’.38
This confirms the view that the significance of the categories, arising
as it does from their applicability within experience, is not secured in
the ‘Transcendental Deduction’ alone.
Kant sometimes characterises the schema as if it were a mental
object; for instance, when he speaks not only of a third thing, but also
of a monogram or a transcendental product of the imagination.39 At
other times the schema counts as a mental process, namely, the tran-
scendental determination of time, ‘a universal procedure of imagina-
tion in providing an image for a concept’.40 This tension is resolved in
his characterisation of it as ‘a rule of synthesis of the imagination’.41
The schema is the temporal elaboration of a category, so that the latter
can be applied within the field of appearances. The schema allows the
concept to synthesise a given object. So, although the claim that the
schema of a triangle ‘can exist nowhere but in thought’, and that it is
a rule for pure figures in space, may make it sound as though the
‘Schematism’ is not yet concerned with empirical appearances, this
is not the case.42 The schemata are pure operations of the imagination
that introduce figure or design into empirical appearances, thus
rendering them knowable: they are not pure images, as Heidegger
suggests.43
Does this genuinely add something new to the account of figura-
tive synthesis in the ‘Transcendental Deduction’? On the occasions
when Kant characterises the schema as a third thing, it appears that
a new element has been introduced. However, on further analysis the
schema is nothing other than the operation of figurative synthesis.
Kant now emphasises the process through which the categories are
applicable within experience. The process no longer appears to be an
immediate or automatic result of the combination of understanding
and intuition through the mediation of imagination. The imagina-
tion must generate a mark, that is, the concept must be temporally
articulated. It is not so much that in the ‘Schematism’ Kant intro-
duces a new element, as that he carries through his analysis of figu-
rative synthesis at greater depth, revealing that it is a temporal
process through which mental operations are capable of taking up
an object, albeit at the pure a priori level.44 This is synthesis in
process, which, I argued in Chapter 4 (pp. 151–6) is the condition of
any cognitive judgement and yet is only held up for our inspection in
aesthetic judgement.
216
The schema, which is characterised as a transcendental product

of the imagination, is also described as the schematism of the
understanding.45 The problematic status of figurative synthesis that
we already discovered in the official ‘Deduction’, in both its versions,
is still evident. We can now see, however, that schematism is of the
understanding strictly in the sense that a schema operates on the
understanding in order to achieve the temporal determination of
the categories. This clarification does not undermine Kant’s remark
that the schematism also belongs to the imagination, for it is the latter
that makes possible the temporal elaboration of the pure concepts of
understanding. The complex cooperation out of which schematism
arises may have been one of the motivations for Kant’s famous
remark that it is: ‘an art concealed in the depths of the human soul,
whose real modes of activity nature is hardly likely ever to allow us
to discover, and to have open to our gaze’.46 The ‘Schematism’
chapter further emphasises the way in which the faculties must coop-
erate if the subsumption of an object under a category is to be possi-
ble, thus deepening our understanding of how the subjective side of
the deduction contributes to the objective deduction. This is easily
lost sight of, however, because Kant is primarily concerned with
establishing the possibility of the determination of an object.
I have emphasised that the ‘Schematism’ chapter is already con-
cerned with the possibility of the application of the categories to empir-
ical appearances. Kant’s aim is to establish how empirical intuitions fall
under the categories. However, as I will argue in what follows, the full
story of synthesis is not yet complete, neither at the a priori nor at the
empirical levels. The schematism operates as a figurative synthesis at
the pure a priori level between a pure category of the understanding
and the pure form of intuition, time. Kant’s aim is to show how a
schema facilitates the application of a category to an appearance, but
he has not yet got so far. In the ‘Principles’ chapter, a potential spatial
schematism emerges as a complement to the official temporal schema-
tism, while in the Introductions to the Critique of Judgement, Kant
introduces a further principle of reflective judgment, which I will argue
is necessary for the full articulation of synthesis at the empirical level.
III The principles and possible empirical knowledge in

general
It might seem, nevertheless, that the applicability of the categories to
intuitions is now established, especially if we return to the passage
217
immediately preceding the ‘Schematism’ where Kant announced the

need for a universal, but sufficient mark.47 In this passage he claimed
that, in the transcendental doctrine of judgement, he would provide:
first, the sensible condition under which the categories can be
employed; and, second, he would lay out ‘the synthetic judgements
which under these conditions follow a priori from pure concepts of
the understanding’.48 The ‘Schematism’ achieved the first of these
tasks, while it might appear that the ‘Principles’ chapter need only
spell out the temporally determined categories.
Nevertheless, at the outset of the ‘System of all principles of pure
understanding’, Kant announces that the principles require a proof.49
Why should this be so if the principles are simply an application of
the temporalised categories? Although the table of categories is ‘the
natural and safe guide’, it would appear that the argument is not yet
complete.50 We will find that the new development entails laying out
the systematic connection of a priori judgements, a step that is now
revealed as necessary for establishing that the categories apply uni-
versally within experience.
Importantly, for my continuing interest in showing that the objec-
tive deduction has a subjective side, Kant remarks that the required
proof can only be carried out by examining the ‘subjective sources of
the possibility of knowledge of an object in general’.51 This should
not, by now, be surprising, for Kant is engaged on an a priori analy-
sis of the form of experience. This form requires a mental input and
can only be understood as arising from a subjective process.
However, Kant could have made his position more clear; for
although the form of experience arises from subjective activity or
synthesis in process, the latter is always intentionally oriented
towards something given in experience. Kant establishes this at the
outset of the ‘Transcendental Aesthetic’, as I discussed in Chapter 3.
In consequence, each of the stages in the B ‘Deduction’ and the
‘Schematism’ are marked by a repeated insistence on how a (subjec-
tive) category must be shown to be capable of synthesising an
object.52 Kant’s a priori analysis is of a process initiated by the mind,
but always exercised on an extra-mental given. We will see that the
‘Principles’ take this process of establishing a fit between mental form
and the given a step further.
In ‘The Highest Principle of all Synthetic Judgements’, Kant con-
tinues to fine-tune the problem of the application of the categories.
The problem now is: how is an a priori synthetic judgement possible?
An analytic judgement requires only the analysis of a concept in order
218
to draw out an implication that it already contains. A synthetic judge-

ment requires us to go beyond what is merely thought in the concept
and to link the latter with an intuition.53 He says this requires a third
thing, which is the medium for all synthetic judgements. Just as in the
‘Schematism’, the third thing is primarily understood as time, but he
goes on to suggest that the synthesis of representations requires the
operation of all three faculties of intuition, imagination and under-
standing.54 This explains his earlier claim that the proof of the prin-
ciples requires a turn to the subjective sources of knowledge. But how
is an examination of the faculties to help in establishing the applica-
tion of a pure concept to a pure intuition and show how an a priori
synthetic judgement is possible?
Kant now returns to the objective side of his argument:
If knowledge is to have objective reality, that is, to relate to an object, and
is to acquire meaning and significance [Bedeutung und Sinn] in respect to
it, the object must be capable of being in some manner given. Otherwise
the concepts are empty; through them we have indeed thought, but in this
thinking we have really known nothing; we have merely played with rep-
resentations.55
The thought or representation must take up something given to it in

empirical experience, as I discussed in Chapter 3. Even space and time
are ‘without objective validity, senseless and meaningless [ohne Sinn
und Bedeutung]’ if they are not applied to objects of experience.56
This is because they, as the forms of intuition, only make receptivity
possible insofar as something is given within them. Kant’s ongoing
project of establishing that the forms of thought are capable of taking
up the empirical given, now leads him to develop his account of the
aesthetic or sensory side of experience beyond its formal conditions
in order to address what is given in intuition. The ‘Principles’ thus
provide further confirmation of the complex identity of intuition
already established in the B ‘Deduction’.57
If a concept is to have objective reality and significance this entails
not merely the temporal determination of the category, but the further
requirement that the ‘representation through which the object is
thought relates to actual or possible experience’.58 While Kant has had
the latter as his goal since the beginning of the ‘Deduction’, it would
appear that he has not yet arrived at his destination. It is not, of course,
that Kant has abandoned his commitment to the temporality of
experience, but rather that the full determination of empirical experi-
ence requires more than temporal schematism.
219
Experience, for Kant, as we already know, is inextricably linked

with knowledge. He now says that experience is only possible as a
synthetic unity of appearances.59 Knowledge does not arise if the
objects that affect us are not related to one another in a systematic
fashion. In order for this to be possible, there must be principles that
establish the form of experience, that is, its most general order.60
Kant now intends to provide an account of the principles that order
experience at its most general level. The principles are elaborations
of the categories, and thus he will finally be in a position to show that
the categories are not merely forms of thought, but also provide the
form of experience. At the conclusion of the principles, he claims:
‘We then assert that the conditions of the possibility of experience in
general are likewise conditions of the possibility of the objects of
experience, and that for this reason they have objective validity in a
synthetic a priori judgment.’61 Thus, although at the end of the B
‘Deduction’ Kant purported to have established that the categories
are the principles of the possibility of experience, it now transpires
that this result was only achieved at the most general level. 62 While,
supposedly, he was already concerned with the objects of experience,
we now discover that he was operating only at a provisional level.
Kant now intends to go further in his analysis of the possibility of
experience insofar as he takes into consideration not only the form
of experience, but also the given to which the latter necessarily stands
in relation.
We can now return to consider Kant’s claim that the latest stage of
his argument cannot be carried out in an objective fashion. He went
on immediately to explain that this is because the principles are the
foundation for knowledge.63 He must mean that because knowledge
arises out of a subjective activity, it rests on the operation of our sub-
jective powers. But clearly, this does not entail that those subjective
faculties do not relate to objects. Indeed, both in the ‘Schematism’ and
in the ‘Principles’, we have seen that Kant’s aim is to show how con-
cepts can relate to objects, and, latterly, not only to the formal con-
ditions of sensibility, but also to what is given through them. The
subjective conditions are the conditions of our access to objects given
in experience. The explanation of this process must start from the
subject and not from the object, because it is knowledge of objects
that is in question. This, and not impositionalism, is the deeper sig-
nificance of the Copernican turn in philosophy. The goal is to show
how the subjective conditions reach their goal, that is, take up the
object given in experience.
220
Towards the end of the introductory account of the principles in

Section 2, Kant clarifies that the systematic unity must be of empiri-
cal experience. Only empirical synthesis will give reality to a priori
synthesis.64 The categories must be shown to apply to empirical
objects. Kant thus acknowledges that although empirical experience
has always been the goal of his epistemology, he has not yet provided
the full analysis of its possibility. The truth of a priori synthesis arises
only insofar as it provides the form of empirical experience.65 Kant
concludes the section saying that synthetic a priori judgements are
only possible insofar as the subjective formal conditions of experience
relate ‘to a possible empirical knowledge in general’.66 We must con-
clude that the ‘Principles’ chapter will show that the objective reality
of the categories is finally achieved by establishing how they apply at
the empirical level. But the last stage of the argument of the ‘Analytic’,
or the canon of understanding, still operates at the level of the possi-
bility of the objects of experience. Although the ‘Principles’ chapter is
more empirically oriented than earlier sections of Kant’s account, it
still falls short of the specific empirical determinations mentioned in
the ‘Transcendental Deduction’. The new element, in what has
emerged as a hierarchical determination of empirical experience, is
the focus on the givenness of intuition.67
We saw in the ‘Deduction’ proper that Kant assumed that the cat-
egories would supply the general form of experience. Nothing at the
level of experience, it would appear, interferes with the framework
arising from the forms of thought, which are merely supplemented by
special empirical laws. However, in the ‘Schematism’ and ‘Principles’,
Kant concedes that if the categories are to legislate within empirical
experience, the pure concepts must be further articulated, first tem-
porally and second as a system of principles oriented to the possibil-
ity of empirical experience as given.
IV The mathematical and dynamic principles as further

articulations of the a priori in anticipation of experience
If schematism transforms the categories into temporalised concepts,
then the principles establish the latter as rules for the synthesis of
objects given in experience. These rules comprise a system that finally
provides the transcendental system of experience. Only under these
rules is judgement capable of determining appearances and this is pos-
sible only insofar as the principles are exercised as a system, that is,
they jointly make possible the determination of objects of experience
221
at the transcendental level. I will show that a progression toward a

material given can be uncovered not just in, but within the ‘Principles’
chapter.
Kant distinguishes four principles, or sets of principles, ultimately
derived from the ‘Table of Categories’. The first two sets count as
mathematical insofar as they are the unconditionally necessary a
priori conditions of any possible experience. They are concerned with
the ‘mere intuition of an appearance in general’.68 The ‘Axioms’ and
‘Anticipations’ are concerned only with the form of experience.
However we will see that, in comparison with the ‘Schematism’, they
are more closely associated with the given in intuition because they
stand in systematic relation to the ‘Analogies’ and ‘Postulates’.
The ‘Axioms of Intuition’ establish that all intuitions have exten-
sive magnitude, that is, they are necessarily extended in space and
time. This extension becomes knowable when it is subject to a rule of
unification arising from understanding, allowing their appearance in
space to be constructed through an exercise of figurative synthesis.69
Kant is here concerned with how synthetic activity generates the form
of appearance, that is, of something given in intuition.
The ‘Anticipations of Perception’ establish that ‘in all appearances,
the real that is an object of sensation has intensive magnitude, that is,
a degree’.70 Appearances are characterised not only by extension in
space, but also by an affect in sensation. This principle also counts as
mathematical insofar as it is necessary for intuitions in general.
Appearances, Kant goes on to explain, contain not only the form of
intuition, but also ‘the matter for some object in general’ or the ‘real
of sensation’.71 Our representation of this is merely subjective, yet it
makes us conscious that we are affected by an object in general.72
Kant here reiterates the position he first laid out at the beginning of
the ‘Transcendental Aesthetic’, namely, that experience arises from
the affect of an object given in experience. This is the force of the dis-
tinction between his position and material idealism, for which there
would be no material given. Any appearance arises from an affect,
which puts us in relation to a given material object. But the given
object counts only as an object in general because an affect is precisely
what is not yet determined. The affect is merely a sensation, that is,
the material in sensation.73 What that sensation will be cannot be
anticipated at the a priori level. We can only await its advent, for the
matter in experience must be given in experience.74 There is, never-
theless, something that can be anticipated a priori, namely, that all
perceptions will have some degree of reality – that is, there will be a
222
degree of sensation or affect.75 This anticipation is the formal condi-

tion of our empirical apprehension of the qualitative or material
aspects of objects. Kant refers not only to the material given in
general, as we have already seen, but also to those aspects of an object
that he considers empirical and have traditionally been referred to as
secondary qualities.76
Our anticipation supplies a ground for the productive imagina-
tion’s syntheses of intensive magnitudes, which count as flowing
insofar as they are temporal.77 Figurative or productive synthesis
begins from a rule, or principle, sanctioning its anticipation of a mate-
rial reality given in experience. Only the possibility of these syntheses
is known – more accurately, objectively cognised – a priori. The syn-
theses of the material given can only occur in experience. Kant con-
cludes the ‘Anticipations’ with: ‘Everything else has to be left to
experience.’78
In passing, Kant comments on the anticipatory character of knowl-
edge in general: ‘All knowledge [Erkenntnis] by means of which I am
enabled to know and determine a priori what belongs to empirical
knowledge [empirische Erkenntnis] may be entitled an anticipation
. . .’79 We can now reinforce the significance of this statement for
Kant’s formal idealism. A priori knowledge as comprehended by
material idealism would not count as an anticipation simply because
it would entail the a priori construction not only of the principle of
knowledge, but also of its content. The principles Kant is in the course
of developing give merely the form of experience and only have objec-
tive reality and significance insofar as they take up something given
in experience. A priori knowledge anticipates the experience of which
it is a form. This is the added emphasis that Kant introduces in the
‘Principles’ chapter and it supports my earlier claim that the formal
structure introduced by mind initiates rather than imposes order.80 To
say that the principles anticipate experience is to say that their valid-
ity requires their applicability, that is, the possibility of application,
within empirical experience. This is not, however, to say that empir-
ical experience can validate its own formal principles. This would
make no sense from a Kantian position. Application of the principles
does not establish their legitimation, which requires, strictly, the
establishment of their applicability at an empirical level.
Despite signalling the anticipatory status of a priori knowledge
(or cognition) as it is understood from the perspective of formal ide-
alism, Kant distinguishes the peculiarly anticipatory character of the
relation in which we stand to matter given in sensation. Nothing can
223
be determined beyond the possibility that there is some material

affect, bearing some degree of intensity or other. This characterisation
marks a further level of discrimination within the mathematical prin-
ciples, which, as we have seen, count as apodeictic insofar as they
concern only the necessary intuitive form of experience.81 The
‘Axioms of Intuition’ establish that we can know a priori that a
spatial intuition can be constructed as an extensive magnitude. The a
priori analysis of spatial construction gives the formal conditions for
empirical syntheses. However no overt role for empirical experience
is developed in the account of the axioms. In contrast, a priori syn-
theses of intensive magnitudes cannot operate without anticipating
the empirical. The second mathematical principle is not just a princi-
ple or form for experience, it is a form of experience, in the sense that
it belongs to the latter and cannot even be expressed at the formal
level without reference to experience.
The second two sets of principles count as dynamic and are equally
a priori and necessary, but, only, Kant adds ‘under the condition of
empirical thought in some experience, therefore only mediately and
indirectly’.82 Their necessity is thus strictly in anticipation of their
application to empirical experience. Kant finally approaches the goal
of empirical application that has structured the composition of the
‘Analytic’ as a whole. The dynamic principles comprise the ‘Analogies
of Experience’ and the ‘Postulates of Empirical Thought in General’.
I will not attempt to enter into the detail of Kant’s elaborations of
these, focusing only on the way in which they reveal the further deter-
mination of the a priori in the interests of the establishing the latter’s
applicability at the empirical level. I will thus comment only selectively
on the principle of the analogies as a whole, on the second analogy and
on the general structure of Kant’s account of the postulates.
The terms in which the general principle of the analogies is presented
in the B edition reveals a progress beyond the intuitions legislated for
by the axioms and the appearances that were the subject of the antici-
pations. Kant now speaks of perceptions.83 The dynamic principles are
distinguished from the mathematical principles in that they deal with
the existence, and not just the possible mental construction in intuition,
of appearances84. To consider an object as existing is to grasp its given-
ness, in addition to its aesthetic and reflective formal properties. And
while the anticipations established merely the form of a material given,
that is, the anticipation that there would be some material given, the
analogies are characterised as dynamic principles insofar as they are
concerned with the possibility of actual given objects.
224
If we are to consider an appearance as an object – that is, as exist-

ing – we must necessarily conceive of it as standing in relations with
other objects. ‘Experience is possible only through the representation
of a necessary connection of perceptions’.85 A mere mental represen-
tation could stand alone for our imagination; for instance, the dagger
that Macbeth saw before him. In contrast, a real dagger would nec-
essarily stand in certain temporal and spatial relations to other per-
ceptions, and our perceptions of one object would be affected by
changes in other objects. For the moment, Kant speaks only of tem-
poral relations but, later, he will retrospectively declare that the rela-
tions in which objects stand to one another are also spatial.86 The
principle of the analogies thus can only be expressed in relation to, or
in anticipation of, possible experience where such relations will be
found.
The three sub-principles of the analogies – the permanence of sub-
stance, succession in time and reciprocal coexistence – only count as
regulative principles.87 Whereas the extensive form of an appearance,
and the possibility (though not, as we have seen, the actuality) of its
intensive quality, can be constructed, the existence of an object
cannot. The analogies of perception do not allow us to construct the
existence of a further object standing in relation to one that we
presently perceive.88 However, we possess a rule that allows us to
search out that missing element in the interests of achieving unity in
experience.89 It would appear, then, that the unity of experience as a
systematic set of relations between appearances is something that we
have an a priori right to search out, but not to assert in advance of
experience.90 Knowledge – at least, in its guise as a priori cognition –
is now firmly established as a task and not merely as a fait accompli.
In this respect, the ‘Analogies’ develop the anticipatory story of Kant’s
account of intensive magnitude.
Kant now makes a claim that sounds odd. He says that, the prin-
ciples in general and the analogies in particular ‘have significance and
validity only as principles of the empirical, not of the transcendental,
employment of understanding’.91 By now, we must take his point that
the principles only have significance insofar as they are applicable to
empirical experience. But can he really intend that they do not count
as transcendental? What Kant must mean here is that the principles
are a priori in status and thus count as part of the transcendental form
of experience, but they must be applicable to empirical experience
and would have no validity were they considered only as mere forms
of thought, that is, of the understanding as a merely subjective
225
process. The categories alone, as examined in the ‘Transcendental

Deduction’, are not yet the principles of empirical experience.92 Kant
adds that only a schematised category qualifies as a principle.93 While
this claim leaves open whether or not the principles count as a further
development of the categories, it is at least clear that the ‘Schematism’
chapter is necessary for the latter to qualify as the form of experience.
The analysis of the principles I have given in this section suggests
strongly that they, too, are needed if the deduction of the categories
as the form of empirical experience is to be completed. The principles
establish the application of the categories, not only to the form of
intuitions but also to existing objects. We must be careful, however,
to remember that even the existence or givenness of objects is, as yet,
considered only at a general level.
The second analogy is central not only for the establishment of a
necessary connection between perceptions – that is, for the analogies
as a whole – but also for the system of principles. In this discussion, I
will merely highlight the way in which Kant’s account of the law of
causality pushes forward his account of the application of the princi-
ples within experience. As is well known, in the second ‘Analogy’ Kant
seeks to provide the rule for distinguishing objective from merely sub-
jective succession in time. The sequence of cause and effect, where the
former determines the latter, cannot simply occur in intuition, but
is an operation of figurative synthesis.94 Objective succession must
capture something in the object and is, thus, not merely a subjective
representation. This requires a concept of the understanding.95
Kant is well aware that the term ‘object’ is problematic and that its
referent is a moveable feast:
Everything, every representation even, insofar as we are conscious of it,
may be entitled object [Objekt]. But it is a question for deeper enquiry
what the word ‘object’ ought to signify in respect of appearances when
these are viewed not insofar as they are (as representations) objects, but
only insofar as they stand for an object.96
Representations must ‘stand for’ objects if they are to qualify as

expressing objective succession in time. We cannot, in experience at
least, distinguish appearances, that is, appearing objects from the rep-
resentations or apprehensions we have of them.97 But despite the fact
that appearances always stand in some relation to our cognitive capac-
ity, Kant nevertheless insists that it is necessary to establish ‘what sort
of a connection in time belongs to the manifold in the appearances
themselves [an der Erscheinungen selbst]’.98 This curious expression is
226
his way of expressing that while appearances are necessarily for us,
they are nevertheless extra-mental insofar as they arise from a mater-
ial given that is taken up by our sensibility and understanding. Kant’s
terminological struggle to express the way in which representations
give access to objects is reminiscent of Merleau-Ponty’s wonderfully
evocative and seemingly contradictory expression: ‘in-itself-for-us’.99
An objective representation has to fulfil the requirement of some-
thing having been given, which is the first condition of knowledge,
as I argued in Chapter 3. The rule of causal succession makes possi-
ble that we can distinguish between mere representations in the imag-
ination and appearances themselves, which are accessed through the
synthesis of imagination.100 Kant is referring to the complex activity
of the imagination through which something given in intuition is
taken up and made available for unification by the understanding.
Access to substances, which was the topic of the first ‘Analogy’, is
only possible insofar as the faculties are used in cooperation with one
another in response to a material given. The law of causality finally
supplies the criterion for distinguishing a mere concept of the under-
standing from a principle. A principle achieves a figurative synthesis
through which our subjective faculties take up something given in
experience. The subjective conditions are necessary but insufficient
conditions of achieving objective knowledge, as I argued in the pre-
vious chapter.
It is, however, tempting to conclude that objective succession arises
from nothing other than a subjective rule: ‘If we enquire what new
character relation to an object confers upon our representations,
what dignity they thereby acquire, we find that it results only in sub-
jecting representations to a rule . . .’101 Kant’s point is that we can
never escape from our representations in order to establish some
‘mysterious kind of objective reality’.102 Grasping objective reality is
indeed the goal of our cognitive engagement with the world, but all
we have is a rule that allows us to make a distinction between sub-
jective and objective succession. We cannot simply abandon ourselves
to the circle of representations, nor can we venture out beyond those
representations. Our position is irrevocably that of seeing objects in
our representations and only thus of being capable of knowing rather
than imagining things in the world.103 All we have as security in our
cognitive project is a law of causal connection on which rest our fig-
urative syntheses of the empirical given.
Only the form of alteration is known a priori.104 Everything else –
for example, ‘how anything can be altered, and how it should be
227
possible that upon one state in a given moment an opposite state

may follow in the next moment’ – has to wait for experience.105
Highlighting the anticipatory status of his epistemological stance,
Kant concludes the second ‘Analogy’ by saying a priori knowledge is
simply the anticipation of our own apprehension, the formal condi-
tions of which alone count as a priori.106 I understand Kant’s use of
‘anticipation’ here as suggesting that the validity of the formal condi-
tions of experience – and even their meaning and sense – requires that
they are implemented within the material given. Formal idealism
anticipates material content, but it does not pre-empt the latter.
Consequently, a priori synthetic syntheses do not impose form on
matter.
The status of the postulates of empirical thought in general is not
only regulative but also subjective. Although Kant says that they
‘concern possible experience and its synthetic unity’, they achieve
only a subjective synthesis in contrast to the objective synthesis of per-
ception achieved by the analogies.107 This is due to the fact that ‘they
add to the concept of a thing (of something real), of which otherwise
they say nothing, the cognitive faculty from which it springs and in
which it has its seat’.108 We can make sense of the status of the pos-
tulates through the distinction between the subjective and objective
sides of the deduction. The last group of principles refers us back to
the cognitive faculties that are the first condition of experiencing
objects. As we have seen, the principles in general are only capable of
a subjective proof insofar as they are the foundation for knowledge
of objects.109 Knowledge is a possible relation in which a subject
stands to a world and the proof of its possibility necessarily entails
going back to its subjective sources.110 The principles, as a whole,
elaborate the way in which a subject takes up something given in
experience. The postulates in particular establish the ways in which
an object can be epistemically significant for a subject, namely as pos-
sible, actual and necessary. The last set of principles thus particularly
focus on the relation in which the object stands to the subject
Given the internal dependence between the subjective and objective
sides of the deduction I have argued for, it should not by now be sur-
prising, though it is striking, that in the heart of the ‘Postulates’, which
supposedly achieve only a subjective synthesis, Kant provides his most
trenchant statement of formal idealism and its commitment to empir-
ical objects. Something is actual if it is ‘bound up with the material
conditions of experience’.111 The formal and universal conditions of
experience give rise to what is possible and what is necessary, which
228
are the subject of the first and third postulates respectively. But if
something is actual then it arises from a material condition given in
experience. This is the subject matter of the second ‘Postulate’ and of
the ‘Refutation of Idealism’. For my present purposes, I simply want
to emphasise that Kant’s critique of Descartes’ material idealism is
specifically of his predecessor’s claim that certainty can arise from a
mere concept. Kant insists that even self-knowledge requires an aes-
thetic element, that is, a temporal intuition and a spatial consciousness
of outer objects.
Knowledge entails a cooperation of a plurality of subjective orien-
tations in response to a material given. In this section, I have shown
how the ‘Principles’ chapter extends the project of establishing the
objective validity of the categories, by showing that their temporally
schematised form must anticipate the given in experience. This entails
recasting the formal conditions of intuition as anticipations of the
empirical material given, that is, empirical objects, as I first argued in
Chapter 3 (p. 107). This finally establishes the relational and antici-
patory character of Kant’s formal idealism. In the next section, I will
consider if a full account of the transcendental structure of the rela-
tion between form and matter requires Kant’s establishing the role of
space in the determination of empirical objects.
V Spatial Schematism?
At the outset of the ‘Principles’, and in both editions, Kant insists on
the necessity of space for the objective validity of pure synthetic
judgements. Were it not for space, our a priori knowledge would
count as ‘nothing but a playing with a mere figment of the brain’.112
Immediately prior to this claim he argues that it is actual or possible
experience that gives ‘objective reality to all our a priori modes of
knowledge’.113 We must, then, conclude that the possibility of expe-
rience is established only when the categories are subjected to a figu-
rative synthesis that takes not only time but also space into account.
Kant goes on to claim that while experience depends on ‘a priori
principles of its form’, the objective reality of those conditions
can only be shown in experience.114 This reinforces the central thesis
of the ‘Principles’ chapter, namely that form must be applicable in
experience. Kant says: ‘Apart from this relation synthetic a priori
principles are completely impossible’.115 This is because there would
be ‘no third something, that is, no object, in which the synthetic unity
can exhibit the objective reality of its concepts’.116 Kant once again
229
uses the notion of a third thing to express the requirement that the
forms of thought apply within experience. The third thing was first
identified as temporal determination. 117 Next it was identified as the
temporal whole of experience arising out of the cooperation of the
faculties of intuition, imagination and understanding.118 Kant’s point
seems to be that the synthesis of a category with an empirical intu-
ition can only arise in relation to the possibility of experience, which
we have now discovered must be understood in spatial terms. This
development coincides with the claim that the third thing, which is
the distinctive characteristic of figurative synthesis, is the object
of experience. Synthesis arising from the productive imagination
requires a bridging term, the most explicit example of which is the
temporal schema. It now appears that the full elaboration of figura-
tive synthesis requires the spatial articulation of the categories.
But Kant’s account does not simply add articulation in respect of
the pure form of space to that of time. He says we can know much
about space in general, and the figures constructed in it, by produc-
tive imagination at the a priori level alone, but only insofar as space
is considered as ‘a condition of the appearances which constitute the
material for outer experience’ can we achieve knowledge and secure
the objective validity of the synthesis grounded on the categories.119
If the categories are to qualify as principles, space has to be consid-
ered not merely as a form of intuition, but as the condition of our
receptivity to a material given.120
Does this mean that the principles supply a spatial schematism? It
might be argued that it is not appropriate to suggest that the spatial
articulation of the categories necessary for their application to mate-
rially given or existent appearances qualifies as a schema. A schema
counts as a pure synthesis and is a transcendental product of the imag-
ination:
which concerns the determination of inner sense in general according to
conditions of its form (time), in respect of all representations, so far as
these representations are to be connected a priori in one concept in con-
formity with the unity of apperception . . .121
Admittedly, Kant consistently characterises the schema only in tem-

poral terms.122 However, I take it that the schema’s principal role is
to establish the possibility of figurative synthesis. If a category were
not temporally articulated, an intuition could not be subsumed under
it. And now it has emerged that a priori knowledge would count as a
fiction, were it not related to space.123 If figurative synthesis is to
230
achieve objective knowledge then the categories must be not only

temporally but also spatially articulated.124
The ‘Schematism’ establishes the possibility of figurative synthesis
and thus counts as transcendental, that is, as part of the necessary
structure of experience. In the ‘Principles’ chapter we discover that
the full articulation of the transcendental structure of experience
requires that figurative synthesis is not only temporal, but also spatial.
In the concluding sections of the ‘Analytic’, Kant argues not that any
actual empirical experience requires a spatial synthesis, but rather
that the general form of experience is spatial. He thus re-establishes
space as a transcendental condition of experience, something that was
first outlined in the ‘Transcendental Aesthetic’. Space must be pre-
supposed if objects are to be established as ‘outside me’ and ‘outside
and alongside one another’ in space.125 Only such objects qualify as
objects of knowledge for Kant. Time is the form of all intuitions
without exception, but the knowledge that Kant seeks to establish is
of intuitions that are also extended in space. If spatial iteration of the
categories is a necessary condition of the figurative synthesis that
gives rise to knowledge, then it makes sense to suggest that a spatial
schematism appears in the ‘Principles’ chapter.
Crucially, though, and as I have argued, space must be understood
not only at the formal level, but also as bound up with the material
given in appearances. For this reason, the spatially figurative synthesis
is not and cannot be purely formal. This may seem to close the discus-
sion on whether or not it makes sense to speak of a spatial schematism.
But I think rather, that the form of experience has now emerged as
entailing an anticipation of a material given. Admittedly, this develop-
ment cannot establish the need for a spatial schema in the narrow terms
of the ‘Schematism’ chapter. However, if a schema can be principally
understood as the formal or transcendental iteration of a concept
through figurative synthesis, what Kant now reveals is that the latter
must be effected spatially, and not only temporally, if knowledge of
extra-mental objects is to be achieved. The purely formal level of analy-
sis adopted in the ‘Schematism’ chapter has been replaced by an
account that is more attuned to the possibility of empirical experience.
Despite a forceful beginning, the role of space slips out of view in
the specific arguments of the first three principles, although, as we will
see, there are some traces of its importance. What remains a contin-
ual focus in both editions is experience as actual or possible. This, as
we have seen, is the new third thing that will make knowledge and
the validation of the categories possible and I have suggested that it
231
is closely associated with space. It appears that the role of space in

establishing the possibility of experience, although initially realised
by Kant as evidenced by our discussion above, only reappears later in
his argument and especially in the B edition. This is revealed by the
increasingly explicit claims about space in the ‘Postulates’, especially
in the B edition and in the ‘General Note’.126
The ‘Axioms’ are concerned with the construction of concepts in
space. Geometry, or the mathematics of space, is based ‘upon the suc-
cessive synthesis of the productive imagination in the generation of
figures’.127 Concepts alone are insufficient for geometric insight,
which always requires construction in space. In drawing a line, I must
draw it in thought, ‘generating from a point all its parts one after
another’.128 This process of spatial synthesis is necessary for a deter-
minate geometric intuition. This is synthesis in process within cogni-
tion. It is to be contrasted to aesthetic synthetic process that is not
directed to any determinate cognition, but reveals the synthetic
process, the cooperation of the faculties, necessary for any knowledge
whatsoever.129
Synthesis of spaces and times is necessary for all knowledge of
outer experience.130 But does space now qualify as a constituent part
of what Kant says is the third something, ‘medium’, or ‘whole’, within
which figurative synthesis is possible?131 Kant’s initial account states
that there is only one such framework, which is time.132 But although
there is no denying that time is a necessary condition of synthesis,
surely the account as it stands is insufficient. Clearly, extensive mag-
nitudes must be predicated of outer objects and can be presented only
in the form of outer objects: that is, space. It would appear, therefore,
that Kant should have revised his account of the medium that is neces-
sary for any possible experience and established that this is spatio-
temporal and not just temporal. This is, in any case, surely evident
from the fact that the knowledge he seeks to legitimate is of outer
things, the form of which is spatial.
Kant makes some progress in the direction of a spatial articulation
of figurative synthesis in the ‘Axioms’, but his account remains strictly
at the level of the formal construction of space and makes no mention
of space as a condition of material appearances. The rule for figura-
tive synthesis introduced in the ‘Axioms’ is the spatial analogue to the
temporal schematism, for it operates at the a priori level. It establishes
that it is only if we are capable of spatially schematising a concept, of
constructing it in space, that our concepts can qualify as applying to
really existing objects.
232
The ‘Anticipations’ add nothing explicitly to the elaboration of the

spatial schematism, concerned as they are with the intensive quality
of any appearance. The anticipations of perception are flowing mag-
nitudes insofar as they are temporal in status. But, as any intensive
magnitude will be presented only in relation to an extensive magni-
tude, at the level of experience, we can conclude that they must finally
relate to space. Intensive magnitude is a characteristic of appearances,
which the ‘Axioms’ establish as spatial.133 Even more importantly,
Kant now insists on the insufficiency of a purely formal account of
the legitimation of the categories. Form alone can only anticipate
what must be given empirically, as we saw in the previous section. The
material given is the necessary counterpart to the formal principle of
quality, and empirical matter can only be understood in spatial terms.
The first and second ‘Analogies’ do not prima facie discuss space.
They are concerned with the temporal determination of an appear-
ance, in accordance with the ‘Schematism’. While the first ‘Analogy’
concerns the temporal determination of a singular appearing object
as a substance, the second turns to the temporal relation in which a
plurality of substances stand to one another. But how can an appear-
ance that is not merely a subjective representation be conceived in
purely temporal terms? We will see that Kant revises his account in
the ‘Postulates’, and first implicitly and then explicitly suggests that
the analogies must be understood as bearing a spatial significance.
The third ‘Analogy’ seems to promise an explicit role for space, for
it is concerned with the necessary whole or ‘community’ within which
all appearances coexist. Moreover, in the second edition statement of
the general principle of the third analogy this whole is identified as
space.134 Surely, here at last Kant must explicitly return to the com-
mitment he made to the necessity of the spatial dimension of experi-
ence at the outset of the ‘Principles’ in both editions. How could the
interaction of a plurality of substances be understood without refer-
ence to their coexistence in space?
Kant undercuts this expectation by distinguishing between two
senses in which community can be understood. First, there is commu-
nio or communio spatii, which Kant considers to count only as ‘local
community’.135 He contrasts this to commercio by which he means a
‘dynamical community’ on which local community is dependent.136
Thus, spatial community – if this is what he means by local commun-
ity – is based on a community of substances. If so, spatial determina-
tion would be only an after-effect of a primary community among
substances. However, this does not seem to be consistent with the
233
internal logic of his account. While a singular substance was analysed

in purely temporal terms in the first ‘Analogy’, even in a commercium
appearances ‘stand outside each other and yet in connection’ and
surely this can only be comprehended as a spatial relation?137
Proliferating the difficulties involved in trying to clarify Kant’s
terms, he also speaks of a ‘community (communio) of appercep-
tion’.138 Apperception is a subjective community insofar as it is a
unity of appearances within the subject. It must stand in a reciprocal
influence with a ‘real community (commercium) of substances’.139 In
the context of the general strategy of the ‘Principles’ we have uncov-
ered so far, this sounds as if subjective unification of representations
must refer to a unity in the appearances themselves. This supports my
insistence on the relation between the subjective and objective sides
of the deduction. And surely it would be plausible to conclude that
such an objective unity would have to be in space, the form of outer
intuitions. But we have to wait until the ‘General Note’ at the con-
clusion of the ‘Postulates’ for confirmation of this suspicion.140
The first ‘Postulate’ is concerned directly with the construction of
a concept in space as the condition for the possibility of an empirical
object.141 But if this is so, then it calls for a re-evaluation of the role
of space in the first, second and third analogies. Kant says that the
concepts of substance, causality and community would contain only
‘a merely arbitrary synthesis’ if they were not related to experience in
general.142 I have already argued that Kant’s introduction of actual or
possible experience should be understood in spatial terms. We can
conclude that if a category is to qualify as the form of experience then
it must be constructible in space. Substance, causal succession and the
reciprocal community of substances must be spatially articulated.
Nevertheless, we can only infer this from Kant’s account.
The role of space in the second ‘Postulate’ comes to the fore in the
‘Refutation of Idealism’. Even self-knowledge must stand in relation
to outer objects in space. The ‘Refutation’ starts from what Descartes
considered indubitable, the Cogito, and argues that the latter is only
certain insofar as we are immediately and not mediately, as Descartes
held, conscious of outer objects. Thus, unless we doubt the Cogito,
we can be confident that we are capable of knowing outer objects.143
If our outer sense were merely a product of imagination, inner intu-
ition would also be annulled! If the internal activity of reflection need
not be doubted – and even Descartes did not doubt that – then there
is no need to doubt the evidence of outer sense. I think, therefore,
there are things in the world.
234
Prior to this striking argument, Kant declares the primacy of the

given object in relation to the concept. The rule that establishes the
actuality of appearances requires not that we construct the concept,
but that we recognise that a ‘perception can precede the concept’.144
The ‘Refutation’ then shows how this precedence can only be the
case if we accept the material existence of appearances in space. The
second ‘Postulate’ establishes the primacy of perception over con-
cepts in Kant’s formal idealism. Kant is committed to the view that
the formal structure must anticipate the possibility of application in
space. Kant sometimes suggests that perception [Wahrnehmung]
coincides with intuition, as he does at this stage of his account of the
second postulate.145 But he sometimes makes the further suggestion
that all perceptions are combined with a concept, or, as he sometimes
says, with consciousness.146 In the passage just cited, he suggests that
a perception coincides with an intuition that is not yet determined
by a concept. We need not conclude that the intuition is wholly
unrelated to a concept, for the intuition anticipates the possibility of
conceptualisation. But this also supports the suggestion that an intu-
ition could, under other circumstances, escape determination by a
concept.147
The third ‘Postulate’ is concerned with necessity, but strictly with
material, that is, causal and not logical necessity.148 In the second
‘Analogy’ Kant had already argued that we cannot know the exis-
tence of things as necessary, but we can know that their state is so,
insofar as they stand in a causal relation to some other appearance.149
Kant does not explicitly mention space in the final ‘Postulate’, but as
he has already insisted in the second ‘Postulate’ that space is the con-
dition of the material existence of objects, we can conclude that mate-
rial necessity arises only in space. Thus, the rearticulation of the
concept of causality as spatial, inferred by the first ‘Postulate’, coin-
cides with the third ‘Postulate’s’ characterisation of causal relation as
equivalent to material necessity.
The ‘General Note’ was added only in the B edition. The devel-
opment that we have already seen in the ‘Postulates’ is now stated
very clearly and counts as a rehearsal of the content of the
‘Principles’ chapter as a whole. Only in relation to an intuition can
something be shown to be permanent, alterable through causality
and coexistent in a community with other permanent things or sub-
stances.150 The concept of substance must be understood as a prin-
ciple for the possibility of experience, if the category is to achieve
objective reality.151 And this entails that the latter will refer to our
235
‘knowledge of an object given in empirical intuition’.152 What was

only inferred in the first and third ‘Postulates’ is now explicitly
stated. It emerges that in order for the concept to relate to objects of
experience, substances must be recognised as spatially existing
appearances or objects. The permanent is not to be found in inner
sense.153 Causal relations, properly understood, hold between spa-
tially existing substances.154 Finally, community is firmly established
as spatial, whereas we found that its status was not entirely clear in
the third ‘Analogy’.155
What I am calling the spatial schematism of the categories is thus
completed. If we are in any doubt that the original goal of the
‘Transcendental Deduction’ has finally been achieved, we need only
turn to the text:
But it is an even more noteworthy fact, that in order to understand the
possibility of things in conformity with the categories, and so to demon-
strate the objective reality of the latter, we need, not merely intuitions, but
intuitions that are in all cases outer intuitions.156
By the end of the ‘Postulates’, even in the A edition, it was becom-

ing clearer to Kant that the possibility of experience, and thus the
validity of the categories, requires the applicability of the latter in
space. By the time of adding the ‘General Note’ to the B edition, he
was adamant on the point. What we find at the conclusion of the
second edition of the ‘Principles’ is not only the elaboration of the
categories as the principles for the possibility of experience and for
appearances themselves as existing, but also a spatial schematism.
The latter is distinguished from the official schematism, not only in
referring to space–time and not simply to time but also insofar as
turning to consider space brings out the dual face of the aesthetic
element of experience. Space is the form of experience that allows
us to take up the empirical given. And the account of figurative
synthesis concludes with its spatial iteration, because it is principally
space that allows for bridging the gap between receptivity and the
reflective forms or categories. An understanding of the complex role
played by space makes possible a comprehension of how the
forms of intuition and the categories are capable of relating to an
empirical object. In achieving this, the often repeated goal of the
‘Deduction’ is now also legitimated, that is, the claim that the cate-
gories of the understanding apply to empirical objects. The final
sentence of the ‘General Note’ sums up Kant’s judgement on the
matter:
236
The final outcome of this whole section is therefore this: all principles of
the pure understanding are nothing more than principles a priori of the
possibility of experience, and to experience alone do all a priori synthetic
propositions relate – indeed, their possibility itself rests entirely on this
relation.157
The spatial schematism of the principles is not merely an addition to
the temporal schematism. Whereas the latter was able to operate
wholly at the level of experience in general, albeit in anticipation of
application to an object in general, the spatial schematism reveals a
necessary preoccupation with the material and empirical given in
experience. This is hardly surprising as space is the form of outer intu-
itions, that is, of objects given to us, which necessarily stand beyond,
though accessible to, the mind. Nevertheless, the spatial schematism
incorporates the official version and reinforces the temporal iteration
of experience.
We have had to excavate Kant’s text and draw out hidden connec-
tions in order to uncover the spatial schematism offered by the
‘Principles’ chapter. We have also seen that the evidence for this is not
continuous and only comes centre-stage in the ‘Refutation’ and
‘General Remark’. However I believe there is sufficient evidence in
both editions to conclude that Kant gradually came to the conclusion
that if the categories are to anticipate empirical application, in the
way that is necessary for their objective validity, they must be spatially
articulated.158 And I further believe that it is helpful to think of this
as a spatial schematism.
Before moving on to the final developments in Kant’s analysis of
the a priori form of empirical experience, it is important to empha-
sise the level at which he is operating. Even though he has moved
a great distance from the initial statement contained in the
‘Transcendental Deduction’ and is now concerned with the material
givenness or existence of objects, he has not strayed from the analy-
sis of a priori determining judgement. He has, however, shown that
transcendental analysis operates in anticipation of a material given.
As he says of the ‘Anticipations of Perception’ in particular: ‘This
anticipation of perception must always, however, appear somewhat
strange to anyone trained in transcendental reflection, and to any
student of nature who by such teaching has been trained to circum-
spection.’159 Despite the strangeness of the thesis, if Kant were not
committed to the anticipatory status of a priori knowledge, his posi-
tion would be condemned to material idealism.
237
VI Approaching empirical experience: material necessity and

the possibility of a gap between transcendental order and
empirical experience
We have seen that the objective validity of the categories will only be
established insofar as they are shown to be applicable within empiri-
cal experience. Kant has argued that this requires a temporal and, I
have suggested, a spatial schematism. We have also seen that the
‘Principles’ chapter offers only a subjective proof and yet this estab-
lishes the possibility of objective experience. Throughout his account
of the principles, Kant highlights the cooperation of the faculties
necessary for the figurative synthesis of objects given in time and
space. The ‘Principles’ chapter displays the two sides of the deduction
working in collaboration with one another so as to give rise to knowl-
edge of empirical objects.
In my account of the ‘Transcendental Deduction’ in Chapter 4, I
insisted that it requires a cooperation of the faculties, although I con-
ceded that there is a hegemony in favour of the understanding.
Examination of the text of the ‘Principles’ chapter shows it to be con-
sistent with the ‘Deduction’ in this regard. The principles are the prin-
ciples of pure understanding, yet we have seen that the articulation of
the concepts of the understanding, qualifying them as the form of
objects of empirical experience, requires a complex account of the
relation in which the understanding stands to the faculties of intuition
and imagination. 160 The concepts must not only be temporally but, I
have argued, also spatially schematised insofar as the understanding
must be supplemented by intuition through the exercise of the figu-
rative synthesis of imagination. When judgement is exercised in the
interests of knowledge, the understanding is always determining at
the end of the day, but it can only succeed in its task if it operates in
cooperation with a plurality of faculties. The complexity of the rela-
tion in which the understanding stands to intuition and imagination
was not yet established fully in the ‘Transcendental Deduction’ and
needs the deeper investigation of the structure of figurative synthesis
provided by the full ‘Analytic’ or canon.
We have come a long way since Kant first claimed that the cate-
gories provide the form of empirical experience in the ‘Transcendental
Deduction’. But are we there yet? We have seen that the ‘Schematism’
and ‘Principles’ chapters provide a successive determination of the
categories in anticipation of application within empirical experience.
Even at the conclusion of the ‘Principles’, however, Kant holds that
238
he has only established the form of empirical experience in general.

Empirical determinations, as he stated in the ‘Deduction’, are not yet
in the frame. Nevertheless, with the conclusion of the ‘Analytic’ we
would appear to have before us the fully articulated version of the a
priori form of empirical experience.
We have seen that Kant’s confidence in his ability to articulate the
form of empirical experience at a purely a priori level is tempered by
his continuing attempts to fine-tune that form to render it capable of
anticipating empirical application. The late discussion of what he
calls the principle of hypothetical necessity in the third ‘Postulate’ is
particularly instructive in this regard.161 Previously, in the ‘Axioms’,
Kant had simply dismissed the ‘idle objections’ that sensible objects
might not conform to their construction in space, on the grounds that
this would undermine the objective validity of space, mathematics
and geometry.162 The very possibility of a failure of fit between cate-
gories and sensible objects was thus dismissed out of hand on the
grounds of the pernicious results that would ensue. But this is, of
course, to beg the question. In contrast, in the third ‘Postulate’, Kant
considers the question somewhat more seriously, although his con-
clusion is no less confident.
In a development of the second ‘Analogy’, Kant addresses the ques-
tion of the material, rather than merely formal necessary connection
between objects. Consistently with his previous account, he concludes
that we can know things only as states following on other states and
not as isolated existing objects.163 The principal shift in the ‘Postulates’
is in the emphasis put on material necessity and, simultaneously, on
the subjective sources of any causal judgement.164 Kant now refers to
the principle of hypothetical necessity, which he says is necessary
for the very possibility of nature.165 This is not so much a new princi-
ple as a further specification of the law of causation. He immediately
proceeds to introduce four propositions, although he gives little sense
of the relation in which these stand to the general principle. A materi-
ally necessary connection between objects is secured by the additional
presuppositions that there will be no hiatus and no leaps in nature;
that nothing happens without a cause; and that necessity in nature is
conditioned and therefore intelligible.166 These jointly:
allow of nothing in the empirical synthesis which may do violence or detri-
ment to the understanding and to the continuous connection of all appear-
ances – that is, to the unity of the concepts of the understanding. For in
the understanding alone is possible the unity of experience, in which all
perceptions must have their place.167
239
Kant’s aim is to establish the systematic unity of empirical nature so

as to make the fit between the latter and the system of the categor-
ies secure. The difference between the treatment here and in the
‘Axioms’ is that while Kant insists in both on the legislation of the
categories, he now finds it necessary to further elaborate the princi-
ples in order to guarantee the empirical, now recognised as identical
with the material, applicability of the categories.168 The introduction
not only of a material and not merely formal articulation of causal
necessity, but also of four supplementary propositions, suggests he is
not yet quite satisfied that he has established that the a priori formal
structure coincides with the form of empirical experience. But if this
is true, he suppresses his anxiety quickly. The principle of hypothet-
ical necessity legislates from its a priori position that there will be
nothing in empirical syntheses that fails to fit with the principles of
understanding. His painstaking and repeated arguments that the
principles can only apply insofar as they anticipate experience is
abruptly short-circuited.
In the next chapter we will find that, on further reflection, Kant
realises that the form of empirical experience cannot simply be estab-
lished by the system of the principles. In the introductions to the
Critique of Judgement he introduces a distinct new principle, though
a reflective not a determinative one, of the purposiveness of nature for
judgement.169 This is also a priori insofar as it arises from the pure
subjective activity of judgement and so is not to be confused with the
empirical determinations Kant discussed in the ‘Deduction’. The
remit of this further a priori principle is to secure that empirical
nature fits with our judgements. Kant says that the new principle
establishes empirical systematicity, in contrast to the systematicity of
the principles. Thus this principle establishes a further link in the
chain between a priori form and empirical givenness. The attempted
Diktat expressed as the principle of hypothetical necessity has not yet
established that the categories provide the form of empirical experi-
ence. This will only be achieved with the addition of a principle that
delves even deeper into the relation between subjective powers and
empirical given in order to show how knowledge is possible. It is also
arguable that Kant is only finally in position to establish the validity
of the categories when he recognises the limitation of the formal
structure and accepts that there must be a gap between the transcen-
dental and empirical levels of analyses. This is the deeper significance
of the subjective status of reflective judgement. We can only seek
order in the empirical world and, in advance of experience, we cannot
240
be apodeictically certain that objects will in fact display the order we

attribute to them.
Conclusion
In this chapter I have shown how synthesis in process is at work
throughout the ‘Analytic’ and have argued that this is the subjective
side of the objective deduction as a whole. I have also brought out the
anticipatory status of a priori knowledge insofar as it stands in a
necessary relation to a material given. However, the ‘Analytic’ stays
strictly at the level of the formal conditions of the possibility of
experience in general. Kant finally comes to the view that if he is to
show that any actual empirical experience necessarily falls under the
rules of the categories of the understanding, a further stage of analy-
sis is required. He offers this final stage of his epistemological argu-
ment in the account of the reflective principle of the systematicity of
empirical nature in the ‘Introductions’ to the Critique of Judgement,
which will be addressed in the next chapter.
Notes
1. CPR, A xvi/xvii and A 93, B 126 respectively. See discussion of the
initial characterisation of the task of the ‘Deduction’ in Chapter 5,
p. 172.
2. B 161–2. See Chapter 4, pp. 143–5.
3. See discussion of Buchdahl’s MPS in Chapter 2, p. 49.
4. Allison, for instance, holds that the ‘Schematism’ chapter is necessary
for the full account of the deduction. He also investigates how the
‘Principles’ contribute to Kant’s core argument. See KTI(1983), Part 3.
5. This is important for forging a link to the discussion in the next chapter
of the principle of the purposiveness of nature, which I will argue is the
completion of Kant’s epistemological argument.
6. A 156, B 195. The relevant phrase is: Erfahrung (es sei wirkliche oder
doch mögliche).
7. A 93, B 125–6.
8. A 93, B 125.
9. A 93, B 126.
10. A 93, B 126.
11. A 92, B 125.
12. A 95.
13. A 95.
14. A 95.
241
15. See, for instance, B 161–2.

16. B 163.
17. B 165. I have altered Kemp Smith’s translation, which makes Kant’s
point opaque. Kant’s own expression is very clear on this issue. Kant
makes the same point in the A ‘Deduction’ at A 127/8.
18. The gap only becomes fully apparent to him in the Critique of
Judgement. See discussion of this development in Chapter 7,
pp. 249–50.
19. In Chapter 7, I will argue for this position (see pp. 254–5).
20. B 167.
21. B 168/9.
22. I discussed indeterminate intuitions in Chapter 4, pp. 130 and 145.
23. A 132, B 171.
24. See CPR, A 12, B 26; A 61, B 85; A 63, B 88.
25. A 796, B 824.
26. A 796, B 824. See also A 63, B 88.
27. A 136, B 175.
28. A95. See previous section above, p. 210.
29. A 136, B 175.
30. A 137, B 176.
31. A 138, B 177.
32. A 138, B 177.
33. Gibbons also sees the Schematism as addressing how intuition can be
schematised with understanding. See Kant’s Theory of Imagination,
pp. 70–1.
34. A 138, B 177.
35. A 136, B 175.
36. A 137, B 176; A 138, B 177; A 139, B 178.
37. A 139, B 178.
38. A 146, B 185.
39. A 138, B 177; A 142, B 181.
40. A 140, B 179–80.
41. A 141, B 180.
42. A 141, B 180.
43. See discussion of Heidegger in Chapter 4, p. 150.
44. And while greater depth is achieved in the ‘Schematism’, we have seen
in Chapter 4, (pp. 151–60) that only aesthetic judgement offers a full
depth analysis of synthesis.
45. CPR, A 142, B 181; A 141, B 180.
46. A 141, B 180–1.
47. A 136, B 175.
48. A 136, B 175.
49. A 148/9, B 188.
50. A 148, B 187.
242
51. A 149, B 188.

52. Indeed, the intentional direction of concepts was already established
in the A ‘Deduction’.
53. Kant talks of linking two concepts at A 155, B 194, but he must mean
that the second is an empirical concept that stands for something given
in intuition.
54. A 155, B 194.
55. A 155, B 194–5.
56. A 156, B 195. Kemp Smith is not consistent in his translation of these
terms, using both significance and meaning to translate Bedeutung (A
146, B 185; A 155, B 194–5, respectively) and both sense and signifi-
cance to translate Sinn (A 156, B 195; A 155, B 194–5, respectively).
But Kant’s intention, which does not require a modern technical use of
these terms, is unambiguous. Both categories and pure intuitions are
mere forms of thought if they are not applied to empirical intuitions.
On the contrast between the role of meaning for Kant and modern ana-
lytical ‘theories of meaning’, see Gardner, Kant and the Critique of
Pure Reason, pp. 288–9.
57. CPR, B 160–1. See discussion in Chapter 4, p. 138.
58. A 156, B 195. The claim that categories have objective reality only in
relation to intuitions comes at the outset of the A ‘Deduction’ at A 95.
See also A 155, B 194–5. We saw at A 146, B 185 that categories only
have significance (Bedeutung) insofar as they are schematised.
59. A 156, B 195.
60. A 156/7, B 196.
61. A 158, B 197.
62. B 168/9.
63. A 149, B 188.
64. See A 157, B 196. Experience is empirical synthesis.
65. A 157/8, B 196/7.
66. A 158, B 197.
67. Clearly, my reading of Kant as laying out the conditions of objectively
valid cognition – or knowledge – in a hierarchical fashion is dis-
tinct from the claim that the resultant account of those conditions
establishes them as anticipatory of the possibility of experience.
Nevertheless, Kant’s style of presentation reinforces the strictly formal
status of transcendental cognition.
68. A 160, B 199.
69. A 162–3, B 202–3.
70. A 166, B 207.
71. A 166, B 207.
72. A 166, B 207/8.
73. See discussion of sensation in Chapter 3, pp. 102–3.
74. A 167, B 209.
243
75. A 168, B 210.

76. A 175, B 217.
77. A 170, B 211/2.
78. A 176, B 218.
79. A 166, B 208.
80. See discussion of Guyer in Chapter 1, pp. 21–2.
81. A 160, B 199.
82. A 160, B 199/200.
83. B 218
84. A 160, B 199; A 178, B 220.
85. B 218.
86. See discussion in next section, pp. 229–37.
87. A 179, B 222.
88. A 179/80, B 222.
89. A 180, B 222.
90. See Chapter 1 (pp. 21–2) for my interpretation of A 125/6 where Kant
says that nature should [soll] be a unity. At A 181, B 224 he says this
unity can be thought only through a schema, not a concept. The initial
account in the A ‘Deduction’ has been revealed to be merely the begin-
ning of a more detailed account.
91. A 180/1, B 223.
92. We could also interpret Kant’s denial of the transcendental status of
the principles in the following way. He is just about to insist that the
Analogies are not concerned with things in themselves. He may simply
have used ‘transcendental’, when ‘transcendent’ would have been
more appropriate. Kant does not always adequately distinguish
between these two terms, as has often been remarked. In any case, the
result of both interpretations is the same. The principles are transcen-
dental in status, as Kant makes clear in the Introductions to the
Critique of Judgement. See Chapter 7, p. 249.
93. CPR, A 181, B 223.
94. B 233–4.
95. B 234. Note the similarity to the tripartite argument of the A
‘Deduction’ and the part played by the faculties.
96. A 189/90, B 234/5. Kant goes on to use the term Gegenstand to refer
to objects of experience, supporting Allison’s reading of the B
‘Deduction’. See discussion of the latter in Chapter 4, pp. 132–3.
97. See my discussion of representation in Chapter 3.
98. A 190, B 235.
99. Merleau-Ponty, Phenomenology of Perception, p. 322 (French edition
p. 372) ‘un véritable en-soi-pour-nous’. However, Merleau-Ponty’s
phrase has a perceptual rather than a cognitive connotation. I discuss
this in my ‘Kant’s Phenomenological Reduction?’.
100. CPR, A 190, B 235.
244
101. A 197, B 242.

102. A 197, B 242.
103. See discussion of the status of representations in Chapter 3,
pp. 96–100.
104. A 207, B 252.
105. A 206/7, B 252.
106. A 210, B 255/6: ‘Wir antizipieren nur unsere eigene Apprehension . . .’
107. A 219, B 267.
108. A 233/4, B 286.
109. A 149, B 188.
110. A 149, B 188.
111. A 218, B 266.
112. CPR, A 157, B 196.
113. A 156, B 195.
114. A 156, B 195.
115. A 157, B 196; my emphasis.
116. A 157, B 196.
117. See discussion of this expression in the section on ‘Schematism’ above,
pp. 213–17.
118. A 155, B 194; discussed in the previous section, p. 219.
119. A 157, B 196. This mention of the materiality of appearances predates
Kant’s introduction of their existence. However, I understand this early
comment as promissory for his later account where he will establish
not only the formal conditions, but also the material givenness of
objects.
120. See my discussion of the two-sided structure of receptivity at the end
of Chapter 3, pp. 105–8.
121. A 142, B 181.
122. See Gardner, Kant and the Critique of Pure Reason, p. 169, on why
time is the key to transcendental schematism. Gardner says this is
because time is ‘the most general unifying condition of intuitions and
concepts; all sensible objects are intuited in time, and all conceptual
activity stands under the condition of self-consciousness, the objects of
which are temporal.’ This is undoubtedly correct. However, my argu-
ment is that, as Kant intends to establish the validity of the categories
to all objects of experience and the latter are necessarily intuited in
space, then it is arguable that he needs to provide not only a temporal,
but also a spatial schematism. See also Makkreel, IIK, pp. 31–2.
Makkreel argues that despite the emergence of what he sees as a role
for space in the Schematism, time is not dependent on space and ‘their
relation is conceived as reciprocal’.
123. See above; A 157, B 196.
124. In Chapter 8 (pp. 292–3 ) I will argue that this is also true of aesthetic
synthesis and despite apparent counter-examples, such as music.
245
125. A 23, B 38.

126. Guyer is one of the few commentators who comments on the role
played by space in the ‘Principles’ chapter. However, he sees this late
development in Kant’s position as leading away from transcendental
idealism, which Guyer believes inevitably leads to impositionalism.
See Guyer, KCK, especially Ch. 16, for the role of space within tran-
scendental idealism. See also pp. 288–9 for the claim that Kant devel-
ops an account of space that entails the failure of transcendental
idealism.
127. CPR, A 163, B 204.
128. A 162/3, B 203.
129. See Chapter 4, pp. 151–6, on synthesis in process.
130. A 165/6, B 206.
131. A 155, B 194.
132. A 155, B 194.
133. A 166, B 207.
134. B 256.
135. A 213, B 260.
136. A 213, B 260.
137. A 215, B 261/2. We have already seen that objects’ standing outside of
one another was shown to require space in the ‘Transcendental
Aesthetic’ at A 23, B 38.
138. A 214, B 261.
139. A 214/15, B 261.
140. B 292–3.
141. A 221, B 268. See also A 224, B 271 on space as the formal a priori
condition of outer experiences.
142. A 221–2, B 268–9.
143. B 276 footnote.
144. A 225, B 272/3.
145. See also A 115.
146. See A 120, B 160.
147. See discussion of undetermined intuitions and, in particular, aesthetic
intuitions in Chapters 3 (p. 102) and 4 (p. 145).
148. A 226/7, B 279.
149. A 179/80, B 222.
150. B 288.
151. B 288.
152. B 289.
153. B 292.
154. B 292.
155. B 292–3.
156. B 291.
157. B 294.
246
158. In the Metaphysical Foundations of Natural Science (= MFNS), first

published in 1786, Kant analyses matter as the movable in space. See
Michael Friedman’s ‘Introduction’ to the English translation, p. xv. As
Friedman comments, in the second edition of the Principles chapter,
Kant claims the categories must refer not only to intuitions, but also
to outer intuitions. See Friedmann, pp. xx–xxii. In contrast, though,
the MFNS operates at a metaphysical level insofar as it is concerned
with the possibility of the further a priori determination of empirically
given objects. See CPR, AA 181. The Principles chapter, meanwhile,
remains at the transcendental level in being concerned only with the
possibility of the empirically given in general. I believe we can make
sense of the apparent coincidence between the ‘Principles’ and MFNS
as to the need for external intuition by making use of my suggestion
that Kant’s method is anticipatory of empirical experience. The
‘Principles’ chapter anticipates the metaphysical analysis of moving
objects in space treated in 1786. I have also shown that there is evi-
dence of the need for a spatial schematism even in the first edition of
CPR.
159. CPR, A 175, B 217.
160. A 148, B 187.
161. A 228, B 280.
162. A 165/6, B 206/7.
163. A 226–8, B 279–280.
164. A 226/7, B 279; A 233–4, B286, respectively.
165. A 228, B 280.
166. A 228/9, B 280/1.
167. A 229/30, B 282.
168. The new emphasis on the material status of causal relation is rein-
forced by Kant’s saying that ‘without material [ohne Stoff] nothing
whatsoever can be thought’ (A 232, B 284).
169. Indeed an intermediate stage is offered in the ‘Regulative Employment
of the Ideas of Pure Reason’, where the systematicity of nature is treated
as an idea of reason. However, Kant’s considered position is that the
systematicity of empirical nature requires a principle of judgement.
247
Empirical Systematicity and its Relation to

Aesthetic Judgement
We have found that the ‘Deduction’ of aesthetic judgement reveals

that the latter is based on the subjective conditions of cognition.1 In
Chapter 5 I argued that the principle of taste or common sense is aes-
thetic in status, insofar as it counts as the principle of the faculty of
judgement as such. Kant makes no mention of any other principle of
judgement other than taste, so I disagreed with Allison’s suggestion
that taste is grounded on a further principle, namely, the principle of
judgement in its subjective employment. In my view, the principle of
taste is the only principle that expresses the autonomous use of judge-
ment. My solution appears to give rise to a worrying result, namely,
that cognition is grounded on an aesthetic principle. There are strong
reasons, both independent and internal to Kant’s philosophy, for
resisting such a conclusion.
A related problem arises for my interpretation in the two
Introductions to the Critique of Judgement insofar as Kant introduces
a principle of judgement that does not at first sight count as aesthetic.
This is the principle of the purposiveness of nature and is the basis for
our presupposition of systematicity across the range of empirical
laws, thus making possible empirical judgements. Kant suggests that
the ‘Analytic of Aesthetic Judgement’ will serve as an exposition
and deduction of this principle, thus it would appear that taste is
grounded in the principle for empirical systematicity.2 We are thereby
faced with the reverse of the problem in the ‘Deduction’. That Kant
should have courted two contrary positions, both leading to conclu-
sions that would eliminate the distinctiveness of the principles
grounding the first and third Critiques, should make us pause for
thought. I will show how both problems can be resolved, once the
exemplary status of taste is established.
In the first section of this chapter, I examine the role played by the
principle of the purposiveness of nature. Although reflective judge-
ment has often been viewed as an optional supplement to Kant’s
248
Empirical Systematicity and Aesthetic Judgement
account of determining judgement, I argue that it is necessary for the

completion of the latter’s task of establishing the possibility of empir-
ical knowledge.
Kant often characterises the principle of the purposiveness of
nature as identical to the presupposition that there is systematicity
across the range of empirical laws. In the second section, I suggest that
we can distinguish between the general question of a fit between mind
and empirical nature and the more specific claim for systematicity. I
argue that the general level of purposiveness makes possible the com-
pletion of the Copernican turn at the empirical level. In contrast, the
concept of empirical systematicity is an instrumental application of
this general principle.
In the third section, I discuss the dual direction of both the purpo-
siveness of nature and the purposiveness of judgement, insofar as they
are directed to subject and object. I further suggest that they are
expressions of one and the same principle of the purposiveness of
nature for our judgement. Judgement is the faculty that allows sensi-
bility to be taken up by the understanding. It is thus that judgement
makes possible a relation between subjectivity and objects in the
world.
Finally, in the fourth section, I turn to Kant’s problematic claim
that the ‘Analytic of Aesthetic Judgement’ will provide an exposition,
and even a deduction, of the purposiveness of nature. I draw on the
distinctions established in the earlier sections in arguing that aesthetic
judgements are exemplary of the possibility of empirical synthesis.
I Understanding goes out into the empirical world

In this section I will argue that formal reflective judgement of nature
is the necessary supplement to the ‘Analytic’ of the first Critique
insofar as it makes possible the application of the categorical system
within the empirical world.3
Retrospectively, in the two Introductions to the third Critique,
Kant reveals that the first Critique established only the transcenden-
tal structure of experience. He suggests that empirical experience
might have been so varied that we would have been incapable of
grasping [fassen] it through the categorial system alone.4 There could
have been a wholly consistent mental order that did not tally with the
world out there.
Having raised this problem at the beginning of the Critique of
Judgement, Kant answers it by saying that the categorical system of
249
the first Critique is supplemented by a further system founded on the

principle of the purposiveness of nature for our judgement. This prin-
ciple establishes that empirical nature is organised as if it were a
system oriented towards a purpose. However, there is no purpose or
final cause and, thus, the principle counts as subjective or reflective.
Within the broad range of reflective purposiveness that is the topic
of the Critique of Judgement, Kant distinguishes between formal pur-
posiveness and real or objective purposiveness. Formal purposiveness
is the systematicity reflection finds in empirical nature viewed as a
whole and is subjective in status.5 Objective purposiveness is teleolog-
ical and is to be found either in the internal organisation of an object
viewed as organic, or insofar as an object is viewed as standing in rela-
tion to the final purpose of nature.6 Teleological judgement is subjec-
tive in the broad sense that it rests on judgement’s principle of
purposiveness and is not determinative of objects. However, while
formal purposiveness arises in relation to intuition, the real purpo-
siveness of nature is judged in respect of concepts.7 Teleological judge-
ment is objective in the weak sense that it applies a subjective principle
in conjunction with a concept in order to make sense of objects. While
formal purposiveness is necessary for any empirical cognition what-
soever, objective purposiveness is necessary only for objects insofar as
we view them as natural purposes. Kant suggests that the teleological
order in nature offers a ‘logical presentation of the purposiveness of
nature’, allowing us to get a grasp on the more indeterminate formal
purposiveness.8 Teleological presentation counts as logical because it
requires a concept of understanding or an idea of reason.
The problem formal purposiveness addresses is what Allison refers
to as the possibility of ‘empirical chaos’, which would arise were the
principles of the first Critique incapable of getting a foothold in the
empirical world.9 Kant solves the problem thus:
Now this principle can only be the following: since universal natural laws
have their basis in understanding, which prescribes them to nature
(though only according to the universal concept of it as a nature), the par-
ticular empirical laws must, as regards what the universal laws have left
undetermined in them, be viewed in terms of such a unity as [they would
have] if they too had been given by an understanding (even though not
ours) so as to assist our cognitive powers by making possible a system of
experience in terms of particular natural laws.10
The principles of the understanding establish the most general order of

nature, but this underdetermines the order that arises over the range of
250
empirical laws. Purely intelligible order arising principally – though, as

I have argued, not exclusively – from the understanding is not sufficient
for establishing order at the empirical level. The strictly mechanical
order of the first Critique must be supplemented by a purposive order
that treats empirical nature as a system for our judgement.
I remarked in the previous chapter that in the ‘Transcendental
Deduction’, Kant sets a limit on the reach of the categories. Special
laws are laws that specify the law-like relations between empirical
objects. Thus Kant makes clear that although empirical laws fall
under the categories, which supply the general structure of experi-
ence, the detail of empirical experience is not derived from its formal
structure.11 Whereas in the first Critique Kant conceded a distinction
between transcendental order and empirical laws, he now introduces
a further transcendental principle for those empirical laws.
While we might momentarily be tempted to think that the intro-
duction of the reflective principle of systematicity finally supplies the
empirical correlate to the universal formal order of the categories, we
cannot sustain such an opinion. The new principle of systematicity
operates formally insofar as it expresses the general organisation of
empirical nature and not the particular contents of the latter.
Nevertheless, the reflective principle is more attuned to the specificity
of empirical nature than were the categories in their initial transcen-
dental form. Kant has added a further layer to his hierarchical recon-
struction of the conditions for the possibility of experience. In so doing
he has narrowed the gap between formal structure and empirical given.
The formal structure now finally supplies an order at the level of the
diversity of forms in which the material given arises. The new princi-
ple of order is transcendental insofar as it only establishes the applica-
bility of the categories and not their actual application.12 The latter
requirement would annul the gap between form and content that
Buchdahl rightly establishes as characteristic of transcendental status.13
The system of the laws of the understanding requires the system of
reflective judgement in order to complete its epistemic task at the
empirical level. Correspondingly, the second system takes the first as
its model and ultimately depends on the system of understanding as
its ground: ‘Now it is clear that reflective judgement, by its nature,
cannot undertake to classify all of nature in terms of its empirical
variety unless it presupposes that nature itself makes its transcenden-
tal laws specific in terms of some principle.’14 This mutual depend-
ency of the two systems can be simply explained by saying that the
second is the specification at the empirical level of the first:
251
The principle by which we reflect on given objects of nature is this: that

for all natural things concepts can be found that are determined empir-
ically. This means that we can always presuppose nature’s products to
have a form that is possible in terms of universal laws which we can
cognize.15
But this is to say specification is not automatic and requires a princi-
ple in addition to those analysed in the ‘Analytic’ of the first Critique.
As Buchdahl argues, the first system gives only the form of the second,
but not the latter’s validating grounds.16 If the categorical system is to
be applied at the empirical level, we must additionally presuppose
that concepts are possible for all empirical objects.
Kant expresses this development in his position in a variety of
ways. Without the principle of reflective judgement, understanding
would find it impossible ‘to discover in nature an order it could
grasp’.17 Only by means of this supplementary principle ‘can we make
progress in using our understanding in experience and arrive at cog-
nition’.18 Finally, without it ‘understanding could not find its way
about in nature’.19
Thus, there are two systematicities and both count as transcen-
dental insofar as they supply the framework conditions of experi-
ence.20 However, whereas one is categorial and arises out of the
unifying power of understanding, the other is reflective and arises
from the power of judgement, which, as we have already seen, can be
characterised as mediating. In this case, reflective judgement mediates
between the transcendental system of the categories and empirical
nature.
I have argued elsewhere that despite the variety of descriptions
Kant uses to capture the status of the principle of the purposiveness
of nature – including heuristic, assumption etc. – it is most helpful to
understand the latter as a presupposition in the hermeneutic sense of
the term. 21 We cannot make sense of empirical nature unless we pre-
suppose that it displays an order accessible to our minds. This pre-
supposition is a condition for the possibility of experiencing empirical
nature, thus it is in no sense optional or psychological as the terms
‘assumption’ and ‘heuristic’ might suggest.
The presupposition of the purposiveness of nature is formal insofar
as it is the structure within which we take in the empirical world in
our intuition of it, but it does not guarantee that on every occasion the
world will make sense in terms of our mental structures.22 It simply
secures the possibility of that success. In this way, the principle of the
systematicity of nature is not simply a projection – as Vaihinger,
252
Buchdahl and Pippin suggest – but rather is an anticipation that makes

possible an exploration of the world or an orientation within it.23 We
presuppose that it is possible to make sense of the complex data of the
world, but we do not yet know how we can do so in any particular
case. The cognitive encouragement given us by judgement does not
encourage the epistemically lazy attitude that the order of nature is
already established by our minds, but rather encourages us to take up
the task of exploring the world with at least the hope that we can make
sense of it. As we will see later, aesthetic experiences are particularly
important in fostering this hope. The reflective task of making sense
of the world is one that permeates the great majority of our cognitive
tasks. It is as relevant to everyday low-level puzzles as it is to the most
abstruse theoretical scientific problems. At both extremes, while the
answer is not supplied a priori, what is transcendentally assured is that
there is some sense in seeking one.
The sceptic may still find that too much assurance is assumed by
this account. However, we should remember that this is a transcen-
dental argument that starts from our everyday practice in the world.
We do in fact operate as if we can make sense of the world. Kant’s
transcendental analysis of the principle of the purposiveness of nature
seeks to make sense of that practice and to show how it can be legit-
imated. The deduction of a principle of systematicity aims to legiti-
mate our practice of judgement by revealing a necessary condition for
the possibility of our experience. Kant aims to achieve this by dis-
playing how this principle is grounded in the subjective conditions of
our cognition.24 In effect, Kant intends to show that the cognitive
hope we display in everyday experience can be legitimated, not to dis-
prove the radical doubts of the sceptic in his or her own terms.
Kant makes a number of different claims for the principle of reflec-
tive judgement. As Allison summarises:
In fact, in various places in the Introductions, Kant suggests that the prin-
ciple of the purposiveness of nature is necessary for the formation of
empirical concepts, the classification of ‘natural forms’ into genera and
species, the unification of empirical laws into a system (theory construc-
tion), the formulation of empirical laws in the first place, and the attribu-
tion of necessity to such laws.25
Allison resolves the diversity of functions by saying that, in all cases,

reflective judgement is concerned to find universals for given particu-
lars. He goes on to argue that as the search for empirical concepts is
inseparable from the search for empirical laws, and the latter entails
253
a hierarchical organisation of such laws, then the attempts to estab-

lish the possibility of empirical concepts and the systematic organisa-
tion of empirical laws ‘are best seen as two poles of a quest for the
conditions of the empirical knowledge of nature qua empirical’.26
This is a very helpful synthesis of Kant’s position.
Allison’s account leaves open the question raised in the previous
chapter in the discussion of Kant’s proviso about special laws in the
B Deduction. Is Kant’s position that the singular empirical object can
be determined at the categorical level, while only special laws cannot?
Now that Kant has introduced a finer tuned account of the formal
structure of experience, he could merely be saying that our consider-
ation of the empirical laws of nature requires an additional reflective
principle, while the principles are sufficient conditions for the deter-
mination of an individual empirical object.
In my view, Kant should have said that the principle of the purpo-
siveness of nature is necessary for the possibility of any empirical
judgement. The form of empirical judgements is their organisation in
a system. For Kant, the necessity characteristic of laws entails unifi-
cation within a system, should that be at the level of the transcen-
dental unity of apperception or with respect to empirical order. Laws
operate within a framework of other laws and empirical concepts
must be understood within that framework.27 Thus empirical laws
and concepts, the classification of natural forms into species and
genera and the establishment of a system of empirical laws all depend
on the principle that nature is specifiable as a system that harmonises
with the categorial system of the understanding. Were we not capable
of presupposing this principle, none of these would be possible. In
short, order at the empirical level requires the principle of the purpo-
siveness of nature. And if we were not able to presuppose order in
empirical nature, we could not even conceive of making a judgement
about singular empirical objects.
Many interpreters, including Buchdahl, have interpreted Kant’s
account of reflective judgement as referring only to the heuristic
system of natural laws, the condition of scientific investigation of
nature.28 However, Kant’s claim that the transcendental system can
only be applied at the empirical level through the mediation of reflec-
tive judgement shows that Kant’s account of empirical systematicity
bears a further significance. If knowledge of a given empirical phe-
nomenon is to be possible, the categories of the understanding must
be applied at the empirical level. If understanding – the faculty giving
rise to the categories – is to find its way about in empirical nature, a
254
presupposition of empirical systematicity is required. Kant’s empha-

sis on law-likeness may seem to suggest that the cognition of partic-
ular empirical objects is not under consideration, but, for Kant,
cognition is only possible within a systematic framework. The cate-
gories of the understanding operate conjointly and give rise to a
system of principles. The principles operate as a system in establish-
ing the possibility of an object in general. Correspondingly, an empir-
ical object is only determinable within the context of a system of
empirical laws. Knowledge of a particular empirical object arises
within a context of lawful relations in which it stands to other objects.
The principle establishing the latter is thus a necessary condition of
the application of the categories to a given empirical object. The very
possibility of empirical synthesis requires that there be an order in
empirical nature that can only be secured by the principle of empir-
ical systematicity.
Kant says that universal natural laws are sufficient for the coher-
ence of objects in terms of their genus ‘as natural things as such’.29 This
encourages the view that transcendental systematicity suffices, at least
for particular natural beings. However, he goes on to say that the prin-
ciples ‘fail to provide them with specific coherence in terms of the par-
ticular natural beings they are’.30 What Kant must mean is that while
transcendental systematicity establishes the determination of an object
in general, it does not yet establish the complete grounds for the deter-
mination of a specific empirical object. Kant concludes that an a priori
principle of judgement is required in order to find a unity in accor-
dance with law within empirical contingency.31 His point clearly is
that were we not able to do so, empirical cognition could not arise.
II A distinction between general purposiveness and

systematicity
In the previous section, I have argued that the empirical application
of the categories is only possible on presupposition of the purposive-
ness of nature for our judgement. The principle of purposiveness
establishes, although strictly reflectively, that we can treat nature as a
system at the empirical level. This presupposition is necessary if we
are to establish knowledge not only of the lawful connections of
empirical nature, but even of the empirical objects that can only be
experienced as part of a wider system.
Towards the end of his discussion of the two ‘Introductions’,
Allison mentions a suggestion by Klaus Düsing, namely, that we
255
should distinguish between a more general and a more specific level

of the purposiveness of nature.32 Only purposiveness at a general level
would be the basis for aesthetic judgement. Allison remarks that
Düsing gives no sense of how even a more general sense of purpo-
siveness could ‘license’ particular claims of taste.33
I now intend to suggest a way in which we can identify a general
sense of the purposiveness of nature in contrast to a more specific
sense of the systematicity of nature’s empirical laws.34 This distinction
is one within the formal purposiveness of nature, which Kant has
argued is strictly reflective or subjective in status. The significance of
this description is not, however, that purposiveness has no role to play
in our knowledge of objects. As I have argued, it is necessary for
knowledge of objects at the empirical level. The point is that this
reflective principle does not determine objects and rather throws us
back on the relation in which the subject stands to an appearing thing.
The general sense of purposiveness concerns the accessibility of the
empirical object to our subjective faculties that cooperate when we
make a judgement.
The more specific expression of the principle of purposiveness, that
there is systematicity across the empirical laws of nature, counts as an
instrument or device that facilitates the general project of establish-
ing a fit between mind and empirical nature. Both levels are necessary
for the completion of the objective Deduction, which requires that the
categories are capable of determining empirical objects. There must
be a fit between mind and empirical nature if cognition is to be pos-
sible and establishing a hierarchical order of laws is the means to
securing this.
There are occasions on which Kant’s presentation of the order in
empirical nature appears to imply just such a distinction between a
more general question of the fit between mind and world and a more
specific device for bringing it about. For instance, early in the ‘First
Introduction’ we are told that if judgement is to be distinguished from
the other faculties by its own concept or rule then this will be one of
‘things of nature insofar as nature conforms to our power of judge-
ment’.35 Kant then goes on to say ‘the only concept we could form of
this character is that [nature’s] arrangement conforms to the ability
we have to subsume the particular laws, which are given, under more
universal laws, even though these are not given’.36 We might say that
the principle of nature, insofar as it conforms to our power of judge-
ment, regards nature understood at the level of general purposiveness
– that is, in terms of the possibility of empirical cognition as such –
256
while our way of grasping this fit requires that we devise a concept of
the systematicity of nature in its empirical laws. Meanwhile, in the
second Introduction Kant says that the principle of reflective judge-
ment is ‘the basis for the unity of all empirical principles under higher
though still empirical principles’.37 This, too, could suggest a distinc-
tion between levels at which purposiveness operates, insofar as the
power of judgement is the basis for the hierarchy of laws.
The purposiveness of nature at its general level takes up the task
of the Copernican revolution. If knowledge claims are to be legiti-
mated, it must be established that concepts are capable of applying
not only to intuitions in general, but, in particular, to empirical intu-
itions. This development of Kant’s argument was established in the
Principles.38 It now transpires that understanding cannot go out into
nature, that is, the concepts are not applicable at the empirical level,
unless there is a further reflective principle of judgement. The final
legitimation of the validity of the categories, first attempted in the
‘Transcendental Deduction’, is only brought to fruition in the
Introductions to the third Critique.
Kant most frequently presents the purposiveness of nature for our
judgement in its more specific sense as the systematicity of nature in
its empirical laws. As Allison says, if there is a distinction between
two levels of purposiveness, there is also a continual slippage bet-
ween them.39 Allison says this makes the task of establishing a link
between the purposiveness of nature and aesthetic judgements
extremely difficult. However, I have suggested that it is possible to dis-
tinguish the general question of purposiveness from its specification,
even if the former cannot be fully stated in isolation from the latter.
The distinction between the two levels of purposiveness is crucial
for making sense of the link Kant draws between the purposiveness of
nature and judgements of taste. But a further clarification is necessary
if I am to establish this.
III Purposiveness of Judgement and the Dual Direction of

Reflective Judgement
We need to establish the relation between the purposiveness of nature
and the purposiveness of judgement. Kant only rarely mentions the
purposiveness of judgement, preferring to talk of the purposiveness
of nature. While this sounds like a quite different notion, perhaps
even a contrary one, I will argue that the two phrases express the same
idea, i.e. that of ‘a formal purposiveness of nature for judgement’.40
257
Purposiveness is the relation in which empirical nature stands to our

faculties, insofar as they cooperate in judging a given thing. This is
the general sense of purposiveness that I established above.
Judgement, in both its determinant and reflective modes, is the
faculty that facilitates a relation between mind and nature, a task for
which it qualifies in that it operates as a mediator between different
faculties, primarily between intuition – for which in the third Critique
Kant often substitutes imagination – and understanding. It is, to a
great extent, the relation in which judgement stands to intuition and
its orientation towards the given in experience that qualifies the
former as fostering not merely the relation between subjective facul-
ties, but also between mind and world. In that judgement relies on
imagination, which is necessarily bound up with the synthesis of
apprehension, this, too, qualifies judgement as intentionally directed
towards an empirical object.41
Allison rightly insists that the principle of the purposiveness of
nature is heautonomous, that is, in exercising it, judgement legislates
only to itself and not to nature per se. This principle allows us to make
sense of the diversity of nature for our own purposes, while falling
short of claiming that nature is so structured in itself. However,
despite this stipulation, Allison still proceeds as if the purposiveness
of nature were principally to do with nature and not with judgement
itself, especially when he tries to prise what he calls ‘logical purpo-
siveness’ apart from judgements of taste. This is justifiable on the
grounds that Kant repeatedly refers to this principle as that of the sys-
tematicity of nature in its empirical laws. Even if the principle is heau-
tonomous, it still refers to our judgement of nature and does not
constitute a mode of self-reflection on the part of the subject, at least
not directly, rather making possible our orientation in the empirical
world. Nevertheless, there appears to be a tension in the status of the
principle of purposiveness insofar as it counts as a self-legislation and
yet makes possible our empirical judgement of nature,
This tension can be resolved insofar as we recognise the relational
status of the principle of the purposiveness of nature. By this, I mean
that the latter principle is concerned with the purposive relation
between mind and nature, a relation without which there would be
no possibility of experience. The principle points in two directions:
towards the subject and towards the object. While its intentional
direction towards objects is emphasised in the Introductions’ discus-
sions of the systematicity of nature, its subjective direction is entailed
by its reflective or subjective status. Reflective judgement in all its
258
forms counts as a mediation between the subjective and objective

dimensions of experience.
Allison believes that the legitimacy of aesthetic judgement can only
be shown by decoupling the principle of the purposiveness of nature
from the strictly subjective purposiveness characteristic of aesthetic
judgement.42 He concludes his initial discussion of the issue saying
that what he refers to as logical purposiveness, and I prefer to call the
formal purposiveness of empirical nature, is not the principle on
which taste is founded. Instead they are both founded on a further
principle, that is, the principle of judgement.43 I agree that both forms
of judgement are grounded in the activity of judgement, but, as I have
already argued in Chapter 5 (pp. 194–6), there is only one pure
expression of the latter, namely, in aesthetic judgements.
In order to show how the two species of formal reflective judgement
are related, I need to bring out the subjective side of the formal pur-
posiveness of empirical nature. The analysis of reflective judgement at
its most general level reveals the subjective conditions that are neces-
sary for knowledge of empirical objects. In other words, reflective
judgement rests on the subjective cooperation of the faculties neces-
sary for cognition.
In the First Introduction, in the section significantly entitled ‘On
the Technic of Judgment as the Basis of the Idea of a Technic of
Nature’, Kant first comments on how the purposiveness characteris-
tic of reflective judgement in general relates to an object. This arises
in one of two ways:
we perceive purposiveness in our power of judgement insofar as it merely
reflects on a given object, whether it reflects on the object’s empirical intu-
ition so as to bring it to some concept or other (which concept this is being
indeterminate), or on the empirical concept itself so as to bring the laws it
contains under common principles.44
The first alternative refers to the formal purposiveness of nature,

which reflects only on the intuition of an object and thus counts as
formal, as we saw above.45 In the second case, reflective judgement
reflects on an empirical concept in its systematic or instrumental use
in order to facilitate unification of empirical nature in a teleological
ordering. In both cases, the faculty of judgement reflects purposively
on a given object or on an empirical concept. This establishes
the intentional direction of reflective judgement towards an object
given in intuition, or at least towards a concept that orders objects
teleologically.
259
Only now does Kant go on to address the subjective side of the

exercise of the power of judgement in response to empirical nature:
But when we merely reflect on a perception we are not dealing with a

determinate concept, but are dealing only with the general rule for reflect-
ing on a perception for the sake of understanding, as a power of concepts.
Clearly, then, in a merely reflective judgment imagination and under-
standing are considered as they must relate in general in the power of judg-
ment, as compared with how they actually relate in the case of a given
perception.46
Kant’s point is that the technic or purposiveness of nature arises out

of a technic or purposiveness of judgement. A technic is a procedure
by which judgement orients itself with respect to natural objects,
which it views as if their possibility rested on art.47 Such a technique
is necessary due to the absence of direction from another faculty. In
establishing the relation between the two technics, Kant turns his
attention away from teleological judgement and addresses ‘merely
reflective judgement’, where the power of judgement is exercised in
its own right without reliance on understanding or reason.48 The
formal purposiveness of nature throws us back on the general condi-
tion of judgement, that is, the cooperation of the faculties of imag-
ination and understanding. This is the subjective direction of the
purposiveness of nature.
Without explanation, Kant moves from a consideration of the
general subjective conditions of judgement of the purposiveness of
nature directly to a discussion of the harmony of the faculties, charac-
teristic of aesthetic reflective judgements.49 It almost appears as if Kant
is suggesting that the immediately preceding discussion of the formal
purposiveness of nature referred to aesthetic judgement. But this surely
cannot be the case, as that discussion is directed to the purposiveness
of nature and not to aesthetic judgements. Moreover, if Kant simply
equated these, it would lead him to the implausible conclusion that one
of the general conditions of empirical cognition is a harmony of the fac-
ulties. We have seen this problem emerge on many occasions. In the
next section, I will finally establish the means for resolving it.
IV The link between the purposiveness of nature and

aesthetic judgement finally revealed
The evidence for Kant’s belief that there is a necessary relation
between purposiveness and aesthetic judgements is to be found in
260
both editions of the Introduction.50 Admittedly, the claims are less

apparent in the main body of the text, although they are not entirely
absent.51 The most strident example of his claims come in the First
Introduction where he says that the ‘Analytic of Beauty’ will provide
‘an exposition and then the deduction of the concept of a purposive-
ness of nature’.52 He also says that the proof that aesthetic judgements
refer intuitions to an idea of the lawfulness of nature will be in the
‘treatise itself’, that is, in the Critique of Aesthetic Judgement.53 While
not quite so dramatically couched, Kant’s commitment to the link
between the two species of formal reflective judgement is also to be
found in the second Introduction, where he says that natural beauty
is ‘the exhibition of the concept of formal (merely subjective) purpo-
siveness’.54 I have argued that formal purposiveness is nothing other
than the purposiveness of nature for our judgement. The question I
now address is whether Kant’s claims should really be construed as
entailing that aesthetic judgements have, as their ground, the princi-
ple of the systematicity of nature.
Allison concedes there is a connection between the principle of
judgement, which is the ground for aesthetic judgement, and the prin-
ciple of the purposiveness of nature. But he insists that this relation is
merely analogical insofar as the deduction of a principle of system-
aticity is a propadeutic, or establishes the possibility that there is a
principle for aesthetic judgement. This analogical connection is based
on the heautonomous status of both principles, that is, their entailing
a self-legislation on the part of judgement.55 But Allison does not
further investigate their systematic connection. While conceding that
Kant tries to systematically connect the two principles, Allison con-
cludes that the attempt failed.56
In contrast to Allison, Makkreel is happy to connect aesthetic
judgements with the purposiveness of nature, but he insists that both
are merely post-cognitive.57 Aesthetic judgements encourage us to
think that we could have success in our enquiries into empirical
nature and in particular into the ordering of the latter according to
genera and species. The latter activity is post-cognitive, Makkreel
concludes, because he identifies cognition strictly with the ‘Analytic’
of the first Critique.58 I have argued that cognition fully elaborated
should be understood as empirical and, as such, requires the princi-
ple of reflective judgement.59 So although the latter is post-categorial,
it is not post-cognitive. Moreover, aesthetic judgements reveal both
the pre- and post-categorial conditions of cognition. The harmony of
the faculties refers us to the initial conditions of cognition and not
261
merely to its successful application at the empirical level, as Makreel

suggests.60 Aesthetic judgements exhibit the initial conditions of cog-
nition, that is, the relation or synthesising activity of the faculties in
a particular empirical application of those conditions.
My solution for making sense of Kant’s claims about the link
between systematicity and aesthetic judgements is as follows: we can
distinguish between two levels of the purposiveness of nature: first, as
the fit in general between mind and nature; and second, as the more
particular systematicity of empirical nature. But it is only the more
general level of purposiveness of nature for judgement that is directly
exhibited in an aesthetic judgement. The exemplary role of the latter is
only possible because both the purposiveness of nature and the purpo-
siveness of judgement characteristic of aesthetic judgements hold a
relational status, mediating between subject and object. A singular aes-
thetic instance shows that objects in the world are, at least in principle,
capable of being taken up by our mental faculties, and that the forms
arising from those faculties are applicable at the empirical level. The
general fit between mind and world is thus established as possible.
With this in view, we can make sense of the passage from the first
Introduction considered at the end of the previous section in which
Kant puzzlingly seems to equate the formal purposiveness of nature
or technic of nature with aesthetic judgements.61 Aesthetic judge-
ments are the only judgements in which the power of judgement is
exercised without relying on the understanding or reason. Thus only
aesthetic judgements reveal the principle of judgement, which
expresses the general level of the purposiveness of nature for our
judgement. Teleological judgements rely explicitly on a concept of
what an object is, that is, its natural purpose. Even judgements aimed
at empirical cognition without any teleological implication, while
based on a strictly formal principle, make use of the more specific
notion of a hierarchy of laws. Their cognitive orientation masks the
activity of the power of judgement that is only revealed in aesthetic
judgements. While the philosopher can analyse reflective judgements
of nature so as to reveal the cooperation of the faculties on which they
are based, only aesthetic judgement gives insight into the synthesising
activity of the power of judgement as such at the experiential level.62
We can find the beginnings of a claim that aesthetic judge-
ment counts as exemplary for empirical cognition in the second
Introduction to the Critique of Judgement. Kant says that although
we once found pleasure in an ability to divide nature into genera and
species, we no longer experience such pleasure.63 We tend to conflate
262
this ordering capacity with ‘mere cognition’ for the reason that even
the most everyday experience would be impossible without such
ordering [der gemeinste Erfahrung ohne sie nicht möglich sein
würde].64 This must mean that we tend to conflate the empirical
ordering of nature, which at one stage in our development led to plea-
sure, with the transcendental structure of cognition, which does not.
Kant suggests that we need something that helps us focus on the pur-
posiveness of nature.65 In the absence of a characteristic pleasure in
purposiveness, we lack awareness of the latter’s distinctive status. The
title of the next section reinforces this insight, for Kant now turns to
what he calls ‘the aesthetic presentation of the purposiveness of
nature’.66 While he does not say in so many words that aesthetic pre-
sentation counts as exemplary of the reflective principle, it is clear
that this is the function it plays. Aesthetic judgements are singular
instances of something that is generally invisible to us. The beautiful
is, as Kant says in the fourth Moment of the ‘Analytic of Taste’, ‘an
example of a universal rule that we are unable to state’.67 In the
Introduction, this rule is identified as the principle of the purposive-
ness of nature for judgement at its formal level. At this level the prin-
ciple can only be exhibited in an exemplary fashion in a particular
case: it cannot be stated in a general proposition. My claim is that
taste is exemplary only of the general level of formal purposiveness.68
The possibility of cognition is encapsulated in one aesthetic
instance, insofar as this particular object, while not under our epis-
temic scrutiny, nevertheless reveals itself as open to the subjective cog-
nitive activity necessary for cognition. It thus displays the subjective
side of knowledge. But we cannot generalise from this case to con-
clude that all nature is like this, as would be necessary if beauty were
to exhibit systematicity. First, all objects are not beautiful and,
second, there are phenomena that defy our cognitive ambitions.69 At
best, aesthetic judgement can suggest but not prove that it is as if
nature were systematic for our judgement. Thus, aesthetic judgements
are not based on the systematicity of nature for our judgement, as the
latter is strictly heuristic for the furthering of our cognitive projects.
But now we need to look more closely at the stronger claims,
namely, that the ‘Analytic of Beauty’ counts as an exposition, a deduc-
tion and an exhibition of the purposiveness of nature and that this is
proved in the Critique of Aesthetic Judgement.70 Aesthetic judge-
ments have as their ground the principle of judgement, which rests on
the subjective conditions of cognition. The harmony such judgements
display is a special case of the cooperation of the faculties necessary
263
for any cognition.71 Such cooperation is thus the necessary subjective

condition of a successful synthesis at the empirical level.72 Aesthetic
judgements thus serve as examples of the possibility of the synthesis
of an empirical intuition under a concept, although they do so
without a determinate concept being in view. Kant says that aesthetic
judgements display a schematism not so much of an intuition under
a concept, but of the power of imagination under that of under-
standing.73 The subsumption of one power under another is what
allows ‘the understanding to proceed in general from intuition to con-
cepts’.74 Kant shifts register from the schematism of an intuition
under a concept, to a schematism of one faculty under another
because he is now concerned with the general possibility of synthesis,
that is, of the cooperation of the faculties that counts as the subjec-
tive side of the Deduction.
But aesthetic judgements do not simply display the subjective side
of the Deduction. If they refer to the general possibility of synthesis,
then they also show how our subjective capacities are capable of
taking up a given empirical object. Kant says that in an aesthetic
judgement the pleasure arises from the object being commensurate
with the cognitive powers.75 This counts as a ‘purposiveness of the
object with regard to the subject’s cognitive powers’.76 It would
appear that the objective orientation of Sections 13 and 14 of the
‘Analytic of Beauty’ does not amount to a slippage from the subjec-
tive orientation of the previous concentration on the harmony of the
faculties.77 Aesthetic judgements reveal the relation between subject
and object necessary for cognition in general.
In the second Introduction Kant reiterates a distinction he made in
the first version, where, as we have already seen, he distinguishes
between the formal purposiveness of nature and teleological purpo-
siveness. I cite only his revised comment on the former:
When an object is given in experience, there are two ways in which we can
present purposiveness in it. We can present it on a merely subjective basis:
as the harmony of the form of the object (the form that is [manifested] in
the apprehension (apprehensio) of the object prior to any concept), with
the cognitive powers – i.e., the harmony required in general to unite an
intuition with concepts so as to produce a cognition.78
It is clear in this passage, as it was in the first Introduction, that formal

purposiveness rests on a relation between the faculties that is neces-
sary for cognition in general.79 As so often before, Kant fails to make
an adequate distinction between the latter relation and the specifically
264
harmonious subjective relation characteristic of aesthetic judgements.

However, unlike the first version, where Kant did not adequately
explain the transition to a discussion of aesthetic judgement, he now
gives a strong hint as to how the relation between the latter and
formal purposiveness should be construed: ‘Hence we may regard
natural beauty as the exhibition of the concept of formal (merely sub-
jective) purposiveness, and may regard natural purposes as the exhi-
bition of the concept of a real (objective) purposiveness . . .’80 In the
face of the indeterminacy of the purposiveness of nature for our
judgement he has already discussed, aesthetic judgements serve as a
presentation of the fit between our subjective faculties and a given
object.81 They thus reveal the hinge between the subjective and objec-
tive sides of the Deduction. Kant speaks of an exhibition, and my clar-
ification is that this must be identified as an exemplary exhibition.82
Aesthetic judgement’s presentation of a harmony between intuition
and concept is an exemplary exhibition of formal purposiveness.
On several occasions, Kant claims that aesthetic judgements reflect
on judgement’s capacity for synthesising intuitions under concepts:
‘For this apprehension of forms by the imagination could never occur
if reflective judgement did not compare them, even if unintentionally,
at least with its ability [in general] to refer intuitions to concepts.’83
Aesthetic appreciation of objects rests on an implicit awareness of our
ability for synthesis. And, crucially, we can see that in aesthetic judge-
ment the synthesising activity takes place at the empirical level
because it arises in response to a particular given object. Something
counts as beautiful insofar as we experience:
only a harmony in reflection, whose a priori conditions are valid univer-
sally, between the presentation of the object and the lawfulness in the
empirical use in general of the subject’s power of judgement (this lawful-
ness being the unity between imagination and understanding).84
Aesthetic judgements display not only the cooperation of the faculties

characteristic of the transcendental schematism (and which give the
formal conditions for all experience), but also the completion of the
Copernican turn in a particular empirical application. The initial
conditions of cognition are combined with the necessary supplement
of the fit between mind and empirical nature as the conditions of cog-
nition in general. This is the form of empirical synthesis. It is thus that
Kant, in conclusion, suggests that this indicates ‘a purposiveness of
objects in relation to the subject’s reflective power of judgement, in
accordance with the concept of nature’.85
265
The general fit between mind and world is not merely inferred: it
is experienced in a particular phenomenon. In contrast, the system-
aticity of nature in its empirical laws could only be inferred from a
particular instance. Thus, aesthetic judgements, far from being
grounded in empirical systematicity, indirectly support the latter by
exhibiting the initial conditions of cognition in an empirical applica-
tion.
The singularity of the object under inspection reveals the general
possibility of cognition, but it does not do so as an explicit proof or
demonstration.86 What we get is a snapshot or intimation of the
general purposiveness between mind and world. A singular judge-
ment about this particular instance does not reveal the order of nature
in general. But it does show us that this object, at least, is conducive
to cognition.87 This encourages us in a cognitive hope that nature in
general may also fit with our subjective faculties. Moreover, it
reminds us that experience offers us evidence for such a fit. This sup-
ports the project of cognition and reveals the educative influence of
aesthetics.88 But it is still the case that the purposive fit between mind
and world can only be proven, if it can be proven at all, in the course
of experience itself. It is often put in question and sometimes we are
faced with the contrary insight that there is no harmony between
mind and world. This is only to be expected, for the fit between mind
and nature is strictly a presupposition, and a priori knowledge is a
task, not a mere fait accompli. Having the capacity to introduce
formal structure into the material given is only a necessary and not a
sufficient condition of experience. A priori cognition requires a mate-
rial given and in this sense counts as an anticipation, not simply an
achievement. It is crucial that aesthetic synthesis remains incomplete
or indeterminate and that no concept completes the synthesis.89 Only
thus is the very process of synthesis necessary for cognition in general
available for reflection and it is this that gives rise to the pleasure char-
acteristic of an aesthetic judgement. The peculiar capacity of aesthetic
judgements for presenting the possibility of cognition must be a
fragile one and, as such, the presentation counts as a deduction of the
principle of purposiveness only insofar as it offers an exemplary exhi-
bition of the possibility of cognition. The specific nature of this exhi-
bition will be explored in the next chapter.
It could be objected, however, that while the a priori conditions of
cognition are anticipatory and that even though the formal condition
of empirical knowledge – that is, the principle of reflective judgement
– also so qualifies, it is not the case that empirical knowledge per se
266
is anticipatory. This would result in a belated recognition that knowl-

edge at the empirical level, at least, is simply an achievement and not
a task. Even then, I would have established that Kant’s formalism is
not impositionalist. However, my further point is that any particular
claim for empirical knowledge necessarily arises as a synthetic process
and that this coming-to-know marks even the conclusion of a cogni-
tive process. While I do not pretend that this is an innocent interpre-
tation of Kant, I do think that his account opens up an understanding
of knowledge as a task. Our experience of objects in the world neces-
sarily arises as a network of inter-related claims, the totality of which
we cannot grasp. While we can have sufficient certainty about the reli-
ability of individual claims, they are therefore revisable in the light of
their wider context. This is not to say that they are merely provisional,
but only that we must be able to review them. It is because empirical
knowledge arises within an anticipatory or initiatory formal frame-
work – both a priori and empirical – that any particular empirical
judgement is marked by its openness to the possibility of revision.
This is what I mean by my claim that knowledge is anticipatory at the
empirical level. And aesthetic judgement has a role to play in this
development in my account. Empirical knowledge makes a claim to
a fit between mind and object that would establish its legitimacy. But
this claim is an anticipation of a proof that cannot be displayed in the
judgement itself; the fit can only be demonstrated in an exemplary
fashion by an aesthetic judgement of taste. Aesthetic harmony of the
faculties makes almost visible a trace of the general systematicity that
is the necessary background to any empirical knowledge claim.
Knowledge (Erkenntnis) is, as Nietzsche says, not ‘a bed to rest on,
or the way to such a bed’.90
Whether or not I have succeeded in convincing my readers that
even empirical knowledge is anticipatory, we can conclude that aes-
thetic judgement is not grounded in systematicity. If it is grounded in
the general level of purposiveness, then it is strictly in the sense that
it is a particular exhibition of the possibility of the latter. And the rela-
tion is reciprocal, for the ground, i.e. the fit between mind and empir-
ical nature, does not become apparent except in its exhibition in a
particular instance. Thus aesthetic judgement is the ratio cognoscendi
of the completion of the Copernican turn, while the fit entailed by the
latter is the ratio essendi of that exemplary exhibition. But as the
ground is exhibited in a singular instance, it can only count as a pos-
sibility or as a hypothetical ground that encourages our acting as if
there were a purposiveness of nature for our judgement.
267
We can now establish the significance of these two sides of judge-

ment for the objective and subjective deductions.91 The principle of
reflective judgement allows for the completion of the objective
deduction insofar as it allows ‘understanding to go out into [empir-
ical] nature’, but it also deepens our grasp of the subjective side of
the deduction as the necessary cooperation of the faculties in the
process of synthesis. Indeed, it shows how the two sides require
one another. Formal reflective judgement achieves the completion of
the objective deduction by making possible the synthesis of under-
standing and intuition at the empirical level. Our response to a beau-
tiful object, where our understanding and our intuition freely
harmonise, exemplifies the general condition of cognition. Addition-
ally, Kant says that judgement supplies a bridge between reason and
understanding.92
Despite the overt insistence on a harmony between faculties in
Kant’s characterisation of aesthetic reflective judgement, it is easy to
lose sight of the importance of his ‘faculty-talk’. We might, as in the
narrowly cognitive terms of the first Critique, be inclined to see it as,
at best, merely a device for expressing the way in which we make
sense of the formative activity of the mind in the face of the natural
world. But in the third Critique, Kant’s pervasive rhetorical trope
emerges as the vehicle for the examination of the relation between
subject and object that is the starting point for transcendental ideal-
ism’s account of experience.
Conclusion
In this chapter I have shown that there is a way of linking the sys-
tematicity of empirical nature to aesthetic judgement via a more
general conception of purposiveness that refers us back to the general
project of the Copernican revolution. I have argued that grasping the
duality within purposiveness, that is, its relational status, reveals how
it is necessary for Kant’s epistemological project. General purposive-
ness is never directly proven, but it is exhibited and as if in a snapshot
in an experience of beauty. In an aesthetic judgement we experience
the subjective conditions of cognition as the cooperation of the fac-
ulties, that is, the subjective side of the Deduction, while at the same
time seeing how a particular empirical object could be determined
by the categories of the understanding. Yet what we actually experi-
ence is not the determination of an intuition under a concept, but the
free play between the faculties of intuition (or imagination) and
268
understanding, a harmony that is conducive to the possibility of ‘cog-

nition in general’. Thus the subjective and the objective sides of the
Deduction are exhibited not so as to give rise to an actual empirical
cognition, but rather by showing how empirical cognition in general
is possible in one singular and ungeneralisable case. The synthesising
activity of the faculties is the key to the way in which reflective judge-
ment relates or mediates both subjectively and between subject and
object.
If I am right, then the principle of judgement in its full form is the
principle of the purposiveness of nature for judgement, understood at
its general level as the condition of the possibility of empirical syn-
thesis. This is not as implausible as it may at first appear. Judgement
is the mediating capacity that allows for the synthesis of concepts
with intuitions. Ultimately Kant seeks to show that concepts apply to
empirical intuitions. In the Critique of Pure Reason, judgement oper-
ates directly under the jurisdiction of understanding, but this is
because the analysis has not yet reached the level of the empirical
given and concerns intuition only as the form of its givenness in space
and time. We saw this in the previous chapter. In his account of taste,
judgement emerges as a faculty in its own right and no longer oper-
ates in the interests of the understanding, as it did in the first Critique.
Judgement is our ability to make sense of objects in the world through
the formal structures our minds initiate: it makes possible a relation
between mind and world. An aesthetic judgement provides an
epiphany of judgement as a power in action in the world, in particu-
lar as our purposive capacity for taking up nature.93 As such, an aes-
thetic judgement is an exemplary exhibition of the principle of
reflective judgement.
The principle of judgement is only expressed in the exercise of
aesthetic judgement. It is not that there is a deeper principle of judge-
ment that grounds aesthetic judgement, but that only as aesthetic
judgement does the faculty of judgement operate autonomously,
that is, as a principle. This makes it a curious principle, not one that
is an independent foundation for experience, but rather one that
depends for its identification on a mode of experience.94 The activ-
ity of judgement would not exist as a principle – in contrast to its
role as the subjective side of all cognition – were there not aesthetic
judgements that displayed it. This principle is constitutive for taste,
but strictly exemplary for cognition and experience in general. In the
final chapter, I will explore the character of taste’s exemplarity for
cognition.
269
Notes
1. I had already argued this in my article ‘The Technic of Nature: What is
Involved in Judging?’; see especially pp. 184, 186–7. This was first pre-
sented at a colloquium on the Critique of Judgement at Cérisy-la-Salle
in 1993. See, also, Allison KTT, p. 169 and my discussion of the
‘Deduction’ in Chapter 5 (pp. 193–201).
2. CJ, AA 251′. See discussion in final section of this chapter, pp. 260–8.
3. For a discussion of the difficulty of establishing Kant’s view on the rela-
tion between empirical systematicity and the determination of empiri-
cal knowledge, see my ‘Technic of Nature’, p. 180; for a positive
conclusion on the matter, see p. 187. For a perceptive comment on the
gap between transcendental principles and their empirical instantiation,
see Rodolphe Gasché, The Idea of Form Rethinking Kant’s Aesthetics,
p. 39. Although reflective judgement allows for the applicability of the
categories, the latter never coincide with their application.
4. CJ, AA 209′, AA 179–80.
5. AA 186 ff; 211′ ff.
6. Kant calls these intrinsic and relative teleological judgements, respec-
tively. For the distinction between these, see AA 378.
7. AA 232′.
8. The phrase is at AA 192, in the title of Section VIII. The confirmation
that teleological judgement counts as logical comes at AA 193. This is
one reason why I prefer to use ‘formal purposiveness’ to refer to the
order that is necessary for any empirical nature whatsoever. Allison, in
contrast, principally uses ‘logical purposiveness’ for the order of empir-
ical nature. See Allison, KTT, pp. 6, 32–3, 169, 354–5 note 11. The
other reason for my preference is that we will see that both judgements
concerning the systematicity of nature and aesthetic judgements display
formal purposiveness according to Kant. Retaining this convergence at
the level of nomenclature makes it easier to go on to make sense of the
substantive link between these two types of reflective judgement.
Admittedly Kant’s own usage is not consistent and he also refers to the
formal purposiveness of nature as logical. See, for instance, CJ AA 219′.
9. Allison, KTT, pp. 38–9.
10. CJ, AA 180.
11. CPR, B 165; see also A 127/8.
12. At CJ, AA 181/2, Kant says that the principle of the purposiveness of
nature is transcendental insofar as it concerns ‘only the pure concept of
objects of possible empirical cognition in general and contains nothing
empirical’. This is contrasted at AA 181 with a metaphysical principle
that establishes the conditions for further determination of the emp-
irically given at the a priori level. The latter is the realm of the
Metaphysical Foundations of Natural Science. The principle of pur-
270
posiveness and the metaphysical principles of MFNS, while occupying

distinctive roles in the broader hierarchy of Kant’s philosophical inves-
tigation of the a priori structure of cognition, both count as further
determinations of the transcendental principles of the ‘Principles’
chapter of CPR.
14. CJ, AA 215′.
15. CJ, AA 211′/12′.
16. Buchdahl, MPS, pp. 501/2. See Chapter 2, pp. 53–4.
17. CJ, AA 185.
18. AA 186.
19. AA 193.
20. See AA 181/2. At 241′, both aesthetic and teleological reflective judge-
ment count as a priori, although the latter does not count as ‘pure’ for
only the former is grounded – we must add ‘directly’ – on the principle
of judgement. See 243′. We can conclude that only aesthetic judgement
is transcendental, being directly founded on one of the higher cognitive
powers. Teleological judgement is only indirectly founded on the same
principle, being also dependent on the principle of reason.
21. See my ‘Technic of Nature’, pp. 179–80.
22. CJ, AA 232′.
23. See Chapter 6 for a discussion of the anticipatory status of knowledge.
24. CJ, AA 182. Importantly in this passage while insisting that a transcen-
dental deduction entails the establishment of the basis for such judge-
ments in the a priori sources of cognition, he also insists that we should
not try to follow ‘the psychological route’. The distinction is based on
the latter being concerned with how we actually judge, while the tran-
scendental deduction is concerned only with how we ‘ought to’ judge,
i.e. with the formal or structural norms of judgement.
25. Allison, KTI, pp. 30–1. Guyer also emphasises the range of roles played
by the principle.
26. Allison, KTI, p. 31.
27. Philip Kitcher argues that explanations of empirical phenomena neces-
sarily require a process of integration within a unified system. See
‘Projecting the Order of Nature’, p. 213. See also pp. 225 and 231.
Kitcher’s reconstruction of the link Kant makes between empirical
explanation and unification in a system of laws is surely correct. His
suggestion that causal laws ‘imply generalisations that legislate for
unactualized possibilities’ (p. 219) is, moreover, congenial to my inter-
pretation of form as an anticipation of empirical instantiation.
However, Kitcher’s aim of discarding ‘Kant’s apriorist lapses’ (p. 232)
clearly is not.
28. See discussion above in Chapter 2, pp. 59–60.
29. CJ, AA 183.
271
30. AA 183.
31. AA 183/4.
32. Allison, KTT, p. 63. Allison treats the formal or logical purposiveness
of nature as identical to the systematicity of empirical nature. In this,
he agrees with Christel Fricke, see Kants Theorie des reinen
Geschmacksurteils, p. 109. However, unlike Allison Fricke sees a way
of making sense of the link between aesthetic judgements and system-
aticity. She argues that the purposiveness without purpose displayed in
aesthetic judgements reveals the general conditions of the application of
schemata at the empirical level. See p. 115. Düsing gives a similar
account of the relation in which purposiveness without purpose stands
to empirical synthesis in his Die Teleologie in Kants Weltbegriff, see
p. 81. I agree that aesthetic judgements display the conditions of empir-
ical synthesis, while insisting with Düsing against Fricke that it is nec-
essary to identify a general level of purposiveness. Meanwhile in
contrast to Düsing, it is my view that purposiveness is principally an aes-
thetic, rather than a teleological, concept.
33. Ibid., p. 63.
34. For a more extended version of this argument, see Hughes 2006a.
35. CJ, AA 202′.
36. AA 202.
37. AA 180.
40. CJ, AA 232/3′.
41. See Chapter 4 (pp. 124–6), for an account of the relation between intu-
ition and imagination that explains why the latter can stand for the
former.
42. See Allison, KTT, p. 365, referring to Christel Fricke, Kants Theorie des
reinen Geschmacksurteils, pp. 109–11. Fricke holds that the purpo-
siveness discussed in the third moment of the Analytic is the same notion
of systematic organisation introduced in the Introductions. Fricke con-
cludes that Kant’s position, as it stands, is incoherent. Allison insists that
the two discussions do not deal with the same issue.
43. Allison, KTT, p. 64. See Chapter 5 (p. 198) where I cited this passage.
44. CJ, AA 220′.
45. AA 232′.
46. AA 220′.
47. Howard Caygill, Art of Judgement, p. 316. See also Howard Caygill, A
Kant Dictionary, pp. 387–8.
48. Though, as Kant says, for the sake of understanding, which (as we have
seen) could not operate at the empirical level without the power of
judgement. See pp. 249–55.
49. AA 221′.
272
50. See my ‘Technic of Nature’ p. 184, where I claim that the two forms of
reflective judgement are different modes of access to the same principle
of judgement. This is close to Allison’s position (see Allison, KTT,
p. 64), although he would disagree with my further claim that the pur-
posiveness of nature reveals the subjective conditions of cognition.
51. See my discussion of the ‘Dialectic of Taste’ in Chapter 8 (pp. 299–302).
At AA 246, Kant states: ‘Independent natural beauty reveals to us a
technic of nature that allows us to present nature as a system in terms
of laws whose principle we do not find anywhere in our understanding:
the principle of a purposiveness directed to our use of judgment as
regards appearances.’ This strongly supports those readings, such as my
own and Fricke’s, that insist on the continuity between discussions of
purposiveness in the Introductions and the main body of the text.
52. CJ, AA 251′. Kant’s claim that the Analytic will provide a deduction of
taste could be seen as evidence for the position that Section 21 counts
as a first attempt at a deduction pace Allison. See Chapter 5
(pp. 189–94).
53. AA 247′.
54. AA 193.
55. See above, p. 258.
56. For a more positive account of Kant’s attempt to connect aesthetic
judgement and the purposiveness of nature, see Douglas Burnham,
Kant’s Philosophies of Judgement, p. 161. Referring to CJ, AA 300,
Burnham brings out how beauty counts as a trace (Spur) or hint (Wink)
of purposive action. See also Burnham, p. 164, where he says that
beauty ‘is not merely the ‘image’ but the symbol of the empirical system
of laws, of the immanent lawfulness of nature’.
57. See Ameriks’ helpful review of Imagination and Interpretation in Kant.
58. Moreover, Makkreel reads cognition as imposing form on matter. See
discussion in Chapter 4, p. 157.
59. See Chapter 6 and Chapter 7, pp. 249–55.
60. See discussion in Chapter 4, p. 159.
61. CJ, AA 220–1′.
62. See Chapter 4 (pp. 151–6) on synthetic process and my discussion of the
significance of the introduction of the power of judgement in Chapter 5
(pp. 194–201).
63. CJ, AA 187.
64. AA 187. Kant’s phrasing here is rather confusing, compounded by his
referring to three feminine nouns, Faßlichkeit, Einheit and Lust.
However, I take it he must be saying that our being able to grasp nature
and its unity in its division into genera and species, not the pleasure once
associated with this ability, is necessary for experience. This interpreta-
tion is supported not only by the fact that he could hardly claim that
something that is no longer the case is necessary, but also that he goes
273
on to say that what is necessary is easily conflated with mere cognition.

As he has already said that the latter does not give rise to pleasure, the
evidence seems overwhelming. Pluhar’s translation seems to suggest that
it is the pleasure that is necessary.
65. AA 187–8. Kant refers here to the purposiveness of nature for our
understanding. He means the purposiveness that is necessary when
understanding goes out into empirical nature.
66. CJ, AA 188, in the title of Section VII. At AA 192, he offers another
exhibition of the indeterminate purposiveness of nature, that is, logical
or teleological presentations. However, those are not presentations of
formal purposiveness, while aesthetic judgement is.
67. AA 237. In the first Critique examples were deemed necessary for those
lacking in the natural talent of judgement. They encourage our using
rules as formulae rather than as principles (A 134, B 173/4). In the third
Critique the artistic genius is exemplary for other geniuses (AA 318).
This shows that exemplarity has taken on a more positive role. This new
development is, however, more indirectly related to the status of taste
than the two passages I refer to here. See, nevertheless, my discussion of
a possible parallel in Hughes, ‘Taste as Productive Mimesis’. Moreover,
examples have a role to play in the development of Kant’s aesthetic
argument. See, for instance, the case of Antiparos, which I discuss in the
next chapter. Nevertheless, the exemplarity I argue for here concerns the
relation between cognitive and aesthetic judgement only.
68. Both Düsing and Fricke also see aesthetic judgements as exemplary of
empirical cognition as I mentioned in a note earlier in this chapter, in
which I also noted the distinction between their positions and my own.
In her article ‘Lawfulness without a Law’, Hannah Ginsborg argues
that aesthetic judgements are exemplary of rule-following. See
pp. 59 ff. As for Düsing and Fricke, aesthetic reflective judgements
reveal ‘a condition of the very possibility of empirical concepts’ (see
p. 66). For Ginsborg, however, this is because they are exemplary of
how our perceptual synthesis ought to be (ibid.), Generally, I think her
approach is much more promising than those readings that focus on
the subjectivity of aesthetic judgement and seek to distinguish the
latter from cognition. However, I am not sure how helpful it is to say
that aesthetic judgements reveal how perceptual judgements should
respond to the object (see pp. 70 and 73). Ginsborg’s account suggests
that perception in an epistemic judgement should aspire to the quality
of perception characteristic of an aesthetic judgement. But, surely, this
cannot be the case for the two types of judgement are distinct from one
another. I prefer to say that aesthetic judgements are exemplary of the
possibility of cognitive judgements, insofar as they display the subjec-
tive conditions of knowledge and the relation in which they stand to
the object.
274
69. On the first issue, see Chapter 8 (pp. 284–90). I intend to argue else-
where that the sublime is a symbol of the limitation on our will to know.
70. CJ, AA 251′, 193 and 247′, respectively. See above, p. 261.
71. See Chapter 4 (pp. 152–6) where I first suggested this.
72. See pp. 255–7.
73. AA 287. See Chapter 5, p. 195.
74. AA 287. See Chapter 4 (pp. 151–6) on synthetic process.
75. AA 189–90.
76. AA 190.
77. Guyer, ‘Formalism and the Theory of Expression in Kant’s Aesthetics’.
I discuss this problem in Chapter 8 (pp. 280–4).
78. AA 192.
79. Compare AA 220′. See discussion on pp. 259–60.
80. AA 193. I have removed Pluhar’s emphasis, which is not in the original
text.
81. AA 187–8.
82. In general, exhibition (Darstellung) consists in placing beside a concept
an intuition corresponding to it, see AA 192. See Pluhar’s note to AA
233. In this case the concept does not subsume the intuition under it. It
is for this reason I speak of an exemplary exhibition.
83. AA 190. See also AA 249′.
84. AA 190. See also AA 192.
85. AA 192. He says, specifically, that judgements of the sublime do not
display this characteristic of judgements of taste.
86. See Chapter 8, pp. 284 and 295.
87. This reveals that aesthetic experience is educative in the sense that it
encourages our cognitive exploration of the world. But it is not didac-
tic, because the insight and encouragement it gives is only ever exem-
plary. We are encouraged to pursue ‘cognition in general’, but we are
given no rules for that task.
88. Although I am mostly here concerned with the pre-cognitive status of
aesthetic judgement, this is not incompatible with the position that aes-
thetic judgement also plays a post-cognitive role in encouraging the
development of our knowledge. One way in which Kant suggests this is
in saying that an aesthetic judgement involves the ‘quickening’ of our
cognitive powers; see AA 222. Makkreel’s account is extremely helpful
in this regard. See discussion in Chapter 4, pp. 158–60.
89. See Chapter 4 (pp. 151–6) on synthesis in process. See Chapter 8
(p. 283) on indeterminacy of aesthetic form.
90. Nietzsche, The Gay Science, Book 4, Aph. 324.
91. See extensive discussion of this in Chapter 5.
92. The bridge between understanding and reason would have to be strictly
propadeutic. That is, the harmony between intuition and understanding
in aesthetic judgement can serve as a symbol for the possibility of the
275
application of moral purposes within the empirical world. However, in

no sense can beauty serve as a guarantee for moral reason’s success at
the empirical level. Just as we have cognitive hope, so can we also have
moral hope. The latter, however, is more fragile even than the former.
93. The epiphany is, of course, perceptual and not divine.
94. Although this characteristic is striking, it is not idiosyncratic. All tran-
scendental principles ultimately rest on the possibility of experience; the
principle of judgement makes explicit what would otherwise remain
implicit.
276
Aesthetic Judgement’s Exemplary Exhibition

of Cognition
While the argument of this book is restricted to showing the way in

which cognitive judgement is dependent on a process of cooperation
that is only explicitly expressed in aesthetic judgement, I believe the
picture of aesthetics that has emerged has a much wider significance.
Reason in general is a plural not a unitary process, for we need to
think with and against our own thought, in addition to recognising
the limitations of thinking. Any logic that views itself as self-
contained, as not open to another position that could introduce a new
and necessary perspective, risks falling into dogmatism. Thinking that
is aware of its own relational status is open to what stands outside
itself. The relation in which aesthetic judgement stands to an aesthetic
object is exemplary for the openness of thinking that is required in
cognitive, moral and political thinking. In this chapter I look ahead
to this wider perspective, while drawing to a conclusion the interpre-
tation of Kant that I have developed in previous chapters.
I argued in the previous chapter that aesthetic judgement provides
an exemplary exhibition of a general purposiveness between our sub-
jective capacities and the empirical world of objects. This aesthetic
exhibition establishes the possibility of synthesis of an empirical
concept with a given empirical intuition, necessary for any cognition
whatsoever. In conclusion I now need to delineate more closely how
exactly aesthetic judgement is exemplary of cognitive synthesis
without counting as a species of the latter. How is it that a mode of
attention to the world that is not aimed at achieving knowledge, can,
nevertheless, illuminate the process of cognition?
Finding an answer to this question is important for both technical
and non-technical reasons. My reading of the relation between the
roles played by judgement in the first and third Critiques requires
an account of the specific identity of aesthetic apprehension and the
general significance it bears for cognition. However, I also want
to establish that Kant’s account of aesthetic judgement reveals an
277
existential insight into the deeper structure of judgement in general

and allows us to reflect on the way in which our experience depends
on a combination of a plurality of different orientations. This is what
I have identified in earlier chapters as the subjective side of the deduc-
tion, which we can now see is necessary for establishing the applica-
bility of the categories within empirical experience. Focusing on
cognitive judgement – or indeed on morality – alone would risk
encouraging a view of human beings as oriented toward one over-
arching goal. I hope I have shown sufficiently by now that Kant’s aes-
thetics presents human judging agents as necessarily coordinating a
plurality of orientations.
Judgement is the ability to connect, not only with respect to a par-
ticular instance, but also, more fundamentally, at the level of the
thinking process. If we were not able to coordinate sensory receptiv-
ity with a reflective capacity for unifying and identifying what we take
in through our senses, we could not experience anything at all.
Neither could we think about anything, for our capacity for thought
always stands in relation to receptivity. Aesthetic judgement reveals
the human condition of finding ourselves in a world we make sense
of through a combination of orientations.
The perspective I have presented in earlier chapters prepares for a
positive characterisation of aesthetic apprehension as directed to the
appearing of the appearance. All objects are appearances for Kant,
but only some exhibit the way in which an object can appear for
human beings who are not merely sensory but also reflective.
Aesthetic phenomena reveal the relation in which the subject stands
to the object and vice versa. My relational interpretation of aesthetic
judgement allows me to reconstrue what might otherwise seem like a
slide from a subjective to an objective determination of form in the
third moment of the ‘Analytic of Beauty’. In my account of the
appearing of the appearance, I implicitly make use of phenomeno-
logical insights from authors including Husserl and Merleau-Ponty.
Having suggested how we can understand the role of aesthetic
apprehension within experience, I turn to address a technical problem
to which my account seems to give rise. If aesthetic objects are char-
acterised by their appearing for us and all perceptual objects are
appearances, then how can aesthetic form be distinguished from per-
ceptual form in general? I address this by arguing that while aesthetic
judgement begins with the spatio-temporal presentation of an object,
only some objects qualify as aesthetic. Unlike objects in general, the
perceptual form of aesthetic things throws us back on the subjective
278
Aesthetic Judgement’s Exemplary Exhibition of Cognition
activity that is necessary for any judgement whatsoever. Thus I need

not arrive at the conclusion often considered the unavoidable corol-
lary of Kant’s formalism, namely, that all objects are beautiful. I make
particular use of Kant’s example of the grotto of Antiparos to explore
the particularity of a merely mechanical form that is also aesthetic.
The connection I draw between perceptual form, in general, and
the appearing of that appearance now requires me to address another
issue. It is usually accepted that Kant restricts his aesthetics to con-
siderations of the spatial and temporal features of aesthetic phenom-
ena, to the exclusion of their qualitative features. I suggest that
although this conclusion need not necessarily be drawn from his
account of aesthetic apprehension, he is nonetheless committed to
what I call the ‘primacy of presentation’.
One strength of the way I interpret aesthetic apprehension as
entailing an implicit reflection on the appearing of the appearance, is
in highlighting the anticipatory nature of the relation in which the
subject stands to the object in cognition in general. Aesthetic judge-
ment makes us aware at one and the same time that judgement
requires not only a combination of different mental orientations, but
is also dependent on something beyond our minds. In the aesthetic
judgement there is a heightened insight into the anticipatory structure
of a priori knowledge, due to the specific character of aesthetic judge-
ment as singular. The result is that there are no rules for beauty and
we cannot know in advance of experience what will or will not give
rise to aesthetic pleasure. We might still suspect, however, that aes-
thetic apprehension is a species of cognition. I argue against this, that
aesthetic judgement displays what I call contrapuntal exemplarity.
Aesthetic appreciation is exemplary by contrast with cognition.
Throughout my account I have emphasised the way in which aes-
thetic judgement requires a combination of different mental orienta-
tions. But what if aesthetic judgement itself were dependent on a
higher order of reason, on which thinking ultimately depends? If the
higher order of reason did not display plurality and, instead, dis-
played an orientation to a unitary end – for instance, a moral end –
the pivotal role I have assigned to aesthetics would be undermined. In
this case, the pluralist activity that I have argued is necessary for
judging in general would turn out to be merely subsidiary to a think-
ing oriented to a unitary end. Now, while I do not deny that judging
strives for unification and that this is a condition of making sense of
the world, I do resist the idea that unity is determining for judging.
This is the significance of Kant’s distinction between reflective and
279
determining judgement for thinking in general. We seek out unity and

sense, but we do so out of a genuine engagement with a multiplicity
of orientations to the world. Unity is an ideal goal or orientation, not
something we can impose upon the world. A textual development
towards the end of the Critique of Aesthetic Judgement requires I
address this problem. In the ‘Dialectic of Taste’, Kant seems to suggest
that aesthetic judgement is ultimately based on morality. I show,
however, that for Kant the plurality of mental activity is conserved
even at the supersensible level and that morality serves only as an
ideal, not a determining, end for aesthetic judgement.
Finally, I return to the question that has haunted formalism from
the outset. How can the subjective faculties make possible access to
an objective given? Drawing on the account of formalism I have
developed in earlier chapters, I conclude that aesthetic judgement
reveals formation as synthesis in process, taking up the material given
and revealing the event of affection as an event. My interpretation
thus finally makes sense of Kant’s claim that his method is one of
formal and not material idealism. Formal idealism is characterised by
a dynamic relation between mind and things in the world, their ulti-
mate context.
I The appearing of appearance is aesthetic form

In the third moment of the ‘Analytic of Beauty’ Kant claims that when
we judge something to be beautiful, we are concerned only with its
form. ‘A pure judgment of taste is one that is not influenced by charm
or emotion (though these may be connected with a liking for the beau-
tiful), and whose determining basis is therefore merely the purpo-
siveness of the form.’1 Many have taken this to raise a problem for
his account insofar as he seems to have shifted from a position where
aesthetic judgements are based on the form of purposiveness in a
judgement to one where they are based on the purposive form of the
object.2 In the first case beauty arises solely from the harmonious
activity of our faculties, whereas in the second case its source lies in
something given to us in experience. Such a switch in position would
fly in the face of many of Kant’s most fundamental commitments,
especially the transfer of impetus from object to subject commonly
held to be expressed by the Copernican revolution. It would also
render incomprehensible Kant’s regular characterisation of aesthetic
judgement as ‘subjective’ and, more precisely, as based on a subjec-
tive principle.
280
It is hardly surprising in the face of these and other considerations

that the great majority of Kant’s readers have concluded that the
development of his argument in Sections 13 and 14 must be taken
with a large pinch of salt. Allison, for one, concludes that the aesthetic
form of the object is not equivalent to its spatio-temporal form.3
Nevertheless we have already seen that the emphasis Kant puts on the
object in the third Moment is not an isolated instance, for in the offi-
cial deduction Kant says that in a pure judgement of taste ‘our liking
for the object is connected with our mere judging of the form of the
object’.4
In the previous chapter, I argued that aesthetic judgement is
capable of displaying the possibility of empirical synthesis and thus
of showing that cognition is, at least in principle, possible. For Kant,
objectively valid cognition is knowledge of objects in space and time.5
If aesthetic judgement is capable of exhibiting the possibility of
knowledge, then that exhibition must itself be in space and time. It is
for this reason that design or spatio-temporal form is inextricably
caught up with aesthetic form. Now, how can this be so, without
entailing the problems I have just raised?
My approach starts from an insistence on the relational status of
aesthetic judgement, by which I mean that the latter reveals how a
subject has access to an external world of objects. As we have seen in
previous chapters, our capacity for standing in relation to something
given in experience rests on sensibility, our capacity for being
affected. However, if we are to become conscious of a given affect, the
imagination is also crucial as it must hold together the manifold in
such a way that it appears to us as a figure. The account of the epis-
temic role played by the imagination is most developed in the
‘Schematism’. If the imagination did not mediate between sensibility
and understanding, the latter would not be able to subsume an intu-
ition under a concept. The mediating role of imagination is crucial for
the pluralist model of mind at work in Kant’s epistemology and in his
system in general. The imagination allows a plurality of faculties to
remain distinctive orientations to the world, while standing in
ongoing relations with one another. In aesthetic judgement the medi-
ating role of the imagination and the figure to which it gives rise are
held up for phenomenological inspection, that is, for reflection within
experience. This is in contrast to the philosophical analysis of the
same phenomena presented in the ‘Transcendental Deduction’ and
‘Schematism’. In the aesthetic case, imagination does not deliver the
schematised intuition for the purposes of cognition. It stops short,
281
while still standing in relation to the formal conditions of cognition

in general, that is, the mutual relation of the faculties that counts as
the subjective side of the deduction. The imagination generates a
figure that holds together the manifold in intuition as if it were about
to be unified under a concept and yet the subsumption does not occur.
Aesthetic apprehension displays a free lawfulness that prepares for
and yet does not conform to a rule of the understanding. The rela-
tional or mediating role of imagination is thus constitutive for aes-
thetic judgement, while in the epistemic case the third faculty operates
under the hegemony of understanding.6
The relational status of aesthetic judgement is the key for under-
standing the mutual dependency of the form of purposiveness and the
purposiveness of form. In an aesthetic judgement the subject becomes
reflectively aware of the relation in which he or she stands to an
object, which would under other circumstances be an object of
knowledge. I am not, of course, arguing that there is something about
the object in-itself that causes an aesthetic response, but rather that
the aesthetic object encountered in experience prompts a heightened
relation between subject and object and thereby reveals the inten-
tional structure of judgement. The object has a role to play, but only
insofar as it is for us or stands in relation to us as an appearance. For
Kant all objects are appearances, but aesthetic objects appear in such
a way as to highlight their appearing.
The purposiveness of form is not an alternative to the form of pur-
posiveness, from Kant’s perspective. Some objects prompt a free play
of our faculties insofar as their appearance resists cognitive conclu-
sion. Due to this resistance we remain preoccupied with such things at
the level of their appearance. We can only sustain an attention to their
appearance insofar as our cognitive powers remain in a free or open
relation. We have to be capable of a high degree of mental mobility
and an ability to tolerate a degree of inconclusiveness if we are to focus
on the mere appearance of the object. Certain objects facilitate this
freedom of mind allowing us to focus on the appearing of the appear-
ance, rather than on the latter as a mode of access to an object that
can be known, valued or acquired. Such contemplation reveals the
general relation between subject and object, that is, the form of cog-
nition in general. This is to say that aesthetic judgement is subjective
in the strict sense that it is based on the activity of the cognitive powers
of the subject, exercised in abstraction from any ulterior purpose.
The aesthetic subject, however, is not detached from the aesthetic
phenomenon and only from the object determined for cognition. In
282
earlier discussions we have seen that the term ‘object’ poses problems
for Kant. Strictly speaking, an object is the result of the synthesis of
an intuition under a concept. But he also uses the term in a more
general sense to express something given to the mind in experience.
We saw that in the ‘Transcendental Aesthetic’ he talks of ‘[t]he effect
of an object [Gegenstand] upon the faculty of representation’, even
when he is referring to mere sensation.7 He also says that it is possi-
ble to experience intuitions without conceptual determination.8 This
suggests that we can be affected by phenomena that do not strictly
count as objects, that is, that are not fully determined.
We can now see the importance of a distinction I have made on
several occasions. In the aesthetic sphere Kant must leave room for
phenomena that are not, or at least not presently, considered as cog-
nitive objects. An aesthetic judgement is characterised by its reflective
status. We are aware of the appearance as appearance and not as the
appearance of a known object, but we still stand in relation to some-
thing given to our minds. The object in the aesthetic case is the
undetermined or underdetermined referent of an indeterminate or
underdetermining judgement, which arises as synthesis in process.
This would not be possible if all intuitions were determined rather
than merely determinable by the categories.
Certain objects present themselves in such a way that we respond
with a free play of the faculties. But these objects could not have an
aesthetic affect on us, were we not capable of exercising such a free
form of judgement.9 While it is true that the object prompts us, it is
also true that the imagination transfigures the empirical givenness of
the object so that appearance as appearance can be apprehended. The
relation between subject and object is reciprocal in the aesthetic case.
This is what it means for there to be a harmony between an object
and our judging power.
Appearances in general are appearing objects and we usually treat
them as such. Insofar as we do so, representations are our means for
experiencing objects that are represented. But in the aesthetic case we
focus on the appearance and are not currently concerned with the
existence of what appears. This contrast is not intended to suggest
that what appears stands behind the appearance, but rather that the
appearance is our mode of access to the appearing thing. When we
look at the appearance for itself we no longer treat the phenomenon
as a point of access to something else: we reflect on the point of access
or relation in which we stand to the thing. Thus an aesthetic appear-
ance reveals the relation in which representations stand to objects: we
283
experience objects in – not through – our representations of them. We

do not have representations, in contrast to experiencing objects; rep-
resentations are nothing other than the subjective conditions of our
cognition of objects.
Our focus in the aesthetic case on the appearing of the phenome-
non for us makes possible a reflection on the process of mental activ-
ity as it takes up something given in experience. Something is received
by intuition, held together by imagination and, potentially, unified by
understanding. Aesthetic form thus displays the characteristics of the
first and second elements of cognitive synthesis (as outlined in the A
edition ‘Deduction’) and the preparatory relation in which they stand
to the third and properly cognitive element.10 This process that is the
precondition of any experience, but usually invisible, is now rendered
visible. Aesthetic judgement thus makes the affective condition of
experience available for reflection.11
The appearing of the appearance that is revealed in aesthetic judge-
ment is not mere semblance (Schein), but rather the relational status
of the object.12 An aesthetically pleasing thing brings to the fore the
ontological status of appearances in general, as objects that are avail-
able for our cognitive faculties. The aesthetic phenomenon affects us
and we feel a pleasure in it which is focused on the presentation of the
thing. We are aware of how an object affects us, but only at the level
of representation per se. Thus the possibility of representation, not
the functional effectiveness of any particular representation, becomes
available as a topic in the aesthetic case.
Aesthetic judgement thus reveals phenomenologically what I have
previously argued philosophically.13 Our powers of representation,
the faculties, do not constitute a screen between ourselves and the
world: they are our modes of discovering the world. And the world
faces us as a task and not as revelation. The aesthetic judgement pre-
sents us with a perceptual epiphany or sudden manifestation of the
meaningfulness of that task in that one particular thing stands out as
being in harmony with our subjective faculties. This is the purposive-
ness of form of the aesthetic object, which is identified only insofar as
our response to it counts as a harmony of the faculties, the form of
purposiveness.
II Not all perceptual forms are aesthetic forms

If an object prompts our power of judgement so that it is exercised
purely as an aesthetic appreciation, then how are we to distinguish
284
beautiful objects from any other empirical object? If it is the percep-

tual form of the object that is in question, why do not all empirical
objects give rise to aesthetic judgements, as they all have perceptual
forms?14 In this section I examine the identity of aesthetic form and
discuss how it can be distinguished from other perceptual forms.15
Breaking the link between perception and aesthetics would lead to
two connected problems regarding judgement’s place in the critical
system and consequently for its status as a distinctive form of judge-
ment. These technical problems would also put in question the valid-
ity of the phenomenological description of aesthetic apprehension
that I have developed through an interpretation of Kant. The tran-
scendental status of aesthetic judgement and the reason it merits a
Critique devoted principally to it, arises from its place within the
system of cognitive powers. This place is earned by the status of
judgement as contributing to the general project of establishing the
possibility of a priori synthetic knowledge.16 It is thus that it con-
tributes to the project of ‘cognition in general’. It is difficult to see
how it could do so unless aesthetic judgement stands in some relation
to the object and our capacity for synthesis of objects at the empiri-
cal level.
While the first problem regards judgement’s systematic status
within the range of cognitive faculties, the second concerns the specific
relation in which judgement stands to reason and morality. While
this topic cannot be addressed in any detail here, it is important to
position my account of the relation in which aesthetic judgement
stands to cognition within a wider perspective of reason in general.17
Judgement, considered as a power in its own right, secures the status
of the third Critique insofar as it acts as mediator between cognition
and morality. Aesthetic judgement in particular reveals that moral
purpose is, at least in principle, applicable within the domain of
nature. Judgements of beauty achieve this by giving a particular,
though indeterminate, presentation of an idea within the phenomenal
realm.18 But aesthetic judgements are only capable of doing so, and
only tacitly, because we can presuppose a fit between mind and nature,
that is, there is a purposiveness of nature for our judgement.19 I have
argued that aesthetic judgements present, in an exemplary fashion, the
possibility of empirical synthesis. Beauty thus serves as a concrete pre-
sentation of the relation within which mind stands to nature and gives
hope that even our higher faculty of reason might successfully be
applied within the empirical domain. If aesthetic and perceptual form
were not intrinsically linked, this would not be possible.
285
But what is the form of the object that is at issue in aesthetic

judgement? Is it the form of intuition? Allison argues in Kant’s
Transcendental Idealism that the form of intuition is distinct from the
formal intuition characteristic of geometry, and which is, in fact,
a hybrid of intuition and understanding. He contends that the
‘Transcendental Aesthetic’ is principally concerned with the form of
the intuited, while conceding that the latter can be traced back to the
form of intuiting.20 Thus, following Allison’s account, if the form of
the aesthetic object coincided with the form of intuition, it would
count as the form of the intuited.
If an aesthetic judgement were prompted merely by the indetermi-
nate form in intuition, then it would entail responding to an intuition
in isolation from any conceptual activity. This would conflict with
Kant’s dualism, that is, his commitment to a plurality of mental activ-
ity characteristic of all experience, falling back into an empiricist posi-
tion with an attendant myth of the given.21 Intuitive form can only be
part of the story, not least because of the analytical status of the argu-
ment in the ‘Transcendental Aesthetic’. The form of an object arises
from the cooperation of intuition with understanding. Consequently,
aesthetic awareness of form is more than a mere affect, being a reflec-
tion or judgement. Were this not the case it could hardly give insight
into the conditions – subjective or otherwise – of cognition, which
necessarily entails the combination of an intuition with a concept
arising from understanding.
While we have seen that aesthetic judgement counts as formal
insofar as it is a reflection on the purposiveness in intuition, Kant more
carefully says that the reflection in question is on intuition ‘to bring it
to some concept or other’.22 The perceptual form at issue in an aesthetic
judgement is not to be understood as the form of the intuited taken in
isolation from our power of concepts, but rather as the experiential
form that arises from the combined operation of the form of intuiting
and the form of conceptualising, the understanding. The beautiful
object has a perceptual form that like every other perceptual form
arises from a cooperation of intuition with understanding.23 However,
in this case the cooperation is brought to our attention because the phe-
nomenon resists determination – that is, cognitive conclusion.
Insofar as a particular phenomenon invites reflection on the sub-
jective conditions of cognition and yet resists cognitive conclusion, it
exhibits the purposiveness of nature for our judgement. We become
aware of the formative process by which the mind stands in relation
to objects in the world and in which judgement plays a principal role.
286
Thus we are aware of the perceptual form of the object as arising from
a process involving both intuition and understanding. As we saw in
the last chapter, Kant says that aesthetic judgement involves a
schematisation of the faculties. What he means is that whereas in
determinate judgement a concept subsumes an intuition, in aesthetic
judgement we become aware of the subjective necessary conditions of
that result. We become aware of this in a singular judgement about a
particular object and thus the exhibition of the purposiveness of
nature is not direct and can only be exemplary.
As we saw in the last section, aesthetic judgement pauses at the
point where imagination has prepared for cognition, but has not yet
achieved it. The imagination is primary here because only it can
sustain the delicate balance that stops short of epistemic resolution.
Imagination both links and holds apart intuition and understand-
ing.24 The form of a beautiful object is the correlate to this imagina-
tive activity and displays a level of integrity and a predisposition to
cognitive resolution without any knowledge actually arising. The
pleasure necessarily associated with the recognition of beauty sustains
the freedom of the mind and the quality of apprehension that are the
prerequisites of aesthetic attention. The imagination is not only the
agent of this balance, but also its subject. In an aesthetic judgement
we can become aware of the mediating role of imagination that is the
precondition for any experience whatsoever.
The object that comes to be judged as beautiful has a form, just as
does any other object, as a result of the cooperative activity of our
faculties. However, in this particular case, the form of the object is
conducive to a play of the faculties. Something about the object gives
rise to a particular proportion of the faculties. Kant makes this point
in Section 21, where he argues that a cooperation of the faculties
occurs when an object induces the imagination and the proportion of
the faculties that ensues depends on what he calls a difference
between objects.25 Only certain objects give rise to a proportion of the
faculties that counts not just as cooperative, but also as harmonious.
What we discover at the level of perception in such cases prompts us
to attend further not only to the object but also to the formative
process that was the condition of its coming to be. Not every object
has a perceptual form that is conducive to aesthetic reflection on the
subjective conditions of cognition. The process of formation, that is,
the relation between mind and object, is displayed in the spatio-
temporal form of these objects insofar as the latter prompt our reflec-
tion on the cognitive process in general.
287
A consideration of Kant’s example of the grotto of Antiparos will

help us clarify the relation between perceptual and aesthetic form.
Shortly before the end of the Critique of Aesthetic Judgement there is
a discussion of free formations of nature or the process through which
fluids take on form. When water freezes, it becomes solid through
crystallisation. The switch between a fluid form and a solid one is not
a gradual transition but rather occurs ‘as it were by a leap’.26 This
transition is strictly mechanical, that is, we need only the laws of
mechanical nature to describe its occurrence and yet some such trans-
itions give rise to forms that we consider beautiful:
Many such mineral crystallizations, e.g., spars, hematite, and aragonite,
often result in exceedingly beautiful shapes, such shapes as art might
invent; and the halo in the grotto of Antiparos is merely the product of
water seeping through layers of gypsum.27
The shape emerging from crystallisation ‘varies in accordance

with what difference in kind there is in the matter (Materien), but is
exactly the same wherever the matter is the same’.28 This comment
echoes his earlier remark on the significance of the difference
between objects for the distinctiveness of the proportion characteris-
tic of aesthetic judgement.29 In this passage Kant reinforces his initial
point saying that different snow-figures – that is, forms – arise
‘depending on what the particular mixture of air is at the time’.30 He
believes this can be accounted for by the theory of ether. Leaving
aside his attachment to a scientific theory no longer tenable, this
passage is significant for the distinctiveness of aesthetic form. As in
the initial passages of the ‘Transcendental Aesthetic’ Kant is com-
mitted to a material given, and as in Section 21 of the third Critique
that given can be differentiated.31 The shape or form of free forma-
tions differs insofar as there are differences in matter. But is this not
to say that transcendent matter shapes our conception of the form, a
position that would run contrary to the Copernican revolution’s shift
of perspective from material given to formal conditions as the con-
dition of experience? Surely, Kant’s transcendental idealism entails
that all differences in matter come from the mind as the initiator of
form? It would appear, however, that – at least in the case of free for-
mations – form is not merely imposed on the material given but
arises in relation to it. I have been arguing throughout this book
that mental form is a mode of access to the material given. The idea
developed in Chapter 3 is that matter is only ever experienced
empirically, yet our experience arises from a condition that is not yet
288
determined. This is a relational or dynamic account of the form–

matter relation.
The specific significance of the relation between form and matter
for our purposes here is that it allows us to see how only some per-
ceptual forms qualify as aesthetic. Kant believes that merely mechan-
ical phenomena can be beautiful and that only some, arising from
certain mechanical circumstances, will be so. It is contingent which
ones are beautiful. Indeed, Kant makes it a defining characteristic of
beauty that it arises contingently. Phenomena that display purpose do
not qualify as beautiful. To say that beautiful objects are characterised
by ‘purposiveness without purpose’ is to say that though they are
merely contingent, it is as if they arose as the result of design. Some
perceptual forms just are beautiful. This is an ineliminable feature of
our experience of the world. The contingency of beauty means that
we have to await its event. This event will typically arise as something
that stands out from the normal range of perceptual forms. It is diffi-
cult to see how beauty could qualify as an exemplary exhibition if it
were not differentiated in this way.
A direct pleasure arising from the appearance itself and with no
additional agenda – or, at least, without one that is determining – is
the necessary mark of a perceptual form that qualifies as aesthetic. We
cannot tell in advance whether pleasure will arise or not. Once a beau-
tiful object appears, a pleasure necessarily arises in response to it.32
We recognise it as beautiful through this feeling of pleasure. But there
is no way of knowing in advance of experience what will give rise to
aesthetic pleasure. Only a retrospective and phenomenological analy-
sis reveals the necessity of the pleasure associated with a particular
object. Thus, the distinctiveness of aesthetic from other perceptual
forms must also await experience.
In his discussion of Antiparos, Kant further brings out the contin-
gent distinctiveness of beautiful things, suggesting that it is as if the
stalactites had been formed by human intervention, that is, by art,
even though we know this is impossible. In fact it is because we know
that they came about naturally that their similarity to artworks gives
rise to a pleasure. They are so beautiful that we can hardly warrant
they arose without some purpose or other. The analogy with art
brings out once again that the distinctive pleasure arising from beau-
tiful phenomena is marked by the appearance, but only the appear-
ance, of their being motivated by a purpose.
While a number of objects may strike us as not yet determined as
to their identity or purpose, only some do so in such a way that, at
289
the same time, they appear to be conducive to our cognitive activity

in general. These exceptional objects or phenomena encourage rather
than thwart reflective activity. And yet this reflective activity does not
lead to a cognitive resolution. Moreover it is only in an aesthetic mode
of attention that our toleration of the suspension of a solution is
accompanied by a pleasure. The hiatus in cognition is the result of the
simultaneous holding apart and relating of our capacities for taking
in the world in intuition and reflecting on it with our understanding.33
Our feeling of pleasure in the phenomenon encourages our capacity
to sustain an open-minded attitude.
III The primacy of presentation

An additional problem that arises from the link Kant makes between
perceptual and aesthetic form must be faced before returning to a
further exploration of the appearing of the appearance. If aesthetic
form is a species of perceptual form, then what of other non-formal
features of phenomena? Are colour and tone, or indeed the sheer
materiality of things, not candidates for aesthetic appreciation? In
addressing this important problem, I develop Kant’s account in a way
I believe is compatible with the spirit and logic of his systematic
account. While I agree with Allison that Kant’s account of aesthetic
form should be more inclusive, I do not agree that we have to break
the link between perceptual and aesthetic form in order to achieve
this.34
My approach would be to insist that any aesthetic phenomenon
must be presented in space and time and thus has some perceptual
form, although that form may well be indeterminate or ill-defined.
Even the most inchoate mass of stuff would have to be presented for
our attention and this would entail that it has some spatio-temporal
extension in space and time, some affect on our senses and stands in
some relation to other things and to our cognitive faculties.35 And
while a particular aesthetic response may be principally concerned
with qualitative rather than quantitative features, it must still present
features such as colour and tone in space and time. This is what I call
the ‘primacy of presentation’. Aesthetic phenomena, whatever their
focus, draw our attention to the appearing of the appearance and to
this extent highlight the formative process through which the object
comes to figure as a phenomenon for the subject. Contrary to the con-
clusion Kant sometimes, although not consistently, draws, quantita-
tively formal qualities are not exhaustive of the range of aesthetic
290
possibilities. Nevertheless, some presentation in space and time is

unavoidable.
But it may be objected that although presentation in space and time
is a necessary condition of aesthetic apprehension, spatio-temporal
form is not determining for aesthetic affect. Extension in space and
time could be an initiating necessary condition and yet remain exter-
nal to the specifically aesthetic quality of an apprehension. Although
I cannot develop my idea of the primacy of presentation extensively
here, I would argue that the aesthetic status of a judgement requires
that presentation in space and time become part of the elaboration of
the aesthetic phenomenon.36 Space and time are co-opted into the
specifically aesthetic quality of the beautiful. The spatial extension of
the canvas is no longer merely a material condition of there being a
painting: the canvas becomes the spatial field within which the aes-
thetic phenomenon appears. And I think it is unimaginable that there
could be an aesthetic phenomenon without such a co-opting and
transfiguration of space and time. Artworks and natural aesthetic
phenomena not only can, but must make us experience space and
time differently.
If aesthetic form is understood as the appearing of the appearance
and not as perceptual form in general, design need not be set in oppo-
sition to features such as colour. As Merleau-Ponty says in ‘Eye and
Mind’:
It is simply a matter of freeing the line, of revivifying its constituting
power; and we are not faced with a contradiction when we see it reappear
and triumph in painters like Klee or Matisse, who more than anyone
believed in color. For henceforth, as Klee said, the line no longer imitates
the visible; it ‘renders visible’; it is the blueprint of a genesis of things.
Perhaps no one before Klee had ‘let a line muse’.37
Space and time at a deeper level are the conditions of the possibility
of the appearance of things and in aesthetic phenomena these condi-
tions become part of the topic rather than merely an initial condition.
This is the case, even if the topic is now predominately the phenom-
enon of colour.
Paul Klee’s watercolour painting ‘Southern Gardens (Tunisian
Gardens)’ invites us to see the way in which line and colour can coop-
erate so that each renders visible the other.38 This and others painted
in the same period present a patchwork of colours expressive of the
Mediterranean ambience of North Africa, representing, it would
appear, irregular plots of land intermittently planted with palm
291
trees.39 The lines, as the limit points of the patches of colour, are not
determinate boundaries, but are defined by the patches of colour they
demarcate. The vibrant colours rely on the indeterminately drawn
lines for the visibility of their appearing to us as patches of colour.
Design and colour are mutually enhancing complementaries and not
contraries. In this aesthetic example, form and content cooperate in
a way that is exemplary of the more general relation between form
and matter.
The insistence I have placed on the formative process rather than
the formed object goes hand-in-hand with the way in which I have
emphasised the relation in which form stands to matter. Form is antic-
ipatory of synthesis with matter. Kant’s formal idealism entails that
form stands in a necessary reciprocity with material. If Kant had
developed his account of formal idealism more explicitly, he would
have been in a position to show that attention to aesthetic form need
not be seen as excluding concern for the material in aesthetic phe-
nomena. If form stands in a dynamic relation with matter, then aes-
thetic response is not only to the formal conditions of experience, but
also potentially to its material possibilities. Appreciation of a display
of colour in nature or in art, for instance, can reveal the relation in
which the judging subject stands to a given phenomenon, and the
dependence of that relation on the combination of sensory affect and
reflective response. Unfortunately for the destiny of Kant’s philo-
sophical project, his account of the reciprocity of form and matter
was insufficiently developed. There are a few exceptions, however,
and we have seen one in his discussion of the grotto of Antiparos.
I have argued that the aesthetic phenomenon reveals the possibility
of empirical synthesis in an exemplary fashion insofar as a particular
empirical object is peculiarly accessible to our faculties and their activ-
ity of synthesis.40 But are there aesthetic phenomena that do not appear
within space and time? Any aesthetic phenomenon appears in time. A
painting takes time to look at. A sculpture requires that we take several
perspectives on a plastic form and this requires time. But do all aes-
thetic phenomena appear in space? Does a sonata or a prelude occupy
a position in space? I believe that it is arguable that they do, insofar as
hearing a sound, and especially a musical sound, entails that we are
within a certain range of it. And although spatial position is not the
primary content of an auditory aesthetic phenomenon, it is not only a
necessary condition of it, but is co-opted within the aesthetic acoustic
experience. The particular aesthetic effect of a sound makes use of the
spatial relation in which we stand to that sound.
292
Nevertheless, might there not be aesthetic phenomena that stand

in no relation whatsoever to the senses? Examples of hidden or invis-
ible works of art to be found in ancient cultures and also among con-
temporary artworks, raise an important question. For instance, in the
Neolithic passage tomb at Newgrange outside Dublin intricate
designs have been placed on the upper side of the ceiling, buried into
the surrounding mound.41 Jochen Gerz, a contemporary artist,
orchestrated a work where the names of 2,146 Jewish cemeteries in
use until 1933 in Germany were inscribed on the underside of the
same number of paving stones in front of Saarbrücken castle, the seat
of the provincial parliament.42 Although this started as an illegal artis-
tic action, the authorities later sanctioned it and the Schloßplatz
was renamed Platz des unsichtbaren Mahnmals (‘The Square of the
Invisible Monument’). If the artwork is positioned so as not to be
seen, heard or otherwise accessed by the senses, must we not conclude
that there is no sensory appearance and thus no spatio-temporal
form? Although I believe such examples force us to extend our per-
spective to the limits of sensory experience, any non-sensible phe-
nomenon will stand in some relation to a possible sensory experience
in imagination. Here imagination is the surrogate for the senses.
Are not these artistic phenomena, however, instances of the
sublime, a topic that has not yet been addressed? I believe that this
may, indeed, be the best way of accounting for the aesthetic affect
achieved by hiding something visual, or even just of suggesting that
something visual is hidden. However, the role of the sublime in Kant’s
analysis of aesthetic apprehension can only be understood as a con-
trastive move, always standing in relation to the beautiful. In the
experience of the sublime the logic of the senses breaks down, but the
lack of a sensory resolution for a phenomenon or, more strictly, a
mental state, can only be felt as a lack. The imagination, failing to find
a harmony between our understanding and our senses, turns to
reason and our capacity for thinking the infinite. But the sublime does
not occupy the rational perspective; it turns towards it through the
intermediary power of imagination. Imagination unavoidably strad-
dles the middle ground between the world of the senses and the per-
spective of pure intellect. Thus the sublime cannot escape a reference
to the senses, even though the relation is a negative or contrastive
one.43
While these remarks are merely suggestive of how I would address
a number of very important issues that arise from the strong link I
draw between aesthetic appreciation and perception, I hope that they
293
will be helpful in showing how Kant’s aesthetics could be further

developed. Kant gives us a rich starting point that allows for the
recognition of the centrality of the aesthetic within experience: we
can critically build on this in ways that Kant himself could not have
imagined.
IV The Anticipation of the Aesthetic Object

Aesthetic judgement reveals not only the combination of a plurality
of faculties necessary for any cognitive judgement, but also the rela-
tion in which the subject stands to an object. The aesthetic thing
encourages a judgement that is exemplary of and yet stops short of
cognition. Beautiful things give rise to a pleasure that arises contin-
gently and yet is necessarily connected with the phenomenon under
consideration. I have argued for this account of aesthetic judgement
in the first two sections of this chapter. I will now develop my account
to show how aesthetic form reveals the dependence of judgement on
the givenness of an object in experience.
We have seen that in Kant’s account of sensible affect at the outset
of the ‘Transcendental Aesthetic’, it emerged that we are capable of
being affected through our power of sensibility, which makes possi-
ble receptivity to the given. Only if such an affect has taken place, is
there anything to take up in sensibility and determine under a
concept.44 When we find something beautiful we dwell on our intu-
ition of it in association with a feeling. Our capacity for understand-
ing has some part to play, as Kant tentatively suggests at the outset
of the ‘Analytic of Beauty’.45 However, the concept that is evoked
is a merely indeterminate one of beauty or purposiveness without
purpose. Intuition is oriented towards the possibility of conceptual
determination, but necessarily falls short of an epistemic conclusion.
This allows us to become aware of the givenness of the object. It is
exactly because of the failure to achieve cognitive determination, that
the aesthetic judgement makes us aware of this, while at the same time
it is capable of revealing the structure of ‘cognition in general’. The
failure or weakness of aesthetic judgement from the cognitive per-
spective is the necessary condition of its success in acting as a dual
reflection on an object and our activity of thought, without resolving
either into the other.
Intuiting while aiming at a concept that is always absent is a form
of anticipation, not of knowledge of the object, but rather of the pos-
sibilities presented by the object. We anticipate these possibilities by
294
exploring the object in our perception of it, envisaging various possi-

ble aspects or developments of the phenomenal presentation. This is
the way in which it makes sense to talk about aesthetic apprehension
as an anticipation of the object.
But not only do we anticipate the aesthetic object once we are faced
with it, we can do nothing but anticipate it in advance. The aesthetic
event is contingent. We are always in the position of awaiting the aes-
thetic object and anticipating how it might (next) appear. The con-
tingency of the event of the beautiful – both within and prior to
aesthetic apprehension – allows us, while in the aesthetic mode, to
glean an awareness of how every experiential judgement is an antici-
pation of something that must be given to finite beings such as us.46
Once again, the aesthetic offers a heightened example of a general
condition of experience.
The contingency of the aesthetic event has further important ram-
ifications for the relation between theory and aesthetic judgement. We
cannot define in advance which properties the object must have in
order that it qualifies as beautiful. Even ruling out that it could be geo-
metric – or regular – simply does not capture the way in which beauty
must be discovered by us and cannot be achieved by following a
formula.47 We have to await the event of beauty.
However, might it not be objected that cognition also awaits the
given and we can nevertheless specify the properties that are charac-
teristic of an object? Indeed I have argued that a priori knowledge is
best understood as an anticipation of the given. Yet Kant presents the
categories of the understanding as the form of an object in general,
that is as the necessary descriptors of quantity, quality, relation and
modality. The givenness of the object does not, it would seem, exclude
the specification of the necessary characteristics of an object. Why
should this not also be the case for the aesthetic object? But, as beauty
gives rise to a free play of the faculties there is no objective criterion
for aesthetic judgement and no rules that prescribe how or what an
aesthetic phenomenon must be. This means that the aesthetic phe-
nomenon must be an event our response to which cannot be deter-
mined in advance. In contrast, in the epistemic case rules are given in
advance. Even here, though, a specific cognition requires that we
await experience to see if anything conforms to those rules.
As a result of the lack of a determining basis for aesthetic judgement,
it would be futile to seek out instances of beauty in the way that we
seek knowledge. It is true that we can put ourselves in a position where
we are likely to experience beauty, be that by going to a spectacular
295
natural environment or to a gallery where we know there to be works

of great aesthetic interest. But we cannot guarantee that just because
we are in a propitious location that we will experience aesthetic feeling.
Indeed the harder we try to control the result, the more likely we are
to fail. Beauty cannot be taken up as a goal, in contrast to cognition.48
In the aesthetic case, theory can only be a reflective anticipation or
a reconstruction of the given. There is no catalogue of rules or prop-
erties of the beautiful. This also means that we cannot determine the-
oretically which perceptual forms will count as beautiful and which
will not. We can be confident that beautiful objects will stand out
from the norm and that they will give rise to the mental freedom char-
acteristic of aesthetic judgement. Aesthetic judgement and perceptual
form are inextricable from one another, but they are not identical.
Only experience will reveal those perceptual forms that qualify as
aesthetic forms and those that do not.
V Contrapuntal Exemplarity
Aesthetic response makes use of the subjective conditions of cognition
in the absence of a cognitive conclusion. Removing the end of cogni-
tive activity transforms the latter into something quite different. So
while we might suspect that Kant’s account renders aesthetic judge-
ment merely a component part of cognition, this is not the case.49
Aesthetic judgement facilitates a reflection on the possibility of cogni-
tion, but this is only possible because the latter does not complete the
cognitive synthesis. The lack of a cognitive end is not to be understood
as failure, but rather as a resistance to what would otherwise be
natural, namely, to bring a synthetic activity to some conclusion. The
aesthetic standpoint interrupts cognition in such a way that not cog-
nition, but its possibility becomes available for inspection.
Aesthetic judgement reveals the form of judgement. This is why
Kant remarks at the outset of the ‘Analytic of Beauty’ that he has used
the ‘Table of Judgements’ to guide his investigation.50 We can draw
out his comment concluding that his investigation is not, however,
based on the ‘Metaphysical Deduction’, but rather uncovers the con-
ditions of the possibility of the latter, that is, the faculty of judgement
per se. It is even less the case that the analysis of beauty could have
been based on the ‘Table of Categories’, because aesthetic judgements
are not derived from the conditions of determinate judgement. They
are, however, derived from the general form of judgement and display
this capacity as a pure activity. This form is the basis not only of
296
judgements in general, but also of our possible knowledge of objects,

as the ‘Transcendental Deduction’ first argued. An aesthetic judge-
ment stands in relation to a phenomenon as appearing and thus is
concerned with the presentation of a thing in space and time. The
intentional object of a judgement of beauty reveals the relation
between subject and object that makes judgements of quantity,
quality, relation and modality possible.
Recognition of the relational quality of aesthetic apprehension
reveals why mere resistance to cognitive determination would not
suffice.51 The beautiful invites our cognitive powers and suggests that
we may come to a happy resolution of our cognitive quest, yet at the
same time, ultimately resists any determinate resolution of it. Only
some objects combine this quality of inviting cognitive activity while
resisting a final resolution of that synthetic process. Aesthetic and
cognitive judgements illuminate one another by contrast, as in a con-
trapuntal keyboard composition where the left and the right hands
develop contrastive independent melodies, creating a complex and
reciprocally reinforcing harmony.52
We have seen that reflection is a condition of the possibility of cog-
nition insofar as it is necessary for empirical cognition, and that the
harmonious reflection characteristic of aesthetic judgement reveals
the mutual relation of the capacities necessary for any cognition
whatsoever. This is the reflective process that is the subjective side of
synthesis. Aesthetic judgement allows us to become aware of our
general capacity for reflection. Now the question is: if cognition is
internally structured so as to depend on reflective activity, is it con-
ceivable that we could exercise our cognitive capacity in the absence
of a capacity for reflecting on the possibility of cognition? And if there
are alternative possible species of reflection on our cognitive power
within experience, do they share the structural form of aesthetic
judgement?53 I will merely suggest that for finite beings such as our-
selves, an absence of the possibility of self-reflection, even at this
highly general level, is unimaginable. I am sure that our grasp of the
possibility of cognition in general is always indistinct and rarely
recognised as such, but, although I cannot argue the case here, we
must surely have the capacity for such an uncanny awareness of the
possibility of thinking as a process. I will offer a provisional answer
to the second question by considering the possibility that the power
of judgement could be exercised in a non-aesthetic guise.
Kant claims that the only pure exercise of judgement is to be found
in aesthetic judgement. But is this right? Surely it is at least possible
297
that we become aware of the activity of judgement on other occa-

sions. For instance, in thinking about a philosophical problem, we
can become aware, not so much of the problem, but of the process of
thought through which we approach that problem. We focus not on
the contents of thought, but on the very process of thinking and
become aware of the appearing of thinking to itself. We will almost
certainly struggle to capture this fleeting glimpse of what appears inef-
fable. Such moments are probably brief and, as they tend to occur
when we are deeply involved with a problem we wish to solve, we
most likely move on to the work in hand. But surely what we are
aware of in such cases is the play of different mental orientations in
thinking. Our reflective powers cooperate freely without being
anchored in a goal. And what is striking about such an insight is that
our minds do not merely wander from content to content, but remain
at the level of the form of thinking as an activity. But does such an
insight count as aesthetic? If not, we have discovered at least one pure
expression of judgement other than the aesthetic case. I would
venture to suggest that such a relation to the thinking process is aes-
thetic in the restricted sense that we are concerned only with the
appearing of thinking to itself.
Leaving aside the difficult – perhaps unanswerable – question of
the status of such introspection, I would suggest that even if it is
impossible to conclude definitively that Kant is right in claiming that
aesthetic judgement is the only exercise of pure judgement in and for
itself, it is at least recognisable as one such exercise. Aesthetic judge-
ment is characterised by its lack of a determinate end and indeed even
by the lack of a desire for such an end. This is not to say that aesthetic
judgement must be wholly void of any cognitive content. This is a
point that Kant did not sufficiently develop. What is necessary is only
that an aesthetic phenomenon cannot be resolved into a cognitive
conclusion. Thus, it is quite feasible for us to respond to a natural or
artistic phenomenon aesthetically, while knowing certain things
about it. I can know which mountains I am looking at, where they are
relative to my own position and that they are dangerous to traverse
during winter. I do not have to forget this information, although I may
abstract from it, in appreciating their beauty. Indeed some informa-
tion – for instance, that they are dangerous – might contribute in an
indirect way to my aesthetic pleasure. All that is necessary is that the
cognitive information does not determine my pleasure. This is espe-
cially important when we consider artworks that are informed by sci-
entific enquiry – be it a work by Leonardo or a modern installation.54
298
In responding aesthetically, our powers of reflection are at play in a

dual sense. At the first level we are using our power of thought in
response to a given affect. This is the characteristic cooperation neces-
sary for any judgement whatsoever, including any cognitive judgement.
At a second level, we reflect on our reflective and affective activity. But
this second level is not a meta-level where we stand outside and look
on. Aesthetic reflection only arises in that we are engaged with an
empirical given object. This is reflection that is necessarily intertwined
with experience. Even in the speculative case I just developed, where
the pure exercise of judgement arises within the context of a philo-
sophical project, it only arises insofar as we are already engaged with
something else.55 Reflection on reflection can only be glimpsed at the
limits of our thought. For this reason it is always liminal and fleeting.
Aesthetic phenomena command our attention and in so doing allow us
to glimpse the usually unfocused on process of attention. Aesthetic phe-
nomena – both natural and artistic – allow us to experience the process
of thinking necessary for any experience whatsoever, but only in the
midst of our involvement with a particular empirical given.56
VI The supersensible harmony of the faculties

I have argued that aesthetic judgement reveals the necessary cooper-
ation of a plurality of capacities as the subjective condition of cogni-
tion. I have also argued that, for Kant, this reveals the way in which
we have access to extra-mental objects in the world. A further hurdle
for my insistence on a pluralist model of mental activity is introduced
at the end of the ‘Analytic of Beauty’. In the ‘Dialectic’ of the Critique
of Aesthetic Judgement Kant introduces the idea that there is a super-
sensible harmony of the faculties on which aesthetic harmony is ulti-
mately grounded. As the supersensible is usually associated with our
moral capacity, does this mean that the plurality revealed by aesthetic
judgement is ultimately grounded in our moral capacity? If this were
so, a pluralist model of mind would ultimately resolve into a complex
monism. In contrast to the impositionalist model within which under-
standing is the dominant mental faculty, this alternative model would
be one in which practical reason presides.57 This result would corre-
spond with the prima facie plausible interpretation of Kant’s position
as entailing the primacy of moral reason in orienting the relation in
which we stand to the world.
Whereas it may appear that the Dialectic makes aesthetic judge-
ment dependent on the moral supersensible, Kant persistently refers
299
to the idea of a purposiveness of nature for judgement as the super-

sensible specifically related to taste.58 The supersensible is tripartite
and consists of three ideas that are ultimately derived from the three
higher faculties of understanding, reflective judgement and reason.
The three ideas are those of nature in general, the purposiveness of
nature for our judgement and our moral freedom.59
Thus when Kant ultimately traces the validity of aesthetic judge-
ment to the supersensible, we need not conclude that he is committed
to the view that the aesthetic is derived from the moral, for the aes-
thetic also has a place in the supersensible. The harmony of the
faculties experienced in aesthetic judgements refers back to the
cooperation between a plurality of faculties at the deepest level of our
rational being. The supersensible structure of reason is not simply
moral, but plurally constituted by three cooperating faculties giving
rise to distinctive, but compatible principles.
Nevertheless the priority given to the moral within the supersensi-
ble seems to put the autonomy of aesthetic judgement at risk.
The morally good is the intelligible that taste has in view . . . for it is with
this intelligible that even our higher cognitive powers harmonize, and
without this intelligible contradictions would continually arise from the
contrast between the nature of these powers and the claims that taste
makes.60
Although the mere relation between the aesthetic and the super-
sensible does not commit Kant to the view that the former is grounded
in morality, it would appear that his account of the internal dynam-
ics of the supersensible finally leads to that conclusion.
But what exactly is Kant’s committment here? He says that the
higher cognitive powers harmonise with the morally good and only
thus are intelligible contradictions avoided. But this is not to say that
the aesthetic supersensible is grounded in the moral, but rather that
the former aims at the latter as a goal. Only insofar as the plurally
constituted mind is capable of aiming at the morally good as an ideal,
can we hope our faculties will not lead us towards mutually incom-
patible goals. My interpretation of the ideal status of morality takes
inspiration from Kant’s philosophy of history, where he argues that
the only hope for an eventual realisation of our moral goals within
the world – and thus for a harmony between morality and nature –
is that a moral end is adopted as the ideal goal by which we orient
our practice, where the latter is understood in a broader sense than
that of the purely practical or moral.61 Taking the ideal status of
300
Kant’s historical teleology as a clue, we can conclude that ideally, but

only ideally, the faculties in their diversity tend towards a moral ideal.
This does not mean that the aesthetic can be derived from the moral,
but rather that the moral idea is only approached through a cooper-
ation of different orientations and aesthetic harmony is preparatory
for moral harmony. Schiller develops Kant’s position, arguing that the
aesthetic is the step that makes the moral possible within experi-
ence.62 Aesthetic judgement is a bridge between reason and under-
standing, in that taste prepares for the realisation of morality within
experience. Beauty signals the possibility of a link between reason and
nature in particular aesthetic cases. The achievement of a moral world
or kingdom of ends would require a generalisation of this result and
for Kant this remains strictly an ideal.63
Kant goes on to remark:
And because the subject has this possibility within him, while outside
[him] there is also the possibility that nature will harmonize with it, judg-
ment finds itself referred to something that is both in the subject himself
and outside him, something that is neither nature nor freedom and yet is
linked with the basis of freedom, the supersensible, in which the theoret-
ical and the practical power are in an unknown manner combined and
joined into a unity.64
It is difficult to give a clear interpretation of the text at this point, but

we can at least be confident that the bridging role played by aesthet-
ics depends on the way in which taste makes possible a relation
between the freedom within us and the nature that lies outside.
Aesthetic judgement gives an exemplary exhibition of the possibility
in principle that our rational projects could be in tune with external
nature. If aesthetic judgements did not reveal the purposiveness of
nature for our judgement, they would be incapable of providing a
transition between reason in its narrow, or moral, role and under-
standing, the faculty that makes possible unification of objects within
mechanical nature.
I can now reinforce the link made in the last chapter between the
purposiveness of nature and aesthetic judgement. Aesthetic judge-
ment is capable of linking the theoretical and practical orientations
just because of the possibility that nature will harmonise with taste.
This possibility of the purposiveness of nature in general is actualised
in the purposiveness of beautiful objects. It is thus that aesthetic
judgement is capable of making the link between what is subjective
or ‘within’ and what is objective or ‘outside’.
301
Taste is the exemplary exhibition not only of the relation between

understanding and intuition, but of the system of reason in its broad
sense. Aesthetic harmony captures the dynamic cooperation charac-
teristic of all rational activity, even at the supersensible level. The
dynamic is now oriented towards an ideal resolution, which can never
be achieved within experience.65 For beings such as ourselves,
dualism or, as we can now see, more precisely, the plural orientation
of our mental makeup cannot be overcome. What counts as knowl-
edge for us starts with something given to us in appearance and only
brought to determinate form through a combination of the faculties.
Aesthetic judgement itself arises out of a complex combination of fac-
ulties and is, at the same time, a specific exemplary exhibition of the
general makeup of our minds and the relation in which we stand to
the world within which we find ourselves.
VI Form and World

Aesthetic judgement is a response to an empirical object and simulta-
neously a revelation of the conditions of possibility of that response.
It is the only point in Kant’s system where the transcendental and the
empirical almost coincide, yet they remain distinct for we are not
faced with a concrete presentation of the general framework.66 The
beautiful gives us a glimpse of the general form of empirical synthe-
sis crystallised in a singular ungeneralisable instance. And although
our awareness of the beautiful object need not be merely formal, the
resultant insight into cognition remains formal as it is only through
the harmony between the form of the object and the form of our
mental activity that this epiphany arises. In what follows I will suggest
that this formal insight reveals the dynamic relation between form
and matter, but this is not to suggest that we have access to matter per
se. The aesthetic phenomenon reveals the formal conditions of access
to materiality; any other considerations of materiality (for instance,
colour or tone) are aspects of our aesthetic appreciation of the par-
ticular thing at the phenomenal level.
Aesthetic judgement finally shows how our minds are capable of
getting at something outside of ourselves in the world by revealing the
process of cooperation that makes possible receptivity to the given.
Although this cooperation arises from our subjective faculties, it is
not a privatively subjective activity. It begins with our capacity to be
affected by something outside us through sensibility. However, this
affect is nothing for us if it is not combined with reflection. As we have
302
seen, in the first Critique Kant sometimes suggests that all intuitions
must be subject to the determining power of understanding. I have
argued that even in these passages we should interpret Kant’s position
as entailing that it must be possible for all intuitions to be taken up
by the understanding, but not that they are in fact necessarily subject
to determining synthesis. This leaves open the option that we could
have intuitions of which we are aware and yet which do not give rise
to knowledge. I now intend to show how this is crucial for the rela-
tion between form and matter characteristic of formal idealism.
At the outset of the ‘Transcendental Aesthetic’ of the Critique of
Pure Reason, Kant all too briefly established that our experience
begins with an affect on our sensibility. I argued in Chapter 3 that this
affect must be understood as arising from an empirical material
object. In Chapter 6 I argued that Kant’s account of the legitimation
of the categories culminates in the Principles’ claim that they apply to
what he calls experience in general and finally, in the Analogies, to the
material givenness or existence of objects. While at first Kant is satis-
fied to show that the categories must be combined with the form of
intuition, he concludes the ‘Analytic’ by requiring that they must
apply to empirically given things. Most notably in the ‘Refutation of
Idealism’, he asserts that these are spatial things. What, if anything,
has the Critique of Aesthetic Judgement to add to this hierarchical
account of the conditions of knowledge?
The aesthetic object is an empirically given object. It is one that is
particularly conducive to the activity of our faculties necessary for
any cognition whatsoever. As an empirical object, it is a composite of
form and matter. In Kant’s account of the grotto at Antiparos it tran-
spires that the material conditions of this beautiful phenomenon play
a role in the emergence of its form. This is surely suggestive for the
general case of beauty. What is distinctive about a beautiful form in
contrast to other perceptual forms is that the beautiful thing invites
our attention and a response, that is, our synthetic activity. The com-
bination of material affect and form that is the condition of any
experience whatsoever is, in this case, particularly congenial to our
cognitive activity. This thing is not only easy on the eye, but also con-
ducive to our reflection: the aesthetic event is distinguished by the one
necessarily leading to the other.
Kant, unfortunately, focuses disproportionately on the formal
qualities of the phenomenon, but his account does not preclude a
more dynamic relation between form and matter. Indeed, as I have
argued in previous chapters, his formal idealism requires a reciprocal
303
relation between form and matter in which both play distinct though
necessary roles. The aesthetic phenomenon displays our access to an
object that is formed by our mental activity and yet is only con-
structed out of a material given. This is the ‘appearance itself’ of the
second Analogy, which is neither a thing in itself nor a mere repre-
sentation. But in the aesthetic case, the objectivity of the thing is not
the issue. The beautiful thing could under other circumstances be con-
sidered as an independently existing object displaying extensive and
intensive magnitude and standing in relation to other things. But in
this case we are not interested in considerations of objectivity. We
suspend the goal of knowledge in a way that is comparable to the phe-
nomenological epoché.67 We are concerned not so much with achiev-
ing cognition as with observing something, which indirectly gives us
access to the subjective activity that makes cognition possible. And
aesthetic judgement does not merely reveal a harmony of the faculties
that are the subjective conditions of cognition: it also shows how
those subjective faculties access the empirically given object. Our
reflection on the beautiful thing is intertwined with a sensory appre-
hension of the qualities of the thing. Aesthetic reflection only arises in
conjunction with aesthetic apprehension. In this way aesthetic judge-
ment not only displays the twin conditions of perception, but also
deepens our awareness of how they cooperate with each other in
response to a material given.
An aesthetic phenomenon not only turns our attention to the form
of the appearing thing, but at the same time reveals that form as our
mode of access to a material given. The event of intentional conscious-
ness – that is, of form taking up matter – is displayed as an event. I
argued in Chapter 3 that the event of affection is only experienced in
relation to reflection. When an intuition is determined by a concept,
affect arises only as sensation, that is, as an inseparable element of our
experience of an empirical object. I have suggested in this case affect
should be viewed as result. But if the intuition is not determined by a
concept, the affective element becomes available for inspection in dis-
tinction not from our capacity for reflection, but from a determining
concept. In this case, the material dimension in empirical experience
becomes available as the affective moment or the event of affectivity. A
merely indeterminate intuition would not achieve this. It is only an inde-
terminate intuition that in harmonising with the conditions of cognition
in general gives rise to a pleasure that is capable of revealing the affec-
tive component of empirical experience. It does so by showing how a
material given is taken up by our subjective formative capacities.
304
Aesthetic judgement finally reveals the structure of experience as

understood by formal idealism. In an aesthetic judgement I am con-
cerned with the form of the object. This form arises first as the form
of the intuited, an organisation in space and time. We saw in Chapter
3 that intuitive form arises insofar as our sensibility, the form of intu-
iting, introduces an order that allows us to take up empirical matter.
This complex operation counts as ‘affect’. There must be a material
given, thus preventing Kant from falling into material idealism. But
the form of the object also arises from our capacity for reflection, the
understanding. The latter makes the unification of the affect under a
concept possible. In aesthetic judgement there is no unification of the
phenomenon, but the imagination nevertheless holds together the
given in such a way that would under other circumstances prepare for
determination, that is, cognition. We thus have a capacity for making
sense of things prior to determination under a concept.68 The combi-
nation of the affect and a process of unification, falling short of unity,
give rise to the form of the object, though not yet a determinate form
or formal intuition.69 Aesthetic form is the crystallisation of the
process of synthetic activity, balanced between sensory affect and
conceptual determination. As such it is neither one nor the other and
symbolises their process of cooperation. Aesthetic judgement’s focus
on the form of the object throws us back on the synthetic activity, now
experienced in a heightened form as a ‘harmony of the faculties’. It is
thus that the form of an object we find beautiful is able to uncover the
possibility of cognition in general, that is, how it is that we are able
to have access to the world of things through the exercise of our
formal subjective capacities. Form, or the formal subjective faculties
that make forms possible, opens up our access to the empirical mate-
rial world.70 Form is not imposed on the world: it is the form of the
world. But, in this case, the world cannot be an alien collection of
things in themselves: the world is a world of things that stand in rela-
tion to us. Our environment is this relation.
Notes
1. CJ, AA 223, in Section 13. In Section 14 he develops this idea further.
2. See, for instance: Guyer, KCT, pp. 220–3 and ‘Formalism and Theory
of Expression’, pp. 59–61.
4. CJ, AA 289. See discussion of the ‘Deduction of Taste’ in Chapter 5
(p. 197).
305
5. See Chapter 6 (pp. 229–37) on the implicit spatial schematism in the

‘Principles’.
6. Earlier chapters establish each of the component parts of this account.
7. CPR, A 19/20, B 34.
8. A 89–90, B 122–3.
9. Just as they could not have any affect on us at all, had we not a power
of sensibility or receptivity.
10. They are preparatory, not only because they are prior in order of pre-
sentation, but because they provide the form for an apprehension,
which would, if it were determined, count as knowledge. As it does not
so qualify, the apprehension counts as indeterminate or underdeter-
mined.
11. See the final section of this chapter (pp. 302–5)for further discussion of
how the relationship between form and matter is revealed by aesthetic
judgement.
12. This is much closer to what Merleau-Ponty calls the ‘in-itself-for-us’,
than are cognitive ‘appearances themselves’ discussed in Chapter 6
(pp. 226–7). See Merleau-Ponty, Phenomenology of Perception, p. 322,
French edition, p. 372: un véritable en-soi-pour-nous.
13. See Chapter 3 (pp. 96–105) on the status of representations as our point
of access to objects.
14. For a discussion of the problem arising from the link Kant makes
between perceptual and aesthetic form, see Paul Guyer, Kant and the
Claims of Taste, especially p. 226; Donald Crawford, Kant’s Aesthetic
Theory, especially pp. 96–110. See also John H. Zammito, The Genesis
of Kant’s Critique of Judgment, pp. 119–21; and Henry Allison, Kant’s
Theory of Taste, pp. 132–5.
15. It is perhaps surprising that Allison accepts Guyer’s conclusion that
there should be no link between perceptual and aesthetic form, given
the distinctiveness of their interpretative approaches. Nevertheless
Allison’s agreement with Guyer about perceptual form arises from
an independent conviction that aesthetic judgement is grounded in
the subjective and not the objective conditions of cognition. See
Chapter 5 (pp. 198–9) on Allison’s interpretation of the ‘Deduction
of Taste’.
16. See Chapter 4 (pp. 151–60) where I argue that aesthetic judgement
qualifies as synthetic.
17. See also the discussion of the relationship between the faculties at the
supersensible level on pp. 299–302 below.
18. In the first instance, the idea is aesthetic, but Kant holds that the latter
can serve as a symbol of a moral idea. See CJ, AA 351–4.
20. Allison, KTI, p. 97. See discussion of Allison’s account of the forms of
intuition in Chapter 2 (p. 69). Contrast Longuenesse’s reading of the
306
‘Transcendental Aesthetic’ as already directed at formal intuition also

discussed in Chapter 2, p. 73.
21. For a discussion of this issue, see McDowell, Mind and World, for
instance, pp. 5–9.
22. See CJ, AA 232′ and AA 220′ respectively. This precision was discussed
in the previous chapter.
23. Even an undetermined, not merely an underdetermined, intuition would
stand in a potential relationship with our capacity for reflection. This is
necessitated by Kant’s dualism. See Chapter 3 (pp. 92–4).
24. Sarah Gibbons makes a similar observation in her Kant’s Theory of
Imagination, p. 29.
25. CJ, AA 238. See discussion in Chapter 5 (pp. 181–3). These two claims
count as Steps 3 and 4 respectively of my reconstruction.
26. CJ, AA 348.
27. AA 349. In his aesthetic appreciation of mineral forms and crystallisa-
tion, Kant anticipated a popular theme in German Romanticism. See,
for instance, Novalis’ use of chemical metaphors in his ‘Miscellaneous
Writings’, 95, reprinted in Kathleen M. Wheeler (ed.), German aesthetic
and literary criticism, p. 92. Later Stendhal uses the idea of crystallisa-
tion to explain the formation of feelings, in particular, love. See De
l’amour, I, Ch. 1, p. 5.
28. AA 348.
29. AA 238; discussed in Chapter 5 (p. 181).
30. AA 349.
31. See Chapter 3 (pp. 87–9) on the materiality of the given and Chapter 5
(p. 181) on the differentiation among the given.
32. This is the sense in which the liking for an aesthetic object is a necessary
one, as Kant argues in the fourth moment of the ‘Analytic of Taste’. See
AA 236–7.
33. Ian Hamilton Finlay’s bridge comprised of two identically inscribed
planks at ‘Little Sparta’ expresses this hiatus in an aesthetic form. See
the discussion in the Introduction (pp. 1–2).
34. Allison, KTT, especially pp. 135–8.
35. This is to say, that it will stand in some relation to all four categories
and to their correlates in the Axioms, Anticipations, Analogies and
Postulates.
36. A discussion with Michael Podro helped me clarify this issue. In a
related but distinct context, see Podro, Depiction, p. 13, where he talks
of ‘the recruitment of the paint surface’ as a way of sustaining recogni-
tion of the subject of the painting.
37. Merleau-Ponty, ‘Eye and Mind’, The Merleau-Ponty Aesthetics
Reader, p. 143. The final phrase comes from Michaux’s Aventures
de lignes. The French is ‘laissé rêver une ligne’: L’Oeil et L’Esprit,
p. 74.
307
38. Paul Klee, ‘Southern Gardens (Tunisian Gardens)’, 1919, watercolour

and pen on paper, 24.4 x 18.7/ 18.9 cm, The Berggruen Klee Collection,
Metropolitan Museum of Art, Paul Klee: catalogue raisonné, vol. 3
(1919–22), p. 86.
39. See also ‘Tunisian Gardens’ (1919), ibid., vol. 3, p. 87 and ‘View from
St. Germain (Tunis) Looking Inland 1’ (1918), ibid., vol. 2 (1913–18),
p. 440.
40. See Chapter 7 (pp. 260–9).
41. See Claire O’Kelly, ‘Passage-grave art in the Boyne Valley’, particularly
pp. 363–4.
42. Jochen Gerz, with the participation of students from Hochschule für
Bildende Kunst, Saarbrücken ‘2146 Steine – Mahnmahl gegen
Rassismus’ (‘2,146 stones, monument against racism’), 1993,
Saarbrücken, Germany. See Stewart Martin, ‘A New World Art?
Documenting Documenta 11’, particularly p. 12.
43. I will argue in my Afterword that beauty stands in a reciprocally neces-
sary relation to the possibility of the breakdown of harmony and order,
expressed most explicitly in experiences of the sublime.
44. These are the general conditions of cognition. See Chapters 3 (p. 94) and
4 (pp. 120–47).
45. CJ, AA 203: ‘ . . . rather, we use imagination (perhaps in connection
with understanding) . . .’
46. I come back to this in the final section, pp. 302–5.
47. I thus derive a rather different conclusion from his position than does
Kant himself. See AA 242–3, where he suggests that regularity would
not qualify as worthy of aesthetic appreciation.
48. Although setting cognition as a goal does not result in its end being com-
pletely within our control.
49. See Avner Baz’s article, ‘Kant’s Principle of Purposiveness and the
Missing Point of (Aesthetic) Judgements’, for the view that aesthetic
judgement is, for Kant, a variety of cognition.
50. CJ, AA 203.
51. Gasché appears to suggest this in The Idea of Form, p. 16.
52. An example would be the last prelude of the first book of Bach’s Well-
Tempered Klavier.
53. Clearly within philosophy our cognitive power can be analysed. But my
question concerns reflection on the possibility of cognition at a phe-
nomenological level.
54. One such contemporary artwork is Richard Wilson’s ‘20:50’, 1987
(Saatchi Gallery, London). The title of this work concerns the specific
density of recycled engine oil, which floods a gallery space. The aesthetic
affect of this work, which has a dramatic reflective quality, is not dimin-
ished by the information the title supplies about the material properties
of oil.
308
55. It could also arise in the context of scientific, for instance, mathematical
reflection or even alongside a practical concern.
56. Kant expresses some hesitation about the purity of an aesthetic judge-
ment in response to an artwork. However the artwork still plays a
central role in his account of aesthetic pleasure.
57. In this section I am only concerned with the significance of the relation
between the aesthetic play of the faculties and supersensible harmony
for my position that aesthetic judgement reveals a plurality of orienta-
tions, which is necessary for experience.
58. See in particular CJ, AA 340, 347–8, 350, 351, 353.
59. AA 346.
60. AA 353.
61. See Kant, ‘Perpetual Peace’ (On History), ‘Appendix I: On the
Opposition (Mißhelligkeit) between Morality and Politics with respect
to Perpetual Peace’ and ‘Appendix II: Of the Harmony (Einhelligkeit)
which the Transcendental Concept of Public Right establishes between
Morality and Politics’, AA Band VIII, 370–86. Harmony only arises in
the face of the distinction between the political and moral orientations
and through the adoption of the latter as an ideal and thus, necessarily
absent, goal for the former. Kant focuses exclusively on how political
practice should be oriented towards the moral, but I believe his account
has a wider relevance for experience in general.
62. Friedrich Schiller, On the Aesthetic Education of Man. See, for instance,
p. 131, Nineteenth letter, paragraph 6: ‘beauty can become a means of
leading man from matter to form, from feeling to law, from a limited to
an absolute existence’.
63. See ‘The End of All Things’, in Kant, On History, pp. 69–84. AA Band
VIII 325–39.
64. CJ, AA 353 (Pluhar’s addition).
65. This is the message of Kant’s essay, ‘The End of All Things’.
66. Kant does not claim that the aesthetic phenomenon counts as an Idee in
Hegel’s sense. The sensory given expresses the universal conditions of
cognition, but only indirectly.
67. I argue elsewhere that Kant’s account of aesthetic judgement shares a
project with Merleau-Ponty’s version of the epoché, which operates in
the interests of uncovering the relation between subject and object.
Although Husserl’s classical version aims to achieve this, arguably even
in his later works he remains trapped in the perspective of the subject.
See my ‘Kant’s Phenomenological Reduction?’.
68. See the discussion of a potentially pre-determinative unity or synopsis
in Section 2 of the A Deduction in Chapter 4 (p. 123).
69. For Kant, unity entails determination under a concept. I am highlight-
ing the way in which unity arises out of a process of unification. When
this process arises without reaching a conclusion, then we ‘hold
309
together’ sensory impressions. This is sufficient for the figurative syn-

thesis of an aesthetic form, but not for a formal intuition.
70. I am not suggesting, however, that we have direct access to the world as
such. Our access is to things that necessarily appear against the back-
drop of the world. See discussion in my ‘Kant’s Phenomenological
Reduction?’.
310
Afterword
I would like to add a few further comments on the character of

harmony in aesthetic and cognitive judgements. I will also tentatively
suggest a way in which judgements of the sublime have a significance
not only for judgements of beauty, but also for Kant’s epistemolog-
ical project.
It is easy to conclude that Kant’s account of aesthetic judgement
underestimates the extent to which the disharmonious plays a role in
our experience. In twentieth century and contemporary art it would
be fair to say that the disharmonious holds priority over the harmo-
nious. This raises questions about the continuing relevance not only
of Kant’s aesthetics but also of his theory of knowledge. For if I am
right in arguing that aesthetic judgement presents an exemplary exhi-
bition of cognition in general, then it might appear that Kant’s
account of the cognitive relation between mind and world suggests
much too unproblematic a ‘fit’ between the subject and the object.
The importance of this is that were Kant’s position to amount to the
view that the mind and world simply stand in harmony with one
another, he would be in severe danger of falling back into something
approaching a ‘pre-established harmony’. I will leave to one side the
question of whether this would amount to the mind imposing order
on the world or vice versa. In either case Kant would be guilty of
falling back into the sort of metaphysics he was intent on avoiding.
Were Kant committed to the view that mind and world immedi-
ately harmonise with one another, this would run counter to his
conviction that we are governed by moral goals, that although imper-
fectly exercised by us, nevertheless count as categorical. Kant’s insis-
tence that morality is an idea that is not to be found in the
mechanically ordered world of nature, and yet which can be intro-
duced into that world as an ideal goal, simply would make no sense
were our minds and the world already standing in harmony with one
another. Our capacity for moral reason allows us to go beyond the
finite world of objects in thought. A possible disharmony between
mind and nature is thus equally necessary in the moral sphere, as
311
is a possible harmony in the cognitive domain. The disharmony

between the senses and reason is expressed aesthetically in the
sublime, especially in its dynamic mode.
The only question I will address here is that of the epistemic relev-
ance of the claim that Kant is committed to a harmony between mind
and world. Kant’s formal idealism has emerged as comprising a series
of stages of determination. The subjective capacities are the sources
of formal structures that allow us to take up given objects in the
world. But our subjective faculties do not stand in harmony with the
world; they are simply the conditions that make possible our knowl-
edge of the world. Knowledge is achieved when the subjective repre-
sentation of the world matches the objective state of affairs. Kant has
revealed that subjective representation and objective states of affairs
necessarily imply one another, and if they did not, he argues, we
would have no chance of knowing the latter. This, broadly, is the
message of the Copernican Revolution. But Kant does not merely
state dogmatically that objects fit with our cognitive powers, he
shows how this is possible through a process of synthesis. This is the
work of the ‘Transcendental Deduction’, the ‘Schematism’ and the
‘Principles’ chapter. Knowledge a priori has emerged as a task, which
I have suggested marks even the achievements of empirical knowl-
edge. The formal structures supplied by the mind anticipate the mate-
rial given so as to make knowledge possible. When and if knowledge
is achieved, then what we can call an epistemic harmony can be
glimpsed. But this knowledge will be of a particular empirical state of
affairs, situated among a variety of such states, some of which have
and some of which have not yet been determined. Even a firm piece
of knowledge will only be partial because our knowledge of the world
always occurs as a system, as the Analogies revealed. And as the cer-
tainty or harmony we have achieved occurs within a context that has
not been fully determined, then the system of knowledge is never
experienced as a totalised harmony. In Sartre’s terms there are totali-
sations, but not totalities.1 The possibility of a harmony of mind and
world may be a working presupposition of the cognitive project, but
it has to be combined with our need to discriminate and differentiate,
tendencies that work against the goal of totality but in the interests of
knowledge.2
But what of the aesthetic case where it appears that Kant is com-
mitted not only to a harmony between the faculties but also one
between mind and world? It is necessary to look more closely at what
Kant means by ‘harmony’ in aesthetic judgement. As I have argued,
312
Afterword
aesthetic judgement reveals the synthetic activity of judgement in

general. Imagination holds a phenomenon together as if it were exer-
cising a rule of the understanding, that is, as if it were determining an
object. But the thing is not viewed as an object of knowledge. It is
apprehended through a free or open combination of different orien-
tations to the world, intuition or apprehension, understanding or a
broadly reflective power and imagination. This ‘harmony of the fac-
ulties’ may appear, for lexical reasons, to signal a rather comfortable
accommodation within the subject and between subject and world.
But the meaning of ‘harmony’ is not self-evident and has to be gleaned
from the way in which it is used in the text. Harmony is a ‘play’ of
the faculties. Now as children know very well, play is a serious and
varied activity. When the faculties are in free play, they do not merely
meld into one another. To say that there is a play is to say that there
is an exchange and an engagement, even a communication, while
insisting that all this activity occurs through the mediation of imagi-
nation. We have already seen that even in the cognitive case the imag-
ination is the mediating faculty par excellence. But in the aesthetic
case, the imagination comes into its own. Play is activity that
embraces its lack of a final resolution and thus stays in the domain of
the imagination.
A harmony of the faculties is not an identity relation, no more is it
a bland compromise. It is rather a tonic exchange where the lawfulness
of the understanding combines with the receptivity of intuition without
either side taking priority over the other. This harmony is one within
which there is difference and yet communication. An aesthetic har-
mony already has gaps within it, even before Kant moves on from the
beautiful to the sublime. The pleasure of an aesthetic phenomenon is of
something that holds together even though it is not bound together.
And this is very important for the cognitive resonance of aesthetic
harmony. Aesthetic judgement is exemplary of the possibility of cog-
nition insofar as it displays synthesis in process and not synthesis as
conclusion. The process or project of knowledge is anticipated in the
aesthetic instance, revealing the viability of aiming at cognition, but
giving us no grounds for epistemic complacence.
The harmony of the faculties is a heightened case of the cooper-
ation of the faculties necessary for any cognition whatsoever.
Cooperation depends on the distinctiveness of the two terms that, in
this case, work together. If sensibility and understanding genuinely
cooperate with one another then it must at least be possible that their
joint purpose could break down. This would be true both in the case
313
of the failure of a cognitive aim and when the aim is other than cog-
nitive. For Kant, morality entails the possible overcoming of the
senses by reason. This discord of the harmonies is displayed in moral
agency and in our aesthetic feeling of the sublime. Thus, the harmony
of the faculties is only one pole of a range of possible relations in
which the faculties can stand to one another and in which mind can
stand to the world. At the other pole is the discord characteristic of
the sublime. And neither of these poles is pure. Harmony, as we have
seen, entails gaps and disharmony is never wholly discordant with our
experience. Between these extremes is the diversity of proportional
relations in which the faculties stand to one another. Judgements of
beauty and of the sublime symbolise the range between harmony and
disharmony within which our experience emerges.
The sublime is not simply external to the beautiful: the two species
of aesthetic judgement stand in an internal relation to one another, as
I will now suggest. The sublime is the moment when the senses fail to
make sense of the world and in this moment the faculties of imagina-
tion and understanding stand in disharmony with one another.3 The
possibility of such a disharmony is the shadow of disruption within
the beautiful. Harmony is an achievement, our pleasure in which
requires an awareness that this result is not automatic. A harmony
between the faculties and between mind and world – and we have
seen that, for Kant, these are necessary corollaries of one another –
can only emerge against the background of a possible disharmony.
Indeed complex harmonies employ disharmony, just as Kant remarks
that the artist is able to transform an ugly scene into a beautiful artis-
tic presentation. Thus the sublime is not a mere parergon to the main
account of aesthetic judgement, although it may count as such in a
deeper sense, where the possibility of disruption is necessary to the
possibility of presentation.4
Ian Hamilton Finlay’s bridge at Little Sparta is not, I think, an
instance of the Kantian sublime. Our senses are not thwarted: they
amplify our reflection on a perceptual phenomenon. However, the
unity that we achieve emerges in and through difference. The differ-
ence within harmony marks the necessary possibility of the sublime
that haunts the beautiful.
Notes
1. I am thinking of the comment in his Critique of Dialectical Reason:
‘la totalité – contrairement à ce que l’on pourrait croire – n’est qu’un
314
Afterword
principe régulateur de la totalisation . . .’ (Sartre 1960, p. 138); ‘total-

ity, despite what one might think, is only a regulative principle of total-
isation . . .’ (Sartre 1976, p. 46).
2. Unity or affinity is only achieved through seeking both homogeneity and
variety. See ‘The Regulative Employment of the Ideas of Pure Reason’,
CPR, A 657, B 685.
3. For the purposes of this discussion, I focus only on the first disruptive
moment of the sublime and not on Kant’s resolution of the sensory
aporia in a turn to our capacity for moral reason.
4. Allison’s characterisation of the sublime as a parergon is not at all
similar to Derrida’s consideration of the same topic. Allison refers to
Derrida’s notion of the parergon in KTT, on p. 8. Derrida discusses this
notion in The Truth in Painting; see, in particular, pp. 37–82.
315
Bibliography
Allison, H. (1983), Kant’s Transcendental Idealism, New Haven, CT, and

London: Yale University Press (2nd rev. enlarged edn, 2004).
Allison, H. (1996), Idealism and Freedom: Essays on Kant’s Theoretical and
Practical Philosophy, Cambridge: Cambridge University Press.
Allison, H. (2001), Kant’s Theory of Taste, Cambridge: Cambridge
University Press.
Ameriks, K. (1982), ‘Recent Work on Kant’s Theoretical Philosophy’,
American Philosophical Quarterly vol. 19, no. 1, January: 1–24.
Ameriks, K. (1982), ‘How to Save Kant’s Deduction of Taste’, Journal of
Value Inquiry 16: 295–302.
Ameriks, K. (1983), ‘Kant and Guyer on Apperception’, Archiv für
Geschichte der Philosophie vol. 65: 174–86.
Ameriks, K. (1983), ‘Kant and the Objectivity of Taste’, British Journal of
Aesthetics vol. 23, no. 1, Winter: 3–17.
Ameriks, K. (1992), ‘Rudolf A. Makkreel, Imagination and Interpretation
in Kant’, Man and World 25: 227–34 [= ‘Review’].
Arendt, H. (1978), The Life of the Mind, Orlando, FL: Harcourt Brace and
Company.
Bauemler, A. [1923] (1967), Das Irrationalitätsproblem in der Ästhetik und
Logik des 18. Jahrhunderts bis zur Kritik der Urteilskraft, Tübingen: Max
Niemeyer Verlag.
Baz, A. (2005), ‘Kant’s Principle of Purposiveness and the Missing Point of
(Aesthetic) Judgements’, Kantian Review vol. 10: 1–32.
Berkeley, G. [1710] (1996) ‘A Treatise Concerning the Principles of Human
Knowledge’ in M. R. Ayers (ed.), Philosophical Works Including the
Works on Vision, London: Everyman.
Bird, G. [1962] (1973), Kant’s Theory of Knowledge, New York:
Humanities Press.
Buchdahl, G. [1969] (1988), Metaphysics and the Philosophy of Science:
The Classical Origins Descartes to Kant, Lanham, MD, and London:
University Press of America.
Buchdahl, G. (1992), Kant and the Dynamics of Reason, Oxford: Blackwell.
Burnham, D. (2004), Kant’s Philosophies of Judgement, Edinburgh:
Edinburgh University Press.
Caygill, H. (1989), Art of Judgement, Oxford: Blackwell.
Caygill, H. (1995), A Kant Dictionary, Oxford: Blackwell.
316
Bibliography
Crawford, D. (1974), Kant’s Aesthetic Theory, Madison, WI: The University

of Wisconsin Press.
Derrida, J. (1987), The Truth in Painting, trans. G. Bennington and
I. McLeod, Chicago and London: University of Chicago Press.
Dickens, C. [1837] (1953), The Pickwick Papers, London and Glasgow:
Collins.
Düsing, K. (1968), Die Teleologie in Kants Weltbegriff (Kantstudien
Ergänzungshefte, 96), Bonn: H. Bouvier & Co. Verlag.
Fricke, C. (1990), Kants Theorie des reinen Geschmacksurteils, Berlin and
New York: Walter de Gruyter.
Gardner, S. (1999), Kant and the Critique of Pure Reason, London:
Routledge.
Gasché, R. (2003), The Idea of Form Rethinking Kant’s Aesthetics, Stanford,
CA: Stanford University Press.
Gibbons, S. (1994), Kant’s Theory of Imagination, Oxford: Clarendon
Press.
Ginsborg, H. (1997), ‘Lawfulness without a Law: Kant on the Free Play of
Imagination and Understanding’, Philosophical Topics vol. 25, no. 1,
Spring: 37–81.
Guyer, P. (1977), ‘Formalism and the Theory of Expression in Kant’s
Aesthetics’, Kant-Studien 68: 46–70.
Guyer, P. (1979), ‘Review of Identität und Objectivität’, Journal of
Philosophy vol. 76, no. 2: 151–67.
Guyer, P. (1979), Kant and the Claims of Taste, Cambridge, MA, and
London: Harvard University Press.
Guyer, P. (1980), ‘Kant on Apperception and A Priori Synthesis’, American
Philosophical Quarterly vol. 17, no. 3: 205–12.
Guyer, P. (1987), Kant and the Claims of Knowledge, Cambridge:
Cambridge University Press.
Heidegger, M. (1962), Kant and the Problem of Metaphysics, Bloomington,
IN: Indiana University Press.
Heidegger, M. (1972), On Time and Being, trans. J. Stambaugh, New York:
Harper & Row. Originally published in 1969 as Zur Sache des Denkens,
Tübingen: Max Niemeyer Verlag.
Heidegger, M. [1929] (1973), Kant und das Problem der Metaphysik,
Frankfurt am Main: Klostermann.
Heidegger¸ M. (1978), Being and Time, trans. J. Macquarrie and E.
Robinson, Oxford: Basil Blackwell.
Heidegger, M. [1927] (1986), Sein und Zeit, Tübingen: Max Niemeyer
Verlag.
Henrich, D. (1969), ‘The Proof-Structure of Kant’s Transcendental
Deduction’, Review of Metaphysics 22: 640–59.
Henrich, D. (1989), ‘The Identity of the Subject in the Transcendental
Deduction’, in E. Schaper and W. Vossenkuhl (eds), Reading Kant: New
317
Bibliography
Perspectives on Transcendental Arguments and Critical Philosophy,

Oxford: Basil Blackwell, pp. 250–80.
Henrich, D. (1992), Aesthetic Judgement and the Moral Image of the World,
Stanford, CA: Stanford University Press.
Henrich, D. (1994), ‘Identity and Objectivity: An Inquiry into Kant’s
Transcendental Deduction’, in R. Velkley (ed.), The Unity of Reason,
Cambridge, MA: Harvard University Press. Originally published in 1976
as Identität und Objectivität, Heidelberg: Carl Winter Universitätsverlag.
Hughes, F. (1998), ‘The Technic of Nature: What is Involved in Judging?’,
in H. Parret (ed.), Kants Ästhetik/Kant’s Aesthetics/L’esthétique de Kant,
Berlin/New York: de Gruyter, 176–91.
Hughes, F. (2006a), ‘On Aesthetic Judgement and Our Relation to Nature:
Kant’s Concept of Purposiveness’, Inquiry vol. 49, no. 6, December 547–72.
Hughes, F. (2006b), ‘Taste as Productive Mimesis’, Journal for the British
Society of Phenomenology vol. 37, no. 3, October: 308–25.
Hughes, F. (2007), ‘Kant’s Phenomenological Reduction?’, Études
phénoménologiques vol. XXII, nos 43–4, January: 157–85.
Hume, D. [1739–40] (1978), A Treatise of Human Nature, 2nd edn, with
text revised and notes by P. H. Nidditch, Oxford: Clarendon Press.
Husserl, E. (2001), Analyses Concerning Passive and Active Synthesis:
Lectures on Transcendental Logic, trans. A. J. Steinbock, Dordrecht:
Kluwer, mainly drawn from Husserliana [=Hu] Band XI, Analysen zur
passiven Synthesis.
Husserl, E. (1950), Cartesian Meditations, trans. D. Cairns, The Hague:
Nijhoff; Husserliana [=Hu] Band I.
Kemp Smith, N. [1918] (1992), Commentary to Kant’s Critique of Pure
Reason, Atlantic Highlands, NJ: Humanities Press.
Kitcher, P. (1990), Kant’s Transcendental Psychology, Oxford: Oxford
University Press.
Kitcher, P. (1986), ‘Projecting the Order of Nature’, in R. E. Butts (ed.),
Kant’s Philosophy of Physical Science, Dordrecht: D. Reidel.
The Paul Klee Foundation (ed.) (1998–2004), Paul Klee: catalogue raisonné,
London: Thames and Hudson.
Llewelyn, J. (1982), ‘Heidegger’s Kant and the Middle Voice’, in R. Bernasconi
and D. Wood (eds), Time and Metaphysics, Warwick: Parousia Press.
Llewelyn, J. (1991), The Middle Voice of Ecological Conscience: A Chiasmic
Reading of Responsibility in the Neighbourhood of Levinas, Heidegger
and Others, London: Macmillan.
Llewelyn, J. (2000), The HypoCritical Imagination, London: Routledge.
Locke, J. [1689] (1975), An Essay Concerning Human Understanding, ed.
P. H. Nidditch, Oxford: Oxford University Press.
Longuenesse, B. (1998a), ‘The Divisions of the Transcendental Logic and the
Leading Thread’, in G. Mohr und M. Willaschek (eds), Immanuel Kant:
Kritik der reinen Vernunft, Berlin: Akademie Verlag.
318
Bibliography
Longuenesse, B. (1998b), Kant and the Capacity to Judge. Sensibility

and Discursivity in the Transcendental Analytic of the Critique of
Pure Reason, trans. C. T. Wolfe, Princeton, NJ: Princeton University
Press.
Lyotard, J-F. (1994), Lessons on the Analytic of the Sublime, trans. E.
Rottenberg, Stanford, CA: Stanford University Press.
Lyotard, J-F. (1991), Leçons sur l’Analytique du sublime, Paris: Éditions
Galilée.
McDowell, J. (1994), Mind and World, Cambridge, MA: Harvard
University Press.
Makkreel, R. (1990) Imagination and Interpretation in Kant, Chicago:
University of Chicago Press.
Makkreel, R. (1992), ‘Response to Guenter Zoeller’, Philosophy Today
Fall: 276–80.
Martin, S. (2003), ‘A New World Art? Documenting Documenta 11’,
Radical Philosophy 122, Nov/Dec: 7–19.
Meerboote, R. (1992), ‘Reflection on Beauty’, in T. Cohen and P. Guyer
(eds), Essays in Kant’s Aesthetics, Chicago: University of Chicago Press.
Melnick, A. (1973), Kant’s Analogies of Experience, Chicago: University of
Chicago Press.
Merleau-Ponty, M. (1945), Phénoménologie de la perception, Paris:
Gallimard.
Merleau-Ponty, M. (1962), Phenomenology of Perception, trans. C. Smith,
London: Routledge.
Merleau-Ponty, M. (1968), The Visible and the Invisible, Evanston, IL:
Northwestern University Press.
Merleau-Ponty, M. (1964), Le visible et l’invisible, Paris: Gallimard.
Merleau-Ponty, M. (1993), The Merleau-Ponty Aesthetics Reader:
Philosophy and Painting, ed. G. A. Johnson, Evanston, IL: Northwestern
University Press.
Merleau-Ponty, M. (1964), L’Oeil et L’Esprit, Paris: Gallimard.
Nietzsche, F. (1974), The Gay Science, trans. W. Kaufmann, New York:
Random House.
Nietzsche, F. [1887] (1988), Die fröhliche Wissenschaft, Kritische
Studienausgabe, Band 3, G. Colli and M. Montinari (eds), Berlin/New
York: De Gruyter.
O’Kelly, C. (1973), ‘Passage-grave Art in the Boyne Valley’, in Proceedings
of the Prehistoric Society vol. 39: 354–82.
O’Neill, O. (1984), ‘Transcendental Synthesis and Developmental
Psychology’, Kant-Studien 75, Heft 2: 149–67.
Pippin, R. (1979), ‘Kant on Empirical Concepts’, Studies in the History of
the Philosophy of Science vol. 10: 1–19.
Pippin, R. (1982), Kant’s Theory of Form, New Haven, CT, and London:
Yale University Press.
319
Bibliography
Pippin, R. (1995), ‘Avoiding German Idealism: Kant and the Reflective

Judgement Problem’, Proceedings of the International Kant Congress vol.
8: 977–97.
Pippin, R. (1996), ‘The Significance of Taste: Kant, Aesthetic and Reflective
Judgement’, Journal of the History of Philosophy vol. 34, no. 4, October:
549–69.
Podro, M. (1998), Depiction, New Haven, CT, and London: Yale University
Press.
Prauss, G. (1989), Kant und das Problem der Dinge an sich, Bonn: Bouvier
Verlag.
Sartre, Jean-Paul (1960), Critique de la raison dialectique, Paris: Gallimard.
Sartre, Jean-Paul (1976), Critique of Dialectical Reason, trans Alan
Sheridan-Smith, ed. Jonathan Ree, London: New Left Books.
Savile, A. (1987), Aesthetic Reconstructions: The Seminal Writings of
Lessing, Kant and Schiller, Oxford: Basil Blackwell.
Schaper, E. and Vossenkuhl, W. (1989), Reading Kant: New Perspectives on
Transcendental Arguments and Critical Philosophy, Oxford: Basil
Blackwell.
Schiller, F. [1795] (1967), On the Aesthetic Education of Man, ed. Elizabeth
M. Wilkinson and L. A. Willoughby, Oxford: Clarendon Press.
Stendhal, (1891), De l’amour, Paris: Calmann Lévy.
Strawson, P. (1966), The Bounds of Sense, London and New York:
Routledge.
Vaihinger, H. (1918), Die Philosophie des Als Ob, Leipzig: Felix Meiner.
Vaihinger, H. (1935), The Philosophy of ‘As if ’, trans. C. K. Ogden, London:
Kegan Paul, Trench, Trubner & Co. Ltd.
Vaihinger, H. [1876] (1976), Kommentar zu Kants Kritik der reinen
Vernunft: zum hundertjährigen Jubiläum derselben, New York: Garland.
Walker, R. C. S. (1978), A Selective Bibliography on Kant, 2nd edn, Oxford:
University of Oxford.
Wheeler, K. M. (ed.) (1984), German Aesthetic and Literary Criticism: The
Romantic Ironists and Goethe, Cambridge: Cambridge University Press.
Zammito, J. H. (1992), The Genesis of Kant’s Critique of Judgment,
Chicago: The University of Chicago Press.
Zoeller, G. (1992), ‘Makkreel on Imagination and Interpretation in Kant’,
Philosophy Today Fall: 266–75.
320
Index
aesthetic formalism see formalism canon, 213, 221, 238

aesthetics circle of representation, 49–78, 96–100,
aesthetic form see form 102, 107, 171, 227
aesthetic ideas see ideas ‘cognition in general’, 76, 152, 154,
definitions of ‘aesthetic’, 3–4, 190–3 174–5, 179–80, 183–4, 187–9,
affect, 55, 57, 61, 65–7, 69–72, 74–5, 194, 198–9, 201, 264–6, 269, 279,
78, 86–7, 95–107, 139, 149–51, 282, 285, 294, 297, 304–5, 311
171, 183, 200, 211, 214, 222–4, cognitive formalism see formalism
281, 290, 292, 294, 299, 302–5 common sense, 169, 178–9, 182, 184,
definition of, 108n 186–93, 196–7, 201, 248
Allison, H., 25–6, 32–3, 38, 49, 61–9, concepts see ideas
73, 75–8, 86, 89, 98–9, 107, constructivism see impositionalism
132–3, 169, 178–9, 189–94, Copernican revolution, 8, 38, 58, 62–3,
198–200, 253–9, 261, 281, 286 65, 86, 89–90, 92–7, 100, 104,
Ameriks, K., 30, 43n, 46n, 159, 135, 171–2, 181, 214, 220, 249,
167n–8n, 201n, 204n, 273n 257, 265, 267–8, 280, 288, 312
appearance (definition of), 92
appearances themselves, 226–7, 234, deduction
236 of aesthetic judgement, 169, 193–4,
appearing of the appearance, 197–9, 248
278–84, 290–1 objective, 5, 107, 169–74, 181, 256,
apperception, 20, 26, 29, 32–4, 127–9, 268
134–6, 144, 234 subjective, 5, 68, 169–201, 268
unity of, 12, 18, 20, 24, 29, 30, 34, transcendental, 5, 20–5, 27–30, 33,
127–9, 133, 135, 143–4, 254 62, 64, 68–9, 74, 91, 101, 112,
120–48, 152, 156–7, 160, 170–2,
Baumgarten, 190–1 176, 179–80, 207–21, 226,
beauty 236–40, 251, 257, 264–5, 268–9,
the beautiful, 6, 263, 280, 291, 293, 281, 297, 312
295–7, 302, 313–14 Descartes, 88, 229, 234
beautiful object, 268, 284–7, 289, design, 216, 281, 289, 291, 292
296, 301–4 dualism, 11, 34–6, 38, 59, 69–70, 73,
judgements of beauty, 160, 178, 193, 112, 114, 116, 121, 128, 131, 142,
285, 297, 311, 314 153, 160, 286, 302
Berkeley, 65, 87–8 Düsing, K., 255–6
Buchdahl, G., 38, 49–61, 62, 64, 67–8,
71, 78, 86, 89, 107, 207, 209, empirical guidedness, 10–15, 17, 39
251–2 event (of affection), 99, 103–4, 106–7,
Burnham, D., 168n, 273n 280, 289, 295, 303–4
321
Index
exemplary exhibition, 265–7, 269, Hegel, 17

277–305 Heidegger, 121–2, 127, 147–51, 157,
216
faculties Henrich, D., 8, 10, 25–35, 39, 132
faculty talk, 10, 13, 35, 37–9, 68–71, hidden art, 293
112, 116–19, 131, 172–3, 268 Hume, 65, 89
harmony of, 113, 118–19, 152, Husserl, 50, 54, 278
155–6, 175, 177, 183, 195, 260–3,
267–8, 284, 299–302, 304–5, idealism
312–14 formal, 5, 14, 19, 24, 62–3, 78,
mutual relation (or cooperation) 86–108, 115, 171, 183, 223,
of, 113, 117–51, 155, 160, 228–9, 235, 280, 292, 303, 305,
169–70, 172–6, 180–2, 184, 312
187–8, 192–3, 195–7, 201, 227, material, 5, 86–9, 96–7, 99, 104–5,
230, 232, 238, 259–60, 262–5, 115, 125, 171, 223, 229, 237, 280,
268, 282, 287, 299–300, 302, 305
313 ideas
plurality of, 78, 113, 118–19, 127, difference from concepts, 12–13, 15
130–2, 137, 146, 176, 178, 187, imagination, 72, 76–7, 121–2, 125,
192, 238, 281, 294 127, 130–6, 140, 146–50, 156–7,
Fichte, 57 214–15, 293, 313
Finlay, I. H., 1–2, 3, 6, 314 free lawfulness of, 76, 156, 282
form, 11, 24–6, 49, 52–6, 59–61, 63, productive, 76, 127–30, 136, 138,
65–7, 69, 70–5, 77, 86, 92–5, 140, 223, 230
97–100, 104, 103–7, 113, 115, synthetic power of, 75, 116, 127,
130, 135, 137, 228–9, 239, 138, 140–4, 146–51, 156–8, 160,
285–9, 292, 302–5; see also 215, 223, 227, 258, 264, 281–2,
matter 287, 293, 313
aesthetic, 71, 93, 104–5, 115, 278, transcendental synthesis of see
280–92, 294, 305 synthesis, figurative
of matter, 56, 61, 98 impositionalism, 5, 8–30, 33–6, 39, 60,
of thought, 32, 34, 70–2, 172, 63, 71, 77–8, 86–91, 93–4, 97,
213–15 100, 106, 114, 121, 130–1, 134,
formalism, 5, 8–39, 49–78, 86, 97, 136, 142, 145, 183, 220
131, 267, 279–80 intuition
aesthetic formalism, 4, 93–5 a priori, 13
Fricke, C., 272n–4n form of, 59, 73–6, 91, 96–8, 104,
107, 113, 120, 122–3, 127,
Gardner, S., 203n, 243n, 245 139–42, 144, 222, 236, 286
Gibbons, S., 110n, 163n, 202n, 242n, formal, 12, 73–4, 123, 139, 141–3,
307n 286, 305
Ginsborg, H., 274n
grotto of Antiparos, 279, 288–9, 292, judgement
303 aesthetic, 1–6, 76–7, 102, 113, 136,
Gurwitsch, A., 68–9 151–9, 169, 174–201, 207, 212,
Guyer, P., 8–13, 17–27, 29–30, 34, 36, 216, 249, 256–7, 259–62, 264–9,
38–9, 60, 71, 88, 95, 99, 107, 130, 277–305, 311–14; deduction of
179–80, 184–5, 193–4 aesthetic judgement see deduction
322
Index
of beauty see beauty objective deduction see deduction

cognitive, 77–8, 151–3, 156, 174, O’Neill, O., 161
176–9, 181–3, 186–8, 191, 197, orientation
199, 207, 277–8, 297, 299, 311 in empirical world, 252–3, 258
constitutive, 55, 60 plurality of different mental
determining, 15, 73, 77, 102, 112, orientations, 38, 78, 92, 112, 114,
151, 156, 177, 182, 237, 249, 258, 116, 118–19, 160, 229, 278–81,
280–1 298, 301–2, 313
empirical, 15, 49, 155, 248, 254,
258, 267 phenomenology, 54, 107, 278–85, 304
power of, 170, 194–9, 201, 252, Pippin, R., 8–17, 20, 38–9, 49, 59,
256–60, 262, 265, 284, 297 60–2, 64, 71, 88, 99, 107
reflective, 15–16, 77, 102, 118, 151–2, principles
155–6, 158, 177, 217, 240, 248–9, analogies, 50, 64, 124, 180, 211,
251–4, 257–62, 266, 268–9, 280–1 222, 224–8, 233–6, 239, 303–4,
of taste, 3, 145, 153, 170, 177, 183, 312
188–9, 193–8, 200, 257–8, 267, anticipations, 222–5, 233, 237
281 axioms, 222, 224, 232–3, 239, 240
teleological, 159, 250, 260, 262 categorical, 51, 249, 252, 254
constitutive, 51, 176, 269
Kemp Smith, N., 11 dynamic, 221, 224
Kitcher, P., 68 mathematical, 221–4
Klee, P., 291 postulates, 222, 224, 228–9, 232–6,
knowledge 239
anticipatory status of, 207–41, regulative, 51, 208, 225
266–7, 279, 295 purposiveness, 6, 198–9, 201, 240,
relation to cognition, 6n–7n 248–61, 265–8, 277, 280, 282,
284–6, 300–1; see also judgement
Leibniz, 28, 31 (aesthetic and reflective)
Llewelyn, J., 111n, 151, 164n–6n formal, 250, 252, 256, 259–61,
Locke, 38 264–5
Longuenesse, B., 49, 64, 69–78, 86, 89, general sense of, 256, 258
107, 122, 124, 127, 133, 140–2, objective, 250
148 of judgement, 249, 257–60, 262
of nature, 198, 201, 240, 248–50,
McDowell, J., 101, 118 252–69
Makkreel, R., 156–60, 261–2
matter, 14, 25, 54–6, 59, 60–1, 65–7, receptivity, 2, 56, 59, 61, 75, 86–7,
75, 77, 86, 98, 99–100, 103–5, 93–4, 96, 100, 104–7, 114, 118,
107, 137, 228–9, 288–9, 292, 147
302–4; see also form reflective judgement see judgement
Merleau-Ponty, M., 227, 278, 291 ‘Refutation of Idealism’, 17, 88, 149,
music, 292 229, 234–5, 237, 303
object ‘Schematism’, 13, 14, 33, 64, 149–50,

affective status of, 66–7, 96–106 157, 172, 208, 211–22, 226,
Kant’s understanding of, 58–9, 65, 229–38, 264–5, 281, 312
90–2, 97, 100–2 spatial, 207–9, 217, 229–38
323
Index
Schiller, 301 recognition (synthesis of), 126,

sensus communis see common sense 146
Strawson, P., 8, 10, 13, 35–9, 60, 68, reproduction in imagination
71, 88, 100, 107 (synthesis of), 20, 124–7, 137–8,
subjective deduction see deduction 141
sublime, 6, 293, 311–14 spatial, 141–2, 230–2
synthesis synthetic power of imagination see
a priori synthesis, 11–13, 18–19, imagination
21–2, 23, 24, 27–8, 30, 160, 221 temporal, 34, 141–2, 150, 231–2
apprehension in intuition (synthesis systematicity, 6, 198, 240–1
of), 73, 122–9, 137–47 empirical, 2, 48–69
empirical, 11, 20, 221, 249, 255, transcendental, 251–5
265, 269, 281, 285, 292, 302
figurative, 132–8, 142, 148, 152, transcendental deduction see
157, 172, 213–7, 222–3, 226–7, deduction
229–32, 236, 238 Trendelenburg, A., 18, 95
in process, 113, 154–5, 159–60, 169,
201, 208, 213, 216, 218, 232, 241, Vaihinger, H., 11
280, 283, 313
intellectual, 132–4, 146, 172, 209 Walker, R., 30
324

Kant's Aesthetic Epistemology - Form and World (2007)

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's Aesthetic Epistemology - Form and World (2007)

Uploaded by

Copyright:

Available Formats

Fiona Hughes is Lecturer in

Philosophy at the University KANT’S KANT’S AESTHETIC

KANT’S AESTHETIC EPISTEMOLOGY

She draws on phenomenological

Kant’s Aesthetic Epistemology

‘Music . . . is not an expression of what life is, but an expression

Daniel Barenboim. Reith Lectures 2006

In memory of George Elder Davie (1912–2007), student of Kemp Smith

FORM AND WORLD

Edinburgh University Press

© Fiona Hughes, 2007

Edinburgh University Press Ltd

ISBN 978 0 7486 2122 4 (hardback)

The right of Fiona Hughes

References to Kant’s Works viii

1 The Centrality of the Problem of Formalism 8

2 Formalism and the Circle of Representation 49

3 Formal Idealism and the Aesthetic Condition

4 The Deep Structure of Synthesis 112

5 The Completion of the Subjective Deduction in

6 A Priori Knowledge as the Anticipation of a Material

7 Empirical Systematicity and its Relation to

8 Aesthetic Judgement’s Exemplary Exhibition

Author/subject index 321

leave from teaching. In conjunction with sabbatical leave from the

References to Kant’s Works

All references to Kant’s works are to Kant’s Gesammelte Schriften

In ‘Little Sparta’, a garden in the Pentland hills outside of Edinburgh,

Kant’s Aesthetic Epistemology

Finlay’s work is particularly pertinent to establishing this role for

Aesthetics or the philosophy of the arts is practised seriously and to

Kant’s Aesthetic Epistemology

referring to any sensory experience whatsoever. In this book I will

Chapters 1 and 2 establish the centrality of cognitive formalism for

Kant’s Aesthetic Epistemology

was addressed in nuce in Chapter 4. Having established the role of

Erkenntnis has a much broader range than ‘knowledge’, I think it is jus-

The Centrality of the Problem of Formalism

What is the nature of Kant’s epistemological formalism, as his critics

The Centrality of the Problem of Formalism

guaranteed – a thesis to which he believes Kant is committed – then

Guyer contrasts this to the view that:

Thus the impositionalist interpretation of Kant attributes to him the

Kant’s Aesthetic Epistemology

independent existence of an extra-mental world accessed by a causal

I Pippin’s Critique of Kant’s Formalism

The Centrality of the Problem of Formalism

the thesis that the forms of experience must be applied to empirical

Kant’s Aesthetic Epistemology

our knowledge of it. Elsewhere Pippin further counters the challenge

The Centrality of the Problem of Formalism

be instances of subjective projection and this is just what Pippin

Kant’s Aesthetic Epistemology

The Centrality of the Problem of Formalism

Kant’s Aesthetic Epistemology

empirical concepts, as it is simply too subjective to render it capable

So while Pippin does not state that Kant’s position is impositionalist,

The Centrality of the Problem of Formalism

substantive or metaphysical foundation for pure intuition, concept,

II Guyer’s critique of impositionalism

Kant’s Aesthetic Epistemology

imposes this form.60 This is contrasted to what Guyer considers a dis-

The Centrality of the Problem of Formalism