Mind and World - John McDowell

MIND AND
·WORLD
• • •
With a New Introduction

• • •
John
McDowell
HARVARD UNIVERSITY PRESS
Cambridge, Massachusetts
London, England
Copyright <0 1994, 1996 by the President and Fellows
of Harvard College
All rights reserved
Printed in the United States of America
Fifth printing, 2000
First Harvard University Press paperback edition, 1996
Library of Congress Cataloging-in-Publication Data

McDowell, John Henry.
Mind and world : with a new introduction I John McDowell.
p. em.
Includes bibliographical references and index.
ISBN 0-674-57610-1
1. Philosophy of mind. 2. Knowledge, Theory of. 3. Concepts.
I. Title.
[BD 418 .3 M 37 1996]
.
121'.4-<lc20 96-22268
Contents
Preface vii
Introduction xi
Lecture I. Concepts and Intuitions 3
Lecture II. The Unboundedness of the Conceptual 24
Lecture III. Non-conceptual Content 46
Lecture N. Reason and Nature 66
Lecture V. Action, Meaning, and the Self 87
Lecture VI. Rational and Other Animals 108
Afterword
Part I. Davidson in Context 129
Part II. Postscript to Lecture III 162
Part III. Postscript to Lecture V 175
Part IV. Postscript to Lecture VI 181
Index 189
Preface
The main text of this book is a sort of record of the John Locke Lec
tures that I delivered in Oxford in Trinity Term, 1 991. I have done
some recasting of the lectures from the form in which I gave them. I
have tried to make improvements in clarity and explicitness. I have also
eliminated phrases like "next week" and "last week", which it seemed
absurd to leave standing in a version meant to be taken in through the
eye, perhaps-at least for the texts of the lectures-at a single sitting.
But apart from correcting one inessential falsehood at the end of the
last lecture, the texts I offer here, headed "Lecture I" and so on through
" Lecture VI", aim to say just what I said in Oxford.
They aim, moreover, to say it with a mode of organization and in a
tone of voice that reproduce those of the lectures as I gave them. There
are at least three points here.
First, even where I have made revisions at the level of phrases and
sentences, I have preserved the order of the lectures, as I delivered
them, at the level of paragraphs and sections. In particular, I have not
tried to eliminate, or even lessen, repetitiveness. I hoped that frequent
and sometimes lengthy recapitulations would be helpful to hearers,
and I hope they will be helpful to readers too.
Second, in a brief set of lectures, it seemed sensible to try to pursue
a reasonably linear train of thought, and I have not tried to make the
revised texts any less two-dimensional. The footnotes, in so far as they
go beyond merely bibliographical information, and the Afterword are
meant to give some indication of what a more rounded treatment of
these issues might look like. But they are no more than an adjunct to
the record of the lectures, more or less as I gave them.
viii Preface
Third, I have tried not to erase an unguardedness that seemed

proper for the lecture format.
I have many substantial debts to acknowledge.

Someone who read these lectures superficially might suppose Don
ald Davidson figures in them, after the first page or so, as an enemy. I
hope it is clear to less superficial readers, even from the texts of the
lectures themselves, that I single out Davidson's work for criticism as
a mark of respect. I define my stance against his by way of a contrast
that it would be easy to relegate to the edges of the picture, with mas
sive agreement in the centre. For my purposes in the lectures, I play up
the contrast. In the Afterword I try to make some amends. The fact is
that Davidson's writing has been an inspiration to me ever since, at
David Wiggins's urging, I first read "Truth and Meaning" or perhaps
"On Saying That" (I am not sure which came first for me).1
I have been more strongly influenced than footnotes can indicate by
P. F. Strawson, especially by his peerless book on Kant's First Cri
tique.2 I am not sure that Strawson's Kant is really Kant, but I a m
convinced that Strawson's Kant comes dose t o achieving what Kant
wanted to achieve. In these lectures I follow Strawson directly when I
exploit Kant in the context of considering the first person (Lecture V);
and my use of Kane in saying how we should conceive experience
the main thing I try to do here-is Strawsonian in spirit and often in
detail.
Strawson's influence operates on me both directly and at one re
move, through Gareth Evans. Evans did not live to write a preface to
his seminal book, The Varieties of Reference;3 if he had, he would
surely have tried to convey the extent to which his teacher had shaped
his thinking at its most central points. Evans's direct importance for
me is incalculable. For a decade or so what mattered most in my intel
lectual life was my collegial life with him. Anyone who knew him will
know what this amounted to: a non-stop barrage of intellectual stim
ulation. I have no idea how I could so m uch as begin to separate out
the difference he made to me; I cannot imagine what sort of philoso-
1. Both now reprinted in Inquiries into Truth and Interpretation (Clarendon Press, Ox
ford, 1984 ).
2. The Bounds of Sense (Methuen, London, 1966). I should also mention Individuals: An
Essay in Descriptive Metaphysics (Methuen, London, 1959).
3. Clarendon Press, Oxford, 1982.
Preface ix
pher (if any) I would have been now if it had not been for him. He is
one of the two people now gone with whom I most wish I could dis
cuss this work.
The other is Wilfrid Sellars. His classic essay "Empiricism and the
Philosophy of Mind"4 began to be central for me long before I ever
thought of coming to the University of Pittsburgh, and it is an abiding
regret for me that I became a colleague of his too late in his life to
profit from talking to him as I have profited from reading him.
Robert Brandom's writings, and conversations with him, have been
very important in shaping my thinking, usually by forcing me to get
clear about the differences, small in themselves, that transform for me
the look of our wide measure of agreement. The way I put things here
bears substantial marks of Brandom 's influence. Among much else, I
single out his eye-opening seminar on Hegel's Phenomenology of
Spirit, which I attended in 1 990. Thoughts Brandom elicited from me
then show up explicitly at a couple of places in these lectures, but the
effect is pervasive; so much so that one way that I would like to con
ceive this work is as a prolegomenon to a reading of the Phenomeno
logy, much as Brandom's forthcoming Making It Explicit: Reaso11ing,
Representing, and Discursive Commitment5 is, among many other
things, a prolegomenon to his reading of that difficult text. I am also
deeply indebted to Brandom for detailed help and support while I was
preparing the lectures.
Many other people have helped me with this work . I try to mention
specific debts in the footnotes, but I am sure there are many places
where I have forgotten who first taught me to say things as I do, and
I am sorry about that. Here I want to thank James Conant, john
Haugeland, and Danielle Macbeth for special help and encourage
ment.
I made my first sketches of the kind of formulation I have arrived at

here during the winter of 1985-86, in an attempt to get under control
my usual excited reaction to a reading-my third or fourth-of Rich
ard Rorty's Philosophy and the Mirror of Nature.6 I think it was an
earlier reading of Rorty that put me on to Sellars; and it will be obvi-
4. In Herbert Feigl and Michael Scriven, eds., Mitmesota Studies in the Philosophy oi
Science, vol. 1 {University of Minnesota Press, Minneapolis, 1 956), pp. 253-329.
5. Harvard University Press, Cambridge, Mass., 1 99+.
6. Princeton University Press, Princeton, 1 979.
ous that Rorty's work is in any case central for the way I define my
stance here.
I used those first formulations in lectures in Oxford in that aca
demic year, my last ther�, and in my Whitehead Lectures at Harvard
in the spring of 1986. That initial work was done while I was a Rad
cliffe Philosophy Fellow, and even though this fruit of my Fellowship
is rather belated, I want to record gratefully that this book owes a
great deal to the generosity of the Radcliffe Trust. I also thank the
Master and Fellows of University College, Oxford, for giving me per
mission to accept the Fellowship.
I am very grateful to the Sub-Faculty of Philosophy in the Univer·
sity of Oxford for doing me the great honour of inviting me to give the
John Locke Lectures, and to many friends in England for their kind
ness during my stay.
Introduction
1. This book first appeared without an Introduction. Since then,

however, I have been made to realize that it is harder to understand
than I thought. I hope an overview, omitting some details in order to
focus on the central theme, will help at least some readers.
My aim is to propose an account, in a diagnostic spirit, of some
characteristic anxieties of modern philosophy-anxieties that centre,
as my title indicates, on the relation between mind and world. Contin
uing with the medicarmetaphor, we mignfSaYth"a t a sati;fa ctory di
agnosis ought to point towards a cure. I aim at explaining how it
comes about that we seem to be confronted with philosophical obliga
tions of a familiar sort, and I want the explanation to enable us to
unmask that appearance as illusion.
It matters that the illusion is capable of gripping us. I want to be
able to acknowledge the power of the illusion's sources, so that we
fin d ourselves able to respect the conviction that the obligations are
genuine, even while we see how we can, for our own part, reject the
appearance that we face a pressing intellectual task.
2. A good way into the picture I offer is to consider the plausibility

of a minimal empiricism.
To make sense of the idea of a mental state's or episode's being
directed towards the world, in the way in which, say, a belief or judge
ment is, we need to put the state or episode in a normative context. A
belief or j udgement to the effect that things are thus and so-a belief
or j udgement whose content (as we say) is that things are thus and
xii Introduction
so--must be a posture or stance that is correctly or incorrectly

adopted according to whether or not things are indeed thus and so. (If
we can make sense of judgement or belief as directed towards the
world in that way, other kinds of content-bearing postures or stances
should easily fall into place.) This relation between mind and world is
normative, then, in this sense: thinking that aims at judgement, or at
the fixation of belief, is answerable to the world-to how things are
for whether or not it is correctly executed.
Now how should we elaborate the idea that our thinking is thus
answerable to the world? In addressing this question, we might re
strict our attention, at least tacitly, to thinking that is answerable to
the empirical world; that is, answerable to how things are in so far as
how things are is empirically accessible. Even if we take it that answer
ability to how things are includes more than answerability to the em
pirical world, it nevertheless seems right to say this: since our cogni
tive predicament is that we confront the world by way of sensible
intuition (to put it in Kantian terms), our reflection on the very idea of
thought's directedness at how things are must begin with answerabil
ity to the empirical world. And now, how can we understand the idea
that our thinking is answerable to the empirical world, if not by way
of the idea that our thinking is answerable to experience? How could
a verdict from the empirical world-to which empirical thinking must
be answerable if it is to be thinking at all-be delivered, if not by
way of a verdict from (as W. V. Quine puts it) "the tribunal of
experience"?1
That is what I mean by "a minimal empiricism": the idea that expe
rience must constitute a tribunal, mediating the way our thinking is
answerable to how things are, as it must be if we are to make sense of
it as thinking at all. And this is one side of a combination of plausibil
ities that promises to account for the philosophical anxieties I alluded
to. The other side is a frame of mind, which I shall come to (§4
below), that makes it hard to see how experience could function as a
tribunal, delivering verdicts on our thinking.
Fully developed, of course, such a combination would amount to
an antinomy: experience both must (minimal empiricism) and cannot
(the line of thought I have yet to broach) stand in judgement over our
1. "Two Dogmas of Empiricism", in W. V. Quine, From a Logical Point of View (Har

vard University Press, Cambridge, Mass., 1961; lsr ed. 1953), pp. 20-46, at p. 41.
Introduction xiii
attempts to make up our minds about how things are. But consider a
stage at which reflection is subject to such a pair of pressures, but not
self-consciously enough for it to be clear that what they generate is an
antinomy. With an inexplicit awareness of the tension between such a
pair of tendencies in one's thinking, one could easily fall into an anx
iety of a familiar philosophical sort, about that directedness of mind
to world that it seemed we would have to be able to gloss in terms of
answerability to how things are. In such a position, one would find
oneself asking: "How is it possible for there to be thinking directed at
how things are?" This would be a "How possible?" question of a fa
miliar philosophical kind; it acquires its characteristic philosophical
bite by being asked against the background of materials for a line of
thought that, if made explicit, would purport to reveal that the
question's topic is actually not possible at all.
3. It may seem surprising that I am associating empiricism with a

philosophical anxiety about the possibility of thought. Surely, it may
be objected, empiricism is an epistemological position, and the rele
vant question should rather be this: "How is it possible for there to be
empirical knowledge?" What that comes to, in terms of Quine's juri
dical image, is something on these lines: "How can experience, stand
ing in judgement over, say, a belief, return a verdict sufficiently fa
vourable for the belief to count as a case of knowledge?"
But suppose we are prone to have our thinking shaped by the sec
ond element in my combination of plausibilities (which l have so far
brought into view only in terms of its effect). That is just to suppose
that we find it hard to see how experience can function as a tribunal,
standing in judgement over our beliefs. This would be a difficulty
about how experience can return any verdicts on our thinking at all,
and that is surely more fundamental than a difficulty about how expe·
rience can return a particular kind of verdict, one that reaches some
high level of favourableness.
It is true that modern philosophy is pervaded by apparent problems
about knowledge in particular. But I think it is helpful to see those
apparent problems as more or less inept expressions of a deeper anx
iety-an inchoately felt threat that a way of thinking we find ourselves
falling into leaves minds simply out, of touch with the rest of reality,
not just questionably capable of getting to know about it. A problem
about crediting ourselves with knowledge is just one shape, and not
the most fundamental, in which that anxiety can make itself felt.
4. What is the pressure oo our .thinking that makes it hard to see how
experience could function as a tribunal? I can bring it to light by re
hearsing a central element in Wilfrid Sellars's attack on "the Myth of
the Given".
Sellars insists that the concept of knowledge belongs in a normative
context. He writes: "In characterizing an episode or a state as that of
knowing, we are not giving an empirical description of that episode or
state; we are placing it in the logical space of reasons, of justifying and
being able to justify what one says."2 It is a way of repeating what I
have just been urging (§3 above) to say this: though Sellars here
speaks of knowledge in particular, that is just to stress one application
of the thought that a normative context is necessary for the idea of
being in touch with �orld at all, whether knowledgeably or not.
..-one way of putting what Sellars is driving at IS to say that epiSte
mology is liable to fall into a naturalistic fallacy.31n the more general
version I have insisted on, the thought is that the risk of a naturalistic
fallacy besets reflection about world-directedness as such, whether
knowledgeable or not. If we put Sellars's point this way, we are iden
tifying the natural-as indeed Sellars sometimes does-with the sub
ject matter of "empirical description"; that is, with the subject matter
of a mode of discourse that is to be contrasted with placing something
in the normative framework constituted by the logical space of rea
sons. Sellars separates concepts that are intelligible only in terms of
how they serve to place things in the logical space of reasons, such as
the concept of knowledge, from concepts that can be employed in
"empirical description". And if we read the remark as a warning
against a naturalistic fallacy, we are understanding "empirigl
description" as placing things in the logical space of nature, to coin a
-
p!lr'aSe' that is Sellarsian at least m spmt.
What would the logical space of nature be? I think we capture the
essentials of Sellars's thinking if we take it that the logical space of
2. "Empiricism and the Philosophy of Mind", in Herbert Feigl and Michael Scriven, eds.,
Minnesota Studies in the Philosophy ofScience, vol. 1 (University of Minnesota Press, Min
neapolis, 1956), pp. 253-329, at pp. 298 -9.
3. See p. 257 of "Empiricism and the Philosophy of Mind" for a formulation on these
lines.
XV
nature is the logical space in which the natural sciences function, as we

have been�nabled to conceive them by a well-charted, clndin itself
admirable, development of modern thought. We might say that to
place something in nature on the relevant conception, as contrasted
with placing it in the logical space of reasons, is to situate it in the
realm of law. But what matters for Sellars's point is not that or any
other positive characterization, but the negative claim: whatever the
relations are that constitute the logical space of nature, they are differ
ent in kind from the normative relations that constitute the logical
space of reasons. The relations that constitute the logical space of na
ture, on the relevant conception, do not include relations such as one
thing's being warranted, or-for the general case-correct, in the light
of another. That is what Sellars is saying when he insists that "empir
ical description" cannot amount to placing something in the logical
space of reasons.
Now, supposing we accept this dichotomy of logical spaces, which
logical s ace would be the home of the concept of ex er· ce?m
ourse it depen s on w a y expenence". But suppose we
want to conceive the course of a subject's experience as made up of
impressions, impingements by the world on a possessor of sensory
capacities. Surely such talk of impingements by the world is "empiri
cal description"; or, to put the point in the variant terms I have intro
duced, the idea of receiving an impression is the idea of a transaction
in nature. On Sellars's principles, then, to identify something as an
impression is to place it in a logical space other than the one in which
talk of knowledge-or, to keep the general case in view, talk of world
directedness, knowledgeable or not-belongs. On these principles, the
logical space in which talk of impressions belongs is not one in which
things are connected by relations such as one thing's being warranted
or correct in the light of another. So if we conceive experience as made
up of impressions, on these principles it cannot serve as a tribunal,
something to which empirical thinking is answerable. Supposing that
it can would just be a case of the naturalistic fallacy that Sellars warns
us against-a case of taking it that "empirical description" can
amount to placing things in the logical space of reasons.
I have here extracted from Sellars a line of thought that at least
potentially stands in tension with a minimal empiricism. (Whether the
tension is actual depends on whether empiricism needs to conceive
"the tribunal of experience" as made up of impressions; I shall come
back to that question (§6 below).) In the lectures that follow, it is
XVI Introduction
mainly Donald Davidson who figures in the role I have here cast Sell
ars in: as someone whose reflection a bout experience disqualifies it
from intelligibly constituting a tribunal. For these purposes, Sellars
and Davidson are interchangeable. Sellars's attack on the Given corre
sponds, in a way I exploit in my first lecture, to Davidson's attack on
what he calls "the third dogma of empiricism"-the d ualism of con
ceptual scheme and empirical "content". And Davidson explicitly sug
gestS that the thought dislodges even a minimal empiricism; he de
scribes the third dogma of empiricism as "perhaps the last", on the
ground that "if we give it up it is not clear that there is anything dis
tinctive left to call empiricism" .4
5 . I have been suggesting that we can trace some distinctive a nxieties

of modern philosophy to a tension between two forces, both of which
have an intelligible tendency to shape our reflection a bout empirical
thinking, and thereby about world-directedness in general. One is the
attractiveness of a minimal empiricism, which makes out that the very
idea of thought's directedness at the empirical world is intelligible
only in terms of answerability to the tribunal of experience, conceived
in terms of the world impressing itself on perceiving subjects. The
other is a frame of mind that makes it seem impossible that experience
could be a tribunal. The idea of a tribunal belongs, together with the
idea of what the tribunal passes its verdicts on, in what Sellars calls
"the logical space of reasons"-a logical space whose structure con
sists in some of its occupants being, for instance, warranted or correct
in the light of others. But the idea of experience, at least construed in
terms of impressions, evidently belongs in a logical space of natural
connections. That can easily make it seem that if we try to conceive
experience as a tribunal, we must be falling into the naturalistic fallacy
that Sellars depicts as a pitfall for would-be epistemologists. Suppose
we are inexplicitly aware that our thinking is subject to both these
forces; that makes it intelligible that we should find thought's being
about the empirical world philosophically problematic.
As I said (§1 above k m aim i dia nosis, with a view to a cure. If
'
philosophical anxiety ab ut the very poss1 ility of being m touch With
the world can be traced to the tension between those two forces, a
cure would require resolving the tension. Obviously the description I
�
4. "On the Very Idea of a Conceptual Scheme", in Donald Davidson, Inquiries into
Truth and Interpretation (Clarendon Press, Oxford, 1984), pp. 183-98, at p. 189.
Introduction xvii
have given leaves various options available for doing that. In this
book I recommend one way of resolving this tension. Here I shall
briefly locate it by distinguishing it from a couple of others.
6. One option would be to renounce empiricism, at least with expe

rience construed in terms of impressions. As I have already mentioned
(§4 above), Davidson explicitly links the fate of empiricism with the
fate of scheme-content dualism, which he effectively demolishes in a
way that parallels Sellars's attack on the Myth of the Given. And Sell
ars himself works at delineating a concept of impressions that is insu
lated from epistemology.
I do not believe a position on these Jines can be genuinely satisfying.
I urge this, specifically in connection with Davidson, in the first of the
lectures that make up the body of this book. Davidson's ground for
giving up empiricism is, in its essentials, the claim that we cannot take
experience to be epistemologically significant except by falling into
the Myth of the Given, in which experience, conceived in such a way
that it could not be a tribunal, is nevertheless supposed to stand in
judgement over our empirical thinking. That certainly has the right
shape for an argument that we must renounce empiricism. The trou
ble is that it does not show how we can. It does nothing to explain
away the plausibility of the empiricist picture1 according to which we
can make sense of the world-directedness of empirical thinking only
by conceiving it as answerable to the empirical world for its correct
ness, and we can understand answerability to the empirical world
only as mediated by answerability to the tribunal of experience, con
ceived in terms of the world's direct impacts on possessors of percep
tual capacities. If we are restricted to the positions Davidson consid
ers, the attractions of empiricism lead only to the incoherence of the
Myth of the Given. But so long as the attractions of empiricism are
not explained away, that fact is only a source of continuing philo
sophical discomfort, not a basis for being content with abandoning
empiricism, however compulsory Davidson's conception of the op
tions makes that seem.
It is true that, on the principles that induce Davidson and Sellars to
reject empiricism, empirical thinking can be seen as rationally con
strained by cases of its perceptually appearing to a subject that things
are thus and so. This might be offered as� concession to the attrac-
tions of empiricism. But the concession does not fill the gap I am
pointing to; it does not explain away the attractions of empiricism,
spelled out in terms of impressions. When it perceptually appears to a
subject that things are tbus and so, that things are thus and so is itself
a case of empirical content. A� long as nothing is done to undermine
the plausibility of the thought that empirical content in general is in
telligible only in terms of answerability to impressions, empirical con
tent will look just as problematic in this context as it does in the con
text of judgements or beliefs.
7. Sellars and Davidson think we are forced to renounce empiricism,

in the relevant sense, partly because they think the logical space of
reasons is sui generis, as compared with the logical space in which
Sellars sees "empirical description" as functioning, which I have iden
tified on Sellars's behalf with the logical space of nature. That is
Sellars's way of putting the claim, but Davidson has a counterpart;
what figures in Sellars as the sui generis character of the logical space
of· reasons figures in Davidson as the sui generis character of what he
calls "the constitutive ideal of rationaliti'. 5
This �o a se!=ond way the tension could be resolved: rejecting
the dichotomy of logical spaces. If we take this course, we can accept
that the concept of experience belongs in the logical space of nature,
but deny that that poses any problem for empiricism. The idea is that
the logical space of reasons-in which, if we are to hold on to empir
icism, experience must be related to empirical thinking-is just part of
the logical space of nature; the normative relations that constitute the
logical space of reasons can be reconstructed out of conceptual mate
rials whose home is the logical space that Sellars, wrongly on this
view, contrasts with the logical space of reasons. This outlook figures
in this book under the label "bald naturalism". Bald naturalism re
fuses to accept that the relations that constitute the logical space of
reasons are anything but natural, in a sense of "natural" that connects
with the logical space that figures in Sellars (and, with different termi
nology, in Davidson) on the other side of a contrast with the logical
space of reasons. According to this naturalism, moves of the sort Sell
ars stigmatizes as committing a naturalistic fallacy are indeed natural-
5. See especially "Mental Events", in Donald Davidson, E ssays on Actions and Events
(Clarendon Press, Oxford, 1 9 8 0), pp. 207-25; the quoted phrase is from p. 223.
istic, but not thereby shown to be fallacious. Thus we can accept that
the concept of experience belongs in the logical space of nature, but
that does not debar experience, so conceived, from being intelligible as
a tribunal. So there is no need to explain away the attractions of em
piricism.
I shall say a little more about bald naturalism shortly (§9 below),
but first I want to outline the different way of resolving the tension
that I recommend.
8. My alternative holds on to the thought rejected by bald natural

ism, that the structure of the logical space of reasons is sui generis, as
compared with the structure of the logical space within which natural
scientific description situates things. Even so, my alternative makes
room for us to suppose, as according to Sellars and Davidson we can
not, both that the very idea of experience is the idea of something
natural and that empirical thinking is answerable to experience. This
requires a different way to avoid the threat of a naturalistic fallacy.
The modern scientific revolution made possible a newly clear con
ception of the distinctive kind of intelligibility that the natural sciences
allow us to find in things. The new clarity consists largely, I claim, in
an appreciation of something close to what underlies Sellars's warning
of a naturalistic fallacy. We must sharply distinguish natural-scientific
intelligibility from the kind of intelligibility something acquires when
we situate it in the logical space of reasons. That is a way of affirming
the dichotomy of logical spaces, as bald naturalism refuses to. Even
so, we can acknowledge that the idea of experience is the idea of
something natural, without thereby removing the idea of experience
from the logical space of reasons. What makes this possible is that we
need not identify the dichotomy of logical spaces with a dichotomy
between the natural and the normative. We need not equate the very
idea of nature with the idea of instantiations of concepts that belong
in the logical space-admittedly separate, on this view, from the logi
cal space of reasons-in which the natural-scientific kind of intelligi
bility is brought to light.
On this view, Sellars is right that the logical space in which natural
scientific investigation achieves its distinctive kind of understanding is
alien to the logical space of reasons. The logical space of reasons is the
frame within which a fundamentally different kind of intelligibility
XX Introduction
comes into view. And (the same point in different terms) Davidson is
right that "the constitutive ideal of rationality" governs concepts that
are for that reason quite special, in comparison with the conceptual
apparatus of the nomothetic sciences. But it is one thing to acknowl
edge this-in Sellarsian terms, to single out a logical space that is to be
contrasted with the logical space of reasons. It is another to equate
that logical space, as Sellars at least implicitly does, with the logical
space of nature. That is what makes it seem impossible to combine
empiricism with the idea that the world's making an impression on a
perceiving subject would have to be a natural happening. The mistake
here is to forget that nature includes second nature. Human beings
acquire a second nature in part by being initiated into conceptual ca
pacities, whose interrelations belong in the logical space of reasons.
Once we remember second nature, we see that operations of nature
can include circumstances whose descriptions place them in the logical
space of reasons, sui generis though that logical space is. This makes
it possible to accommodate impressions in nature without posing a
threat to empiricism. From the thesis that receiving an impression is a
transaction in nature, there is now no good inference to the conclu
sion drawn by Sellars and Davidson, that the idea of receiving an im
pression must be foreign to the logical space in which concepts such as
that of answerability function. Conceptual capacities, whose interre
lations belong in the sui generis logical space of reasons, can be oper
ative not only in judgements-results of a subject's actively making up
her mind about something-but already in the transactions in nature
that are constituted by the world's impacts on the receptive capacities
of a suitable subject; that is, one who possesses the relevant concepts.
Impressions can be cases of its perceptually appearing-being
apparent-to a subject that things are thus and so. In receiving im
pressions, a subject can be open to the way things manifestly are. This
yields a satisfying interpretation for the image of postures that are
answerable to the world through being answerable to experience.
9. The position I call "bald naturalism" figures in this book only as

a competitor with the outlook I have just sketched, in the project of
exorcizing some philosophical anxieties. The shared aim is to see how
we need not seem obliged to set about answering the questions that
express the anxieties. I have suggested that we can bundle the sup-
Introduction xxi
posed problems together in a question on these lines: "How is empir

ical content possible?" Empirical content looks problematic, in the
way I aim to deal with, when one becomes inexplicitly aware of an
apparent tension between empiricism and the fact that the idea of an
impression is the idea of an occurrence in nature. If we can achieve a
way of seeing things in which there is after all no tension there, the
question, taken as a way of expressing that philosophical puzzlement,
should lapse; that needs to be distinguished from its seeming to have
been answered. I concern myself with bald naturalism only as sharing
the wish to achieve that effect.6
There is a possible confusion to avoid here. Much contemporary
work sets out, in a naturalistic spirit, to answer (not exorcize) ques
tions that can be framed in the "How possible?" form, about empiri
cal content or other aspects of mindedness. The work I mean aims to
give perspicuous descriptions of the material constitution of, say, per
ceivers, in such a way as to make it intelligible that things composed
of mere matter can possess the relevant complex of capacities. A ques
tion to which this would be an appropriate response is not a "How
possible?" question of the sort I am concerned with. As I said (§2
above), a "How possible?" question of the sort I am concerned with
expresses a distinctive kind of puzzlement, issuing from an inexplicit
awareness of a background to one's reflection tha, if made explicit,
would yield an argument that the topic of the question is not possible
at all. To respond to a "How possible?" question of this kind in, so to
speak, engineering terms, with a perspicuous description of the requi
site material constitution, would be plainly unhelpful; it would be like
responding to Zeno by walking across a room. That leaves it open
that investigations of the "engineering" sort might be fine for other
purposes. In this book I consider bald naturalism only as a way to
exorcize (not answer) the questions that give expression to that dis
tinctively philosophical kind of puzzlement, the kind that issues from
a frame of mind that, when fully explicit, would purport to display an
impossibility in what the questions are asked about. Bald naturalism
figures here as, I claim, a less satisfying way to do that than my alter
native. I do not concern myself with those different questions and an-
6. The label "bald naturalism" is perhaps infelicitous for a position with a sophisticated
motivation on these lines; that is what I acknowledge � the footnote on pp. 88-9. I took
myself to be stuck with the label even so, since I had given it a thematic prominence in the
lectures of which the book is a version.
swers, the ones that figure in inquiries into the machinery of minded
ness.
I have tried to make it plausible that the anxieties I aim to exorcize
issue from the thought..Loften no doubt only inchoate-that the struc
ture of the logical space of reasons is sui generis, as compared with
the logical framework in which natural-scientific understanding is
achieved. On this view, it emerges as unsurprising that the period in
which dealing with these supposed difficulties came to seem the dom
inant obligation of philosophy coincides with the rise of modern sci
ence, in which natural-scientific understanding, as we are now
equipped to conceive it, was being separated out from a hitherto un
differentiated conception of understanding in general. According to
my picture, an important element in this clarification of the proper
target of natural science was an increasingly firm awareness that we
must sharply distinguish natural-scientific understanding from the
kind of understanding achieved by situating what is understood in
the logical space of reasons; that is, precisely, that the structure of the
logical space of reasons is sui generis, as I have read Sellars, and in
different terms Davidson, as claiming it is.
Now bald naturalism has it that that perhaps inchoate sense of a
conceptual divide was simply wrong; it would be revealed to be wrong
by a reconstruction of the structure of the logical space of reasons in
terms that belong in the logical space of natural-scientific understand
ing. This claim is programmatic, but that is not my ground for finding
bald naturalism unsatisfying. The point is rather one that I mentioned
at the beginning of this Introduction (§1 above). It is easy-and not
because we are merely stupid-to be captivated by the kind of philo
sophy I aim to exorcize. That means that a proposed exorcism is more
satisfying to the extent that it enables us to respect, as insights, the
driving thoughts of those who take the familiar philosophical anxie
ties to pose real intellectual obligations (our driving thoughts when we
find ourselves beset by the anxieties), even while we unmask the sup
posed obligations as illusory. Now my picture is unlike bald natural
ism in just that way. I acknowledge as an insight the basic conviction
that generates the anxieties, in combination with the �car5f1r.sues
tionable conception of impressions as occurrences in nature. In my
picture, those who take it that philosophy has to answer (rather than
exorcize) questions about how minds can be in touch with the world
are not wrong in supposing that the logical space of reasons is sui
generis, in just the way that seems to lead to problems about how
1 ntroattctron XXIII
responsiveness to reasons can fit into the natural world. (Indeed I

think that was a prime lesson our ancestors learned at the time of the
rise of modern science.) We can disown an obligation to try to answer
the characteristic questions of modern philosophy, without needing to
deny, as bald naturalism does, that a real insight is operative in seem
ing to be faced with that obligation.
To make my point here, I need not pretend to have an argument
that the bald naturalist programme-reconstructing the structure of
the logical space of reasons in terms that belong in the logical space of
natural-scientific understanding-cannot be executed. The point is
just that the availability of my alternative and, I claim, more satisfying
exorcism undercuts a philosophical motivation, the only one relevant
to my concerns in this book, for supposing the programme must be
feasible. It is not philosophically threatening to suppose there is in
sight in the thought that reason is not natural, in the only sense of
"natural" countenanced by bald naturalism.
10. It should be clear that reflection about perceptual experience has

served in this Introduction, as it does in this book, as just one example
of a type. Parallel puzzlements will be prone to arise wherever we
want to speak of responsiveness to reasons. "�esponsiveness tQ.
reasons" is a good gloss on one notion of freedom. So the puzzlement
iliit'Sgeneral form is about how freedom, in that sense, fits into the
natural world. That seems hard to comprehend if we are tempted to
eq uate the natural with exemplifications of concepts whose home is
the logical space that Sellars contrasts with the logical space of rea
sons. In the fifth lecture I talk briefly about another instance of the
type.
Some friendly readers have taken issue with the negative attitude I
profess in this book to what I call "constructive philosophy" (see, e.g.,
p. 95). Let me bring the point I mean to make in those terms into
connection with the way I have set things out in this Introduction.
What I mean by "engaging in constructive philosophy" is attempting
to answer philosophical questions of the sort I have here singled out:
"How possible?" questions whose felt urgency derives from a frame
of mind that, if explicitly thought through, would yield materials for
an argument that what the questions are asked about is impossible.
Evidently it can seem sensible to embark on such a project only if one
xxiv Introduction
does not quite understand the predicament that seems to motivate it.
If the frame of mind is left in place, one cannot show how whatever it
is that one is asking about is possible; if the frame of mind is dis
lodged, the "How possible?" question no longer has the point it
seemed to have. Either way, there is no prospect of answering the
question as it was putatively meant. So if I am right about the charac
ter of the philosophical anxieties I aim to deal with, there is no room
for doubt that engaging in "constructive philosophy", in this sense, is
not the way to approach them. As I have put it, we need to exorcize
the questions rather than set about answering them. Of course that
takes hard work: if you like, constructive philosophy in another sense.
And of course that is what I offer in this book.
Mind and World
LECTURE
Concepts and Intuitions
1. The overall topic I am going to consider in these lectures is the

way concepts mediate the relation berween minds and the world. I
shall focus the discussion in terms of a familiar philosophical outlook,
which Donald Davidson has described as a dualism of scheme and
content.1 That will get us quickly to Kant. One of my main aims is to
suggest that Kant should still have a central place in our discussion of
the way thought bears on reality.
When Davidson talks about a dualism of scheme and content,
"scheme" means "conc;eptual scheme". If content is dualistically op
posed to what is conceptual, "content" cannot mean what it often
means in contemporary philosophy, namely, what is given by a "that"
clause in, for instance, an attribution of a belief: just to have a label,
we can call content in this modern sense "representational content".
Representational content cannot be dualistically set over against the
conceptual. That is obviously so, however hospitable we are to the
idea that some representational content is non-conceptual. (I shall
come to that issue in my third lecture.)
So why is it content that is supposed to stand over against concepts,
in the dualism Davidson criticizes? We can arrive at an understand
ing of the terminology from the way it figures in Kant's remark
"Thoughts without content are empty" .2 For a thought to be empty
1. "On the Very Idea of a Conceptual Scheme", in his Inquiries into Truth and Interpre
tation (Clarendon Press, Oxford, 1 9 8 4), pp. 1 83-98. See especia.lly p. 1 87, "a dualism of
total scheme (or language) and uninterpreted content", and p. 1 8 9, "the dualism of concep
tual scheme and empirical content".
2. Critique of Pure Reason, trans. Norman Kemp Smith (Macmillan, London, 1 929),
A51/B75.
4 MIND A ND W ORLD
would be for there to be nothing that one thinks when one thinks it;
that is, for it to lack what I am calling "representational content".
That would be for it not really to be a thought at all, and that is surely
Kant's point; he is n ot, absurdly, drawing our attention to a special
kind of thoughts, the empty ones. Now when Kant says that thoughts
without content are empty, he is not merely affirming a tautology:
"without content" is not just another wording for "empty", as it
would be if "content" simply meant "representational content".
"Without content" points to what would explain the sort of empti
ness Kant is envisaging. And we can spell out the explanation from the
other half of Kant's remark: "intuitions without concepts are blind."
Thoughts without content-which would not really be thoughts at
all-would be a play of concepts without any connection with intu
itions, that is, bits of experiential intake. It is their connection with
experiential intake that supplies the content, the substance, that
thoughts would otherwise lack.
So the picture is this: the fact that thoughts are not empty, the fact
that thoughts have representational content, emerges out of an inter
play of concepts and intuitions. "Content" in Davidson's dualism cor
responds to intuitions, bits of experiential intake, understood in terms
of a dualistic conception of this interplay.
2. This Kantian background explains why what stands over against

the conceptual, in the dualism Davidson considers, is often described
as the Given. In fact "dualism of scheme and Given" is a better label
than "dualism of scheme and content", because it does not resonate
confusingly with the idea of representational content. It also suggests
a specific understanding of why the dualism is tempting.
Kant makes his remark about intuitions and concepts in the course
of representing empirical knowledge as the result of a co-operation
between receptivity and spontaneity, between sensibility and under
standing.3 Now we should ask why it seems appropriate to describe
the understanding, whose contribution to this co-operation is its com·
3. To give the lead-up to the passage I have already quoted: "If the receptivity of our
mind, its power of receiving representations in so far as it is in any wise affected, is to be
entitled sensibility, then the mind's power of producing representations from itself, the
spontaneity of knowledge, should be called the understanding. Our nature is so constiruted
that our intuition can never be other than sensible; that is, it contains only the mode in
which we are affected by objects. The faculty, on the other hand, which enables us to think
the object of sensible intuition is the understanding. To neither of these powers may a pre·
mand of concepts, in terms of spontaneity. A schematic but suggestive

answer is that the topography of the conceptual sphere is constituted
by rational relations. The space of concepts is at least part of what
Wilfrid Sellars calls "the space of reasons". 4 When Kant describes the
understanding as a faculty of spontaneity, that reflects his view of the
relation between reason and freedom: rational necessitation is not just
compatible with freedom but constitutive of it. In a slogan, the space
of reasons is the realm of freedom.5
But if our freedom in empirical thinking is total, in particular if
it is not constrained from outside the conceptual sphere, that can
seem to threaten the very possibility that judgements of experience
might be grounded in a way that relates them to a reality external
to thought. And surely there must be such grounding if experience
is to be a source of knowledge, and more generally, if the bearing
of empirical judgements on reality is to be intelligibly in place in
our picture at all. The more we play up the connection between
reason and freedom, the more we risk losing our grip on how exer
cises of concepts can constitute warranted judgements about the
world. What we wanted to conceive as exercises of concepts
threaten to degenerate into moves in a self-contained game. And
that deprives us of the very idea that they are exercises of concepts.
Suiting empirical beliefs to the reasons for them is not a self-con
tained game.
The dualism of conceptual scheme and "empirical content", of
ference be given over the other. Without sensibility no object would be given to us, without
understanding no object would be thought. Thoughts without content are empty, intuitions
without concepts are blind."
4. "In characterizing an episode or a state as that [better: one] of knowing, we are not
giving an empirical description of that episode or state; we are placing it in the logical space
of reasons, of justifying and being able to justify what one says." This is from pp. 298-9 of
Sellars's classic attack on the Myth of the Given, "Empiricism and the Philosophy of Mind",
in Herbert Feigl and Michael Scriven, eds., Minnesota Studies in the Philosophy of Science,
vol. 1 (University of Minnesota Press, Minneapolis, 1956), pp. 253-329. In much of the rest
of these lectures, I shall be concerned to cast doubt on Sellars's idea that placing something
in the logical space of reasons is, as such, to be contrasted with giving an empirical descrip
tion of it. But the theme of placing things in the space of reasons is of central importance
for me.
I say that the space of concepts is at least part of the space of reasons in order to leave it
open, for the moment, that the space of reasons may extend more widely than the space of
concepts; see the text below for this idea.
5. For a thoughtful discussion of this idea, see Robert Brandom, "Freedom and Con
straint by Norms", American Philosophical Quarterly 16 (1979), 187-96.
6 M I N D A N D W O R L D
scheme and Given, is a response to this worry. The point of the dual
ism is that it allows us to acknowledge an external constraint on our
freedom to deploy our empirical concepts. Empirical justifications de
pend on rational relations, relations within the space of reasons. The
putatively reassuring idea is that empirical justifications have an ulti
mate foundation in impingements on the conceptual realm from out
side. So the space of reasons is made out to be more extensive than the
space of concepts. Suppose we are tracing the ground, the justifica
tion, for a belief or a judgement. The idea is that when we have ex
hausted all the available moves within the space of concepts, all the
available moves from one conceptually organized item to another,
there is still one more step we can take: namely, pointing to something
that is simply received in experience. It can only be pointing, because
ex hypothesi this last move in a justification comes after we have ex
hausted the possibilities of tracing grounds from one conceptually or
ganized, and so articulable, item to another.
I began with the thought that is expressed in Kant's remark: the
very idea of representational content, not just the idea of judgements
that are adequately justified, requires an interplay between concepts
and intuitions, bits of experiential intake. Otherwise what was meant
to be a picture of the exercise of concepts can depict only a play of
empty forms. I have modulated into talking about how the idea of the
Given figures in a thought about the grounding that entitles some em
pirical judgements to count as knowledgeable. But this explicitly epi
stemological idea is straightforwardly connected with the more gen
eral idea I began with. Empirical judgements in general-whether or
not they reflect knowledge, and even whether or not they are justified
at all, perhaps less substantially than knowledge requires-had better
have content of a sort that admits of empirical justification, even if
there is none in the present case (say in a quite unsupported guess).
We could not begin to suppose that we understood how pointing to a
bit of the Given could justify the use of a concept in judgement
could, at the limit, display the judgement as knowledgeable-unless
we took this possibility of warrant to be constitutive of the concept's
being what it is, and hence constitutive of its contribution to any
thinkable content it figures in, whether that of a knowledgeable, or
less substantially justifiable, judgement or any other.
This supposed requirement would bear immediately on observa
tional concepts: concepts suited to figure in judgements that are di-
reedy responsive to experience. The supposed requirement is reflected

in a familiar picture of the formation of such concepts, a picture that
is a natural counterpart to the idea of the Given. The idea is that if
concepts are to be even partly constituted by the fact that judgements
in which they figure are grounded in the Given, then the associated
conceptual capacities must be acquired from confrontations with suit
able bits of the Given: that is, occasions when pointing to an ultimate
warrant would have been feasible. But in any ordinary impingement
on our sensibility, it would have to be a manifold Given that is pre
sented to us. So in order to form an observational concept, a subject
would have to abstract out the right element in the presented multi
plicity.
This abstractionist picture of the role of the Given in the formation
of concepts has been trenchantly criticized, in a Wittgensteinian spirit,
by P. T. Geach.6 I shall come back to Wittgenstein's thought about
this sort of question later in this lecture (§7).
Once we have equipped ourselves with this picture of how empiri
cal substance is infused into concepts at the ground level, the level of
observational concepts, it will seem straightforward to extend the pic
ture from there. The idea is that empirical substance is transmitted
from the ground level to empirical concepts that are further removed
from immediate experience, with the transmission running along
channels constituted by the inferential linkages that hold a system of
concepts together.
3. I have tried to explain what makes the idea of the Given tempting.
But in fact it is useless for its purpose.
The idea of the Given is the idea that the space of reasons, the space
of justifications. or warrants, extends more widely than the conceptual
sphere. The extra extent of the space of reasons is supposed to allow
it to incorporate non-conceptual impacts from outside the realm of
thought. But we cannot really understand the relations in virtue of
which a judgement is warranted except as relations within the space
of concepts: relations such as implication or probabilification, which
hold between potential exercises of conceptual �apacities. The attempt
to extend the scope of justificatory relations outside the conceptual
sphere cannot do what it is supposed to do.
6. Mental Acts (Routledge and Kegan Paul, London, �957), SS6-l l .

What we wanted was a reassurance that when we use our concepts

in judgement, our freedom-our spontaneity in the exercise of our
understanding-is constrained from outside thought, and constrained
in a way that we can• appeal to in displaying the judgements as
justified. But when we make out that the space of reasons is more
extensive than the cobceptual sphere, so that it can incorporate extra
conceptual impingements from the world, the result is a picture in
which constraint from outside is exerted at the outer boundary of the
expanded space of reasons, in what we are committed to depicting as
a brute impact from the exterior. Now perhaps this picture secures
that we cannot be blamed for what happens at that outer boundary,
and hence that we cannot be blamed for the inward influence of what
happens there. What happens there is the result of an alien force, the
causal impact of the world, operating outside the control of our spon
taneity. But it is one thing to be exempt from blame, on the ground
that the position we find ourselves in can be traced ultimately to brute
force; it is quite another thing to have a justification. In effect, the idea
of the Given offers exculpations where we wanted justifications.7
It can be difficult to accept that the Myth of the Given is a myth. It
can seem that if we reject the Given, we merely reopen ourselves to the
threat to which the idea of the Given is a response, the threat that our
picture does not accommodate any external constraint on our activity
in empirical thought and judgement. It can seem that we are retaining
7. In the lecture as delivered, I said "excuses" where I now say " exculpations"; Zvi
Cohen pointed out that that does not make the contrast I want. What I want is an analogue
to the sense in which if someone is found in a place from which she has been banished, she
is exculpated by the fact that she was deposited there by a tornado. Her arriving there is
completely removed from the domain of what she is responsible for; it is not that she is still
responsible, but there is a basis for mitigating any sanctions.
When we are tempted by the Myth of the Given, we carefully ensure that relations across
the envisaged outer boundary of the space of concepts, relations between bits of the Given
and the most basic judgements of experience, can be reason-constituting; that is the point of
taking the space of reasons to extend more widely than the space of concepts. But we forget
to consider how things look at the new outer boundary of the space of reasons, where it
makes contact with independent reality. What we wanted was to see our exercises of spon
taneity as subject to a constraint imposed by the world itself, but in such a way as not ro
undermine the applicability of the idea of spontaneity. We wanted to be able to credit our
selves with responsible freedom, so that we are within the scope of possible justifications, all
the way out to the ultimate contact between our mental life and the world. My main point
in this lecture is to bring out how difficult it is to see that we can have both desiderata: both
rational constraint from the world and spontaneity all the way out. The Myth of the Given
renounces the second, and the Davidsonian response that I consider below (§6) renounces
the first.
a role for spontaneity but refusing to acknowledge any role for recep
tivity, and that is intolerable. If our activity in empirical thought and
judgement is to be recognizable as bearing on reality at all, there must
be external constraint. There must be a role for receptivity as well as
spontaneity, for sensibility as well as understanding. Realizing this,
we come under pressure to recoil back into appealing to the Given,
only to see all over again that it cannot help. There is a danger of
falling into an interminable oscillation.
But we can find a way to dismount from the seesaw.
4. The original Kantian thought was that empirical knowledge re

sults from a co-operation between receptivity and spontaneity. (Here
"spontaneity" can be simply a label for the involvement of conceptual
capacities.) We can dismount from the seesaw if we can achieve a firm
grip on this thought: receptivity does not make an even notionally
separable contribution to the co-operation.
The relevant conceptual capacities are drawn on in receptivity. (It is
important that that is not the only context in which they are opera
tive. I shall come back to this in §5.) It is not that they are exercised
on an extra-conceptual deliverance of receptivity. We should under-·
stand what Kant calls "intuition"-experiential intake-not as a bare
getting of an extra-conceptual Given, but as a kind of occurrence or
state that already has conceptual content. In experience one takes in,
for instance sees, that things are thus and so. That is the sort of thing
one can also, for instance, judge.
Of course one can be misled into supposing that one takes in th�t
things are thus and so when things are not thus and so. But when one
is not misled, one takes in how things are. It does not matter much
that one can be misled. I shall not talk about it until my final lecture,
and not much then.
In the view I am urging, the conceptual contents that sit closest to
the impact of external reality on one's sensibility are not already, qua
conceptual, some distance away from that impact. They are not the
results of a first step within the space of reasons, a step that would be
retraced by the last step in laying out justifications, as that activity is
conceived within the dualism of scheme and Given. This supposed
first step would be a move from an impression, conceived as the bare
reception of a bit of the Given, to a judgement justified by the impres
sion. But it is not like that: the conceptual contents that are most basic
in this sense are already possessed by impressions themselves, im

pingements by the world on our sensibility.
This makes room for a different notion of givenness, one that is
innocent of the confusion between j ustification and exculpation. Now
we need not try to make out that the space of reasons is more exten
sive than the space of concepts. When we trace the ground for an em
pirical judgement, the last step takes us to experiences. Experiences
already have conceptual content, so this last step does not take us out
side the space of concepts. But it takes us to something in which sen
sibility-receptivity-is operative, so we need no longer be unnerved
by the freedom implicit in the idea that our conceptual capacities
belong to a faculty of spontaneity. We need not worry that our pic
ture leaves out the external constraint that is required if exercises of
our conceptual capacities are to be recognizable as bearing on the
world at all.
5. I said (§4) that when we enjoy experience conceptual capacities

are drawn on in receptivity, not exercised on some supposedly prior
deliverances of receptivity. And it is not that I want to say they are
exercised on something else. It sounds off key in this connection to
speak of exercising conceptual capacities at all. That would suit an
activity, whereas experience is passive.8 In experience one finds oneself
saddled with content. One's conceptual capacities have already been
brought into play, in the content's being available to one, before one
has any choice in the matter. The content is not something one has put
together oneself, as when one decides what to say about something. In
fact it is precisely because experience is passive, a case of receptivity in
operation, that the conception of experience I am recommending can
satisfy the craving for a limit to freedom that underlies the Myth of the
Given.
Because experience is passive, the involvement of conceptual capac
ities in experience does not by itself provide a good fit for the idea of
8. Of course this is not to deny that experiencing the world involves activity. Searching
is an activity; so are observing, watching, and so forth. (This sort of thing is usefully stressed
by people who think we should not conceive experience as passive reception at all, such as
J. ]. Gibson; see, for instance, The Senses Considered as Perceptual Systems (George Allen
and Unwin, London, 1 968).) But one's control over what happens in experience has limits:
one can decide where to place oneself, at what pitch to tune one's attention, and so forth,
but it is not up to one what, having done all that, one will experience. This minimal point
is what I am insisting on.
a faculty of spontaneity. That may make it seem that I am not really

disarming the Myth of the Given, but merely rejecting the terms that
pose the apparent problem it responds to. What generates the tempta
tion to appeal to the Given is the thought that spontaneity character
izes exercises of conceptual understanding in general, so that sponta
neity extends all the way out to the conceptual contents that sit closest
to the impacts of the world on our sensibility. We need to conceive
this expansive spontaneity as subject to control from outside our
thinking, on pain of representing the operations of spontaneity as a
frictionless spinning in a void. The Given seems to supply that exter
nal control. And now it may seem that when I stress that experience is
passive, I am dissolving the temptation by simply denying that spon
taneity extends all the way out to the content of experience, even
though I claim that conceptual capacities are operative in experience.
But it is not like that. The craving for external friction in our picture
of spontaneity is not something we can satisfy in that way, by simply
restricting the scope of spontaneity, making it less extensive than the
sphere of the conceptual.
We would not be able to suppose that the capacities that are in play
in experience are conceptual if they were manifested only in experi
ence, only in operations of receptivity. They would not be recogniz
a ble as conceptual capacities at all unless they could also be exercised
in active thinking, that is, in ways that do provide a good fit for the
idea of spontaneity. Minimally, it must be possible to decide whether
or not to judge that things are as one's experience represents them to
be. How one's experience represents things to be is not under one's
control, but it is up to one whether one accepts the appearance or
rejects it.9 Moreover, even if we consider only judgements that register
experience itself, which are already active in that minimal sense, we
must acknowledge that the capacity to use concepts in those judge
ments is not self-standing; it cannot be in place independently of a
capacity to use the same concepts outside that context. That is so even
with the concepts that are most immediately linked to the subjective
character of experience itself, the concepts of secondary qualities.
Q uite generally, the capacities that are drawn on in experience are
recognizable as conceptual only against the background of the fact
9. The point here is well illustrated by familiar illusions. In the Miiller-Lyer illusion,
one's experience represents the two lines as being une�ally long, but someone in the know
will refrain from judging that that is how things are.
that someone who has them is responsive to rational relations, which

link the contents of judgements of experience with other judgeable
contents. These linkages give the concepts their place as elements in
possible views of the world.
For example, consider j udgements of colour. These j udgements in
volve a range of conceptual capacities that are as thinly integrated into
understanding ·of the world as any. Even so, no one could count as
making even a directly observational j udgement of colour except
against a background sufficient to ensure that she understands colours
as potential properties of things. The ability to produce "correct" col
our words in response to inputs to the visual system (an ability pos
sessed, I believe, by some parrots) does not display possession of the
relevant concepts if the subject has no comprehension of, for instance,
the idea that these responses reflect a sensitivity to a kind of state of
affairs in the world, something that can obtain anyway, independently
of these perturbations in her stream of consciousness. The necessary
background understanding includes, for instance, the concept of visi
ble surfaces of objects and the concept of suitable conditions for tell
ing what colour something is by looking at it.10
Of course the concepts that can figure in the content of experience
are not restricted to concepts of secondary qualities. Once we take
that into account, it becomes even clearer that the passive operation of
conceptual capacities in sensibility is not intelligible independently of
their active exercise in judgement, and in the thinking that issues in
judgement.
The conceptual capacities that are passively drawn into play in ex
perience belong to a network of capacities for active thought, a net
work that rationally governs comprehension-seeking responses to the
impacts of the world on sensibility. And part of the point of the idea
that the understanding is a faculty of spontaneity-that conceptual
capacities are capacities whose exercise is in the domain of responsible
freedom-is that the network, as an individual thinker finds it govern
ing her thinking, is not sacrosanct. Active empirical thinking takes
place under a standing obligation to reflect about the credentials of
the putatively rational linkages that govern it, There must be a stand-
1 0. For an elaboration of points of this kind, see Sellars, " Empiricism and the Philosophy
of Mind", § § 10-20.
Lecture l. Concepts and Intuitions 13
ing willingness t o refashion concepts and conceptions i f that is what

reflection recommends. No doubt there is no serious prospect that we
might need to reshape the concepts at the outermost edges of the sys
tem, the most immediately observational concepts, in response to
pressures from inside the system. But that no-doubt unreal prospect
brings out the point that matters for my present purpose. This is that
although experience itself is not a good fit for the idea of spontaneity,
even the most immediately observational concepts are partly consti
tuted by their role in something that is indeed appropriately conceived
in terms of spontaneity.1 1
So we cannot simply insulate the passive involvement of conceptual
capacities in experience from the potentially unnerving effects of the
freedom implied by the idea of spontaneity. If we think that the way
to exploit the passivity of experience is to deny that spontaneity ex
tends all the way out to the content of experience, we merely fall back
into a misleadingly formulated version of the Myth of the Given. If we
try to keep spontaneity out of the picture but nevertheless talk of con
ceptual capacities operating in experience, the talk of conceptual ca
pacities is mere word-play. The trouble about the Myth of the Given
is that it offers us at best exculpations where we wanted j ustifications.
That trouble shows up again here, in connection with impingements
on spontaneity by the so-called conceptual deliverances of sensibility.
If those impingements are conceived as outside the scope of spontane
ity, outside the domain of responsible freedom, then the best they can
yield is that we cannot be blamed for believing whatever they lead us
to believe, not that we are justified in believing it.
I am not, then, proposing a cheap defeat of the Given, to be
achieved by exploiting the fact that experience is passive so as to keep
experience out of the scope of spontaneity. The view I am recom
mending is that even though experience is passive, it draws into oper
ation capacities that genuinely belong to spontaneity.
6. It need not be a mere superficial oversight if someone fails to see a

possibility here-if someone cannot understand how capacities that
1 1 . I intend the imagery of this paragraph to be reminiscent of the well-known closing

section of W. V. Quine's classic paper "Two Dogmas of Empiricism", in his From a Logical
Point of View (Harvard University Press, Cambridge, Mass., 1961; 1st ed. 1 953), pp. 20-
46.
belong to spontaneity could be inextricably implicated in an operation

of mere receptivity. It can be difficult to see a way out here, and the
roots of the difficulty lie deep.
I want to provide a first illustration of this from Davidson himself.
In a well-known paper in which he recommends a coherence ap
proach to truth and knowledge, 12 Davidson shows a blind spot for the
way out that I h ave described. He does not argue against it; it simply
does not figure among the possibilities that he contemplates.
Davidson is clear that if we conceive experience in terms of impacts
on sensibility that occur outside the space of concepts, we must not
think we can appeal to experience to justify judgements or beliefs.
That would be to fall into the Myth of the Given, with its confusion
of justification and exculpation. The space of reasons does not extend
further than the space of concepts, to take in a bare reception of the
Given. So far, this is just what I have been urging.
But Davidson thinks experience can be nothing but an extra-con
ceptual impact on sensibility. So he concludes that experience must be
outside the space of reasons. According to Davidson, experience is
causally relevant to a subject's beliefs and judgements, but it has no
bearing on their status as justified or warranted. Davidson says that
"nothing can count as a reason for holding a belief except another
belief" (p. 310), and he means in particular that experience cannot
count as a reason for holding a belief.
Of course I agree with the point this train of thought starts from.
But the conclusion is quite unsatisfying. Davidson recoils from the
Myth of the Given all the way to denying experience any j ustificatory
role, and the coherentist upshot is a version of the conception of spon
taneity as frictionless, the very thing that makes the idea of the Given
attractive. This is just one of the movements in the oscillation that I
have spoken of. There is nothing to prevent it from triggering the fa
miliar recoil in its tum. Davidson's picture depicts our empirical
thinking as engaged in with no rational constraint, but only causal
influence, from outside. This just raises a worry as to whether the pic
ture can accommodate the sort of bearing on reality that empirical
content amounts to, and that is just the kind of worry that can make
an appeal to the Given seem necessary. And Davidson does nothing to
12. "A Coherence Theory of Truth and Knowledge", reprinted in Ernest LePore, ed.,
Truth and Interpretation: Perspectives on the Philosophy ofDonald Davidson (Basil Black
well, Oxford, 1 98 6), pp. 307-1 9.
allay the worry. I think we should be suspicious of his bland

confidence that empirical content can be intelligibly in our picture
even though we carefully stipulate that the world's impacts on our
senses have nothing to do with justification.
Of course Davidson believes that his position is a place where
thought can come to rest, not a movement in an interminable oscilla
tion. But I think he contrives to make it seem so only by going insuffi
ciently deeply into the motivation of the Myth of the Given.
He remarks that a foundationalist conception of experience "leads
to skepticism" (p. 3 14). Of course it is true that appealing to the
Given gets us nowhere in epistemology. But it is not true that philo
sophical worries about scepticism arise out of the failure of the idea of
the Given. And the idea of the Given is not something that comes to
us in calm reflection, as if from nowhere, as a possible basis for the
epistemology of empirical knowledge, so that we can cheerfully drop
it when we see that it does not work . Rather, the idea of the Given is
a response to a way of thinking that underlies the familiar philosoph
ical anxiety about empirical knowledge, and that way of thinking is
precisely what Davidson endorses.
We can seem to be forced into the idea of the Given; that is what
happens when we are impressed by the thought that conceptual ca
pacities belong to a faculty of spontaneity, and fall into worrying that
our picture deprives itself of the possibility that exercises of concepts
could be what it depicts, because it leaves out any rational constraint
from outside the sphere of thought. One form that this worry takes is
the fear that we have no convincing way to credit ourselves with em
pirical knowledge. The recoil to the Given that results from this
worry-whether in its generic form (how can it be that exercises of
spontaneity bear on a reality outside the sphere of thinking at all?) or
in its specifically epistemological form (how can exercises of sponta
neity amount to knowledge?)-is a natural response to the very sort
of "coherence theory of truth and knowledge" that Davidson recom
mends. Such theories express precisely the unnerving idea that the
spontaneity of conceptual thinking is not subject to rational con
straint from outside. Coherentist rhetoric suggest� images of confine
ment within the sphere of thinking, as opposed to being in touch with
something outside it. To one who finds such imagery both appropriate
and worrying, the idea of the Given can give the appearance of rein
'
stating thought's bearing on reality. And at this point in the dialectic,
it is no good pointing out that the appearance is illusory, that the idea
of the Given does not fulfil its apparent promise, if the worry that can
make the idea nevertheless seem inescapable remains urgent, or is even
exacerbated. The effect: is simply to bring out that neither of the two
positions that we are being asked to choose between is satisfying.
Davidson does nothing to discourage us from taking his coherentist
rhetoric in terms of confinement imagery. On the contrary, he posi
tively encourages it: At one point he says, "Of course we can't get
outside our skins to find out what is causing the internal happenings
of which we are aware" (p. 3 1 2). This is, as it stands, a very unsatis
factory remark. Why should we suppose that to find out about exter
nal objects we would have to get outside our skins? (Of course we
cannot do that.) And why should we suppose that we are interested in
finding out what is causing internal happenings of which we are
aware, rather than that we are simply interested in the layout of the
environment? Of course getting outside our skins is not the same as
getting outside our thoughts. But perhaps we can understand how D a
vidson can be so casual in this remark if we take it that our literal
confinement inside our skins strikes him as an analogue to a meta
phorical confinement inside our beliefs, which he is happy to let his
coherentism imply. Davidson's picture is that we cannot get outside
our beliefs.
Of course Davidson knows that such confinement imagery tends to
prompt a recoil to the idea of the Given, the idea that truth and
knowledge depend on rational relations to something outside the con
ceptual realm. He thinks he can allow free rein to confinement imag
ery, but pre-empt the recoil by arguing, within his coherentist frame
work, for the evidently reassuring thesis that "belief is in its nature
veridical" (p. 3 14). Davidson argues for that thesis by connecting be
lief with interpretation, and urging that it is in the nature of interpre
tation that an interpreter must find her subjects mostly right about the
world with which she can observe them causally interacting.
I do not want to dispute that argument. But I do want to raise the
question how effectively it can reassure us, if we are worried about
whether Davidson's coherentist picture can incorporate thought's
bearing on reality. Suppose one feels the worry in this familiar form:
so far as the picture goes, one might be a brain in a mad scientist's vat.
The Davidsonian response seems to be that if one were a brain in a
vat, it would be correct to interpret one's beliefs as being largely true
Lecture 1. Concepts ana lntutttons ll
beliefs about the brain's electronic environment.13 But is that the reas
surance we need if we are to be immunized against the attractions of
the Given? The argument was supposed to start with the body of be
liefs to which we are supposed to be confined, in our active efforts to
suit our thinking to the available justifications. It was supposed to
make the confinement imagery unthreatening by reassuring us that
those beliefs are mostly true. But the response to the brain-in-a-vat
worry works the wrong way round. The response does not calm the
fear that our picture leaves our thinking possibly out of touch with the
world outside us. It j ust gives us a dizzying sense that our grip on what
it is that we believe is not as firm as we thought.14
I think the right conclusion is chis: whatever credence we give to
Davidson's argument that a body of belief is sure to be mostly true,
the argument starts too late to certify Davidson's position as a genuine
escape from the oscillation.
The only motivation for the Myth of the Given that figures in
Davidson's thinking is a shallow scepticism, in which, taking it for
granted that one has a body of beliefs, one worries about their creden
tials. But the Myth of the Given has a deeper motivation, in the
thought that if spontaneity is not subject to rational constraint from
outside, as Davidson's coherentist position insists that it is not, then
we cannot make it intelligible to ourselves how exercises of spontane
ity can represent the world at all. Thoughts without intuitions are
empty, and the point is not met by crediting intuitions with a causal
impact on thoughts; we can have empirical content in our picture only
if we can acknowledge that thoughts and intuitions are rationally con-
1 3 . We have this on the testimony of Richard Rorty; see p. 340 of his "Pragmatism,
Davidson, and Truth" , in LePore, ed., Truth and Interpretation, pp. 3 33-55.
14. It takes care to say precisely why the response is unsatisfying. It is not that we are
being told we may be egregiously wrong about what our beliefs are about. If I protest that
some belief of mine is not about electronic impulses or whatever but a bout, say, a book, the
reply can be: "Certainly your belief is about a book-given how 'a book' as you use the
phrase is correctly interpreted. " The envisaged reinterpretation, to suit the hypothesis that I
am a brain in a vat, affects my higher-level beliefs about what my first-level beliefs are about
in a way that precisely matches its effect on my first-level beliefs. The problem is that in the
argument Rorry attributes to Davidson, we ring changes on the actual environment (as seen
by the interpreter and brought into the interpretation) without changing how things strike
the believer, even while the interpretation is supposed to capture how the believer is in
touch with her world. This strikes me as making it impossible to claim that the argument
traffics in any genuine idea of being in touch with something in particular. The objects that
the interpreter sees the subject's beliefs as being about become, as it were, merely noumenal
so far as the subject is concerned.
nected. By rejecting that, Davidson undermines his right to the idea

that his purportedly reassuring argument starts from, the idea of a
body of beliefs. In that case his attempt to disarm the confinement
imagery does not work, and his position is exposed as a version of one
phase in the oscillation. A genuine escape would require that we avoid
the Myth of the Given without renouncing the claim that experience
is a rational constraint on thinking.
I have suggested that we can do that if we can recognize that the
world's impressions on our senses are already possessed of conceptual
content. But there is a block that prevents Davidson from seeing
any possibilities in this direction. I shall come back to this in later lec
tures.
7. The Myth of the Given expresses a craving for rational constraint

from outside the realm of thought and judgement. This craving is
most familiar in connection with empirical knowledge of the world
about us: knowledge yielded by what Kant calls "outer sense" .15 But
it is instructive to see that the spatial phrase that I have j ust used,
"outside the realm of thought and judgement", is only metaphorical.
Exactly the same temptation arises in connection with what Kant calls
"inner sense" .16 The realm of thought and j udgement includes judge
ments about the thinker's own perceptions, thoughts, sensations, and
the like. The conceptual capacities that are operative in such judge
ments must belong to spontaneity j ust as much as any other concep
tual capacities do, and that can generate the spectre of a frictionless
spinning in a void for this region of thought too. Then, in the way that
should by now be familiar, ensuring friction, which is required for
genuine content, can seem to oblige us to take exercises of concepts in
this region to be rationally grounded in something extra-conceptual,
bare presences that are the ultimate grounds of j udgements.
This description of an apparently compulsory way of thinking fits
Wittgenstein's target in the so-called Private Language Argument. If
we understand that polemic as applying a general rejection of the
Given, we make available to ourselves a sharp appreciation of its co
gency. And perhaps we can also acquire a richer understanding of the
1 5 . For instance at B67.
1 6. For instance at A22/B37. For the two phrases together in a single discussion, see the
footnote at Bxxxix-xli.
general point by considering the way it shows up in those familiar

passages of Wittgenstein. 17
I say " so-called Private Language Argument" because according to
this reading the main point of the conception Wittgenstein attacks is
to claim that "judgements of inner sense" are ultimately grounded on
bare presences, rather than to devise a way to put the bare presences
into words. If someone in the grip of the conception was convinced by
an argument that language could not embrace the supposed items she
insists on, she might reply that that is really just her point. If language
could embrace them, that would mean they were within the concep
tual sphere, and the point of acknowledging them is to acknowledge
something that constrains spontaneity, which moves within that
sphere, from outside. So certainly language cannot capture them; but
still, it can seem necessary to insist, they are there to be pointed to as
the ultimate justifications for judgements of "inner sense". The funda
mental thrust of Wittgenstein's attack is not to eliminate the idea of a
private language, which by itself would merely push the line of
thought that he opposes to this point. Wittgenstein's attack under
mines even this position, which has already given up the idea of a
private language, by applying the general moral: a bare presence can
not be a ground for anything.
However, if one becomes convinced that the ultimate grounds for
judgements of experience must be bits of the Given, one will naturally
take oneself to be committed to the possibility of concepts that sit as
closely as possible to those ultimate grounds, in the sense that their
content is wholly determined by the fact that judgements involving
them are warranted by the right sort of bare presence. These concepts
will be the concepts that are supposed to be expressible by the words
of a private language. Only one person could be the subject to whom
a particular bit of the Given is given. So any concept that was consti
tuted by a justificatory relation to a bare presence would have to be,
to that extent, a private concept. It would be natural to suppose that
these private concepts are acquired by abstraction from a manifold
Given, as in the story about concept-formation that I mentioned ear-
1 7. I have said a little more about this reading of Wittgenstein in "One Strand in the
Private Language Argument", Grazer Phi/osophische Studien 33/34 ( 1 989), 285-303. See
also my " Intentionality and Interiority in Wittgenstein", in Klaus Puhl, ed., Meaning Scep
ticism (De Gruyter, Berlin, 1991), pp. 148-69.
tier. The work of abstraction is the private ostensive definition that

figures in Wittgenstein's polemic.
So situating the Private Language Argument in the context of a gen
eral rejection of the Given does not dispense us from considering the
argument against private language, or at any rate against private con
cepts, as such; Still, I think we put those of Wittgenstein's remarks
that are specifically about private language in the right light if we see
them as efforts to insist on the consequences for language of the gen
eral point: that a bare presence cannot supply a j ustificatory input into
a conceptual repertoire from outside it, the sort of thing the connec
tion between concepts and spontaneity made us hanker for. If a con
cept is constituted by a justificatory linkage to a bare presence, which
is what its being a private concept would amount to, then spontaneity
does not extend as far as it. In fact it is really the point of the concep
tion to exempt exercises of these supposed concepts from the respon
sibility that comes with spontaneity. What we have here is a version of
a structure I mentioned earlier (§5), in connection with a misconcep
tion of the fact that experience is passive. Calling something to which
spontaneity does not extend "a concept", and calling the linkage "ra
tional", is fraudulent labelling: in effect, labelling a mere exculpation
a justification, in the vain hope that that could make it be one.
I mentioned Geach's Wittgensteinian attack on the abstractionist
account of concept-formation ( §2). I have been suggesting that the
Private Language Argument applies the general rejection of the Given.
But it would be misleading to represent the Private Language Argu
ment as a particular application of a more general thought. As I said,
any concept that was constituted by a j ustificatory relation to a bare
presence would have to be a private concept. Making the abstraction
that would be necessary to form such a concept would be giving one
self a private ostensive definition. In effect the idea that concepts can
be formed by abstraction from the Given just is the idea of private
ostensive definition. So the Private Language Argument j ust is the re
jection of the Given, in so far as it bears on the possibilities for lan
guage; it is not an application of a general rejection of the Given to a
particular area. What is an application of the general point is the re
jection of bare presences as what sensations and so forth are.
It makes no difference if the right occasions for performing a pri
vate ostensive definition are supposed to be signalled by other people.
That is a way in which one might hope to integrate a private element,
rational responsiveness to a bare presence, into a composite concept

that has a public aspect, a rational linkage into a shareable conceptual
repertoire, as well. Wittgenstein expresses the idea in this passage:
" Or is it like this: the word 'red' means something known to everyone;
and in addition, for each person, it means something known only to
him ? (Or perhaps rather: it refers to something known only to
him. ) " 1 8 If the idea of rational responsiveness to a bare presence is a
confusion, this linkage into a shareable repertoire cannot save these
supposedly composite concepts from being vitiated by their private
ingredients. The confusion between j ustification and exculpation just
shows up at the j oint between the supposed ingredients of the com
posite concept.
The Myth of the Given is especially insidious in the case of " inner
sense " . In the case of "outer sense", the idea is that the Given mediates
between the experiencing subject and an independent outer reality, of
which the subject is aware through this mediation. If we reject the
Given, we are not thereby abolishing the outer reality, but merely
obliging ourselves not to suppose that awareness of it is mediated in
that way. But the objects of "inner sense" are internal accusatives to
the awareness that " inner experiences" constitute; they have no exis
tence independently of that awareness.19 That means that if we let
bare presences into the picture, they figure as the only objects of
awareness in play; they cannot figure as mediating an awareness of
something else beyond them, at any rate if the mediated awareness is
itself to be conceived in terms of " inner sense" at work.20 And the
result is that when we reject the Given here, we can seem to be reject
ing " inner" awareness altogether. There seems to be nothing else for
"inner experience" to be experience of.
How can we reject the Given without thus obliterating " inner"
awareness? To give the impressions of " inner sense" the right role in
justifying judgements, we need to conceive them, like the impressions
of "outer sense ", as themselves already possessing conceptual content;
to supply the necessary limit to the freedom of spontaneity, we need
1 8. Philosophical Investigations, trans. by G. E. M. Anscom be (Basil Blackwell, Oxford,

1 9 5 1 ), §273.
19. See P. F. Strawson, The Bounds of Sense (Methuen, London, 1 966), pp. 100-1.
20. Not that "inner experience" cannot mediate awareness. For instance, a certain sen
sation might yield a mediated awareness of a certain bodily condition. But here the object
of the mediated awareness is not "inner" in the Kantian sense. See the text below.
to insist that they are indeed impressions, products of receptivity. So

the impressions of " inner sense" must be, like the impressions of
"outer sense" , passive occurrences in which conceptual capacities are
drawn into operation. But if we are to respect the point about internal
accusatives, we cannot conceive these passive operations of concep
tual capacities exactly on the model of the impressions of "outer
sense " . We cannot suppose that these operations of conceptual capac
ities constitute awareness of circumstances that obtain in any case,
and that impress themselves on a subject as they do because of some
suitable relation to her sensibility. No doubt there are circumstances
that obtain in any case, and figure in the aetiology of impressions of
"inner sense" : for instance, bodily damage in the case of feelings of
pain. But if we are to respect the point about internal accusatives, we
cannot suppose that such circumstances are objects of an awareness
that is constituted by the impressions of "inner sense". (Although one
can no doubt learn to find out about such circumstances from the
impressions of " inner sense".) If we can make out that judgements of
"inner sense" are about anything, it has to be that they are a bout the
impressions of " inner sense" themselves, not about something inde
pendent of which the impressions constitute awareness.
This is a very difficult area. Wittgenstein himself sometimes seems
to betray an understandable wish to duck the difficulties. What I have
in mind here is the fact that he sometimes seems to toy with denying
that self-ascriptions of sensation are assertions, articulations of judge
ments about states of affairs, at all.21
I have claimed that we should connect "inner experience" with con
ceptual capacities, so as to think about "inner sense" in parallel with
"outer sense" to the fullest extent that is possible. One obvious source
of difficulty about this is that creatures with no faculty of spontaneity
can surely, for instance, feel pain. (Remember that "spontaneity" al
ludes to conceptual capacities. I am not trying to obliterate the self
moving character of merely animal life.) I shall come back to this in
later lectures. My aim here has been not to wrap the matter up, but
21. For instance, p. 68 of The Blue Book (in The Blue and Brown Books [Basil Black
well, Oxford, 1958]): "The difference between the propositions 'I have pain' and 'he has
pain' is not that of 'L. W. has pain' and 'Smith has pain'. Rather it corresponds to the
difference between moaning and saying that someone moans.» I would not dispute what is
said in the first of these sentences. But sentences like the second have suggested, at least to
some commentators, a doctrine that assimilates "avowals" to other modes of expression, so
as to distance them from assertions, and that strikes me as a cop-out.
only to introduce the suggestion that we should read the Private Lan
guage Argument as an attack on the Given.
8. In this lecture, I have claimed that we are prone to fall into an

intolerable oscillation: in one phase we are drawn to a coherentism
that cannot make sense of the bearing of thought on objective reality,
and in the other phase we recoil into an appeal to the Given, which
turns out to be useless. I have urged that in order to escape the oscil
lation, we need a conception of experiences as states or occurrences
that are passive but reflect conceptual capacities, capacities that be
long to spontaneity, in operation. In the next lecture I shall start to
consider some difficulties about this conception.
L E CTUR E I I
The Unboundedness
of the Conceptual
1 . In my first lecture I talked about a tendency to oscillate between a

pair of unsatisfying positions: on the one side a coherentism that
threatens to disconnect thought from realiry, and on the other side a
vain appeal to the Given, in the sense of bare presences that are sup
posed to constitute the ultimate grounds of empirical judgements. I
suggested that in order to escape the oscillation, we need to recognize
that experiences themselves are states or occurrences that inextricably
combine receptivity and spontaneity. We must not suppose that spon
taneity first figures only in judgements in which we put a construction
on experiences, with experiences conceived as deliverances of recep
tivity to whose constitution spontaneity makes no contribution. Expe
riences are indeed receptivity in operation; so tht:y can satisfy the need
for an external control on our freedom in empirical thinking. But con
ceptual capacities, capacities that belong to spontaneity, are already at
work in experiences themselves, not just in judgen' •:nts based on
them; so experiences can intelligibly stand in rational relations to our
exercises of the freedom that is implicit in the idea of spontaneity.
In this second lecture I am going to start considering problems
about this conception.
It can be hard to see that there is room for such a conception. In my
first lecture (§6), I introduced Davidson's coherentism to illustrate
that; I suggested that Davidson's position is representative of a style of
thinking within which what I am urging does not even come into view
as an option. Later (starting in Lecture IV), I shall try to say something
about why the conception is hard to achieve, so that we are tempted
to suppose we are stuck with a choice between the original pair of
tect rtre I I . I 'he Lllllm1111ded11ess of the Co11 C/lllwt 2 .:i
positions. My aim will be to suggest that our diffic11lties in this area

result from an intelligibly powerful influence over the cast of our
thoughts, from which, however, we can liberate ourselves.
But that is not on my agenda for this lecture. I do not want to sug
gest that the objection I am going to consider now is itself revelatory
of the deep roots of our difficulties. At best it may not be unconnected.
But I do hope that discussing it will put the conception I am recom
mending in a clearer light.
2. What I want to consider in this lecture is an objection on the score

of idealism.
We seem to need rational constraints on thinking and judging, from
a reality external to them, if we are to make sense of them as bearing
on a reality outside thought at all. Davidson denies that there is any
such need, and proposes that we make do with nothing but causal
constraints. My suggestion has been that he can contrive to be com
fortable with that only because he thinks there is no alternative, since,
as he clearly realizes, the Myth of the Given is hopeless. And here, I
claim, he is wrong. There is an alternative, and this fact removes the
only apparent reason to deny the need for rational constraints from
outside: the only apparent reason to deny that thoughts without a ra
tional connection with intuitions would be empty.
When we try to acknowledge the need for external rational con
straint, we can find ourselves supposing there must be relations of ul
timate grounding that reach outside the conceptual realm altogether.
That idea is the Myth of the Given, and of course the conception I
have described makes no concession to it. The Myth of the Given is
precisely one of the two opposing pitfalls from which the conception
is intended to liberate us.
In the conception I am recommending, the need for external con
straint is met by the fact that experiences are receptivity in operation.
But that does not disqualify experiences from playing a role in
justification, as the counterpart thought in the Myth of the Given
does, because the claim is that experiences themselves are already
equipped with conceptual content. This joint involvement of receptiv
ity and spontaneity allows us to say that in experience one can take in
how things are. How things are is independent of one's thinking (ex
cept, of course, in the special case in which how things are is that one
thinks such-and-such). By being taken in in experience, how things
anyway are becomes available to exert the required rational control,

originating outside one's thinking, on one's exercises of spontaneity.
Certainly one can be misled, at least in the case of "outer experi
ence". I have postponed any discussion of that until my final lecture.
But I insist in advance that when we acknowledge the possibility of
being misled, we do not deprive ourselves of "taking in how things
are" as a description of what happens when one is not misled. In a
particular exp'!rience in which one is not misled, what one takes in is
that things are thus and so. That things are thus and so is the content
of the experience, and it can also be the content of a judgement: it
becomes the content of a judgement if the subject decides to take the
experience at face value. So it is conceptual content. But that things
are thus and so is also, if one is not misled, an aspect of the layout of
the world: it is how things are. Thus the idea of conceptually struc
tured operations of receptivity puts us in a position to speak of expe
rience as openness to the layout of reality. Experience enables the lay
out of reality itself to exert a rational influence on what a subject
thinks.
This image of openness to reality is at our disposal because of how
we place the reality that makes its impression on a subject in experi
ence. Although reality is independent of our thinking, it is not to be
pictured as outside an outer boundary that encloses the conceptual
sphere. That things are thus and so is the conceptual content of an
experience, but if the subject of the experience is not misled, that very
same thing, that things are thus and so, is also a perceptible fact, an
aspect of the perceptible world.
Now it can seem that this refusal to locate perceptible reality out
side the conceptual sphere must be a sort of idealism, in the sense in
which to call a position " idealism" is to protest that it does not genu
inely acknowledge how reality is independent of our thinking. Jf that
were right, my affirmation of reality's independence would be disin
genuous, mere lip-service. But though this objection is easy to under
stand, and even to sympathize with, it is wrong. It reflects the convic
tion that we have to choose between a coherentist denial that thinking
and judging are subject to rational constraint from outside, on the one
hand, and an appeal to the Given as what imposes the constraint, on
the other. If someone takes it that those are the only options, and if
she has a firmer grip on the defects of unconstrained coherentism than
she has on the uselessness of the Given, then anything short of believ-
ing in the Given will strike her as slighting the independence of reality.
Bur the point of the third option, the option I am urging, is precisely
that it enables us to acknowledge that independent reality exerts a
rational control over our thinking, but without falling into the confu
sion between justification and exculpation that characterizes the ap
peal to the Given.
3. I find it helpful in this connection to reflect on a remark of

Wittgenstein's: "When we say, and mean, that such-and-such is the
case, we-and our meaning-do not stop anywhere short of the fact;
but we mean: this-is-so. "1 Wittgenstein calls this a paradox. That
is because, especially in conjunction with the fact that "thought can be
of what is not the case", 2 it can prompt a reaction in which our minds
boggle over what seems a miraculous power of thinking in the most
general sense, in this case meaning what one says, to "catch reality in
its net". 3 But Wittgenstein also says, rightly, that the remark " has the
form of a truism,.
We can formulate the point in a style Wittgenstein would have been
uncomfortable with: there is no ontological gap between the sort of
thing one can mean, or generally the sort of thing one can think, and
the sort of thing that can be the case. When one thinks truly, what one
thinks is what is the case. So since the world is everything that is the
case (as he himself once wrote),4 there is no gap between thought, as
such, and the world. Of course thought can be distanced from the
world by being false, but there is no distance from the world implicit
in the very idea of thought.
But to say there is no gap between thought, as such, and the world
is just to dress up a truism in high-flown language. All the point comes
to is that one can think, for instance, that spring has begun, and that
very same thing, that spring has begun, can be the case. That is truis
tic, and it cannot embody something metaphysically contentious, like
slighting the independence of reality. When we put the point in the
high-flown terms, by saying the world is made up of the sort of thing
1. Philosophical Investigations, �95.
2. S95 continues: "But this paradox (which has the form of a truism) can also he e x ·
pressed in this way: Thought can be of what is 11ot the case."
3 . See Philosophical Investigations, §429: "How was it possible for thought to deal wirh
the very object itself? We feel as if by means of it we had caught reality in our net."
4. Tractatus Logico·Philosophiws, trans. D. F. Pca rs.and B. F. McGuin ness ( Rou rled�e
and Ke�,;an Paul, London, 1 9 6 1 ), § 1 .
one can think, a phobia of idealism can make people suspect we are
renouncing the independence of reality_:_as if we were representing
the world as a shadow.of our thinking, or even as made of some men
tal stuff. But we might just as well take the fact that the sort of thing
one can think is the same as the sort of thing that can be the case the
other way round, as an invitation to understand the notion of the sort
of thing one can think in terms of a supposedly prior understanding of
the sort of thing that can be the case. 5 And in fact there is no reason to
look for a priority in either direction.
If we say that there must be a rational constraint on thought from
outside it, so as to ensure a proper acknowledgement of the indepen
dence of rea lity, we put ourselves at the mercy of a familiar kind of
ambiguity. "Thought" can mean the act of thinking; but it can also
mean the content of a piece of thinking: what someone thinks. Now if
we are to give due acknowledgement to the independence of reality,
what we need is a constraint from outside thinking and ;udging, our
exercises of spontaneity. The constraint does not need to be from out
side thinkable contents. It would indeed slight the independence of
reality if we equated facts in general with exercises of conceptual ca
pacities-acts of thinking-or represented facts as reflections of such
things; or if we equated perceptible facts in particular with states or
occurrences in which conceptual capacities are drawn into operation
in sensibility-experiences-or represented them as reflections of such
things. But it is not idealistic, as that would be, to say that perceptible
facts are essentially capable of impressing themselves on perceivers in
states or occurrences of the latter sort; and that facts in general are
essentially capable of being embraced in thought in exercises of spon
taneity, occurrences of the former sort.
The fact that experience is passive, a matter of receptivity in opera
tion, should assure us that we have all the external constraint we can
reasonably want. The constraint comes from outside thinking, but not
from outside what is thinkable. When we trace j ustifications back, the
last thing we come to is still a thinkable content; not something more
5. The Tractatus is often read on these lines; for a recent version, see David Pears, The
False Prison, vol. I (Clarendon Press, Oxford, 1 9 87). Opponents of the kind of reading
Pears gives sometimes tend to find in the Tractatus a thesis of priority in the opposite dircc·
tion, or at least not to distinguish their interpretations clearly from this kind of thing. (That
might merit a protest of idealism.) But I doubt w hether either claim of priority is to be foun d
in t h e Tractatus.
Lectltre II. The Unbounded11ess uf tiJe Cm:t"eptual 29
ultimate than that, a bare pointing t o a bit of the Given. But these final
thinkable contents are put into place in operations of receptivity, and
that means that when we appeal to them we register the required con
straint on thinking from a reality external to it. The thinkable con
tents that are ultimate in the order of justification are contents of ex
periences, and in enjoying an experience one is open to manifest facts,
facts that obtain anyway and impress themselves on one's sensibility.
(At any rate one seems to be open to facts, and when one is not misled,
one is.} To paraphrase Wittgenstein, when we see �at such-and-such
is the case, we, and our seeing, do not stop anywhere short of the fact.
What we see is: that such-and-such is the case.
4. Wittgenstein's aphorism can be reworked like that for any concep

tual shaping of subjectivity. It is not this general possibility by itself
that underwrites the image of openness. The image of openness is ap
propriate for experience in particular; and to bring the image into
play, we need to appeal to the distinctive passivity of experience. But
the general context matters for the availability of the image.
To bring out how, I want to recall something I said in my first lec
ture ( § 5 ) , to counter a misunderstanding of the idea that conceptual
capacities are passively operative in experience. The misunderstand
ing is to suppose that when we appeal to passivity, we insulate this
invocation of the conceptual from what makes it plausible to attribute
conceptual capacities in general to a faculty of spontaneity. Against
this, I urged that we could not recognize capacities operative in expe
rience as conceptual at all were it not for the way they are integrated
into a rationally organized network of capacities for active adjustment
of one's thinking to the deliverances of experience. That is what a
repertoire of empirical concepts is. The integration serves to place
even the most immediate judgements of experience as possible ele
ments in a world-view.
We can see this even if we restrict ourselves to concepts of second
ary qualities, which cannot be understood in abstraction from the
subjective character of experience. What it is for something to be red,
say, is not intelligible unless packaged with an understanding of what
it is for something to look red. The idea of being red does not go
beyond the idea of being the way red things look in the right circum
stances. That has an implication that I can put like this: although judg
ing that something is red is active, an exercise of spontaneity, it is as
little removed from the passivity of experience as any j udging could

be. Concepts of colour are only minimally integrated into the active
business of accommodating one's thinking to the continuing deliver
ances of experience, and hence only minimally integrated into poss:ble
views of the world. Still, they are so integrated, even if only minimally.
No subject could be recognized as having experiences of colour except
against a background understanding that makes it possible for judge
ments endorsing such experiences to fit into her view of the world. She
must be equipped with such things as the concept of visible surfaces of
objects, and the concept of suitable conditions for telling what
something's colour is by looking at it.
Those remarks, which restate something I said in my first lecture,
are about experiences and judgements chat locate colours in the ap
parent environment. There is another kind of colour experience that
we also need to make room for: the label "colour experience" can fit
mere sensations, operations of " i nner sense" . 6 For instance, a blow to
the head can cause one to "see red" without the ex perie nce s referring
'
the " seen" colour to the apparent environment. Now I have been urg
ing that experiences in general are states or occurrences in which con
ceptual capacities are passively drawn into operation. That had better
hold for these "inner experiences" of colour as much as for any other
experiences. And I believe we should understand the role of colour
concepts in these " inner experiences" derivatively from their role in
"outer experience" . The concept of red gets a grip, in characterizing
an "inner experience" of "seeing red " , because the experience is in the
relevant respect subjectively like the experience of seeing chat some
thing-some " outer" thing-is red, or at least seeming to.
It can be tempting to take things the other way round: to suppose
that the "inner" role of colour concepts is autonomously intelligible,
and to try to explain their " outer" role in terms of the idea that for an
" outer" object to fall under a colour concept is for it to be such as to
cause the appropriate visual " inner experience" in suitable viewing
conditions. We can be encouraged into this by the thought that if
being red and looking red are intelligible only in terms of each other,
that makes it a mystery how anyone can break into this circle: a mys
tery we might hope to dissipate by explaining both being red and
looking red in terms of the " inner" experience of " seeing red" .
6 . See lecture I , S7.

But we should resist this temptation. If the " inner" role o f colour
concepts were a self-standing starting point, " outer experience" of
colour would become impossible to comprehend. By what alchemy
could an " inner experience" be transmuted into an " outer experi
ence " ? If a colour first figures, in the development of our understand
ing, as a feature of " inner experience", not an apparent property of
objects, how could our understanding contrive to project that out into
the world? Starting from there, we might manage to externalize at best
a propensity to induce the relevant feature of " inner experience" in us.
But it is quite doubtful whether the idea of possessing such a propen
sity adds up to the idea of being appropriately coloured: that requires
precisely that our experience and thought locate something phenome
nal in the external world, whereas the "propensity" conception keeps
what is phenomenal in the mind.7 And anyway the circle-the mutual
dependence of the concepts of being red and looking red-is quite
innocent. It is no threat, for instance, co a sane view of how colour
concepts are acquired; we simply have to suppose they come only as
elements in a bundle of concepts that must be acquired together.8 So I
propose to focus on the role of colour concepts, and more generally
concepts of secondary qualities, in " outer experience " , taking it that
that is what is fundamental.
In "outer experience" , a subject is passively saddled with concep
tual contents, drawing into operation capacities seamlessly integrated
into a conceptual repertoire that she employs in the continuing activ
ity of adjusting her world-view, so as to enable it to pass a scrutiny of
its rational credentials. It is this integration that makes it possible for
us to conceive experience as awareness, or at least seeming awareness,
of a reality independent of experience. We can a ppreciate this point by
continuing to consider the way colours figure in the content of experi
ence. Even here, where the linkages into the whole system are mini
mal, the relevant conceptual capacities are integrated into spontaneity
at large, in a way that enables the subject to understand experiences in
which those conceptual capacities are drawn into operation as
7. It is one thing to gloss being red in terms of being such as to look red, and quite
another to gloss it in terms of being such as to induce a certain "inner experience" in us.
Note that "redft in " looking red" expresses a concept of "outer experience" no less than
does "red" in " being redft, in fact the very same concept. (Sellars insists on this point in
"Empiricism and the Philosophy of Mind".)
8. See Sellars, "Empiricism and the Philosophy of Migd", S S 1 8-ZO.
glimpses, o r at least seeming glimpses, of the world: takings in, at least

seemingly, of aspects Qf a reality that goes beyond what is manifest in
the experiences themselves. If a colour concept is drawn into opera
tion in an experience {with the concept playing its "outer" role), the
rational connections of the concept enter into shaping the content of
the appearance, so chat what appears co be the case is understood as
fraught with implications for the subject's cognitive situation in the
world: for instance, that she is confronted by an object with a facing
surface illuminated in such-and-such ways.
The notion of a glimpse is distinctively visual, but we can generalize
it to embrace non-visual experiences. By virtue of the way in which the
conceptual capacities that are drawn into operation in an experience
are rationally linked into the whole network, the subject of the expe
rience understands what the experience takes in {or at least seems to
take in) as part of a wider reality, a reality that is all embraceable in
thought but not all available to this experience. The object of experi
ence is understood as integrated into a wider reality, in a way that
mirrors how the relevant concepts are integrated into the repertoire of
spontaneity at large. Even in the case of colour experience, this inte
gration allows us to understand an experience as awareness of some
thing independent of the experience itself: something that is held in
place by its linkage into the wider reality, so that we can make sense
of the thought that it would be so even if it were not being experienced
to be so.
We have all this, I have been insisting, even if we restrict ourselves
to the way concepts of secondary qualities figure in the content of
experience. And the point would not be secured for this restricted case
by saying that we have to understand any particulat secondary-qual
ity experience against a background of other secondary-quality expe
riences, possible or actual. We cannot make a world, of which such
experiences might intelligibly be glimpses, entirely out of the distinc
tive topics of judgements that are only minimally removed from the
passivity of experience, as judgements attributing secondary qualities
are.9 We have to understand the experienceable world as a subject
matter for active thinking, rationally constrained by what experience
9. This is a way of putting something Gareth Evans urges in "Things without the Mind
a Commentary upon Chapter Two of Strawson'$ Individuals", in Zak van Straaten, ed.,
Philosophical Subjects: Essays Presented to P. F. Strawson (Clarendon Press, Oxford,
1 980), pp. 76- 1 1 6.
Lecture II. The Unbound, dness of the C mceptual 33
reveals. The capacities that are pas ively drawn ir to operation in ex
perience can be recognized as cone :ptual capacit es only because we
can get the idea of spontaneity to fit . And we do r.ot genuinely get the
idea of spontaneity to fit if we try to picture a practice of thinking that
distances itself from actual cases of the passivity of experience only so
far as to contemplate more of the same, including merely possible
cases. That way, we entitle ourselves at most co the idea of an orderly
sequence of "inner experiences". In fact I do not believe we entitle
ourselves even to that, because we cannot make sense even of " inner
experience" in the absence of a world; but it will not be clear why I
say that until much later (Lecture V), and perhaps not even then. The
point for now is just this: if we try to picture a mode of, say, colour
experience with only this slight an integration into a practice of active
thinking and judgement, it is mysterious how what we are picturing
could amount to "outer experience" of colour-how the colour expe
rienced could be experienced as a feature of an "outer" reality.
The point about how experience is rationally linked into the activity
of adjusting a world-view is even clearer when we stop restricting our
selves to concepts of secondary qualities. Of course other concepts
figure in experience too. It would be quite wrong to suppose that ex
perience takes in only features of reality whose concepts are inextrica
bly tied to concepts of modes of appearance, in the way that is exem
plified by the tie between what it is to be red and what it is to look red.
(As if other aspects of the world could come to mind not in experi
ence, but only in theoretical thinking.) Experiences themselves take in
more of the thinkable world than qualities that are phenomenal in
that sense.
Generalized, the remark of Wittgenstein that I started from says
that thinking does not stop short of facts. The world is embraceable in
thought. What I have been urging is that that constitutes a back
ground without which the special way in which experience takes hold
of the world would not be intelligible. And the dependence is not only
in that direction. It is not that we could first make sense of the fact that
the world is thinkable, in abstraction from experience, and proceed
from there to make sense of experience. What is in question could not
be the thinkable world, or, to put it another way, our picture of the
understanding's equipment could not be what it needs to be, a picture
of a system of concepts and conceptions with substantial empirical
content, if it were not already part of the picture that the system is the
34 ' M I N D A N D W cf R L D
medium within which one engages in active thought that is rationally

responsive to the deliverances of experience. Thoughts without intu
itions would indeed be empty. To understand empirical content in
general, we need to see it in its dynamic place in a self-critical activity,
the activity by which we aim to comprehend the world as it impinges
on our senses.
5. This talk of impingements on our senses is not an invitation to

suppose that the whole dynamic system, the medium within which we
think, is held in place by extra-conceptual links to something outside
it. That is just to stress again that we must not picture an outer bound
ary around the sphere of the conceptual, with a reality outside the
boundary impinging inward on the system. Any impingements across
such an outer boundary could only be causal, and not rational; chat is
Davidson's perfectly correct point, and he urges that we should settle
for holding that in experience the world exerts a merely causal
influence on our thinking. But I am trying to describe a way of main
taining that in experience the world exerts a rational influence on our
thinking. And that requires us to delete the outer boundary from the
picture. The impressions on our senses that keep the dynamic system
in motion are already equipped with conceptual content. The facts
chat are made manifest to us in those impressions, or at least seem to
be, are not beyond an outer boundary that encloses the conceptual
sphere, and the impingements of the world on our sensibility are not
inward crossings of such a boundary. My point is to insist that we can
effect this deletion of the outer boundary without falling into ideal
ism, without slighting the independence of reality.
We find ourselves always already engaging with the world in con
ceptual activity within such a dynamic system. Any understanding of
this condition that it makes sense to hope for must be from within the
system. It cannot be a matter of picturing the system's adjustments to
the world from sideways on: that is, with the system circumscribed
within a boundary, and the world outside it. That is exactly the shape
our picture muse not take.
Of course we can initially find another thinker opaque. It may take
work to make the conceptual contents of someone else's engagements
with the world available to us. And in the meantime the world she
engages with is surely already within our view. I have said nothing
that threatens that obvious fact. What I do mean to rule out is this
idea: that, when w e work at making someone else intelligible, we ex

ploit relations we can already discern between the world and some
thing already in view as a system of concepts within which the other
person thinks; so that as we come to fathom the content of the initially
opaque conceptual capacities that are operative within the system, we
are filling in detail in a sideways-on picture-here the conceptual sys
tem, there the world-that has been available all along, though at first
only in outline. It must be an illusion to suppose that this fits the work
of interpretation we need in order to come to understand some peo
ple, or that a version of it fits the way we acquire a capacity to under
stand other speakers of our own language in ordinary upbringing.
This picture places the world outside a boundary around the system
we have supposedly come to understand. That means it cannot depict
anything genuinely recognizable as an understanding of a set of con
cepts with empirical substance. These supposed concepts could be
bound up with impacts from the world only causally, not rationally
(Davidson's point again); and I have been urging that that leaves their
status as concepts with empirical substance, potential determinants of
the content of j udgements that bear on the empirical world, a mystery.
(I think these considerations tell against some appeals to Davidson's
notion of radical interpretation, the procedure by which one would
work one's way into comprehension of a foreign language without
external resources such as dictionaries and the like . )10
The illusion is insidious; so much so that it can entice us into aspir
ing to a sideways-on understanding of our own thinking, which we
take to be the condition of someone else who understands us. Some
sideways-on picture must be innocuous in the case of a thinker who is
opaque to us, and then it can seem obvious that overcoming opaque
ness is just filling in blanks in that sideways-on picture, leaving its
orientation unchanged. But that must be wrong. The mistake is not to
give proper weight to this fact: in the innocuous sideways-on picture,
the person we do not yet understand figures as a thinker only in the
most abstract and indeterminate way. When the specific character of
her thinking starts t o come into view for us, w e are not filling i n
blanks in a pre-existing sideways-on picture of how her thought bears
on the world, but coming to share with her a standpoint within a sys-
10. I have in mind !tony's exploitation ol Da v idson, in '"Pragmatism, Davidson, and

Truth".
tern of concepts, a standpoint from which we can join her in directing

a sha red attention at the world, without needing to break out through
a boundary that encl0$es the system of concepts. 11
6. I have been talking about how the conceptual capacities that are
drawn into operation in experience are integrated into spontaneity at
large. I have suggested that it is this integration that makes it possible
for a subject to understand an " outer" experience as awareness of
something objective, something independent of the experience itself.
The object of an experience, the state of affairs experienced as obtain
ing, is understood as part of a whole thinkable world. Since the whole
is independent of this particular experience, we can use the linkage
into the mostly unexperienced whole to hold the object of this partic
ular experience in place, while we ask how things would have been if
the experience had not occurred. This depends on a specific way in
which concepts are integrated into spontaneity at large, a way that, as
I have claimed, is minimally exemplified by colour concepts.
Now in my first lecture (§7) I said that the object of an " inner ex
perience" does not have this independence from the experience itself.
An object of " inner experience" , I said, has no existence indepen
dently of the awareness that the experience constitutes.
This can put a strain on our understanding of " inner experience " .
lt is easy to think we cannot preserve the claim that a n " inner experi
ence" has no object independent of the experience, unless we construe
the objects of " inner experience" as bits of the Given, somehow con
stitutively related to the occurrences of their reception-that is, a s
" private objects " . I f we are persuaded by Wittger'stein's polemic
against that supposed conception, we come under pressure either to
deny that " inner experience" is a matter of awareness at all, which
would dispense us from worrying about a relation between events in
the stream of consciousness and putative objects of such events, but
which looks like the embarrassing philosophical strategy of " feigning
anaesthesia";12 or else, keeping awareness in play, to give up the claim
1 1 . I mean this to be reminiscent of the concept of a " fusion of horizons"' exploited by

Hans-Georg Gadamer in Truth and Method (Crossroad, New York, 1992; rev. trans. by
joel Weinsheimer and Donald Marshall), pp. 306-7.
12. I took this phrase from A. J. Ayer, p. 1 0 1 of "The Concept of a Person", in The
Cot�cept of a Person and Other Essays (Macmillan, London, 1 964), pp. 82-1 28. Ayer cred
its the idea to C. K. Ogden and I. A. Richards.
Lect11re 11. The llnfwww: ; ; · ; .·,;,, of the Conceptual 37
that an object o f " inner experience" i s not independent o f the experi

ence. On this last view, " inner experience " is after all awareness of
circumstances that obtain anyway, independently of this awareness of
them. Where the relevant " inner experience" is a sensation, suitably
related bodily circumstances will seem to fit the bill. This assimilates
" inner experience", in so far as it constitutes awareness of something,
to " outer experience "; it is just that the object of the experience is not
very far out. All these positions are so unsatisfactory that one_ can
sympathize with Wittgenstein in the inclination that at least some
commentators find him succumbing to, at least sometimes: the incli
nation to deny that self-ascriptions of sensation and so forth give ex
pression to judgements at all.
I think it is helpful here to exploit the fact that the concepts of
" outer experience" are integrated into spontaneity at large in a
specific way. We can find a different specific manner of integration for
the concepts of " inner experience" . Up .to a point, impressions of
" inner experience" are like impressions of "outer experience" . They
are all passive occurrences in which conceptual capacities are drawn
into operation. And it is not that the conceptual capacities that are
drawn into operation in " inner experience" are not integrated into
spontaneity in general, or, what comes to the same thing, that the ob
ject of an " inner experience" is not understood in a way that places it
as a possible element in a view of the world. That would merely pre
vent us from recognizing conceptual capacities here at a ll. But the
mode of integration in this case is not such as co confer independence
on the objects of awa reness.
We could not credit a subject with a capacity to use, say, the c o n
cept of pain in j udgements of " inner experience" if she did not under
stand how the circumstance that those judgements concern fits into
the world at large. What that requires is that the subject must under
stand her being in pain as a particular case of a general type of state of
affairs, someone 's being in pain. So she must understand that the c o n
ceptual capacity drawn on in the relevant " inner experiences" is not
restricted to its role in "inner experience" and ju dgements of " inner
experience" : not restricted, that is, to its first-person present-tense
role.13 This yields what we can think of as a limiting case of the struc-
13. See chap. 3 of P. F. Strawson, Individuals: An Essay in Descriptive Metaphysics
(Methuen, London, 1 959), and Gareth Evans's exploitation of Strawson's thought in chap.
7 of The Varieties of Reference (Clarendon Press, Oxford, 1 982).
M I N D A N D W O R L D
ture o f awareness and o bject. We can understand an impression of

" inner sense" in which the concept of, say, pain is drawn into play as
an awareness of the circumstance that the subject is in pain. The struc
ture of awareness and object is appropriately in place just because the
subject does not conceive what it is for her to be in pain-the circum
stance that is the object of her awareness-exclusively in tctrms of the
"inner" or first-person angle on that circumstance that constitutes her
awareness of it. She understands that the very same circumstance is
thinkable-by someone else, or by herself at different times-ot.her
wise than in a thought expressive of " inner experience" . This givcf. the
idea of the circumstance an independence from her awareness of it.
But though her conception of the circumstance does not come at it
exclusively through the awareness that is the " inner" operation of
sensibility, the circumstance is nothing other than the operation of
sensibility itself.
It is worth comparing objects of " inner experience", in their limit
ing-case substantiality, with secondary qualities of " outer experi
ence" . In one way objects of " inner experience" are less firmly rooted
in objective reality than secondary qualities, but in another way they
are more independent of the distinctive experience in terms of which
they are identified, and so they are more substantial. On the one hand,
secondary qualities are there independently of any particular experi
ence in which their presence is revealed, whereas an object of " inner
experience" is not there independently of the experience. But on the
other hand, a subject understands the sensuous specificity of a second
ary quality exclusively through what it is like to experience it, whereas
a subject must understand that the potential objects of her " inner ex
perience" are essentially also thinkable otherwise than from the stand
point of her experiencing them.14
14. The second half of this contrast is somewhat fragile. We might say that the sensuous
specificity of an object of "inner experience" is equally understood exclusively through
what it is like to experience it. It cannot be someone else's pain that I embrace in thought
unless what I have in mind is what her experience is like for her. When I say that the circum
stance of someone's being in pain is thinkable otherwise than through her "inner experi
ence" of it, I do not mean to imply that the circumstance is thinkable in abstraction from
her "inner experience" of it, for instance in behaviouristic terms. But in the case of second·
ary qualities there is no parallel to the play between first-person and third-person stand
points. This footnote responds to a comment by Danielle Macbeth.
Lecture 11. The Unboundedness of the Conceptual 39
7 . I have been urging that, i n judgements of experience, concep

tual capacities are not exercised on non-conceptual deliverances of
sensibility. Conceptual capacities are already operative in the deliv
erances of sensibility themselves. In my first lecture ( §5), I suggested
that it is best to disallow the question what conceptual capacities arc
exercised on in experience. If we talk of an exercise of conceptual
capacities fn this context, we put at risk our ability to insist that ex
perience is passive; and that is the essential thing if we are to extin
guish the hankering after the Given. But in disallowing the question
what those conceptual capacities are exercised on, I do not disallow
the question what the conceptual contents that are passively received
in experience bear on, or are about. And the obvious answer, if the
question is asked in that general form, is: they are about the world,
as it appears or makes itself manifest to the experiencing subject,
or at least seems to do so. That ought not to activate a phobia of
idealism.
When we reject the Myth of the Given, we reject the idea that trac
ing back the ground for a judgement can terminate in pointing to a
bare presence. Now this can generate some discomfort. It can seem
that we are depriving ourselves of a justificatory role that it must be
possible to attribute to pointing, if we are to be able to assure our
selves that our conception of thinking sufficiently acknowledges the
independence of reality. It must be possible for justifications of judge
ments to include pointing out at features of the world, from what would
otherwise risk looking like a closed circle within which our exercises of
spontaneity run without friction.
But now that we are on guard against the ambiguity in phrases like
"outside the sphere of thought" ( § 3 ), it is easy to deal with this. There
are two different conceptions of a justificatory role for pointing in
play. On the conception I am recommending, justifications can per
fectly well include pointing out from the sphere of thinking, at fea
tures of the world. We fall into the Myth of the Given only if we sup
pose that this pointing would have to break out through a boundary
that encloses the sphere of thinkable content.
8. People sometimes object to positions like the one I have been urg
ing on the ground that they embody an arrogant anthropocentrism, a
baseless confidence that the world is completely within the reach of
our powers of thinking. This looks at least akin to an accusation of

idealism. Why should we be so sure of our capacity to comprehend
the world if not becau� we conceive the world as a shadow or reflec
tion of our thinking?15
But an accusation of arrogance would not stick against the position
I am recommending. In my first lecture (§5) I said that the faculty of
·
spontaneity carries with it a standing oblig�tion to reflect on the ere-
dentials of the putatively rational linkages that, at any time, one takes
to govern the active business of adjusting one's world-view in re
sponse to experience. Ensuring that our empirical concepts and con
ceptions pass muster is ongoing and arduous work for the under
standing. It requires patience and something like humility. There is no
guarantee that the world is completely within the reach of a system of
concepts and conceptions as it stands at some particular moment in its
historical development. Exactly not; that is why the obligation to
reflect is perpetual.
There is a tendency to stop short of accepting that the obligation is
perpetual. One imagines the obligation's ceasing to apply if one con
templates a state of affairs that would deserve to be called "the end of
inquiry" .1 6 That would be a state of affairs in which inquiry, including
reflective inquiry into the credentials of what at present passes for in
quiry itself, is no longer necessary. It might well be argued that, even
as a mere ideal of reason, this conception is suspect: what can it be if
not a vestigial reflection of the unwarranted confidence in our powers
that the objection complains of? But the idea of an end to inquiry is no
part of the position I am recommending.
9. In these first two lectures, I have introduced the tendency to oscil

late between embracing the Myth of the Given and denying that expe
rience has a rational bearing on thought. I have claimed that to escape
this alternation, we must hold that, in experience, spontaneity is inex-
1 5 . For this form of the accusation of idealism, see chap. 6 of Thomas Nagel, The View
{rom Nowhere (Oxford University Press, New York, 1 986).
1 6. The idea is implicit in such well-known remarks of C. S. Peirce as this: "The opinion
which is fated to be ultimately agreed to by all who investigate is what we mean by the
truth, and the object represented in this opinion is the real." "How to Make Our Ideas
Clear", in Writings of Charles S. Peirce, val. 3 (Indiana University Press, Bloomington,
1 986), pp. 257-76, at p. 273; originally in Popular Science Monthly 12 Uanuary 1878),
.
286-302.
tricably implicated in deliverances of receptivity. We must not sup

pose that receptivity makes an even notionally separable contribu
tion to its co-operation with spontaneity. In this lecture I have been
discussing an accusation of idealism, in the sense of a failure to
acknowledge that reality is independent of thinking. This is a good
context in which to raise the question how my description of the
escape route relates to Kant, whose terminology I am of course ex
ploiting.
Does Kant credit receptivity with a separable contribution to its co
operation with spontaneity? It seems that the answers "No" and
"Yes" are both correct.
From the standpoint of experience, the answer is "No". If one pos
its an empirically separable contribution from receptivity, one com
mits oneself to something Given in experience that could constitute
the ultimate extra-conceptual grounding for everything conceptual,
and it is a way of putting a central Kantian thought to say that that
idea must be rejected. For Kant, experience does not take in ultimate
grounds that we could appeal to by pointing outside the sphere of
thinkable content. In experience we take in, through impacts on the
senses, elements in a reality that is precisely not outside the sphere of
thinkable content.
But Kant also has a transcendental story, and in the transcendental
perspective there does seem to be an isolable contribution from recep
tivity. In the transcendental perspective, receptivity figures as a suscep
tibility to the impact of a supersensible reality, a reality that is sup
posed to be independent of our conceptual activity in a stronger sense
than any that fits the ordinary empirical world.
If we restrict ourselves to the standpoint of experience itself, what
we find in Kant is precisely the picture I have been recommending: a
picture in which reality is not located outside a boundary that encloses
the conceptual sphere. It is no accident that I have been able to put
what I am urging in Kantian terms. The fact that experience involves
receptivity ensures the required constraint from outside thinking and
judging. But since the deliverances of receptivity already draw on ca
pacities that belong to spontaneity, we can coherently suppose that
the constraint is rational; that is how tl;le picture avoids the pitfall of
the Given.
But the transcendental perspective embeds this potentially liberat-
ing picture within a peculiar version of the sideways-on view I men

tioned earlier (§5), with the space of concepts circumscribed and
something-the supersensible in this version, not the ordinary empir
ical world-outside its outer boundary. And in that frame, the liber
ating thought cannot take proper shape. Once the supersensible is in
the picture, its radical independence of our thinking t�nds to present
itself as no more than the independence any genuine reality must have.
The empirical world's claim to independence comes to seem fraudu
lent by comparison. We are asked to suppose that the fundamental
structure of the empirical world is somehow a product of subjectivity,
in interaction with supersensible reality, which, as soon as it is in the
picture, strikes us as the seat of true objectivity. But how can the em
pirical world be genuinely independent of us, if we are partly respon
sible for its fundamental structure? It does not help to be told that it is
only transcendentally speaking that the fundamental structure of the
empirical world is of our making. 17
Crediting experience with ordinary empirical passivity meets our
need: it ensures that when we invoke spontaneity in connection with
the employment of concepts in empirical thinking, we do not con
demn ourselves to representing empirical thinking as rationally un
constrained, a frictionless spinning in a void. The idea of a transcen
dental passivity is at best problematic anyway, in a familiar way; by
Kant's own lights we are supposed to understand causation as some
thing that operates within the empirical world. Adding this problem
atic idea only undermines the reassurance that empirical passivity
could afford.
Kant comes within a whisker of a satisfactory escape from the os
cillation. He points the way to undermining the central confusion in
the Myth of the Given. According to the Myth of the Given, the obli
gation to be responsibly alive to the dictates of reason lapses when we
come to the ultimate points of contact between thinking and reality;
the Given is a brute effect of the world, not something justified by it.
But in fact the obligation must be in force all the way out to reality.
The world itself must exert a rational constraint on our thinking. If we
suppose that rational answerability lapses at some outermost point of
the space of reasons, short of the world itself, our picture ceases to
17. For a viv �d expression of a dissatisfaction on these lines, see Barry Stroud's contribu
tion to the symposium "The Disappearing 'We'" (commenting on a paper by Jonathan
Lear), Proceedings of the Aristotelian Society, supp. vol. 58 ( 1 984), 243-5 8.
depict anything recognizable as empirical judgement; we have obliter

ated empirical content altogether. If it were not for the transcendental
framework, we could credit Kant with a clear formulation of that in
sight. This is how I have interpreted his almost explicit claim that
thoughts without intuitions would be empty.
The idea of a faculty of spontaneity is the idea of something that
empowers us to take charge of our lives. Kant points the way to a
position in which we can satisfyingly apply that idea to empirical
thinking: we can hold that empirical inquiry is a region of our lives in
which we exercise a responsible freedom, and not let that thought
threaten to dislodge our grip on the requirement that empirical think
ing be under constraint from the world itself. But the transcendental
framework forces a qualification. Transcendentally speaking, our re
sponsible freedom in empirical thinking seems to fall short of the gen
uine article. It is as if Kant were saying that although an exculpation
cannot do duty for a justification, and although, empirically speaking,
we can have justifications for empirical judgements, still the best we
can have for empirical j udgements, transcendentally speaking, is ex
culpations.
This is a profoundly unsatisfactory aspect of Kant's philosophy. 1 8
And what I have said about it so far is much too simple, in particular
in giving the impression that it would be straightforward to excise the
1 8. Contrast Henry E. Allison, Kant's Transcendental Idealism: An Interpretation and

Defense (Yale University Press, New Haven, 1 983 ). Allison defends transcendental idealism
on the ground that it is the only alternative to a psychologistic phenomenalism. His basic
thought is encapsulated in this passage (p. 13): " Indeed, one can claim that the fundamental
issue raised by the Critique is whether it is possible to isolate a set of conditions of the
possibility of knowledge of things . . . that can be distinguished from conditions of the pos·
sibility of the things themselves. Since the former kind of condition would count as a condi·
tion of things as they appear and the latter of things as they are in themselves, an affirmative
answer to this question entails the acceptance of the transcendental distinction [berween
things as they appear and things as they are in themselves], and with it of transcendental
idealism. If, on the other hand, the question is answered in the negative, as it is by the
standard picture, then any purportedly 'subjective' conditions are inevitably construed in
psychological terms. The subjectivistic, psychologistic, phenomenalistic reading of Kam,
which is characteristic of the standard picture, is thus a direct consequence of its negative
answer to this question." I agree that we should not find a psychologistic phenomenalism in
Kant. (However, Allison's representative of "the standard picture", which finds a psy·
chologistic phenomenalism in Kant, is Strawson's The Bounds ofSense, and such a reading
of Strawson strikes me as absurd.) I do not agree that a negative answer to Allison's ques·
tion inevitably imports psychologism. That ma kes no sense of the responses to Kant of
•
Fichtc: and, especially, Hegel.
transcendental framework. I shall say a little more about this later

(Lecture V).
I think it has to be asfmitted that the effect of the transcendental
framework is to make Kant's philosophy idealistic in the sense I have
been considering. This is quite contrary to Kant's intentions, but in
spite of his staunch denials, the effect of his philosophy is to slight the
independence of the reality to which our sensttS give us access. What is
responsible for this is precisely the aspect of Kant's philosophy that
struck some of his successors as a betrayal of idealism: namely, the
fact that he recognizes a reality outside the sphere of the conceptual.
Those successors urged that we must discard the supersensible in
order to achieve a consistent idealism. In fact that move frees Kant's
insight so that it can protect a commonsense respect for the indepen
dence of the ordinary world.
As I said, if we abstract from the role of the supersensible in Kant's
thinking, we are left with a picture in which reality is not located out
side a boundary that encloses the conceptual. What I have been urging
here is that such a picture does not slight the independence of reality.
The picture is not offensive to common sense, but precisely protective
of it.
It is central to Absolute Idealism to reject the idea that the concep
tual realm has an outer boundary, and we have arrived at a point from
which we could start to domesticate the rhetoric of that philosophy.
Consider, for instance, this remark of Hegel's: "In thinking, I am free,
because I am not in an other." 19 This expresses exactly the image I
have been using, in which the conceptual is unbounded; there is no
thing outside it. The point is the same as the point of that remark of
Wittgenstein's that I spent some time discussing (§§3 and 4 above) :
"We-and our meaning-do not stop anywhere short of the fact. " I
1 9. Phenomenology of Spirit, tra ns. A. V. Miller (Oxford University Press, Oxford,

1 977), S I 97 (p. 120). Robert B. Pippin, at p. 164 of Hegel's Idealism: The Satisfactions of
Self·Consciollsness (Cambridge University Press, Cambridge, 1989), reads this remark as
primarily an expression of stoicism, though he says that "ir is also clearly indicative of the
position toward which Hegel is heading". The latter is ali i need, but I wo u ld suggest char it
is actually the first thing to say about the remark. The opening paragraphs of a section in
rhe Phenomenology typically adumbrate "the position toward which Hegel is heading" .
Only later in the section does an insufficiency come co l i ght. This remark comes in rhe
sectio n headed " Stoicism", but it belongs co the phase of that section that marks progress,
before the emergence of something co keep the dialectical pendulum swinging.
should like to take this further, but for several reasons, of which the
fact that I have said enough for one lecture is perhaps the least serious,
I cannot do that now.
10. I have been claiming that the content of experience is conceptual,

and many people think that cannot be right. In my next lecture I am
going to discuss that question.
LECT U RE I I I
Non-conceptual Content
1. I have been talking about a pair of opposing pitfalls: on the one

side a coherentism that does not acknowledge an external rational
constraint on thinking and therefore, I c laim, cannot genuinely make
room for empirical content at all; and on the other side a recoil into
the Myth of the Given, which offers at best exculpations where what
we need is justifications. I have urged that the way to stop oscillating
between those pitfalls is to conceive empirical knowledge as a co-op
eration of sensibility and understanding, as Kant does. To avoid mak
ing it unintelligible how the deliverances of sensibility can stand in
grounding relations to paradigmatic exercises of the understanding
such as judgements and beliefs, we must conceive this co-operation in
a quite particular way: we must insist that the understanding is al
ready inextricably implicated in the deliverances of sensibility them
selves. Experiences are impressions made by the world on our senses,
products of receptivity; but those impressions themselves already have
conceptual content.
This unqualified claim that the content of perceptual experience is
conceptual will have been raising some eyebrows since my first lec
ture. Here I am going to defend it against some doubts.
Before I start, let me note that the issue cannot be defused as just a
matter of idiosyncratic terminology on my part-as if I am merely
affixing the label " conceptual" to the content of experience, although
I regard the content of experience in the very way that my opponents
express by saying that it is not conceptual, at least not through and
through. It is essential to the picture I am recommending that experi
ence has its content by virtue of the drawing into operation, in sensi-
bility, o f capacities that are genuinely elements in a faculty of spon

taneity. The very same capacities must also be able to be exercised
in judgements, and that requires them to be rationally linked into a
whole system of concepts and conceptions within which their pos
sessor engages in a continuing activity of adjusting her thinking to
experience. Indeed, there can be other elements in the system that
are not capable of figuring in experience at all. In my last lecture, I
claimed that it is only because experience involves capacities be
longing to spontaneity that we can understan d experience as
awareness, or apparent awareness, of aspects of the world at all.
The way I am exploiting the Kantian idea of spontaneity commits
me to a demanding interpretation for words like "concept" and
"conceptual". It is essential to conceptual capacities, in the de
manding sense, that they can be exploited in active thinking, think
ing that is open to reflection about its own rational credentials. 1
When I say the content of experience is conceptual, that is what I
mean by "conceptual".
2. To focus the discussion, I am going to consider what Gareth Evans

says about this question.
Evans makes the equally unqualified claim that the content of per
ceptual experience is non-conceptual. According to Evans, conceptual
content first comes into play, in the context of perception, in judge
ments based on experience. When one forms a judgement on the basis
of experience, one moves from non-conceptual content to conceptual
content.
The informational states which a subject acquires through perception
are non-concep tual, or non-conceptualized. Judgements based upon
such states necessarily involve conceptualization: in moving from a per
ceptual experience to a judgement about the world (usually expressible
in some verbal form), one will be exercising basic conceptual skills . . . .
The process of conceptualization or judgement takes the subject from
his being in one kind of informational state (with a content of a certain
kind, namely, non-conceptual content) to his being in another kind of
1. It is worth noting, since it helps to bring out how demanding the relevant idea of the
conceptual is, that this openness to reflection implies self-consciousness on the part of the
thinking subject. But I relegate the point to a footnote at this stage, since issues about self
consciousness will not be in the foreground until later (l�ture V).
cognitive state (with a content of a different kind, namely, conceptual

contcnt ) .2
These non-conceptual informational states are the results of

perception's playing its role in what Evans calls "the informational
system" (p. 122). The informational system is the system of capacities
we exercise when we gather information about the world by using our
senses (perception), receive information from others in communica
tion (testimony), and retain information through time (memory).3
It is central to Evans's view that "the operations of the informa
tional system" are "more primitive" than the rationally intercon
nected conceptual skills that make room for the notion of judgement
and a strict notion of belief (p. 1 24).4 To put the thought in the terms
I have been using: the operations of the informational system are more
primitive than the operations of spontaneity. This point is straightfor
ward in the case of perception and memory, which, as Evans says,
"we share with animals" (p. 124); that is, with creatures on which the
idea of spontaneity gets no grip. Strikingly, he insists on the point for
testimony too: "the mechanism whereby we gain information from
others . . . is already operative at a stage of human intellectual devel
opment that pre-dates the applicability of the more sophisticated no
tion" (p. 1 24). His thought here is that for much of the knowledge
that one has by virtue of having been exposed to statements of it, one
was not in a position to understand the statements at the time.
Evans, then, identifies perceptual experiences as states of the infor
mational system, possessing content that is non-conceptual.5 Accord
ing to Evans, conceptual capacities are first brought into operation
only when one makes a judgement of experience, and at that point a
different species of content comes into play. Contrast the account I
have been urging. According to the picture I have been recommend
ing, the content of a perceptual experience is already conceptual. A
judgement of experience does not introduce a new kind of content,
2. The Varieties of Reference, p. 227 (emphasis in the original). Unless otherwise indi
cated, all citations from Evans in this lecture will be from this work. I want ro make it clear
i mmediately that I believe the thesis on which I shall be taking issue with Evans here is
inessential to the main claims of Evans's profoundly important book. I shall come back to
Evans's main claims in Lecture V (§6).
3. See pp. 122-9.
4. I shall say something about the strict notion of belief in S6.
5. lr would be easy to complicate Evans's account so as to accommodate the fact that
�experience" can fit occurrences as well as states.
Lecture lll. Non-conceptual Content 49
but simply endorses the conceptual content, or some of it, that is al
ready possessed by the experience on which it is grounded.6
It is important not to misconceive this divergence. In Evans's view,
experiences are states of the informational system, and as such they
have content that is non-conceptual. But he does not equate the idea
of an experience with the idea of a perceptual informational state,
produced independently of spontaneity by the operations of the infor
mational system. On the contrary, he insists that perceptual informa
tional states, with their non-conceptual content, "are not ipso facto
perceptual experiences-that is, states of a conscious subject"
(p. 1 57). According to Evans, a state of the perceptual informational
system counts as an experience only if its non-conceptual content is
available as "input to a thinking, concept-applying, and reasoning
system" (p. 158); that is, only if its non-conceptual content is avail
able to a faculty of spontaneity, which can rationally make or with
hold judgements of experience on the basis of the perceptual state. So
a non-conceptual informational state, produced by the perceptual ele
ment of the informational system in a creature that lacks a faculty of
spontaneity, does not count as a perceptual experience, even though a
state that does count as a perceptual experience, by virtue of its avail
ability to spontaneity, is in itself just such a non-conceptual informa
tional state, endowed with its non-conceptual content independently
of the coming into play of the faculty of spontaneity.
3. Near the end of my first lecture (§7), I noted a difficulty it would

be natural to feel when one sets out to apply to "inner experience" the
picture I have been recommending of experience in general, according
to which experiences are states or occurrences in which conceptual
capacities are drawn into operation. I have been claiming that it is
essential to conceptual capacities that they belong to spontaneity, that
is, to a faculty that is exercised in actively self-critical control of what
one thinks, in the light of the deliverances of experience. But that
6. Note that grounding need not depend on an inferential step from one content to an·
other. The judgement that things are thus and so can be grounded on a perceptual appear·
ance that things are thus and so. This does not obliterate the characteristic richness of expe
rience (especially visual experience). A typical j �dgement of experience selects from the
content of the experience on which it is based; the experience that grounds the judgement
that things are thus and so need not be exhausted by its affording the appearance that things
are thus and so. Selection from among a rich supply of already conceptual content is not
what Evans takes judgement to effect, a transition from one kind of content to another.
means we cannot attribute the conceptual capacities that would figure

in the account of "inner experience " I have endorsed-for instance, a
capacity to use the concept of pain-to many creatures of which it
would be outrageous to deny that they can feel pain. It is not just
active self-critical thinkers that can feel pain. Whatever it may be that
is true of a creature without spontaneity when it feels pain, it cannot
be that it has " inner experience", according to the picture of experi
ence I have been recommending.
This point is obviously not peculiar to "inner experience" . The ap
plication to "outer experience" is similar: " outer experience" that
purports to disclose that things are thus and so is, according to the
account I am recommending, a state or occurrence involving the oper
ation of the conceptual capacities that would be actively exploited in
judging that things are thus and so. In that case " outer experience"
can be attributed only to a creature that could engage in such active
thinking. So we have a parallel point in this case: I am committed to
denying " outer experience" of features of their environment to some
creatures, even though it would be outrageous to deny that they
are perceptually sensitive to those features. It is not just active self
critical thinkers that are perceptually sensitive to features of their en
vironment.
At this stage I am simply acknowledging this twin discomfort, not
aiming to alleviate it; I shall make an attempt at that later (Lecture
VI). One straightforward response would be to conclude that the no
tion of experience needs to be completely detached from anything on
the lines of the notion of spontaneity. Then we would not be commit
ted to having different stories to tell about the sentient lives of rational
and non-rational animals. My point now is just that anyone who is
tempted by this course cannot easily enrol Evans as an ally. In his
picture as in mine, the concept of experience has a restricted use, gov
erned by a link of a broadly Kantian sort to what is in effect the idea
of spontaneity.
4. I have been concerned in these lectures with a standing threat of

falling into philosophical anxiety. If we focus on the freedom implied
by the notion of spontaneity, what was meant to be a picture of think
ing with empirical content threatens to degenerate into a picture of a
frictionless spinning in a void. To overcome that, we need to acknowl
edge an external constraint on the exercise of spontaneity in empirical
thinking. But now w e come t o the other side o f the standing difficulty:
we must avoid conceiving the external constraint in such a way that it
could at best yield exculpations where we needed justifications. One
might simply refuse to address this difficulty, by refusing to give any
place, in an account of experience, to anything like the idea of sponta
neity. But as I have just stressed, that is not a line that Evans takes.
To acknowledge the required external constraint, we need to ap
peal to receptivity. I have urged that the way to introduce receptivity
without merely tipping the seesaw back to the Myth of the Given is
this: we must not suppose that receptivity makes an even notionally
separable contribution to its co-operation with spontaneity.
Now Evans does not respect this rule. In Evans's account of experi
ence, receptivity figures in the guise of the perceptual element of the
informational system, and his idea is that the perceptual system pro
duces its content-bearing states independently of any operations of
spontaneity. It is true that the content-bearing states that result count
as experiences, in the somewhat Kantian restricted sense that Evans
employs, only by virtue of the fact that they are available to spontane
ity. But spontaneity does not enter into determining their content. So
the independent operations of the informational system figure in
Evans's account as a separable contribution made by receptivity to its
co-operation with spontaneity.
In that case, the way experiences are related to conceptual capaci
ties in Evans's picture is j ust the way intuitions are related to concepts
in the picture of empirical knowledge that Kant, as I am reading him,
displays as hopeless, at least as a picture of how things are from the
standpoint of experience. It is true that Kant tries to allow a kind of
correctness for a picture with that shape at the transcendental level,
but Evans's account of experience is not meant to be only transcen
dentally correct, whatever indeed that might mean. So unless there is
something wrong with the Kantian considerations I rehearsed in my
first two lectures, Evans's account of experience ought to be demol
ished by them.
It may be hard to believe that Evans's view of experience is a ver
sion of the Myth of the Given. Evans's smoothly naturalistic account
of perceptual informational states shows no sign of the epistemologi
cal obsessions that are usually operative in motivating the Myth of the
Given. What usually underlies the Myth is a worry that spontaneity's
involvement in our picture of empirical thO\lght makes it mysterious
52 .\1 I .-.: iJ A S CJ \t 0 f<. L iJ
how we can be picturing something that is in touch with reality at all,

and there is no sign of that in Evans.
Moreover, there may*seem to be a more specific problem about at
tributing the Myth of the Given to Evans. If experiences as Evans con
ctives them are intuitions without concepts, in a sense that would
make his position vulnerable to the Kantian attack on the Myth of the
Given, they ought to be blind. But Evans takes care to credit experi
ences with representational content, even independently of the avail
ability to spontaneity in virtue of which they count as experiences.
The content is non-conceptual, certainly, but one might wonder how
that could warrant the image of blindness. Surely, one might think,
something that is blind would have to be totally devoid of representa
tional content?
The structure of Evans's position is comparable to the structure of
a position I considered in my first lecture (§5), when I was trying to
avert a misinterpretation of what I was recommending. That position
purported to accept that conceptual capacities are drawn into opera
tion in experience. But it treated the states and occurrences that it de
scribed in those terms as insulated from spontaneity. The aim was to
ensure that they were not subject to the potentially unnerving effects
of the freedom that the idea of spontaneity implies.
What I said about that position was that in the context of the insu
lation from spontaneity, the talk of concepts is mere word-play. The
point of the claim that experience involves conceptual capacities is
that it enables us to credit experiences with a rational bearing on em
pirical thinking. But the point of the strategy of insulation is that it
confines spontaneity within a boundary that leaves experiences out
side it. That means that the putatively rational relations between ex
periences, which this position does not conceive as operations of
spontaneity, and judgements, which it does conceive as operations of
spontaneity, cannot themselves be within the scope of spontaneity
liable to revision, if that were to be what the self-scrutiny of active
thinking recommends. And that means that we cannot genuinely re
cognize the relations as potentially reason-constituting. We cannot
put limits on the self-scrutiny of reason. If we want to be able to take
it that the operations of conceptual capacities in experience impinge
rationally on our thinking, as we must if they are to be recognizable as
operations of conceptual capacities at all, we must acknowledge that
those rational relations fall within the scope of spontaneity. And it is
Lecture Ill. Non-conceptttdl Content 53
hard to see how we could acknowledge that while refusing to accept

that the perceptual states and occurrences that lie at one end of the
relations involve capacities of spontaneity in operation.
Evans's account of experience is not guilty of that fraudulent label
ling; exactly not, since he keeps concepts out of the content of experi
ence. But the word "content" plays just the role in Evans's account
that is played in that position by the fraudulent use of the word "con
ceptual": that is, to make it seem that we can recognize rational rela
tions between experiences and judgements, so that we can say, as
Evans does, that judgements of experience are "based upon" experi
ence (p. 227), even though these relations are supposed to hold across
a boundary that encloses spontaneity. The same point should apply
here too. If these relations are to be genuinely recognizable as reason
constituting, we cannot confine spontaneity within a boundary across
which the relations are supposed to hold. The relations themselves
must be able to come under the self-scrutiny of active thinking.
Evans's position has a deceptively innocent look. It can seem obvi
ous that a possessor of one piece of representational content, whether
conceptual or not, can stand in rational relations, such as implication
or probabilification, to a possessor of another. But with spontaneity
confined, we lose the right to draw the conclusion, as a matter of rou
tine, that one term in such a relation can be someone's reason for
another. If experience is pictured as input to spontaneity from outside,
then it is another case of fraudulent labelling to use the word "con
tent" for something we can even so take experience to have, in such a
way that reason-constituting relations can intelligibly hold between
experiences and judgements. The label serves to mask the fact that the
relations between experiences and judgements are being conceived to
meet inconsistent demands: to be such as to fit experiences to be rea
sons for judgements, while being outside the reach of rational inquiry.7
I am claiming that although Evans does take care to credit experi
ences with content, that does not save them from being intuitions in a
sense that entitles us to apply the Kantian tag to them: since they are
7. Why can we not acknowledge that the relations berween experience and judgements
have to be rational, and therefore within the scope of spontaneity, without being thereby
committed to a concession about experience itself? I have claimed that it is hard to see how
this combination could work, but as long as Evans's position looks innocent it will seem
quite easy. Rather than radically recasting this lecture from the form in which I delivered it,
I postpone further discussion of this matter to the Afterword.
without concepts, they are blind. And actually there is an interpreta

tion of that claim under which Evans would not dispute it. It was
wrong to suggest that experiences as he conceives them cannot possi
bly be blind, since he equips them with content. It makes all the differ
ence that the content is supposed to be non-conceptual.
How should we cash out. the image of blindness? To say that an
experience is not blind is to say that it is intelligible to its subject as
purporting to be awareness of a feature of objective reality: as a seem
ing glimpse of the world. And Evans himself insists that that can be so
only against the background of an understanding of how perception
and reality are related, something sufficient to sustain the idea that the
world reveals itself to a perceiving subject in different regions and as
pects, in a way that depends on the subject's movement through the
world. 8 Such a background can be in place only for a subject with a
self-conscious conception of how her experience relates to the world,
and we cannot make sense of that in the absence of conceptual capac
ities in a strong sense, a faculty of spontaneity.9
So when Evans says that experiences, considered in themselves,
have non-conceptual content, he is not thereby pre-empting my sug
gestion that experiences as he conceives them are blind, because they
are intuitions without concepts. What makes it intelligible, in his view,
that the eyes of empirical thought are opened is not the claim that,
8. See p. 176: "Any subject at all capable of thought about an objective spatial world
must conceive of his normal experiences as simultaneously due to the way the world is and
his changing position in it. . . . The capacity to chink o f oneself as located in space, and
tracing a continuous path through it, is necessarily involved in the capacity to conceive the
phenomena one encounters as independent of one's perception of them-to conceive rhe
world as something one 'comes across' ." See also p. 222: "Any thinker who has an idea of
an objective spatial world-an idea of a world of objects and phenomena which can be
perceived bur which are not dependent on being perceived for their existence-must be able
to think of his perception of the world as being simultaneously due to his position in the
world, and to the condition of the world ar that posicion. The very idea of a perceivable,
objective, spatial world brings with it the idea of the subject as being in the world, with rhe
course of his perceptions due to his changing position in the world and co the more or less
stable way the world is." Evans elaborates this thought in "Things without the Mind ". The
thought is central to Strawson's reading of Kant; see chap. 2 of The Bounds of Sense, espe
cially p. 104.
9. In "Things without the Mind" Evans argues that the idea of an object of experience
"cannot stand on its own, stand without any surrounding theory" (p. 88). The required
theory is a theory of the conditions under which something perceptible is actually perceived
(pp. 88-9). If we make sense of the Kantian notion of spontaneity, we must surely suppose
that the possession of spontaneity marks the difference between creatures that can intelligi
bly be thought to have such a theory, even i f only implicitly, and creatures that cannot.
Lecture Ill. Non-conceptual Content 55
even considered in abstraction from any connection with spontaneity,

experiences have (non-conceptual) content. It is the claim that that
content is available to spontaneity: that it is a candidate for being in
tegrated into the conceptually organized world-view of a self-con
scious thinker. I am only stressing an aspect of Evans's own view
when I say that, according to him, the item that an experience is, con
sidered in itself (in abstraction from the availability to spontaneity in
virtue of which it acquires the title "experience"), is blind.
This would be all right if we could make sense of the potential for
rational linkage with a world-view that is supposed to make the item
an experience, something that is not blind. But although crediting the
items that experiences are with content, independently of their avail
ability to spontaneity, gives the appearance of making room for this
connection, I have urged that when the content is said to be non-con
ceptual, the appearance stands revealed as illusory.
I am not saying there is something wrong with just any notion of
non-conceptual content. It would be dangerous to deny, from a philo
sophical armchair, that cognitive psychology is an intellectually re
spectable discipline, at least so long as it stays within its proper bounds.
And it is hard to see how cognitive psychology could get along without
attributing content to internal states and occurrences in a way that is
not constrained by the conceptual capacities, if any, of the creatures
whose lives it tries to make intelligible. But it is a recipe for trouble if
we blur the distinction between the respectable theoretical role that
non-conceptual content has in cognitive psychology, on the one hand,
and, on the other, the notion of content that belongs with the capacities
exercised in active self-conscious thinking-as if the contentfulness of
our thoughts and conscious experiences could be understood as a well
ing-up to the surface of some of the content that a good psychological
theory would attribute to goings-on in our cognitive machinery.10
10. For a clear and engaging exposition of this welling-up picture, see Daniel Dennen,
"Toward a Cognitive Theory of Consciousness", in his Brainstorms: Philosophical Essays
on Mind and Psychology (Bradford Books, Montgomery, Vt., 1978), pp. 149-73. Dennett
suggests that the role of content at the personal level is to be understood in terms of our
access to some of the content that figures in a sub-personal story about our internal ma
chinery. I think Dennen's own discussion strongly suggests that something is wrong with
this picture: it leads Dennett into the highly implausible claim that perceptual awareness is
a matter of presentiments or premonitions, differing from what are ordinarily so called only
in that they are not isolated (see pp. 165-6). I discuss this in "The Concept of Perceptual
Experience", Philosophical Quarterly 44 (1994), 190-2()5.
5. Why does Evans think he has to locate experiences outside the

sphere of the conceptual? If his position is indeed a case of lapsing into
the Myth of the Given; it is a special case. As I said, it does not issue
from the usual epistemological motivation, the recoil from a picture
that threatens to leave empirical thinking out of touch with reality,
and so not recognizable as empirical thinking at all.
One consideration that impresses Evans is the determinacy of detail
that the content of experience can have. He claims that this detail can
not all be captured by concepts at the subject's disposal. "Do we really
understand the proposal that we have as many colour concepts as
there are shades of colour that we can sensibly discriminate?" 1 1 Oth
ers besides Evans have taken this kind of consideration to require
crediting experience with non-conceptual content. This formulation
includes people who do not follow Evans in relegating the content of
experience completely to the non-conceptual, but aim to accommo
date the phenomenological point Evans makes here by saying that the
content of experience is partly non-conceptual. 1 2
When Evans suggests that our repertoire of colour concepts is
coarser in grain than our abilities to discriminate shades, and there
fore unable to capture the fine detail of colour experience, what he has
in mind is the sort of conceptual capacities that are associated with
colour expressions like "red", "green", or " burnt sienna ". Such
words and phrases express concepts of bands on the spectrum,
whereas Evans's thought is that colour experience can present proper
ties that correspond to something more like lines on the spectrum,
with no discernible width.
But why should we accept that a person's ability to embrace colour
within her conceptual thinking is restricted to concepts expressible by
words like "red" or "green" and phrases like "burnt sienna" ? It is
possible to acquire the concept of a shade of colour, and most of us
have done so. Why not say that one is thereby equipped to embrace
shades of colour within one's conceptual thinking with the very same
determinateness with which they are presented in one's visual experi
ence, so that one's concepts can capture colours no less sharply than
one's experience presents them? In the throes of an experience of the
kind that putatively transcends one's conceptual powers-an experi-
I I . The Varieties of Reference, p. 229. Obviously colour figures here as representative of

a number of features of experience.
1 2. Christopher Peacocke takes this view in recent work; for an overview see A Study of
Concepts (MIT Press, Cambridge, Mass., 1 992).
ence that ex hypothesi affords a suitable sample-one can give linguis

tic expression to a concept that is exactly as fine-grained as the expe
rience, by uttering a phrase like "that shade", in which the demonstra
tive exploits the presence of the sample.
We need to be careful about what sort of conceptual capacity this
is. We had better not think it can be exercised only when the instance
that it is supposed to enable its possessor to embrace in thought is
available for use as a sample in giving linguistic expression to it. That
would cast doubt on its being recognizable as a conceptual capacity at
all. Consider undertaking to give expression to a thought in a way
that exploits the availability of a sample, by saying (possibly to one
self) something like " My visual experience represents something as
being of that shade". Suppose we try to hold that this attempted ex
pression of a thought contains an expression of a colour concept that
is restricted to this occasion of utterance. This looks like
Wittgenstein's case of the person who says "I know how tall I am",
putting his hand on top of his head to prove it.n The putative
thought-"! am this tall", "It looks to me as if something is of that
shade''-is being construed so as to lack the distance from what
would determine it to be true that would be necessary for it to be
recognizable as a thought at all.
We can ensure that what we have in view is genuinely recognizable
as a conceptual capacity if we insist that the very same capacity to
embrace a colour in mind can in principle persist beyond the duration
of the experience itself. In the presence of the original sample, "that
shade" can give expression to a concept of a shade; what ensures that
it is a concept-what ensures that thoughts that exploit it have the
necessary distance from what would determine them to be true-is
that the associated capacity can persist into the future, if only for a
short time, and that, having persisted, it can be used also in thoughts
about what is by then the past, if only the recent past.14 What is in play
here is a recognitional capacity, possibly quite short-lived, that sets in
with the experience. It is the conceptual content of such a recognitio
nal capacity that can be made explicit with the help of a sample, some
thing that is guaranteed to be available at the time of the experience
with which the capacity sets in. Later iri the life of the capacity it can
13. Philosophical Investigations, S279.

14. Obviously people differ in the retentiveness of their memory for precise shades. No
doubt it can be cultivated, as a memory for flavours is cultivated by aspiring connoisseurs
of food or wine.
be given linguistic expression again, if the course of experience is fa

vourable; that is, if experience again, or still, presents one with a suit
able sample. But even in the absence of a sample, the capacity goes on
being exploitable as long as it lasts, in thoughts based on memory:
thoughts that are not necessarily capable of receiving an overt expres
sion that fully determines their c.o ntent.
If such recognitional capacities are conceptual, Evans's question
does not have the answer he thinks it does. It is true that we do not
have ready, in advance of the course our colour experience actually
takes, as many colour concepts as there are shades of colour that we
can sensibly discriminate. But if we have the concept of a shade, our
conceptual powers are fully adequate to capture our colour experi
ence in all its determinate detail.
What reason could there be for refusing to accept that such
recognitional-capacities are conceptual? They seem perfectly suited to
figure in an understanding of how experience takes hold of aspects of
the world, on the lines I sketched in my first two lectures. I claimed
that we can make sense of this image of experience taking hold of the
world, or being open to it, if we suppose that experience involves the
operation of capacities that are conceptual, in the sense that they are
rationally integrated into spontaneity at large. Evans's phenomeno
logical point is that the world as experience takes hold of it is more
finely grained than we could register by appealing only to conceptual
capacities expressible by general colour words and phrases. Now it is
true that the fine-grained capacities I have appealed to have a special
character, which is marked by how demonstrative expressions would
have to figure in linguistic expressions of them. But why should that
prevent us from recognizing them as rationally integrated into sponta
neity in their own way, so that they can simply take their place in my
general framework? Why, in fact, are they not so much as considered
in Evans's argument, and in the appeal by many others to the consid
eration about fineness of grain that drives Evans's argument? 1 5
1 5. Peacocke is an exception; see A Study of Concepts, pp. 83-4. But notice that even
though Peacocke there in effect acknowledges that fineness of grain is no threat to the thesis
that the content of experience is conceptual, he is not above claiming, in an earlier passage,
an advantage for his different view in the fact that "writers on the objective content of
experience have often remarked that an experience can have a finer-grained content than
can be formulated by using concepts possessed by the experiencer" (p. 67). If this claim that
writers have often made is false, why should accommodating it be an advantage for
Peacocke's view?
The very identity of one of these possibly short-term recognitional

capacities is tied to a particular case of the kind of impact on sensibil
ity that is supposed to be captured by the associated concept. A capac
ity to embrace a shade within one's thinking (as that shade, we can say
in favourable circumstances) is initiated by the figuring of an instance
of the shade in one's experience.16 There is no saying which capacity it
is in abstraction from the activating experience itself. That is how
these capacities permit the fine-grained sensuous detail that figures in
the actual course of visual life to be taken up into the conceptual con
tent of visual experience.
This means that from the standpoint of a dualism of concept and
intuition, these capacities would seem hybrids. There is an admixture
of intuition in their very constitution, and that might explain why they
do not even figure as candidates for being recognized as conceptual.
But if that is why Evans's thinking takes the course it does, it is obvi
ously damaging. Evans is trying to enforce a distance between the con
ceptual, on the one hand, and the world's impact on the senses, on the
other. If it is assumed in advance that the role of intuition in their
constitution prevents us from counting these capacities as (purely)
conceptual, the distance is being presupposed, not argued for. And
obviously this ground for refusing to accept that these capacities are
conceptual is equally illicit for those who use the fine-grained charac
ter of experience to recommend a mixed position, in which the con
tent of experience is partly conceptual and partly non-conceptual.
Evans thinks intuition and concept, dualistically conceived, need to
be shared out between experience and judgement. Proponents of the
mixed position differ from Evans in thinking intuition and concept
can be juxtaposed in experience. In so far as the claim that intuitions,
conceived as non-conceptual, must figure in experience is based on the
argument from fineness of grain, this mixed position shares its shaki
ness of foundation with Evans's view. Moreover, by simply juxtapos
ing the two species of content in experience, the mixed position makes
it difficult for itself to accommodate the strong point of Evans's view,
16. In another sense, the capacity to have that particular shade in mind is a standing one,
which requires no more than possession of the concept of a shade together with the subject's
standing powers of discrimination. Experience raises this standing potential to a degree of
actuality; the capacity to have that shade in mind as that shade is actually operative in the
experience, and potentially operative subsequently in thoughts that exploit recall of the
experience.
rhe Kanrian insight that we need to appeal to conceptual capacities in

order to make it intelligible that experience is not blind.17
•
6. A second consideration that Evans appeals to is that states of the

informational system are, as he puts it, "belief-independent" (p. 123 ).
The content of a perceptual experience cannot be explained as the
content of an appropriate actual belief, sine� there may be no belief
with a suitable content; a familiar visual illusion continues to present
its illusory appearance even though the subject does not believe that
things are as they look. Some people try to preserve a definitional con
nection between informational content and belief content, while ac
knowledging this point, by suggesting that the content of an experi
ence can be captured as something that the experience gives its subject
"a prima facie inclination to believe" (p. 124). Evans responds (ibid.):
"I cannot help feeling that this gets things the wrong way round. It is
as well to reserve 'belief' for the notion of a far more sophisticated
cognitive state: one that is connected with (and, in my opinion,
defined in terms of) the notion of judgement, and so, also, connected
with the notion of reasons. " That is, to put it in the terms I have been
using: we should reserve the idea of belief for something that can be
understood only in the context of the idea of spontaneity, the idea of
an active undertaking in which a subject takes rational control of the
shape of her thinking. Not that all one's beliefs are the result of ac
tively making up one's mind. But there is a point in reserving the title
of belief for a kind of cognitive state that is essentially within the scope
of one's powers of actively making up one's mind; even in the case of
a belief that one simply finds oneself with, the question of one's enti
tlement to it can always be raised. We can sum up what Evans is sug

gesting about belief by saying that belief is a disposition to make
judgements, and judging is essentially an act of spontaneity.
Evans is here insisting that the active business of making up one's
mind is the proper context in which to place conceptual capacities,
and that is something I have been urging throughout these lectures.
1 7. The main lines of what I say in this section date from a seminar I gave in Oxford in
1 986 (with Colin McGinn). But my thinking on these issues has since been enriched by
discussions with Sonia Sedivy, who independently arrived at similar thoughts, in reaction to
the Sellarsian idea that the sensuous specificity of perceptual experience needs to be ac
counted for in terms of impressions as opposed to concepts. See her 1990 University of
PittSburgh dissertation, "The Determinate Character of Perceptual Experience" .
Lecture III. Noll-conceptual Content 61
But he uses the point as an argument that the content of experience

cannot be conceptual, and by my lights that betrays a blind spot. The
point does not tell against the conception of experience I have been
recommending: a conception according to which capacities that be
long to spontaneity are already operative in receptivity, rather than
working on something independently supplied to them by receptivity.
Evans does not argue against that conception; it simply does not figure
among the possibilities he contemplates.
Someone who holds that the content of experience is conceptual,
and who places the idea of the conceptual in the right context, must
register a link between the conceptual capacities that she takes to be
at work in perception and the active exercise of spontaneity in judge
ments. Evans considers only one way in which one might try to regis
ter the link: namely, by identifying experiences with dispositions to
make judgements. The idea would be that these dispositions are real
ized in actual judgements only when an "other things being equal"
clause is satisfied. That is how the position accommodates the fact
that experience is "belief-independent": other things fail to be equal in
the presence of known illusions, and in any other circumstances in
which there is experience without the associated belief.
Evans objects that this picture falsifies the phenomenology of per
ception (p. 229): "The proposal is implausible, because it is not the
case that we simply find ourselves with a yen to apply some concept
a conviction that it has application in the immediate vicinity. Nothing
could more falsify the facts of the situation. " The picture interposes a
distance between experience itself and the active employment of con
cepts in judgement, a distance that is supposed to be bridged by the
idea of dispositions. Evans's protest is that, even so, the picture con
nects the content of experience too closely to active thinking for it to
be able to do justice to experience. Suppose the "other things being
equal" clause is satisfied, and there is an inclination to apply some
concept in judgement. This inclination does not just inexplicably set
in. If one does make a judgement, it is wrung from one by the experi
ence, which serves as one's reason for the judgement. In a picture in
which all there is behind the judgement is a disposition to make it, the
experience itself goes missing.
This is very perceptive, and I think it is devastating to the suggestion
that we can bring experience within the scope of a faculty of sponta
neity by conceiving experiences as dispositions to make judgements.
But the point does not touch the position I am recommending. Ac
cording to the position I am recommending, conceptual capacities are
already operative in experience itself. It is not that actual operations
of conceptual capacities first figure only in actualizations of disposi
tions to judge, with which experiences are identified-so that experi
ence is connected with concepts only by way of a potentiality. Having
things appear to one a certain way is already itself a 'mode of actual
operation of conceptual capacities.
This mode of operation of conceptual capacities is special because,
on the side of the subject, it is passive, a reflection of sensibility. In the
context of that claim, it takes work to ensure that the capacities are
recognizable as genuinely conceptual capacities-that the invocation
of the conceptual is not mere word-play. What is needed is that the
very same capacities can also be exploited in active j udgements. And
what secures this identification, between capacities that are operative
in appearances and capacities that are operative in j udgements, is the
way appearances are rationally linked into spontaneity at large: the
way appearances can constitute reasons for judgements about objec
tive reality-indeed, do constitute reasons for j udgements in suitable
circumstances ("other things being equal" ).
Now this link between experience and spontaneity is similar in
some ways to the link that is effected, in the position Evans attacks, by
conceiving experiences as dispositions to judge. But the link I envis
age, unlike that one, is a link that connects experiences to judgements
as reasons for them. That means my picture does not have the feature
that Evans complains of: that when there is an inclination to make a
judgement of experience, the inclination seems to float mysteriously
free of the situation, taking on the look of an unaccountable convic
tion that some concept " has application in the immediate vicinity" .
On the contrary, when one does have such a conviction, m y picture
allows it to be satisfactorily grounded in how things appear to one.
In my first lecture ( § 6 ), I suggested that Davidson's coherentism
reflects an obstacle in the way of seeing that operations of conceptual
capacities can be passive. The same obstacle seems to be at work in
Evans's argument from the fact that experience is " belief-indepen
dent" . In fact Davidson and Evans represent the two horns of a di
lemma posed by that obstacle. If one fails to see that conceptual ca
pacities can be operative in sensibility itself, one has two options:
either, like Davidson, to insist that experience is only causally related
to empirical thinking, not rationally; or else, like Evans, to fall into the
Myth of the Given, and try to credit experience, conceived as extra
conceptual, with rational relations to empirical thinking. Davidson
holds that the Myth of the Given can be avoided only by denying that
experience is epistemologically significant. Evans, for good reasons,
cannot stomach that denial, and he shows that he shares Davidson's
view of the possibilities by accordingly embracing a form of the Myth
of the Given. My point is that we need not confine ourselves within
this framework of possibilities. I shall return to this in the next lecture.
7. I have already mentioned a third consideration Evans appeals to.

This is the fact that we share perception (like memory) with " ani
mals" (p. 124); that is, with creatures that cannot be credited with
conceptual capacities, in the demanding sense Evans and I agree on.
This brings me back to the area of the twin discomforts I mentioned
some time ago ( § 3 ) . Both Evans and I are committed to there being
different stories to tell about perceptual goings-on in creatures with
spontaneity and in creatures without it. In the one case we can apply
the notion of experience, in a strict sense that connects it with concep
tual capacities, and in the other case we cannot. But it may seem that
Evans's position makes this implication less embarrassing, because the
position supplies us with something we can conceive as straightfor
wardly common to the two cases: namely, states of the informational
system, with their non-conceptual content.
We can shake this impression somewhat by considering the third
element of the informational system, the testimony system. According
to Evans, we have knowledge derived from operations of the testi
mony system that took place before we were in a position to under
stand the linguistic performances in question. So the testimony system
is, in itself, " more primitive" than understanding. And that is a partial
parallel to the fact that we share perception and memory with mere
animals. Now suppose our sharing perception with creatures that lack
spontaneity were a good reason for crediting our perceptual experi
ence with content that is not conceptual, on the ground that if we said
that the content of our experience is conceptual, that would put this
kind of content out of reach of those other perceivers. In that case, by
· parity of reasoning, the partially parallel fact about the primitive op-
erations of the testimony system should be a good reason for suppos
ing that non-conceptual content is involved in our mature dealings
•
with the testimony system, when we do understand the linguistic per

formances we witness. But understanding language is surely a matter
of conceptual capacities if anything is. So what is the role of these
conceptual capacities in our mature dealings with the testimony sys
tem, if the content involved is non-conceptual? A straightforward par
allel to Evans's picture of the role of conceptual capacities in experi
ence would be this: the conceptual s;apacities exercised in
understanding a linguistic performance do not enter into determining
the content with which one takes oneself to be presented, but serve
only to account for one's access to that content, which is indepen
dently determined by the operations of the informational system. But
that is surely a quite unattractive idea.
If we share perception with mere animals, then of course we have
something in common with them. Now there is a temptation to think
it must be possible to isolate what we have in common with them by
stripping off what is special about us, so as to arrive at a residue that
we can recognize as what figures in the perceptual lives of mere ani
mals. That is the role that is played in Evans's picture by informa
tional states, with their non-conceptual content. But it is not compul
sory to attempt to accommodate the combination of something in
common and a striking difference in this factorizing way: to suppose
our perceptual lives include a core that we can also recognize in the
perceptual life of a mere animal, and an extra ingredient in addition.
And if we do take this line, there is no satisfactory way to understand
the role of the supposed core in our perceptual lives. We are con
fronted with the dilemma whose horns are embraced by Davidson and
Evans, each of them, I claim, deceived in the thought that his position
is satisfactory.
We can avoid the dilemma. We do not need to say that we have
what mere animals have, non-conceptual content, and we have some
thing else as well, since we can conceptualize that content and they
cannot. Instead we can say that we have what mere animals have,
perceptual sensitivity to features of our environment, but we have it in
a special form. Our perceptual sensitivity to our environment is taken
up into the ambit of the faculty of spontaneity, which is what dis
tinguishes us from them.
I think we ought ultimately to be able to take something on those
lines in our stride. But perhaps such formulations begin to uncover the
character of the obstacle I have been alluding to: the intelligibly pow-
erful influence over the cast of our thinking that tends to obliterate the
very possibility of the right picture. The difficulty comes out in ques
tions like this: how can spontaneity permeate our lives, even to the
extent of structuring those aspects of them that reflect our natural
ness-those aspects of our lives that reflect what we share with ordi
nary animals? The thought is that the freedom of spontaneity ought to
be a kind of exemption from nature, something that permits us to
elevate ourselves above it, rather than our own special way of living
an animal life. I shall come back to these issues in the next lecture.
L E C T U R E I V
Reason and Nature
1. So far, I have been concerned with difficulties we can fall into, in

our reflection about empirical judgement and knowledge, if we try to
accommodate the point that Kant makes by talking of spontaneity.
To accept that point is to acknowledge that judging is an active em
ployment of capacities that empower us to take charge of our thinking.
But that threatens us with oscillating between two unpalatable alterna
tives. The initial threat is that we lose a connection between empirical
thinking and independent reality, a connection that there must be if
what is in question is to be recognizable as bearing on independent
reality at all. The idea of spontaneity is an idea of freedom, and that
threatens to make what was meant to be empirical thinking degenerate,
in our picture, into a frictionless spinning in a void. Recoiling from
that, we are tempted to suppose we can reinstate friction between
thought and the world by making out that justifications of empirical
judgements stop at objects of pure ostension, uncontaminated by con
ceptualization. But when we think this alternative through, we realize
that these supposed stopping points for justification cannot intelligibly
serve as a subject's reasons for her judgements. Now we are tempted to
recoil back into renouncing the need for friction.
I have been urging that we must conceive experiences as states or
occurrences in which capacities that belong to spontaneity are in play
in actualizations of receptivity. Experiences have their content by vir
tue of the fact that conceptual capacities are operative in them, and
that means capacities that genuinely belong to the understanding: it is
essential to their being the capacities they are that they can be ex
ploited in active and potentially self-critical thinking. But when these
Lecture IV. Reason and Nature 67
capacities come into play in experience, the experiencing subject is

passive, acted on by independent reality. When experience makes con
ceptual content available to one, that is itself one's sensibility in oper
ation, not understanding putting a construction on some pre-concep
tual deliverances of sensibility. At least with "outer experience" ,
conceptual content i s already borne by impressions that independent
reality makes on one's senses. This allows us to acknowledge an exter
nal constraint on the freedom of spontaneity without falling into inco
herence. So we can exorcize the spectre of frictionless spinning, which
deprives us of anything recognizable as empirical content.
The position I am urging appeals to receptivity to ensure friction,
like the Myth of the Given, but it is unlike the Myth of the Given in
that it takes capacities of spontaneity to be in play all the way out to
the ultimate grounds of empirical judgements. That is what enables us
to reinstate friction without undermining the very idea of ultimate
grounds, as the Myth of the Given does.
I have suggested (Lecture I, §6; Lecture III, § §6, 7) that there is a
difficulty in the way of accepting this conception, one whose roots
must lie deep. We can appreciate this by comparing the views of Da
vidson and Evans, as I did at the end of the last lecture. Both Davidson
and Evans aim to accommodate the point of Kant's talk of spontane
ity. Neither of them is tempted by a bald naturalism that would opt
out of this area of philosophy altogether, by denying that the sponta
neity of the understanding is sui generis in the way suggested by the
link to the idea of freedom. But they do not so much as consider the
possibility that conceptual capacities might be already operative in ac
tualizations of sensibility. It is not that they a rgue that there is no such
possibility; it simply does not figure in their thinking. And the result of
this absence is that they are faced by the choice I have described. So
long as a passive operation of conceptual capacities is not in view as
an option, one cannot even try to cast experience as a rational con
straint on empirical thinking without falling into the Myth of the
Given. But that is to picture what are supposed to be ultimate
justifications in a way that makes it unintelligible that they might exert
. rational influence at all. And now the only alternative is to stop trying
to make out that empirical thinking is rationally constrained by expe
rience. I have been urging that this is intolerable: within this concep
tion of the possibilities, there is no way to credit thought with friction
against independent reality, but we must have that if we are to have
•
empirical content in our picture at all. There is no comfortable resting

place here, and something deep must be influential in persuading Da
vidson and Evans, in tKeir different ways, to the contrary.
Davidson embraces a version of one horn of this dilemma, the re
nunciation of rational control from independent reality. He thinks a
merely causal, not rational, linkage between thinking and indepen
dent reality will do, as an interpretation of the idea that empirical con
tent requires friction against something external to thinking. But it
will not do. Thoughts without intuitions would be empty, as Kant
almost says; and if we are to avert the threat of emptiness, we need to
see intuitions as standing in rational relations to what we should
think, not just in causal relations to what we do think. Otherwise the
very idea of what we think goes missing. The items that were meant to
be thoughts are still without intuitions in the relevant sense, and so
empty. Davidson manages to be comfortable with his coherentism,
which dispenses with rational constraint on thinking from outside it,
only because he does not see that emptiness is the threat. He thinks the
only point of wanting a rational connection between intuitions and
thoughts is reassurance that we are justified in endorsing the thoughts,
as if we could take it for granted in any case that they are thoughts,
that they possess content. But if we do not let intuitions stand in ratio
nal relations to them, it is exactly their possession of content that is
put in question. When Davidson argues that a body of beliefs is sure
to be mostly true, he helps himself to the idea of a body of beliefs, a
body of states that have content. And that means that, however suc
cessfully the argument might work on its own terms, it comes too late
to neutralize the real problem for this horn of the dilemma.
Evans embraces a version of the other horn of the dilemma. Like
Davidson, Evans takes experiences to be extra-conceptual (in them
selves, we have to add in Evans's case). But unlike Davidson, Evans
thinks experiences can both he extra-conceptual and constitute a ra
tional constraint on the operations of spontaneity.. He thinks judge
ments can be "based upon" experiences even though experiences are
external to spontaneity. Evans is right, as against Davidson, in the
implicit belief that motivates this view: if thoughts are not to be
empty, that is, if they are to be thoughts at all, they must be rationally
responsive to intuitions. But Evans's position cannot be made to co
here. Davidson is right, as against Evans, that if experiences are extra
conceptual, they cannot be what thoughts are rationally based on.
When Evans invokes non-conceptual content, that serves merely to

mask the fact that by excluding experiences from the province of
spontaneity, he has made it impossible to see how an experience could
be someone's reason for a paradigmatic exercise of spontaneity such
as a judgement.
So the situation is as follows. Given an assumption that Davidson
and Evans share, they are confined to the pair of positions between
which they choose. And each has what looks like a completely cogent
argument against the other.
2. Now I want to start uncovering the presumably deep-rooted men

tal block that produces this uncomfortable situation. This is a task I
have had hanging over me since my first lecture.
Evans offers a number of grounds for holding that the content of
our perceptual experiences is non-conceptual. One of them, which I
discussed briefly at the end of my last lecture (§7), is the fact that we
share perception with creatures that lack conceptual capacities, in the
demanding sense Evans accepts: we share perception with creatures
incapable of active and self-critical thinking.
I insisted that this point cannot dislodge the picture I have been
urging, in which spontaneity permeates our perceptual dealings with
the world, all the way out to the impressions of sensibility themselves.
It cannot compel us to substitute Evans's picture, in which the spon
taneity that is distinctive of our perceptual lives is extra to a part of the
truth about us, our sensibility, that must be independent of spontane
ity, since it is common to us and more primitive perceivers. What we
share with dumb animals is perceptual sensitivity to features of the
environment. We can say there are two species of that, one permeated
by spontaneity and another independent of it. This accommodates the
combination of likeness and difference between us and dumb animals,
but not as Evans does, by factorizing the truth about us into indepen
dent components corresponding to the respects of likeness and differ
ence. And it is just as well that this alternative is available, since
Evans's factorizing move takes us back to the dilemma I began by
elaborating.
So comparing ourselves with dumb · animals cannot require us to
separate sensibility from understanding, to exclude intuitions from
the scope of spontaneity. But as I suggested at the end of my last lec
ture, the comparison can put us on the way to seeing why it is so easy
to think the separation is compulsory. It can begin to explain why we

tend to miss the very possibility that conceptual capacities, in the de
manding sense, might be operative in actualizations of our sensibility.
Dumb animals are natural beings and no more. Their being is en
tirely contained within nature. In particular, their sensory interactions
with their environment are natural goings-on. Now we are like dumb
animals in that �e, too, are perceptually sensitive to our environment.
Sentience is a feature of their animal life, and it should be something
animal in our case too. The sentience of dumb animals is one way in
which their animal being, their purely natural being, actualizes itself,
and our sentience, as an aspect of our animal life, should equally be
one way in which our natural being actualizes itself. (Even if we are
willing to suppose that our being is not purely natural, it is at least
partly natural.)
But it can seem impossible to reconcile the fact that sentience be
longs to nature with the thought that spontaneity might permeate our
perceptual experience itself, the workings of our sensibility. How
could the operations of a bit of mere nature be structured by sponta
neity, the freedom that empowers us to take charge of our active
thinking? If we see no possibility here, we are forced to suppose intu
itions must be constituted independently of the understanding, by the
senses responding naturally to the world's impacts on them. And then
we are in the space of options that Davidson and Evans locate them
selves in.
3. What is at work here is a conception of nature that can seem sheer

common sense, though it was not always so; the conception I mean
was made available only by a hard-won achievement of human
thought at a specific time, the time of the rise of modern science. Mod
ern science understands its subject matter in a way that threatens, at
least, to leave it disenchanted, as Weber put the point in an image that
has become a commonplace. The image marks a contrast between two
kinds of intelligibility: the kind that is sought by (as we call it) natural
science, and the kind we find in something when we place it in relation
to other occupants of "the logical space of reasons", to repeat a sug
gestive phrase from Wilfrid Sellars. 1 If we identify nature with what
1. See Lecture I, §2. Of course depictions of nature are linked by relations of justification.
The point is that there are no such linkages in what is depicted.
Lecture I V. Reason and Nature 71
natural science aims to make comprehensible, we threaten, at least, to

empty it of meaning. By way of compensation, so to speak, we see it
as the home of a perhaps inexhaustible supply of intelligibility of the
other kind, the kind we find in a phenomenon when we see it as gov
erned by natural law.2 It was an achievement of modern thought when
this second kind of intelligibility was clearly marked off from the first.
In a common mediaeval outlook, what we now see as the subject mat
ter of natural science was conceived as filled with meaning, as if all of
nature were a book of lessons for us; and it is a mark of intellectual
progress that educated people cannot now take that idea seriously,
except perhaps in some symbolic role.3
Now if we conceive the natural as the realm of law, demarcating it
by the way its proper mode of intelligibility contrasts with the intelli
gibility that belongs to inhabitants of the space of reasons, we put at
risk the very idea that spontaneity might characterize the workings of
our sensibility as such. The faculty of spontaneity is the understand
ing, our capacity to recognize and bring into being the kind of intelli
gibility that is proper to meaning. We disclose this kind of intelligibil-
2. The crucial contrast here is between the internal organization of the space of reasons
and the internal organization of nature, on a conception that modern natural science invites
.us to hold. This contrast echoes the Kantian contrast between the realm of freedom and the
realm of nature. It sets the agenda for much post-Kantian philosophy, and it is central to
Sellars's thinking.
In the text, I avoid a gloss that some of Sellars's followers put on what stands opposed to
the space of reasons: Rorty, for instance, speaks on Sellars's behalf of a distinction between
the logical space of reasons and the logical space of "causal relations to objects" (Philoso
phy and the Mirror of Nature [Princeton University Press, Princeton, 1979), p. 1 57). I think
this reflects a disputable picture of how modern natural science most fundamentally orga
nizes its subject matter: one that Russell protested against in his essay "On the Notion of
Cause", in Bertrand Russell, Mysticism and Logic (George Allen and Unwin, London,
1 9 1 7), pp. 1 32-51 in the 1 963 paperback edition. Russell suggested that the idea of causa
tion should be replaced, in the role of basic organizing principle for the world as viewed by
natural science, with something like the idea of law-governed processes. So the right con
trast for the space of reasons is not the space of causes but, as in my text, the realm of law.
(This leaves untouched the fact, which I exploited in my Sellarsian explanation of why the
Myth of the Given is a myth [Lecture I, S3), that a merely causal relation cannot do duty for
a justificatory relation.)
It is not just that this reading of the contrast is wrong about science; it is also disputable
in its implication that the idea of causal connections is restricted to thinking that is not
framed by the space of reasons. On my reading, the contrast leaves it possible for an area of
discourse to be in the logical space of causal relations to objects without thereby being
shown not to be in the logical space of reasons. Contrary to what Rorty's contrast implies,
reasons might be causes.
�
3. See chap. 1 of Charles Taylor, Hegel ( Cambridge Un versity Press, Cambridge, 1 975).
ity by placing things in a logical space that is sui generis, by compari

son with the realm of •law. But sensibility, as I said, is part of our
nature, part of what we share with mere animals. If that means its
operations are what they are by virtue of their positions in the realm
of law, it can seem incoherent to suppose that they might be shaped
by concepts. That would imply that their being what they are is also a
matter of positions in the contrasting logical 6pace.
Moreover, we had better not aspire to put the lost enchantment
back into the merely natural world. According to the picture I have
been recommending, our sensibility yields states and occurrences with
conceptual content. That enables us to see an experiencing subject as
open to facts. The conceptual sphere does not exclude the world we
experience. To put it another way: what we experience is not external
to the realm of the kind of intelligibility that is proper to meaning. (See
Lecture II.) But in so far as what we experience includes merely natu
ral facts, this can look like a call to regress into a pre-scientific super
stition, a crazily nostalgic attempt to re-enchant the natural world.
Let me stress that the issue here cannot be confined to our under
standing of one another, something that must involve " space of rea
sons" intelligibility. If we acquiesce in the disenchantment of nature,
if we let meaning be expelled from what I have been calling "the
merely natural", we shall certainly need to work at bringing meaning
back into the picture when we come to consider human interactions.
But it is not just in our comprehension of language, and in our making
sense of one another in the other ways that belong with that, that
conceptual capacities are operative. I have urged that conceptual ca
pacities, capacities for the kind of understanding whose correlate is
the kind of intelligibility that is proper to meaning, are operative also
in our perception of the world apart from human beings. The question
is how we can take that view without offering to reinstate the idea
that the movement of the planets, or the fall of a sparrow, is rightly
approached in the sort of way we approach a text or an utterance or
some other kind of action.
4. As I remarked ( § 1 ), neither Evans nor Davidson is tempted by

what I called "a bald naturalism": a style of thinking that would sim
ply dismiss the fuss I have made over spontaneity. In effect, they hold
that we cannot understand the idea of spontaneity naturalistically, if
"naturalistically" is used in accordance with the conception of nature

I have been describing.
That conception obviously raises a question about the status of .
spontaneity, and I am going to distinguish three styles of response.
First, there is bald naturalism, which aims to domesticate concep
tual capacities within nature conceived as the realm of law. This ap
proach need not deny that conceptual capacities belong to a faculty of
spontaneity, a faculty that empowers us to take charge of our lives.
But the idea is that if there is any truth in talk of spontaneity, it must
be capturable in terms whose fundamental role lies in displaying the
position of things in nature so conceived. Perhaps we should grant
that the relations that constitute the structure of the space of reasons,
relations of justification and the like, are not visibly there, as such, in
nature as the paradigmatic natural sciences depict it. But according to
this approach, we can reconstruct the structure of the space of reasons
out of conceptual materials that already belong in a natural-scientific
depiction of nature. And then modes of thought that place their sub
ject matter in the space of reasons, for instance, reflection that brings
spontaneity into view as such, can after all count as natural-scientific
too. No doubt they are not paradigmatically natural-scientific, but
only because it takes work to show how their distinctive concepts
serve to place things in nature.
In the most straightforward version of this approach, the task is to
reduce the structure of the space of reasons to something that is al
ready unproblematically natural on the relevant conception. But I do
not want to limit the approach to such reductionism. What matters is
just that ideas whose primary home is the space of reasons are de
picted as, after all, serving to place things in nature in the relevant
sense. On these lines, we can equate nature with the realm of law, but
deny that nature so conceived is utterly disenchanted. The claim is
that, even so conceived, naturalness does not exclude the intelligibility
that belongs to meaning.
Opponents of this kind of naturalism hold that the contrast that
poses our difficulties, the contrast of logical spaces, is genuine. The
structure of the space of reasons stubbornly resists being appropriated
within a naturalism that conceives nature as the realm of law. So if
something's way of being natural is its position in the realm of law,
what is natural, as such, cannot be anything but disenchanted. I want
to distinguish two ways in which this can be elaborated: the second
and third of the three styles of response to the question about the sta
tus of spontaneity.
One is the way of thinking that I have recommended. Even though
the logical space that is the home of the idea of spontaneity cannot be
aligned with the logical space that is the home of ideas of what is
natural in the relevant sense, conceptual powers are nevertheless oper
ative in the workings of our sensibility, in actualizations. of our animal
nature, as such. As I have acknowledged, this can seem to express a
nostalgia for a pre-scientific world-view, a call for a re-enchantment
of nature. And certainly it requires that we resist the characteristically
modern conception according to which something's way of being nat
ural is its position in the realm of law.
The third approach diverges in just that respect. What I have in
mind here is a way of thinking that is almost explicit in Davidson.
Davidson opposes a baldly naturalistic domestication of, in effect,
the idea of spontaneity. He urges that concepts of "propositional atti
tudes" make sense only as governed by a "constitutive ideal of ratio
nality" .4 In the terms I have been using, the claim comes to this: the
fundamental point of those concepts is to subserve the kind of intelli
gibility that is proper to meaning, the kind of intelligibility we find in
something when we place it in the space of reasons.5 On that basis
Davidson argues that we cannot reduce those concepts to concepts
governed by a different "constitutive ideal " , or, to put it in Sellarsian
terms, concepts whose home is a different logical space. Specifically,
and again putting it in the terms I have been using: the intellectual role
of those spontaneity-related concepts cannot be duplicated in terms
of concepts whose fundamental point is to place things in the realm
of law.
So far, this is something like common ground for opposition to bald
naturalism. What is distinctive a bout Davidson's approach is an onto
logical claim: the very things that satisfy the sui generis concepts, the
concepts whose applicability signals the presence of spontaneity, are
4. See "Mental Events", in his Essays on Actions and Events (Clarendon Press, Oxford,
1 980), pp. 207-25; especially pp. 221-3.
S. I am taking Davidson's thought to concern what I have been calling, in Kantian terms,
"the spontaneity of the understanding". I base this on the obvious convergence berween
Davidson's i nvocation of "the constitutive ideal of rationality", to explain what is special
about the conceptual apparatus he is concerned with, and rhe Sellarsian image of the space
of reasons, which I have exploited to gloss Kant's idea of spontaneity.
Lecture IV. Reason ami N,1 ture 75
already in principle available to an investigation whose concern is the

realm of law. The constitutive focus on the two kinds of intelligibility
separates two batches of conceptual equipment, but it does not sepa
rate their subject matter. Davidson makes this ontological claim
specifically about events: every event, even those that fall under the
concepts that subserve " space of reasons" intelligibility, can in princi
ple be made intelligible in terms of the operations of natural law.
Davidson's purpose here is to make room for holding that the
satisfiers of the sui generis concepts stand in causal relations, to one
another and to other things, without threatening the thesis that causal
relations hold only between occupants of the realm of law. Given that
thesis, satisfiers of the sui generis concepts can be causally linked only
if they are also occupants of the realm of law; and Davidson says they
are, even though they are not revealed as such by their satisfaction of
the sui generis concepts.6 But we come closer to my concerns if we
consider a counterpart purpose, in which the thesis that causal rela
tions hold between occupants of the realm of law is replaced by the
thesis that to be natural is to have a position in the realm of law. In
this context, the point of the ontological claim would be to make
room for holding that the satisfiers of the sui generis concepts are
items in nature, even though their satisfaction of the sui generis con
cepts does not disclose their positions in the realm of law.
I do not {here at least) want to question the truth of the ontological
claim. I just want to note how this approach excludes the conception
of experience I have recommended. So long as we do not dispute that
something's way of being natural is its place in the realm of law, the
fact that sensibility is natural works together with the fact that the
concept of spontaneity functions in the space of reasons, so as to rule
out the possibility that spontaneity might permeate the operations of
sensibility as such-at least if we set our faces against a baldly natu
ralistic integration of the space of reasons within the realm of law.
According to the ontological thesis, the items that instantiate the sui
generis spontaneity-related concepts have a location in the realm of
law. But the concepts are sui generis precisely in that it is not by virtue
of their location in the realm of law that things instantiate those con-
6. So a reason can be a cause, though it is not by virtue of its rational relationships that
it stands in causal relations.
cepts. So if we go on equating something's place in nature with its

location in the realm of law, we are debarred from holding that an
experience has its conc�tual content precisely as whatever natural
phenomenon it is.
It is common ground that impressions of sense are manifestations
of sentient life and hence natural phenomena. The strategy I am con
sidering ensures that it cannot be as the natW"al phenomena they are
that impressions are characterizable in terms of spontaneity. Their
place in nature is their location in the quite different structure of the
realm of law. So actualizations of a natural capacity of sensibility,
considered as such, can only be intuitions on a dualistic conception:
products of disenchanted nature operating independently of sponta
neity. And that locks us into the framework of possibilities that Da
vidson and Evans move in. I have argued that this is intolerable.
5. I have distinguished three conceptions of how spontaneity relates

to nature. If one is intolerable, we are left with the other two.
One alternative is to revert to bald naturalism. It should be begin
ning to emerge how that can be attractive, not just because it con
forms to a scientism that shapes much contemporary thinking, but
also because it offers what can seem to be the only escape from a
philosophical impasse. I have been exploiting the appearance of a
forced choice between the Myth of the Given and a coherentism that
renounces external rational constraints on thinking. A tempting diag
nosis of this philosophical difficulty traces it to the idea that we can
not have thought in our picture unless we secure an application for sui
generis notions of rational justification and the like: notions that
function in their own logical space, which is alien to the structure of
the realm of law. If we can discard that idea, this whole sea of philo
sophy can subside. Bald naturalism tells us not to go on being nagged
by these anxieties; instead we should set about saving whatever is
worth saving in our conception of ourselves, by reconstructing it in
terms of conceptual equipment that is already unproblematically nat
uralistic.
The only other option is to see our way to accepting what I have
been urging: that a concept of spontaneity that is sui generis, in the
way that threatens to pose the problem, can nevertheless enter into
characterizing states and occurrences of sensibility as such-as the ac
tualizations of our nature that they are. Bald naturalism blames the
trouble on the idea that spontaneity is sui generis, but that is not the
only possible target of suspicion. There is also the naturalism that
equates nature with the realm of law. We need that background to
make it appear that we can acknowledge a sui generis character for
spontaneity only by locating ourselves in the framework of possibili
ties that Davidson and Evans move in.
It would be a cheat, a merely verbal manoeuvre, to object that nat
uralism about nature cannot be open to question. If we can rethink
our conception of nature so as to make room for spontaneity, even
though we deny that spontaneity is capturable by the resources of
bald naturalism, we shall by the same token be rethinking our concep
tion o f what it takes for a position to deserve to be called "nat
uralism".
6. The rethinking requires a different conception of actualizations of

our nature. We need to bring responsiveness to meaning back into the
operations of our natural sentient capacities as such, even while we
insist that responsiveness to meaning cannot be captured in natura
listic terms, so long as "naturalistic" is glossed in terms of the realm
of law.
It can easily seem that there is no space to move here. Setting our
faces against bald naturalism, we are committed to holding that the
idea of knowing one's way about in the space of reasons, the idea of
responsiveness to rational relationships, cannot be reconstructed out
of materials that are naturalistic in the sense that we are trying to
supersede. This can easily seem to commit us to a rampant platonism.7
It can seem that we must be picturing the space of reasons as an au
tonomous structure-autonomous in that it is constituted indepen
dently of anything specifically human, since what is specifically
human is surely natural (the idea of the human is the idea of what
pertains to a certain species of animals), and we are refusing to natu
ralize the requirements of reason. But human minds must somehow be
able to latch on to this inhuman structure. So it looks as if we are
picturing human beings as partly in nature and partly outside it. What
7. I usc the lower case to stress that I mean the label "platonismh in something like the
sense that it bears in the philosophy of mathematics. I imply no connection with Plato,
beyond the general resemblance in imagery that underlies the usc of the term in the mathe·
matical context. I shall say something against connecring the position with Plato in Lecture
VI (SlJ.
we wanted was a naturalism that makes room for meaning, but this is
no kind of naturalism at all. 8
But there is a way out. We get this threat of supernaturalism if we
interpret the claim that the space of reasons is sui generis as a refusal
to naturalize the requirements of reason. But what became available
at the time of the modern scientific revolution is a clear-cut under
standing of the realm of law, and we can refuse to equate that with a
new clarity about nature. This makes room for us to insist that spon
taneity is sui generis, in comparison with the realm of law, without
falling into the supernaturalism of rampant platonism.
To reassure ourselves that our responsiveness to reasons is not su
pernatural, we should dwell on the thought that it is our lives that are
shaped by spontaneity, patterned in ways that come into view only
within an inquiry framed by what Davidson calls "the constitutive
ideal of rationality" . Exercises of spontaneity belong to our mode of
living. And our mode of living is our way of actualizing ourselves as
animals. So we can rephrase the thought by saying: exercises of spon
taneity belong to our way of actualizing ourselves as animals. This
removes any need to try to see ourselves as peculiarly bifurcated, with
a foothold in the animal kingdom and a mysterious separate involve
ment in an extra-natural world of rational connections.
This does not require that we blur the contrast between the space o f
reasons a n d the realm of law. T o see exercises o f spontaneity as natu
ral, we do not need to integrate spontaneity-related concepts into the
structure of the realm of law; we need to stress their role in capturing
patterns in a way of living. Of course there would be no contrast here
if the idea of lives and their shapes belonged exclusively or primarily
within the logical space of the realm of law, but there is no reason to
suppose that that is so.
7. The best way I know to work into this different conception of

what is natural is by reflecting on Aristotle's ethics.
For Aristotle, virtue of character in the strict sense is distinct from
a merely habitual propensity to act in ways that match what virtue
8 . Davidsonian monism is no help here. It is no comfort to reflect that all the items we
speak of figure in nature, if we are still stuck with truths about some of them that look
supernatural. The problem posed by the contrast berween the space of reasons and the
realm of law, in the context of a naturalism that conceives nature as the realm of law, is not
ontological but ideological.
Lecture 1 V. Reason and N.l t u rc 7 ')
would require.9 Virtue of character properly so called includes a

specifically shaped state of the practical intellect: " practical wisdom",
in the standard English translation. 1 0 This is a responsiveness to some
of the demands of reason (though that is not Aristotle's way of putting
it). The picture is that ethics involves requirements of reason that are
there whether we know it or not, and our eyes are opened to them by
the acquisition of " practical wisdom". So " practical wisdom" is the
right sort of thing to serve as a model for the understanding, the fac
ulty that enables us to recognize and create the kind of intelligibility
that is a matter of placement in the space of reasons.
Modern readers often credit Aristotle with aiming to construct the
requirements of ethics out of independent facts about human nature. 1 1
This i s t o attribute t o Aristotle a scheme for a naturalistic foundation
for ethics, with nature playing an archaic version of the role played by
disenchanted nature in modern naturalistic ethics. Given the preva
lence of readings on these lines, it is difficult to get Aristotle's picture
into view in the way I want, as a model for a radical rethinking of
nature. Read like this, Aristotle's picture of ethical understanding
counts rather as a peculiar sort of bald naturalism.
But I think this kind of reading is a historical monstrosity. This re
assuring role for nature can seem to make sense only as a response to
a kind of anxiety about the status of reasons--ethical reasons in this
case-that is foreign to Aristotle. What underlies the anxiety is pre
cisely the conception of nature that I have represented as distinctively
modern. When the kind of intelligibility that belongs to the realm of
law is clearly demarcated, the other side of the coin is an appreciation
that the structure of the space of reasons is special by comparison.
And there is an intelligible tendency to set up the study of nature as the
very exemplar of what it is to investigate how things are. Thus, when
nature threatens to extrude the space of reasons, philosophical wor
ries are generated about the status of rational connections, as some
thing that we can be right or wrong about. One response to these
9. Nicomachean Ethics 6. 1 3.
1 0. That of Sir David Ross, The Nicomachean Ethics of Aristotle (Oxford University
Press, London, 1 954). Terence Irwin's translation (Hackett, Indianapolis, 1985) has "intd·
ligence". (Aristotle's word is "phronesis ".)
1 1. For a reading of this sort, see chap. 3 of Bernard Williams, Ethics a11d tl>e Limits of
Philosophy (Harvard University Press, Cambridge, Mass., 1985). There is something similar
in chap. 9 of Alasdair Macintyre, After Virt11e (Duckworth, London, 1 98 1 ) .
•
worries is to resist the extrusion in the manner of bald naturalism,

leaving the conception qf nature unquestioned but insisting that after
all the putative rational requirements that we want to defend can be
founded on, or constructed our of, independent facts of nature. If I am
right about the genesis of the worries, it must be anachronistic to read
something like this into Aristotle. 12
If a person conceives her practical situation in terms provided for
her by a specific ethical outlook, that will present her with certain
apparent reasons for acting. On a better understanding of Aristotle's
picture, the only standpoint at which she can address the question
whether those reasons are genuine is one that she occupies precisely
because she has a specific ethical outlook. That is a standpoint from
which those seeming requirements are in view as such, not a founda
tional standpoint at which she might try to reconstruct the
demandingness of those requirements from scratch, out of materials
from an independent description of nature.
Aristotle scarcely even considers that doubts might arise about the
specific ethical outlook he takes for granted.13 On the reading I am
attacking, this shows confidence that he could validate the demands of
that ethical outlook by appealing to nature. But I think it shows his
immunity to our metaphysical anxieties; he is simply not vulnerable to
the sort of worries such a conception would address. It may also show
12. Compare the worries of epistemology as we know it, which are widely acknowledged
to be distinctively modern. At root the point is the same. Sellars traced the anxieties of
modern epistemology to the fact that the idea of knowledge is the idea of a position in a
justificatory network; this was the context in which he mentioned the space of reasons.
Anxiety about knowledge, of the familiar modern kind, results when that fact is juxtaposed
wirh the threatened extrusion of the space of reasons from nature. It is not that the idea of
knowledge as a position in the space of reasons was new-as if it was not until around the
seventeenth century that people hit on the thought that becomes so pregnant in modern
epistemology, thar knowledge is a normative status. But before the modern era, the idea that
knowledge is a normative status was not felt to stand in tension with, say, the idea that
knowledge might be the result of an exercise of natural powers. A naturalism that responds
to this tension by setting out to ground the normative connections that constitute the space
of reasons, after all, within nature, conceived in the very way that threatens the tension, is
quite different from a naturalism like Aristotle's, which never feels a tension here, and has
no need for imagery of grounding or foundations. (l discuss this further in �Two Sorts of
Naturalism • , forthcoming in a Festschrift for Philippa Foot edited by Rosalind Hursthouse
and Gavin Lawrence.)
1 3 . He shows no interest in addressing such doubts; he stipulates that he is addressing
only people in whom that ethical outlook has been inculcated (Nicomachean Ethics 1 .4,
1 095 b4-6 ).
something less interesting: a tendency to smugness, for which we can

easily correct.
Like any thinking, ethical thinking is under a standing obligation to
reflect about and criticize the standards by which, at any time, it takes
itself to be governed. (For this point applied to empirical thinking, see
Lecture I, §5; Lecture II, §6.) Aristotle may be less than duly sensitive
to the obligation in the case of ethics, but it is implicit in the very idea
of a shaping of the intellect, and that is what "practical wisdom" is.
Now it is a key point that for such reflective criticism, the appropriate
image is Neurath's, in which a sailor overhauls his ship while it is
afloat. This does not mean such reflection cannot be radical. One can
find oneself called on to jettison parts of one's inherited ways of think
ing; and, though this is harder to place in Neurath's image,
weaknesses that reflection discloses in inherited ways of thinking can
dictate the formation of new concepts and conceptions. But the essen
tial thing is that one can reflect only from the midst of the way of
thinking one is reflecting about. So if one entertains the thought that
bringing one's current ethical outlook to bear on a situation alerts one
to demands that are real, one need not be envisaging any sort of vali
dation other than a Neurathian one. The thought is that this applica
tion of one's ethical outlook would stand up to the outlook's own
reflective self-scrutiny.
No doubt the self-scrutiny of an ethical outlook can take note of
independent facts about the layout of the realm of law, when they are
relevant. But that is not to say that we can reconstruct the idea of
genuine ethical demands on us from materials that are naturalistic in
that sense. The idea of getting things right in one's ethical thinking has
a certain autonomy; we need not conceive it as pointing outside the
sphere of ethical thinking itself.
Of course the fact that a thought passes muster so far, in reflective
examination of a way of thinking from within, does not guarantee
that it is acceptable. The way of thinking, including its implicit stan
dards for self-scrutiny, may have hitherto unnoticed defects, such as
parochialism or reliance on bad prejudice. 14 But we can only make
honest efforts to eliminate the sorts of defects we know our thinking
risks, and perhaps to expand our conception of ways things might go
14. "Bad prejudicew is not a pleonasm. For the idea that prejudice, so far from being
necessarily a bad thing, is a condition for understanding, see Gadamer, Truth and Method,
pp. 277-85.
wrong, so as to be on guard against other potential sources of error.

The best we can achieve is always to some extent provisional and in
conclusive, but that is no reason to succumb to the fantasy of an ex
ternal validation.
If we enrich it, then, to include a proper place for reflectiveness,
Aristotle's picture can be put like this. The ethical is a domain of ra
tional requirements, which are there in any case, whether or not we
are responsive to them. We are alerted to these demands by acquiring
appropriate conceptual capacities. When a decent upbringing initiates
us into the relevant way of thinking, our eyes are opened to the very
existence of this tract of the space of reasons. Thereafter our appreci
ation of its detailed layout is indefinitely subject to refinement, in
reflective scrutiny of our ethical thinking. We can so much as under
stand, let alone seek to justify, the thought that reason makes these
demands on us only at a standpoint within a system of concepts and
conceptions that enables us to think about such demands, that is,
only at a standpoint from which demands of this kind seem to be
1n vtew.
In my second lecture ( §4), I talked about a sideways-on picture of
understanding and the world, a picture that places reality outside a
boundary enclosing the conceptual. A scientistic naturalism encour
ages a version of that picture, in which what is outside the boundary
is the realm of law. And if conceiving reality as the realm of law leaves
it disenchanted, what is outside the boundary can contain no demands
of reason or the like. Hence the very idea of sensitivity to real demands
of reason looks spooky, unless we can reconstruct it from materials
that are naturalistic in the relevant sense.
This picture tries to cast the realm of law in a naturalized version of
the role Kant gives to the supersensible. But this is not how to correct
what is unsatisfactory in Kant's thinking about the supersensible:
keeping its basic shape, and merely naturalizing what lies outside the
conceptual. This way we lose the insight that Kant spoils by putting it
in the framework of his talk of the supersensible; we lose what we
discern through obstacles in Kant, a way to make sense of empirical
thinking as rationally answerable to the reality that it aims to be
about. And if we lose that, we put in question the very possibility of
empirical thinking. This kind of naturalism tends to represent itself as
educated common sense, but it is really only primitive metaphysics.
The way to correct what is unsatisfactory in Kant's thinking about the

supersensible is rather to embrace the Hegelian image in which the
conceptual is unbounded on the outside. I have urged {Lecture II, §8)
that that is precisely not threatening to common sense, to the convic
tion that the world is independent of our thinking.
In any case, this pseudo-Kantian naturalism has nothing to do with
Aristotle. It responds to a philosophical anxiety whose sources post
date Aristotle by a couple of millennia. In Aristotle's conception, the
thought that the demands of ethics are real is not a projection from,
or construction out of, facts that could be in view independently of the
viewer's participation in ethical life and thought, so that they would
be available to a sideways-on investigation of how ethical life and
thought are related to the natural context in which they take place.15
The fact that the demands bear on us is just, irreducibly, itself. It is
something that comes into view only within the kind of thinking that
conceives practical situations in terms of such demands.
In crediting this self-standing character to ethical demands and
thoughts about them, the outlook I am attributing to Aristotle looks
like a kind of platonism. But it is not what I called " rampant platon
ism" {§6). We fall into rampant platonism if we take it that the struc
ture of the space of reasons is sui generis, but leave in place the equa
tion of nature with the realm of law. That makes our capacity to
respond to reasons look like an occult power, something extra to our
being the kind of animals we are, which is our situation in nature. But
in Aristotle's conception, the rational demands of ethics are not a lien
to the contingencies of our life as human beings. Even though it is not
supposed that we could explain the relevant idea of demandingness in
terms of independently intelligible facts about human beings, still or
dinary upbringing can shape the actions and thoughts of human be
ings in a way that brings these demands into view.
1 5. According to the kind of reading I am disputing, Aristotle attributes this founda

tional status to facts about what it would be for a human life to be fulfilling. But the nmion
of a fulfilling life figures in Aristotle in a way that is already ethical through and through.
The relevant motivations are shaped by concerns that are already ethical. (See Nicomachean
Ethics 1 .7, 1098a 1 6-17.)
A different candidate to ground demands of reason is social interactions, conceived as
describable without presupposing a context structured by just such demands, those onl's or
others that would replace them in our thinking after better reflection. I shall say something
about this sort of position in the next lecture.
To focus the way this conception can serve as a model for us, con
sider the notion of second nature. The notion is all but explicit in
Aristotle's account of how ethical character is formed.16 Since ethical
character includes dispositions of the practical intellect, part of what
happens when character is formed is that the practical intellect ac
quires a determinate shape. So practical wisdom is second nature to its
possessors. I have been insisting that for A,ristotle the rational de
mands of ethics are autonomous; we are not to feel compelled to val
idate them from outside an already ethical way of thinking. But this
autonomy does not distance the demands from anything specifically
human, as in rampant platonism. They are essentially within reach of
human beings. We cannot credit appreciation of them to human na
ture as it figures in a naturalism of disenchanted nature, because
disenchanted nature does not embrace the space of reasons. But
human beings are intelligibly initiated into this stretch of the space
of reasons by ethical upbringing, which instils the appropriate shape
into their lives. The resulting habits of thought and action are second
nature.
This should defuse the fear of supernaturalism. Second nature
could not float free of potentialities that belong to a normal human
organism. This gives human reason enough of a foothold in the realm
of law to satisfy any proper respect for modern natural science.
The point is clearly not restricted to ethics. Moulding ethical char
acter, which includes imposing a specific shape on the practical intel
lect, is a particular case of a general phenomenon:. initiation into con
ceptual capacities, which include responsiveness to other rational
demands besides those of ethics. Such initiation is a normal part of
what it is for a human being to come to maturity, and that is why,
although the structure of the space of reasons is alien to the layout of
nature conceived as the realm of law, it does not take on the remote
ness from the human that rampant platonism envisages. If we gener
alize the way Aristotle conceives the moulding of ethical character, we
arrive at the notion of having one's eyes opened to reasons at large by
acquiring a second nature. I cannot think of a good short English ex
pression for this, but it is what figures in German philosophy as
Bildung.
1 6 . See Book 2 of Nicomachean Ethics. For an excellent discussion, see M. F. Burnyeat,
" Aristotle on Learning to Be Good", in Amelie Oksenberg Rorty, ed., Essays on Ari.;tot/e's
Ethics (University of California Press, Berkeley, 1 980), pp. 69-92.
8. In these lectures so far, I have taken perceptual experience as an

object lesson, in order to describe a kind of predicament we tend to
fall into when we think about aspects of the human condition. I pro
mised to try to uncover a deep-rooted but, as we can come to realize,
non-compulsory influence on our thinking that accounts for the pre
dicament. I have now introduced my candidate for that role: the nat
uralism that leaves nature disenchanted. We tend to be forgetful of the
very idea of second nature. I am suggesting that if we can recapture
that idea, we can keep nature as it were partially enchanted, but with
out lapsing into pre-scientific superstition or a rampant platonism.
This makes room for a conception of experience that is immune to the
philosophical pitfalls I have described.
We need to recapture the Aristotelian idea that a normal mature
human being is a rational animal, but without losing the Kantian idea
that rationality operates freely in its own sphere. The Kantian idea is
reflected in the contrast between the organization of the space of rea
sons and the structure of the realm of natural law. Modern naturalism
is forgetful of second nature; if we try to preserve the Kantian thought
that reason is autonomous within the framework of that kind of nat
uralism, we disconnect our rationality from our animal being, which
is what gives us our foothold in nature. The upshot is a temptation to
drop the Kantian thought and naturalize our rationality in the manner
of bald naturalism. I have described that as opting out of this region
of philosophy. If we want to combine avoiding the problems with a
more substantial acknowledgement of them, we need to see ourselves
as animals whose natural being is permeated with rationality, even
though rationality is appropriately conceived in Kantian terms.
What 1 am suggesting could be put in terms of a task for philoso
phy, to effect a reconciliation. That may make it seem that 1 am hold
ing out for a conception of philosophy that Richard Rorty has tried to
represent as outmoded.17 But l do not feel threatened by Rorty's ob
jections, for two reasons. First, the need for reconciliation that 1 envis
age arises at a particular period in the history of ideas, one in which
our thought tends, intelligibly, to be dominated by a naturalism that
constricts the idea of nature. My suggestion does not involve an idea
that Rorty persuasively attacks, that there is a timeless set of obliga
tions for philosophy. Second, the task I envisage is not the one that
1 7. See Philosophy and the Mirror of Nature.

Rorty deconstructs, the reconciling of subject and object, or thought

and world. My proposal is that we should try to reconcile reason and
nature, and the point of doing that is to attain something Rorty him
self aspires to, a frame of mind in which we would no longer seem to
be faced with problems that call on philosophy to bring subject and
object back together. If we could·achieve a firm hold on a naturalism
of second nature, a hold that could not be shaken by any temptation
to lapse back into ordinary philosophical worries about how to place
minds in the world, that would not be to have produced a bit of con
structive philosophy of the sort Rorty aims to supersede. In
Wittgenstein's poignant phrase, it would be to have achieved " the dis
covery that gives philosophy peace" . u
1 8 . Philosophical investigations, S 1 33 .
L E C T U R E V
Action, Meaning, and the Self
1. I have been discussing how to accommodate the point of Kant's

remark "Thoughts without content are empty, intuitions without con
cepts are blind". According to the view I have recommended, concep
tual capacities are in one sense non-natural: we cannot capture what
it is to possess and employ the understanding, a faculty of spontane
ity, in terms of concepts that place things in the realm of law. But
spontaneity is inextricably implicated in receptivity, and our capaci
ties of receptivity, our senses, are part of our nature. So in another
sense conceptual capacities must be natural. Otherwise, if we ac
knowledge that the idea of spontaneity works in a sui generis concep
tual framework, we commit ourselves to picturing the deliverances of
sensibility as intuitions without concepts. That leaves us oscillating
between, on the one hand, a coherentism that makes no sense of how
thoughts could fail to be empty and, on the other hand, a vain appeal
to bare presences. This awkward choice looks like the only alternative
to a bald naturalism, a way of thinking that would not even let these
philosophical difficulties get started.
It can be hard to see any way out, short of denying that the idea of
spontaneity works in a sui generis conceptual framework. Sensibility
is one of our natural powers. How can spontaneity be non-natural, in
any sense, and yet be inextricably implicated when our sensory capac
ities are activated?
But in my last lecture I suggested that the difficulty is illusory. Our
nature is largely second nature, and our second nature is the way it is
not just because of the potentialities we were born with, but also be
cause of our upbringing, our Bildung. Given the notion of second na-
'
ture, we can say that the way our lives are shaped by reason is natural,
even while we deny that the structure of the space of reasons can be
integrated into the layotlt of the realm of law. This is the partial re-en
chantment of nature that I spoke of.
This is not to lapse into a rampant platonism. In rampant platon
ism, the structure of the space of reasons, the structure in which we
place things when we find meaning in them� is simply extra-natural.
Our capacity to resonate to that structure has to be mysterious; it is as
if we had a foothold outside the animal kingdom, in a splendidly non
human realm of ideality. But thanks to the notion of second nature,
there is no whiff of that here. Our Bildung actualizes some of the po
tentialities we are born with; we do not have to suppose it introduces
a non-animal ingredient into our constitution. And although the
structure of the space of reasons cannot be reconstructed out of facts
about our involvement in the realm of law, it can be the framework
within which meaning comes into view only because our eyes can be
opened to it by Bildung, which is an element in the normal coming to
maturity of the kind of animals we are. Meaning is not a mysterious
gift from outside nature.
These considerations should undermine one attraction of what I
have called " bald naturalism" . If we refuse to naturalize spontaneity
within the realm of law, it can seem that we are trapped in the philo
sophical impasse I began with, the forced choice between coherentism
and the Myth of the Given. But the refusal to naturalize spontaneity
does not generate the impasse all by itself. There is also the naturalism
that equates disclosing how something fits into nature with placing it
in the realm of law. Apart from that naturalism, we would not have
to conclude that since the operations of sensibility are, as such, natu
ral goings-on, considered in themselves they can only be intuitions
without concepts. So once the idea of second nature is in place, the
impasse need no longer look like a recommendation for bald natural
ism. We can claim both that the notion of spontaneity functions in a
conceptual framework that is alien to the structure of the realm of
law, and that it is needed for describing actualizations of natural pow
ers as such. If nature had to be identified with the realm of law, that
combination would be incoherent. But once we allow that natural
powers can include powers of second nature, the threat of incoherence
disappears.1
I. Perhaps "bald naturalism" is nO[ a good label for a position adopted on the grounds I
consider here. Someone who thought on these lines would be alive to the case for supposing
2. In order to introduce the attractions of a relaxed naturalism, I

have exploited philosophical difficulties about perceptual experience.
But this focus was not essential; the difficulties exemplify a type.
I have stressed that experience is passive (Lecture I, SS). In that re
spect the position I have been recommending coincides with the Myth
of the Given. The passivity of experience allows us to acknowledge an
external control over our empirical thinking, if passivity will cohere
with an involvement of spontaneity. But it is difficult to see how the
combination is possible; my last lecture aimed to uncover the source
of the difficulty.
Now the difficulty concerns not the passivity of experience as such,
but its naturalness. The problem is that operations of sensibility are
actualizations of a potentiality that is part of our nature. When we
take sensing to be a way of being acted on by the world, we are think
ing of it as a natural phenomenon, and then we have trouble seeing
how a sui generis spontaneity could be anything but externally related
to it. But passivity is not part of the very idea of what it is for a natural
potentiality to be actualized. So we should be able to construct a train
of thought about actualization of active natural powers, duplicating
the difficulties I have exploited in the case of passive natural powers.
Consider, say, the capacity to move one of one's limbs, as it appears
in the framework of a naturalism that leaves nature disenchanted. I
have been discussing the way such a naturalism distances a sui generis
spontaneity from a subject's enjoyment of sensibility, given that sensi
bility is a natural capacity. In a parallel way, it distances a sui generis
spontaneity from exercises of the power to move one's limbs, given
that that power is natural. The result is a similar difficulty in our
reflection about bodily action.
Kant says "Thoughts without content are empty, intuitions without
concepts are blind". Similarly, intentions without overt activity are
idle, and movements of limbs without concepts are mere happenings,
not expressions of agency. I have urged that we can accommodate the
point of Kant's remark if we accept this claim: experiences are actual
izations of our sentient nature in which conceptual capacities are in-
that the structure of the space of reasons is sui gene�is, alien to the structure of the realm of
law, but would suppose that all the same it cannot be so, on pain of philosophical impasse.
Here naturalism is motivated by a reflective avoidance of unprofitable philosophy. My label
"bald naturalism" better fits an unreflective scientism: not a principled avoidance of
unprofitable philosophy, but a way of thinking that does not explicitly appreciate what
threatens to lead to it. Perhaps people who think like this should be congratulated on their
immunity, but it ought not to be mistaken for an intellectual achievement.
extricably implicated. The parallel is this: intentional bodily actions

are actualizations of our active nature in which conceptual capacities
are inextricably implicated.
But just as a naturalism that disenchants nature excludes the under
standing from actualizations of our sentient nature as such, so here it
excludes our mastery of concepts from what ought to �e recognizable
as actualizations of our active nature as such: goings-on in which nat
ural things, like limbs, do natural things, like moving. And this exclu
sion has characteristic results when we reflect about action. Shut out
from the realm of happenings constituted by movements of ordinary
natural stuff, the spontaneity of agency typically tries to take up resi
dence in a specially conceived interior realm. This relocation of spon
taneity may be seen as a renunciation of naturalism, or the interior
realm may be conceived as a special region of the natural world.2 Ei
ther way, this style of thinking gives spontaneity a role in bodily ac
tion only in the guise of inner items, pictured as initiating bodily go
ings-on from within, and taken on that ground to be recognizable as
intentions or volitions. The bodily goings-on themselves are events in
nature; in the context of a disenchanting naturalism, combined with a
conviction that the conceptual is sui generis, that means that they can
not be imbued with intentionality.3 They are actualizations of natural
powers, and for that reason they can figure in this style of thinking
only as mere happenings. (No doubt they can be singled out from
mere happenings at large, but only in that they are the effects of those
interior operations of spontaneity.)
2 . The second option, with its counterpart i n philosophical reflection about other aspects
of the mental, fits Cartesian philosophy of mind, at least on a certain familiar reading (the
one given currency by Gilbert Ryle in The Concept of Mind [Hutchinson, London, 1 949]).
We can understand Cartesian philosophy of mind, on that reading, as reflecting an inchoate
awareness of what only later comes into focus as the s11i generis character of the space of
reasons. It is intelligible that at an early stage in the formation of the conception of place
ment in nature by comparison with which the space of reasons is s11i generis-the concep
tion that, when it comes into focus, is seen as excluding what is done by concepts that
function in the space of reasons-there should be an inclination to suppose that what is
special about those concepts is that they place their satisfiers in a special traer of nature,
with nature understood according to a rudimentary form of the very conception that sets up
the strain.
3. Like anything else, they can conform to a specification, which might give the content
of an intention. But on this view intention cannot be more intimately involved in move
ments of an agent's limbs than it is in, say, some failings of trees. In both cases, we can have
an evenr conforming to a specification framed by an agent, and occurring in consequence of
the framing of the specification. Intention has only an external bearing on the event itself.
This withdrawal of agency from nature, at any rate from the ordi
nary nature in which the movements of our bodies occur, strains our
hold on the idea that the natural powers that are actualized in the
movements of our bodies are powers that belong to us as agents. Our
powers as agents withdraw inwards, and our bodies with the powers
whose seat they are-which seem to be different powers, since their
actualizations are not doings of ours but at best effects of such do
ings-take on the aspect of alien objects. It comes to seem that what
we do, even in those of our actions that we think of as bodily, is at
best to direct our wills, as it were from a distance, at changes of state
in those alien objects.4 And this is surely not a satisfactory picture of
our active relation to our bodies. Just as the exclusion of spontaneity
from sentient nature obliterates anything we can recognize as empir
ical content, so here the withdrawal of spontaneity from active na
ture eliminates any authentic understanding of bodily agency.
Here too, we can return to sanity if we can recapture the Aristote
lian idea that a normal mature human being is a rational animal, with
its rationality part of its animal, and so natural, being, not a mysteri
ous foothold in another realm. The way to do that is to realize that
our nature is largely second nature.5
3. In my last lecture ( §7) I claimed that Aristotle's ethics contains a

model case of a naturalism that would not stand in the way of a sat
isfactory conception of experience (and of action, I can now add).
The position is a naturalism of second nature, and I suggested that
we can equally see it as a naturalized platonism. The idea is that the
dictates of reason are there anyway, whether or not one's eyes are
opened to them; that is what happens in a proper upbringing. We
need not try to understand the thought that the dictates of reason are
objects of an enlightened awareness, except from within the way
of thinking such an upbringing initiates one into: a way of thinking
4. "At best" because such a picture is unstable. What are supposed to be willings are
distanced from any occurrence in ordinary nature, in a way that ultimately undermines the
very idea of willing.
5. My aim in these remarks about agency is only ro bring out that the philosophical
anxieties I have been exploiting are general: the application to experience is just one case.
Much more could be said about the application ro action. In particular, I think the way
certain bodily goings-on are our spontaneity in action, nor just effects of it, is central to a
proper understanding of rhe self as a bodily presence in the world, something that comes to
the fore later in this lecture (§5). But I shall not expanq on that.
92 .\ 1 I :\ D .\ :-.; D \\ 0 R L D
that constitutes a standpoint from which those dictates are already

In VIeW.
This naturalized platonism is quite distinct from rampant platon
ism. In rampant platbnism, the rational structure within which mean
ing comes into view is independent of anything merely human, so that
the capacity of our minds to resonate to it looks occult or magic:J,l.
Naturalized platonism is platonistic in that the structure of the space
of reasons has a sort of autonomy; it is not derivative from, or reflec
tive of, truths about human beings that are capturable independently
of having that structure in view. But this platonism is not rampant: the
structure of the space of reasons is not constituted in splendid isola
tion from anything merely human. The demands of reason are essen
tially such that a human upbringing can open a human being's eyes to
them.
Now rampant platonism figures as a pitfall to avoid in Wittgen
stein's later writings about meaning and understanding.6 And I think
naturalized platonism is a good way to understand what Wittgenstein
is driving at there.
I want to stress how different this is from many readings of
Wittgenstein. Many readers implicitly attribute to Wittgenstein a phi
losophical stance that I described in my last lecture (§7): a stance in
which one finds a spookiness in the very idea that requirements of
reason are there for subjects to have their eyes opened to them, unless
the idea can be reconstructed out of independent facts. This poses a
philosophical task, and the thought is that Wittgenstein points to a
way of executing it by appealing to social interactions, described in a
way that does not presuppose the material to be reconstructed.7
If we try to construct something that can pass for possession of
meaning, the kind of intelligibility that is constituted by placement in
6. Rampant plawnism is clearly operative in the idea of a '"superlative fact" (Philosoph

ical Investigations, § 1 92), which Wittgenstein registers as an idea that one can be tempted
into when one reflects on "how meaning [something] can determine the steps in advance"
(§ 1 90). In Wittgenstein's presentation of the syndrome, the supposed '"superlative fact"
tends to be pictured in terms of a super-mechanism, like an ordinary mechanism except that
it is made of some inconceivabl)· rigid material; see §97.
7. See Saul A. Kripke, Wittgenstein on Rules and Private Language (Basil Blackwell,
Oxford, 1 982). Crispin Wright has an independent and partly similar reading. Both start
from Wingenstein's rejecting the mythology of rampant platonism. Kripke's Wittgenstein
concludes that there is nothing that constitutes our sensitivity to the requirements that
Lecture V. Action, Meani71g, and the Self 93
the space of reasons, in this communitarian or "social pragm:.l tist"

style, we cannot see meaning as autonomous. Indeed, that is the point
of this kind of reading: the sense of spookiness reflects the conviction
that any platonism about meaning, any position that credits meaning
with autonomy, must be a rampant platonism, with its characteristic
trafficking in the occult. When the autonomy of meaning is renounced
on these grounds, that puts in question a seemingly commonsense
conception of the objectivity of the world, the reality that our com
mand of meaning enables us to think and talk about. If there is no
thing to the normative structure within which meaning comes into
view except, say, acceptances and rejections of bits of behaviour by
the community at large, then how things are-how things can be said
to be with a correctness that must partly consist in being faithful to
the meanings one would exploit if one said that they are thus and
so-cannot be independent of the community's ratifying the judge
ment that things are thus and so. The most dear-sighted proponents
of this sort of reading explicitly embrace this consequence on
Wittgenstein's behalf. 8
I think the consequence is intolerable, but I am not going to work
at justifying that assessment. What I want to bring out now is a point
about the philosophical orientation of this sort of approach: it is out
of tune with something central to Wittgenstein's own conception of
what is to be done in philosophy, namely, his "quietism", his rejection
of any constructive or doctrinal ambitions.9 I believe that ensures that
this sort of approach must miss Wittgenstein's point.
Modern philosophy has taken itself to be called on to bridge dual
istic gulfs, between subject and object, thought and world. This style
of approach to meaning sets out to bridge a dualism of norm and
nature. The claim might be that this is a deeper dualism, the source of
meaning places on us; instead, we must understand the role of that idea in our lives in terms
of our participation in a community. Wright reads the same materials as gesturing towards
a substantive account of what constitutes our grasp of meaning. (Only gesturing:
Wittgenstein's "official" renunciation of substantive philosophical theory, his "quietism",
prevents him from acknowledging that he is in the business of constructive philosophy.) See
Wright's "Critical Notice of Colin McGinn, Wittgenstein on Meaning", Mind 98 (1 989),
289-305.
8. See Wright's discussion of "ratification-t.ranscendence", in chap. 11 of Wittgenstein
on the Foundations of Mathematics (Duckworth, London, 1980).
9. Wright is self-conscious about this; his response is to deplore Witrgensrein's affirma·
rions of "quietism".
the familiar dualisms of modern philosophy. So far, so good; this fits

the picture I have been urging. But what is debatable is how we ought
to respond to the deeper dualism.
Ordinary modern philosophy addresses its derivative dualisms in a
characteristic way. It takes its stand on one side of a gulf it aims to
bridge, accepting without question the way its target dualism con
ceives the chosen side. Then it constructs something as close as possi
ble to the conception of the other side that figured in the problems,
out of materials that are unproblematically available where it has
taken its stand. Of course there no longer seems to be a gulf, but the
result is bound to look more or less revisionist. (How revisionist de
pends on how urgent the original apparent problems were: how firmly
entrenched was the way of thinking that gave the appearance of an
uncrossable chasm.) Phenomenalism is a good example of a philo
sophical construction with this traditional shape; it aims to overcome
a nxiety about a gap between experience and the world by construct
ing the world out of experience, still conceived in j ust the way that
gives rise to the anxiety.
This putatively Wittgensteinian style of approach to meaning ap
proaches its deeper dualism in exactly that way. The impetus to the
approach is that one finds a spookiness in norms if they are conceived
platonistically. This reflects looking at norms from nature's side of the
duality of norm and nature; nature is equated with the realm of law,
and that poses the familiar threat of disenchantment. Now any
platonism has the effect that norms are on the far side of a gulf, and
that sets a philosophical task with a familiar shape: to build as close a
likeness as possible to what threatened to seem out of reach, using
only materials that are reassuringly present on the hither side of the
threatening gulf. The aim is that the gulf should disappear. If our con
struction lacks features that what posed the problem seemed to pos
sess, that is a revisionism that was only to be expected.
This is one more spasm of ordinary modern philosophy, touting
itself as the last; not what Wittgenstein aspires to, which is that we
should see through the apparent need for ordinary philosophy. And
that is not just a quirk of Wittgenstein's self-conception, to be set
aside while we proceed to read him as just another ordinary philoso
pher. The aspiration is not fantastic. The naturalism of second nature
that I have been describing is precisely a shape for our thinking that
would leave even the last dualism not seeming to call for constructive
philosophy. The bare idea of Bildung ensures that the autonomy of
meaning is not inhuman, and that should eliminate the tendency to be
spooked by the very idea of norms or demands of reason. This leaves
no genuine questions about norms, apart from those that we address
in reflective thinking about specific norms, an activity that is not par
ticularly philosophical. There is no need for constructive philosophy,
directed at the very idea of norms of reason, or the structure within
which meaning comes into view, from the standpoint of the natural
ism that threatens to disenchant nature. We need not try to get mean
ing into view from that standpoint.
Of course the category of the social is important. Bildung could not
have its place in the picture if that were not so. But the point is not
that the social constitutes the framework for a construction of the
very idea of meaning: something that would make the idea safe for a
restrictive naturalism, the sort that threatens to disenchant nature.
Wittgenstein says, " Commanding, questioning, recounting, chatting,
are as much part of our natural history as walking, eating, drinking,
playing" . 10 By "our natural history", he must mean the natural history
of creatures whose nature is largely second nature. Human life, our
natural way of being, is already shaped by meaning. We need not con
nect this natural history to nature as the realm of law any more tightly
than by simply a ffirming our right to the notion of second nature.
I am crediting Wittgenstein with an " ism" , naturalized platonism.
Am I myself flouting his insistence that he is not in the business of
offering philosophical doctrine? No. Recall what I said at the end of
the last lecture { § 8 ) about Rorty and "the discovery that gives philoso
phy peace". "Naturalized platonism" is not a label for a bit of con
structive philosophy. The phrase serves only as shorthand for a "re
minder", an attempt to recall our thinking from running in grooves
that make it look as if we need constructive philosophy . 1 1
4. At the end of my second lecture ( § 8 ) , I said that if we take Kant's

conception of experience out of the frame he puts it in, a story about
a transcendental affection of receptivity by a supersensible reality, ir
1 0. Philosophical Investigations. §25.

1 1 . Cf. Philosophica/ Jnvestigations, S 127.
becomes just what we need. Outside that frame, Kant's conception is

a satisfactory way to avoid our dilemma, the apparently forced choice
between the Myth qf the Given and a coherentism that renounces ex
ternal constraints on thinking. But the frame spoils the insight, be
cause the radical mind-independence of the supersensible comes to
seem exemplary of what any genuine mind-independence would be,
and then when Kant purports to attribute mind-independence to the
ordinary empirical world, as it figures in his thinking, that looks
merely disingenuous. I left hanging the question why Kant locates his
attempt at the necessary insight in this unsatisfactory context.
Now there are familiar features of Kant's thinking that can help
explain why he is attracted by the idea of an unknowable supersensi
ble reality, apparently in violation of his own standards for what
makes sense. The transcendental framework gives the appearance of
explaining how there can be knowledge of necessary features of expe
rience. And Kant thinks acknowledging the supersensible is a way to
protect the interests of religion and morality. This last point, indeed,
relates quite directly to my question. There are pressures within ethi
cal thinking that tend to distort ethics into what Bernard Williams
distinguishes as "the morality system" .12 One feature of "the morality
system" is an appearance that one could be genuinely responsible
only for exercises of a completely unconditioned freedom. That can
help explain why Kant is prone to suppose genuine spontaneity
would have to be wholly unconstrained. The best we could have in
experience, empirically conceived, is a naturally constrained sponta
neity. That would be bound to look second-rate in comparison with
the unconditioned freedom that moral responsibility is supposed to
require.
But none of these explanations operates within Kant's thinking
about experience itself. These extraneous considerations, the interests
of philosophy itself, religion, and morality, can take us some distance
towards understanding why Kant's thinking about experience distorts
its own best insight, but surely they cannot be the whole explanation.
We can construct an internal explanation in terms of the pressures
of modern naturalism. No doubt the notion of Bildung is at Kant's
disposal, but not as the background for a serious employment of the
idea of second nature. For Kant, the idea of nature is the idea of the
1 2. See chap. 10 of Ethics and the Limits of Philosophy.

realm of law, the idea that came into focus with the rise of modern
science. Consider Kant's response to Hume. Hume had responded
with excessive enthusiasm to the disenchanting effect of modern nat
uralism; he thought nature had to be denied not only the intelligibility
of meaning but also the intelligibility of law. Against Hume, Kant
aims to regain for nature the intelligibility of law, but not the intelligi
bility of meaning. For Kant, nature is the realm of law and therefore
devoid of meaning. And given such a conception of nature, genuine
spontaneity cannot figure in descriptions of actualizations of natural
powers as such.
The point here is one of some delicacy. For Kant, the ordinary em
pirical world, which includes nature as the realm of law, is not exter
nal to the conceptual. In view of the connection between the concep
tual and the kind of intelligibility that belongs to meaning, l have
suggested that defending that Kantian thought requires a partial re
enchantment of nature. (See Lecture IV, §§3, 4. ) But it does not re
quire us to rehabilitate the idea that there is meaning in the fall of a
sparrow or the movement of the planets, as there is meaning in a text.
It is a good teaching of modernity that the realm of law is as such
devoid of meaning; its constituent elements are not linked to one an
other by the relations that constitute the space of reasons. But if our
thinking about the natural stops at an appreciation of that point, we
cannot properly comprehend the capacity of experience to take in
even the meaningless occurrences that constitute the realm of law. We
cannot satisfactorily splice spontaneity and receptivity together in our
conception of experience, and that means we cannot exploit the Kant
ian thought that the realm of law, not just the realm of meaningful
doings, is not external to the conceptual. The understanding-the
very capacity that we bring to bear on texts-must be involved in our
taking in of mere meaningless happenings.13
Kant's lack of a pregnant notion of second nature explains why the
right conception of experience cannot find a firm position in his think
ing. But it does not explain how, even so, he comes so close to the
right conception. At this point I think we must simply marvel at his
insight, especially in view of how the transcendental framework pre
vents the insight from taking proper form. And it is not that the tran
scendental framework is a gratuitous afterthought. In the absence of a
1 3. I am responding here to comments by Robert Brandom and Michael Lockwood.

98 & M I N D A N D W 0 R L D
pregnant notion of second nature, the insight can appear only in that
distorted form.
If we conceive intuitions as products of disenchanted nature, and
spontaneity as non-natural, the closest we can get to the conception
we need is the Davidsonian position I discussed in my last lecture (§4):
spontaneity characterizes what are in fact operations of sentient na
ture, but it does not characterize them as such. That leaves us in our
familiar dilemma: either we must see our way to supposing that, even
so, operations of sentient nature can stand in rational relations to
thought (the Myth of the Given), or we must accept that sensibility
has no epistemological significance at all (a radical coherentism). In
effect Kant sees that this choice is intolerable. So spontaneity must
structure the operations of our sensibility as such. Since he does not
contemplate a naturalism of second nature, and since bald naturalism
has no appeal for him, he cannot find a place in nature for this re
quired real connection between concepts and intuitions. And in this
predicament, he can find no option but to place the connection out
side nature, in the transcendental framework.
Kant is peculiarly brilliant here. Even though he has no intelligible
way to deal with it, he manages to hold on to the insight that a merely
notional connection of concepts with intuitions will not do. That
forces him into a way out that is unintelligible by his own lights. The
real connection has to be that spontaneity is involved in the transcen
dental affection of receptivity by the supersensible. And now the good
thought that our sensibility opens us to a reality that is not external
to the conceptual can show up only in a distorted form, as if the
ordinary empirical world were constituted by appearances of a reality
beyond.
Alongside this strain to which Kant's thinking is subject, when he
tries to find a place for the essential insight about experience in the
lethal environment of a naturalism withou t second nature, we should
note another historical influence: the rise of Protestant individualism.
That brings with it a loss or devaluation of the idea that immersion in
a tradition might be a respectable mode of access to the real. Instead
it comes to seem incumbent on each individual thinker to check every
thing for herself. When particular traditions seem ossified or hide
bound, that encourages a fantasy that one should discard reliance on
tradition altogether, whereas the right response would be to insist that
Lecture V. A ction, Meaning, and the Self 99
a respectable tradition must include an honest responsiveness to

reflective criticism.
What results from this devaluation of tradition is an outlook in
which individual reason is sovereign. And that is hard to combine
with the idea that reason might be operative in states or occurrences
of sheer passivity, which would make it indebted to the world. So
what starts as a loss of the idea that reason might owe its being to a
place in a tradition shows up as a strain on the idea that reason might
owe anything to impacts from the world. I shall say a little more about
the significance of tradition in my final lecture.
5. If we could equip Kant with the idea of second nature, that would
not only free his insight about experience from the distorting effect of
the framework he tries to express it in; it would also allow the connec
tion between self-consciousness and consciousness of the world,
which figures in an equivocal way in his thinking, to take a satisfac
tory shape.
In the Transcendental Deduction, Kant seems to offer a thesis on
these lines: the possibility of understanding experiences, " from
within", as glimpses of objective reality is interdependent with the
subject's being able to ascribe experiences to herself; hence, with the
subject's being self-conscious.14
Now it would be satisfying if the self that is in question here were,
in the end at least, the ordinary self. But it is hard to make that cohere
with what Kant actually says. When he introduces the self-conscious
ness that he argues to be correlative with awareness of objective real
ity, he writes of the "I think" that must be able "to accompany all my
representations" .15 In the Paralogisms of Pure Reason, he claims that
if we credit this "I" with a persisting referent, the relevant idea of
identity through time is only formal. It has nothing to do with the
substantial identity of a subject who persists as a real presence in the
world she perceives. 1 6 The subjective temporal continuity that is a
14. For this reading of the Transcendental Deduction, see pp. 72-1 1 7 of Strawson, Tbe
Bounds of Sense.
1 5 . Critique of Pure Reason, B 1 3 1 .
1 6 . A363: "The identity of the consciousness o f myself a t different times i s . . . only a
formal condition of my thoughts and their coherence, and in no way proves the numerical
identity of my subject."
1 00 M I N D A N D W O R L D
counterpart to experience's bearing on objective reality shrinks to the

continuity of a mere point of view, not, apparently, a substantial con
tinuant. 17
Kant has a reas�n for the thesis of the Paralogisms, that only a
formal idea of persistence through time is available for the "I" in
the I think" that can "accompany all my representations". He
"
thinks anything else would commit him to a Cartesian conception· of

r
the ego.
Consider Locke's account of what a person is: "a thinking intelli
gent being, that has reason and reflection, and can consider itself as
itself, the same thinking thing, in different times and places" . 1 8 Locke
is talking about what he calls "consciousness" ; we might call it "self
consciousness" . "Consciousness" can hold together, in a single sur
vey, states and occurrences that are temporally separated; they are
conceived as belonging to the career of a continuant, a thinking thing.
To put the point in Kant's terms: in the "I think" that can "accom
pany all my representations", the reference of the "I" is understood as
reaching into the past and future. But Kant's point in the Paralogisms
is that the flow of what Locke calls "consciousness" does not involve
applying, or otherwise ensuring conformity with, a criterion of iden
tity.19 In the continuity of "consciousness" through time, there is what
appears to be knowledge of an identity, the persistence of an object
over a period; part of the content of the flow of "consciousness" is the
idea of a persisting referent for the "I" in the "I think" that can "ac
company all my representations". But when a subject makes this ap
plication of the idea of persistence, she needs no effort to ensure that
her attention stays fixed on the same thing. For a contrast, consider
keeping one's thought focused on an ordinary object of perception
over a period. That requires the ability to keep track of things, a skill
whose exercise we can conceive as a practical substitute for the ex
plicit application of a criterion of identity. Continuity of "conscious
ness" involves no analogue to this, no keeping track of the persisting
self that nevertheless seems to figure in its content.20
1 7. See Quassim Cassaro, "Kant and Reductionism", Review of Metaphysics 43 ( 1 989),

72-106, especially pp. 87-8.
1 8. An Essay concerning Human . Understanding, ed. P. H. Nidditch (Clarendon Press,
Oxiord, 1 975), 3.27.9.
1 9 . This is Strawson's reading of the Paralogisms: The Bounds ofSense, pp. 1 62-70.
20 . For an elaboration on these lines of the point Srrawson finds in Kant, see Evans, The
Lecture V. Action, Meaning, and the Self 1 01
Now suppose we assume that, in providing for the content of this

thought of a persisting self, we must confine ourselves within the flow
of "consciousness" itself. If the topic of the thought is a substantial
continuant, what its continuing to exist consists in must be peculiarly
simple. The notion of persistence applies itself effortlessly; there is
nothing to it except the flow of "consciousness" itself. This looks like
a recipe for arriving at the conception, or supposed conception, of the
referent of "I" that figures in Descartes.
That is essentially Kant's account of how the Cartesian conception
of the ego arises. And it can easily seem that we had better draw
Kant's conclusion: the idea of persistence that figures in the flow of
"consciousness" had better be only formal. If we allowed that it is an
idea of substantial persistence, the continuing to exist of an objective
item, we would be committed to understanding self-consciousness as
awareness of a Cartesian ego.
But as I have stressed, this depends on assuming that when we pro
vide for the content of this idea of persistence, we must confine our
selves within the flow of "consciousness" . And this assumption is not
sacrosanct. Indeed, it is deeply suspect; I think it is the real root of the
Cartesian conception. If we discard it, we make room for supposing
that the continuity of the "I think" involves a substantial persistence,
without implying that the continuant in question is a Cartesian ego.
We can say that the continuity of "consciousness" is intelligible only
as a subjective take on something that has more to it than "conscious
ness" itself contains: on the career of an objective continuant, with
which the subject of a continuous "consciousness" can identify itself.
It is true that the continuity within the subjective take does not involve
keeping track of a persisting thing, but this effortlessness does not re
quire us to agree with Kant that the idea of identity here is only for
mal. Even "from within", the subjective take is understood as situated
Varieties of Reference, p. 237. Evans (or more probably his editor) seems to slip when he
suggests that the point can be captured in terms of the idea of "immunity to error through
misidentification" , which he exploits elsewhere (pp. 179-9 1 ), in connection with perceptu
ally based demonstrative thinking. We have " immunity to error through misidentification"
when a predication is not attached to its subject by way of a judgement of identity. But as
Evans points out on p. 236, " identification-freedom" in that sense is consistent with a
j udgement's depending on keeping track of its object, the very thing he denies in the case of
self-consciousness. Keeping track serves as it were instead of an " identification compo
nent", in what underlies continuing demonstrative thought about objects of perception. The
point about the self is a peculiarly strong form of "identification-freedom".
in a wider context; so there can be more content to the idea of persis

tence it embodies. The wider context makes it possible to understand
that the first person, the continuing referent of the "I" in the "I think"
that can "accompany all my representations " , is also a third person,
something whose career is a substantial continuity in the objective
world: something such that other modes of continuing thought about
it would indeed require keeping track of it. That is a way of putting
the gist of Gareth Evans's brilliant treatment of self-identification,
which builds on P. F. Strawson's brilliant reading of the Paralogisms.21
I think something on these lines is the right frame for the Kantian
thought that self-awareness and awareness of the world are inter
dependent. And there may be hints of this in Kant.22 But I do not see
how it can be his official view. If we situate self-consciousness in a
wider context, we can avoid the Cartesian ego, without needing to say
that the idea of a persisting self that figures in the continuity of "con
sciousness" is merely formal. But that is exactly what Kant thinks he
has to say. It looks as if Kant leaves in place the suspect assumption
that when we set out to provide for the content of the idea of a persist
ing self, we may not go outside the flow of "consciousness" . Only so
can it seem that to avoid the Cartesian ego, he needs to adjust the
content of the i dea of persistence. If I am right a bout the assumption,
Kant's diagnosis of Cartesian thinking does not go to the root.
The result of Kant's move is that the subjective continuity he ap
peals to, as part of what it is for experience to bear on objective real
ity, cannot be equated with the continuing life of a perceiving animal.
It shrinks, as I said, to the continuity of a mere point of view: some
thing that need not have anything to do with a body, so far as the
claim of interdependence is concerned.
This is quite unsatisfying. If we begin with a free-standing notion of
an experiential route through objective reality, a temporally extended
point of view that might be bodiless so far as the connection between
subjectivity and objectivity goes, there seems to be no prospect of
building up from there to the notion of a substantial presence in the
world. If something starts our conceiving itself as a merely formal re-
21. The Varieties of Reference, chap. 7 . Evans's thinking here can be seen as an elabora
tion of Strawson's remark, in his reading of the Paralogisms (The Bounds of Sense, p. 1 65 ) :
"'I' can b e used without criteria o f subject-identity a n d yet refer t o a subject because, even
in such a use, the links with those criteria are not severed."
22. This is what Strawson finds, at least in germ, in the Paralogisms.
ferent for "I" (which is already a peculiar notion), how could it come
to appropriate a body, so that it might identify itself with a particular
living thing? Perhaps we can pretend to make sense of the idea that
such a subject might register a special role played by a particular body
in determining the course of its experience. But that would not pro
vide for it to conceive itself, the subject of its experience, as a bodily
element in objective reality-as a bodily presence in the world.
If Kant's connection between self-awareness and awareness of the
world is to leave it open to us to regain the idea that the subjects of
our experience are our ordinary selves, then the merely formal persis
tence of the I, in the "I think" that can "accompany all my represen
tations" , had better be only an abstraction from the ordinary substan
tial persistence of the living subject of experience. 23 It had better not
be something free-standing, which we might hope to build on in re
constructing the persistence of the ordinary self. But this does not
seem to fit Kant's conception of what he is doing. Kant takes himself
to be laying bare a necessary connection that is knowable a priori.24
And it would be hard to make the idea of an abstraction from the
persistence of the ordinary self cohere with the temporal connotations
he gives to "a priori", as when he suggests that transcendental self
consciousness "precedes all data of intuitions" (A107).
It is not surprising if Kant cannot get his thinking into the right
light. Why can there not be a free-standing idea of formal subjective
continuity? The answer is this: the idea of a subjectively continuous
series of states or occurrences in which conceptual capacities are im
plicated in sensibility-or, more generally, the idea of a subjectively
continuous series of exercises of conceptual capacities of any kind,
that is, the idea of a subjectively continuous series of " representa
tions" , as Kant would say-is just the idea of a singled out tract of a
life. The idea of a subjectively continuous series of " representations"
could no more stand alone, independent of the idea of a living thing in
whose life these events occur, than could the idea of a series of diges
tive events with its appropriate kind of continuity. But in the absence
of a serious notion of second nature, this exploitation of the concept
of life, which is a quintessentially natural phenomenon, to make sense
23. See Strawson's appeal to abstraction at pp. 103-4 of The Bounds of Sense.
24. See, for instance, Critique of Pure Reason, Al l6: "We are conscious a priori of the
complete identity of the self in respect of all representations which can ever belong to our
knowledge, as being a necessary condition of the pos!ii bility of all representations."
.. : _,
of a unity within the domain of spontaneicy·, which by Kant's lights

has ro be non-natural, is not within Kant's grasp.
Kant's aim is to exorcize Cartesian temptations about the self, and
he gets to the very brink of success. He wants to acknowledge the
peculiarities of self-awareness that encourage Cartesian philosophy,
but without letting them seem to show that the object of self-aware
ness is a Cartesian ego. But he thinks the only alternative is a transcen
dental self-awareness, something that has no object substantially pres
ent in the world. If we insist on supplying this self-awareness with an
object, we can locate the object in the world only geometrically, as a
point of view. This avoids the familiar Cartesian problems about the
relation between a peculiar substance and the rest of reality. But it
leaves us with what look like descendants of those problems. If we
start from a putative sense of self as at most geometrically in the
world, how can we work up from there to the sense of self we actually
have, as a bodily presence in the world? (When I say that that is what
self-awareness is, I am not implying that one's bodily presence in the
world is always borne in on one in self-awareness.)25 Kant's insight
would be able to take satisfactory shape only if he could accommo
date the fact that a thinking and intending subject is a living animal.
But with his firm conviction that conceptual powers are non-natural,
in the sense that equates nature with the realm of law, and with his
lack of a seriously exploitable notion of second nature, he is debarred
from accommodating that fact.
6. Kant sometimes associates the idea of the conceptual especially

with generality.26 That can suggest that we should explain the relation
of concepts to intuitions in terms of the relation of predicates to sub
jects.27
This might seem to reveal a Kantian shape in a certain trend in re-
25. Cases of sensory deprivation bring out that one's bodily being need not always im·
press itself on one. In "The First Person" (in Samuel Guttenplan, ed., Mind and Language
]Clarendon Press, Oxford, 1 975], pp. 45-{)5), G. E. M. Anscom be exploits that fact to
argue that "the use of 'I' as subject" (as distinguished by Wittgenstein from "the use as
object"; The Blue and Brown Books, pp. 66-7) functions independently of a sense of one
self as a bodily presence in the world. Anscombe concludes that "the use of 'I' as subject"
should not be taken to refer, on pain of a Cartesian conception of what it refers to. But the
argument is flawed in j usr rhe same way as the argument of the Paralogisms.
26. See, for instance, A320/B377.
27. This thought is central to Strawson's reading of the First Critique. See The Bounds of
Smse, pp. 20, 72.
l.t•cturt• V. .-\ t'tiPII, ,\l,·,min.t.:. , ,, [ tl•t• .'\1·/f I OJ
cent thinking about singular reference. There was a time when the
standard view of reference was inspired by Russell's Theory of De
scriptions. The idea was that whenever a thought is directed at a par
ticular object, part of its content is given by a specification of the ob
ject in general terms: conceptual terms, the equation I am considering
would lead us to say. The trend is to recoil from this.28 There are kinds
of object-directedness in thoughts that cannot easily be made to fit
that mould. For instance, a perceptual demonstrative thought surely
homes in on its object not by containing a general specification, with
the object figuring in the thought as what fits the specification, but by
virtue of the way this sort of thinking exploits the perceptible presence
of the object itself. If the conceptual is equated with the predicative,
this resistance to the general application of the Theory of Descriptions
takes the form of saying that in the cases that warrant the resistance,
singular reference is, or rests on, an extra-conceptual relation between
thinkers and things.29 So the picture is that the conceptual realm does
have an outside, which is populated by particular objects. Thought
makes contact with objects, from its location within the conceptual
realm, by exploiting relations such as perception, which are conceived
as penetrating the outer boundary of the conceptual.
This picture fits a contemporary view of singular reference that per
haps deserves to be counted as an orthodoxy. As I said, it can seem to
be Kantian. But in fact it is not Kantian at all. In Kant, the conceptual
realm has no outside; not unless we shift to the transcendental story,
and nobody thinks the objects that, say, demonstrative thoughts focus
on are noumenal. And in any case, the picture is really incoherent,
unless "conceptual" is serving as a mere synonym for "predicative" .
Circumscribing the conceptual realm can seem t o make interesting
sense only if the circumscribed realm is singled out as the realm of
28. Influential early proponents of this trend include Saul A. Kripke, "Naming and Ne
cessity", in Donald Davidson and Gilbert Harman, eds., Semantics of Natural Language
( Reidel, Dordrecht, 1 972), pp. 253-355, 763-9, reissued as a monograph by Basil Black
well, Oxford, 1980; and Keith S. Donnellan, " Proper Names and Identifying Descriptions",
ibid., pp. 356-79. See also, predating the trend, Ruth Barcan Marcus, " Modalities and In
tensional Languages", Synthese 27 ( 1962), 303-22.
29. For a striking exposition of a position on these lines, long predating the works usu
ally cited as initiating the contemporary trend, see. Geach, Mental Acts, S 15. For a more
recent expression of the idea that the relation between thought and individual things, in the
relevant sorts of case, is extra-conceptual, see Tyler Burge, "Belief De Re", Journal of Phi
losophy 74 (1 977), 338-62.
thought. This picture has predications in the conceptual realm, but

thought supposedly has to break out of the conceptual to make con
tact with the objects of which predications are to be made. And that
leaves no room for a coherent conception of how a predication, lo
cated within the circumscribed conceptual realm, could be brought
into connection with an object.
Given this conception of the anti-Russellian revolution, it is easy to
sympathize with counter-revolutionary partisans of the generalized
Theory of De�riptions, such as John Searle. 30 If this picture is the only
recourse for someone who wants to be able to conceive thought as
focusing on particulars otherwise than by way of specifications, it
would be better to give up wanting that, and go back to the task of
trying to spell out the unobvious specifications by way of which, after
all, the sorts of thought that trigger the recoil must make contact with
their objects.
In my third lecture I attacked Gareth Evans's view that the content
of perceptual experience is non-conceptual. That was for purposes
somewhat remote from Evans's fundamental concern, which is with
singular reference. Indeed, I do not believe Evans's appeal to a non
conceptual content for perception is central to his thinking about sin
gular reference. It is easy to recast Evans's main contentions, even
about perceptual demonstrative thought, without mentioning non
conceptual content.
Evans's master thought is that Frege's notion of sense, which
Frege introduces in terms of modes of presentation, can accommo
date the sorts of connection between thinkers and particular ob
jects that have been recognized to make trouble for the generalized
Theory of Descriptions.31 The detail of Evans's work spells this
out case by case. It explains the various ways in which thoughts
focus on particular objects, always by placing thinking in its proper
context, the thinker's competent self-conscious presence in the
30. See chap. 8 of Intentionality (Cambridge University Press, Cambridge, 1983). I dis·
cuss Searle's view of singular thought in "Intentionality De Re", in Ernest LePore and Rob
ert Van Gulick, eds., John Searle and His Critics (Basil Blackwell, Oxford, 1 99 1 ), pp. 2 1 5-
25. See also my "Singular Thought and the Extent of Inner Space", in Philip Pettit and John
McDowell, eds., Subject, Thought, and Context (Clarendon Press, Oxford, 1986), pp. 1 37-
68, for a more extensive consideration of the issues that are in play here.
3 1 . So it is quite wrong to lump Frege in with Russell as a target for the recoil from the
generalized Theory of Descriptions.
Lecture V. Action, Meaning, and the Self 1 07
world.32 If we want to identify the conceptual realm with the realm

of thought, the right gloss on "conceptual" is not " predicative" but
"belonging to the realm of Fregean sense". (The stupid idea that those
come to the same thing is unfortunately still widespread.) Evans's
achievement is to show how we can avoid an apparently forced choice
between, on the one hand, the implausibilities of the generalized The
ory of Descriptions, which are increasingly palpable despite the
rearguard efforts of philosophers like Searle, and, on the other, the
incoherence of the pseudo-Kantian picture, in which thought has to
break out of its own proper sphere in order to make contact with
particulars otherwise than by specification. By invoking Frege, Evans
makes it clear how the non-specificatory relations between thinkers
and objects that proponents of the latter picture rightly insist on need
not be conceived as carrying thought outside an outer boundary of the
conceptual realm.
I think describing Evans's thinking like this reveals it to be deeply
right, at least in general outline. It is common for philosophers to
think they can dismiss Evans's position, without attention to the
wider context I have placed it in, on the ground that they find its im
plications counter-intuitive. This j ust reveals the depressing extent to
which his ground-breaking work has not been understood. That such
work can be so little appreciated is a mark of degeneracy in our philo
sophical culture.
32. I stress this to bring out how important Strawson's Kant is for the main lines of
Evans's thinking in The Varieties of Reference. Strawson's influence goes far beyond where
it is visible on the surface.
L E C T U R E V I
Rational and O ther Animals
1 . I have been considering the tendency to oscillate between two un

palatable positions: a coherentism that loses the bearing of empirical
thought on reality altogether and a recoil into a vain appeal to the
Given. I have proposed a diagnosis of this tendency: it reflects an in
telligible distortion undergone by the Aristotelian idea that normal
mature human beings are rational animals. Animals are, as such, nat
ural beings, and a familiar modern conception of nature tends to ex
trude rationality from nature. The effect is that reason is separated
from our animal nature, as if being rational placed us partly outside
the animal kingdom. Specifically, the understanding is distanced from
sensibility. And that is the source of our philosophical impasse. In
order to escape it, we need to bring understanding and sensibility, rea
son and nature, back together.
One way to avoid the dilemma is to leave unquestioned the concep
tion of nature that threatens to extrude reason from nature, but to
reconceive reason in naturalistic terms, on a corresponding under
standing of what it is for a term to be naturalistic. This position is
what I have been calling "bald naturalism". It allows us to conceive
ourselves as rational animals, but I think the conception is not
Aristotle's. Admittedly, though, bald naturalism is like Aristotelian
thinking in that it does not address the philosophical worries I have
been considering, but simply refuses to feel them.
The threat is that an animal endowed with reason would be meta
physically split, with disastrous consequences for our reflection about
empirical thinking and action. I have claimed that we can avoid the
threat even while we maintain, unlike bald naturalism, that the struc-
Lecture VI. Rational and Other Animals 1 09
ture of the space of reasons is sui generis, in comparison with the or
ganization of the realm of law. The spontaneity of the understanding
cannot be captured in terms that are apt for describing nature on that
conception, but even so it can permeate actualizations of our animal
nature. If we can see our way to accepting that, we can avoid the
philosophical difficulties while fully appreciating what makes them
gripping.
In Aristotle's conception of human beings, rationality is integrally
part of their animal nature, and the conception is neither naturalistic
in the modern sense (there is no hint of reductiveness or foundational
ism) nor fraught with philosophical anxiety. What makes. this possible
is that Aristotle is innocent of the very idea that nature is the realm of
law and therefore not the home of meaning. That conception of na
ture was laboriously brought into being at the time of the modern
scientific revolution.
I am not urging that we should try to regain Aristotle's innocence.
It would be crazy to regret the idea that natural science reveals a spe
cial kind of intelligibility, to be distinguished from the kind that is
proper to meaning. To discard that part of our intellectual inheritance
would be to return to mediaeval superstition. It is right to set a high
value on the kind of intelligibility we disclose in something when we
place it in the realm of law, and to separate it sharply from the intelli
gibility we disclose in something when we place it in the space of
reasons.
But instead of trying to integrate the intelligibility of meaning into
the realm of law, we can aim at a postlapsarian or knowing counter
part of Aristotle's innocence. We can acknowledge the great step for
ward that human understanding took when our ancestors formed the
idea of a domain of intelligibility, the realm of natural law, that is
empty of meaning, but we can refuse to equate that domain of intelli
gibility with nature, let alone with what is real.
The notion of second nature needs no particular emphasis in the
context of Aristotle's innocence, but it takes on a special significance
in this attempt to achieve a knowing counterpart. We are looking for
a conception of our nature that includes a capacity to resonate to the
structure of the space of reasons. Since we are setting our faces against
bald naturalism, we have to expand, nature beyond what is counte
nanced in a naturalism of the realm of law. But the expansion is lim
ited by the first nature, so to speak, of human animals, and by plain
facts a bout what happens to human animals in their upbringing. We

are not irresponsibly cutting the concept of nature loose from the
realm of law, as we would be if we pretended that the capaciry postu
lated by rampant platonism, a capacity to resonate to structures of
reason constituted in utter isolation from anything human, might
count as a natural capacity of h uman minds.
Rampant platonism has what intelligibiliry it has as a desperate at
tempt to keep meaning, conceived as able to come into view only
within a sui generis logical space, while acquiescing in the disenchant
ment of nature. It is not so much as intelligible that a prelapsarian
thinker might feel that temptation. My naturalized version of platon
ism is not usually in the picture, and what I am calling "rampant
platonism" is usually called just " platonism ". But if I am right about
its historical context, naming the position after Plato must do him an
injustice.
2. A rypical form of modern philosophy confronts a familiar predic

ament. It takes itself to be called on to explain how, starting from
independently available data of consciousness, we work out into a
j ustified confidence that there is an objective world. In a part of his
thinking that I discussed in my last lecture (§5), Kant aims to super
sede that conception of philosophy's task. He tries to make it plausi
ble that the very idea of data for consciousness is interdependent with
the idea that at least some states and occurrences of consciousness
constitute glimpses of an objective world. In that case, it makes no
sense to think we might start with contents of minds and build up to
objective realiry. Kant does not have occasion to consider the inverted
form of traditional philosophy that I mentioned in my last lecture (§3)
i n connection with some readings o f Wittgenstein: that is, a philoso
phy in which the project is to start with the natural world and make a
place in it for minds and their contents. But I think he would be simi
larly dismissive of that.
I suggested that Kant's insight could take satisfactory form only in
the context of a naturalism of second nature, a conception that Kant
himself does not achieve. Kant wants subjects of experience and inten
tional action to be already, just as such, in possession of objective re
ality. He wants exercises of conceptual powers to be intelligible only
as undertaken by subjects who do not need philosophy to regain the
world for them. But since he lacks a pregnant notion of second nature,
and has no inclination to naturalize spontaneity within the realm of

law, the best he can provide in the way of an experiencing and acting
subject is the merely formal referent he allows to "I", in the "I think"
that must be able to "accompany all my representations" . Such a sub
ject could not be something substantially present in the world; it is at
best a point of view. That means, I suggested, that Kant cannot succeed
in his admirable aim, to supersede traditional philosophy. Cartesian
thinking confronts familiar difficulties about how to relate a subjective
substance to objective reality, and Kant's conception is beset by what
look like descendants of those difficulties. If we start with a referent for
"I" that is only geometrically in the world, it seems impossible to build
up to a substantial presence, an embodied perceiver and agent.
Things look different if we equip Kant with a seriously exploitable
notion of second nature. Now we can give a satisfactory form to the
insight he is straining after. We can conceive exercises of capacities
that belong to spontaneity as elements in the course of a life. An expe
riencing and acting subject is a living thing, with active and passive
bodily powers that are genuinely her own; she is herself embodied,
substantially present in the world that she experiences and acts on.
This is a framework for reflection that really stands a chance of mak
ing traditional philosophy obsolete.
In that recapitulation of something I said in my last lecture, 1 have
described a philosophical project: to stand on the shoulders of the
giant, Kant, and see our way to the supersession of traditional philo
sophy that he almost managed, though not quite. The philosopher
whose achievement that description best fits is someone we take al
most no notice of, in the philosophical tradition I was brought up in,
although I have mentioned him a couple of times before: namely,
Hegel.1
3. I have urged that experience can be conceived as openness to the

world (Lectures I and II). I promised to return to the fact that experi
ence can mislead us (Lecture I, §4; Lecture II, §2). There is a tendency
to conclude that even a non-misleading experience cannot genuinely
be a case of openness to reality. In that case we cannot exploit the
1. In view of how I exploited Strawson's reading of Kant in my last lecture, this remark
implies that Strawson's Kant is more Hegel than Kant. For a reading of Hegel that takes
very seriously Hegel's own idea that his philosophy completes a Kantian project, see Rohert
•
B. Pippin, Hegel's Idealism.
image of openness in the way I have suggested, so as to bypass the

anxieties of traditional epistemology. The objection might be put like
this. "You grant that e.¥perience can be misleading. That is to grant
that what you are pleased to call 'glimpses of the world' can be sub
jectively indistinguishable from states or occurrences that cannot be
glimpses of the world, since they would lead one astray if one took
them at face value. So surely the problems of traditional epistemology
are just as pressing as they ever were. In your terminology, they come
out like this: how can one know that what one is enjoying at any time
is a genuine glimpse of the world, rather than something that merely
seems to be that?"
But this misses the point. An objection on these lines would be ap
propriate if I were aiming to answer traditional sceptical questions, to
address the predicament of traditional philosophy. That is the predic
ament in which we are supposed to start from some anyway available
data of consciousness, and work up to certifying that they actually
yield knowledge of the objective world. Of course if that is our predic
ament, we need to answer the traditional sceptical questions before we
can talk of openness to the world. But my talk of openness is a rejection
of the traditional predicament, not an attempt to respond to it.
Traditional epistemology accords a deep significance to the fact that
perception is fallible. It is supposed to show something like this: how
ever favourable a perceiver's cognitive stance may be, we cannot make
sense of subjective states of affairs constituted by a subject's letting the
layout of the objective world reveal itself to her. As the objector in
sists, something that is not a glimpse of reality, since the subject would
be misled if she took it at face value, can be subjectively indistinguish
able (at least at the time) from experiences that are veridical. That is
supposed to show that the genuinely subjective states of affairs in
volved in perception can never be more than what a perceiver has in a
misleading case.
This strains our hold on the very idea of a glimpse of reality. If we
are confined to such materials when we build a conception of the best
cognitive stance perception could yield, the most we can aspire to is
something like this: an explicably veridical presentiment of some fact
about the layout of the environment.2 We cannot have the fact itself
2. In fact we cannot even have that; empirical content as such (even as possessed by mere
presentiments) is intelligible only in a context that allows us to make sense of direct rational
constraint on minds from the world itself.
impressing itself on a perceiver. This seems off key, phenomenologi

cally at least, and we can resist it if we can so much as comprehend the
idea of a direct hold on the facts, the sort of position that the image of
openness conveys. It is true that we could not establish that we are
open to facts in any given case; at any rate not to the satisfaction of a
determined sceptic, who can always insist on exploiting fallibility to
give bite to the question how we know the present case is one of the
non-misleading ones. But that is beside the point. It would matter if it
showed that the very idea of openness to facts is unintelligible, and it
does not show that. For my present purposes, the sheer intelligibility
of the idea is enough. If the idea is intelligible, the sceptical questions
lack a kind of urgency that is essential to their troubling us, an ur
gency that derives from their seeming to point up an unnerving fact:
that however good a subject's cognitive position is, it cannot consti
tute her having a state of affairs directly manifest to her. There is no
such fact. The aim here is not to answer sceptical questions, but to
begin to see how it might be intellectually respectable to ignore them,
to treat them as unreal, in the way that common sense has always
wanted to.
Insisting on the image of openness is a way to give vivid expression
to this point: there is no good argument from fallibility to what I call
"the highest common factor conception" of our subjective position
the idea that even when things go well, cognitively speaking, our sub
jective position can only be something common between such cases
a nd cases in which things do not go well. That is a way of putting the
traditional picture of our epistemic predicament. It is not compulsory,
and the fact of fallibility cannot make it so. Traditional epistemology
cannot be vindicated by the sheer possibility of asking, "How do you
know that what you are enjoying is a genuine glimpse of the world ? ",
as was suggested by the objection at the beginning of this section. If
someone insists on asking that, on some particular occasion, an
appropriate response might start like this: "I know why you think
that question is peculiarly pressing, but it is not." If the question
still stands, nothing particularly philosophical is called for in answer
ing it.3
3. I discuss the "highest common factor conception" in "Criteria, Defeasibility, and

Knowledge", Proceedings of the British Academy 68 ( 1982), 455-79; and in "Singular
Thought and the Extent of Inner Space".
4. We can put part of the Kantian thesis that I discussed in my last

lecture, and recalled earlier in this one (§2), like this: the objective
world is present only to a self-conscious subject, a subject who can
ascribe experiences to herself; it is only in the context of a subject's
ability to ascribe experiences to herself that experiences can constitute
awareness of the world. Now that rakes us back to a restriction I drew
attention to in my third lecture (§3). It is the spontaneity of the under
standing, the power of conceptual thinking, that brings both the
world and the self into view. Creatures without conceptual capacities
lack self-consciousness and-this is part of the same package-expe
rience of objective reality.
I acknowledged that this restriction raises a question about the per
ceptual capacities of mere animals. Mere animals do not come within
the scope of the Kantian thesis, since they do not have the spontaneity
of the understanding. We cannot construe them as continually reshap
ing a world-view in rational response to the deliverances of experi
ence; not if the idea of rational response requires subjects who are in
charge of their thinking, standing ready to reassess what is a reason
for what, and ro change their responsive propensities accordingly. It
follows that mere animals cannot enjoy " outer experience" , on the
conception of "outer experience" I have recommended. And that can
seem to commit me to the Cartesian idea that brutes are automata.
The thought here is a form of one I found in Evans (Lecture III, §7).
It is a plain fact that we share perception with mere animals. Partly on
that basis, Evans suggests that when we make judgements about the
perceptible world, we must be converting experiential content of a
kind we share with mere animals, so that it must be non-conceptual,
into conceptual form. When I discussed this in my third lecture, I had
in place some of the Kantian framework that disallows Evans's con
clusion. My claim• was that when Evans argues that judgements of
experience are based on non-conceptual content, he is falling into a
version of the Myth of the Given, which is one side of the useless
oscillation from which Kant tries to rescue us. And we now have an
other angle on why Evans's conclusion will not fit into a Kantian
framework: why the framework precludes supposing that sensibility
by itself yields content that is less than conceptual but already world
involving. In the absence of spontaneity, no self can be in view, and by
the same token, the world cannot be in view either.
What about the fear that this implies an obvious falsehood: that
mere animals are not genuinely sentient? To deal with this, I want to
borrow from Hans-Georg Gadamer a remarkable description of the
difference between a merely animal mode of life, in an environment,
and a human mode of life, in the world.4 For my purposes, the point
of this is that it shows in some detail how we can acknowledge what
is common between human beings and brutes, while preserving the
difference that the Kantian thesis forces on us.
In mere animals, sentience is in the service of a mode of life that is
structured exclusively by immediate biological imperatives. That is
not to imply that the life is restricted to a struggle to keep the individ
ual and the species going. There can be immediate biological impera
tives that are at most indirectly connected with survival and reproduc
tion: for instance, the impulse to play, which is found in many
animals.5 But without falling into that kind of restrictiveness, we can
recognize that a merely animal life is shaped by goals whose control of
the animal's behaviour at a given moment is an immediate outcome of
biological forces. A mere animal does not weigh reasons and decide
what to do. Now Gadamer's thesis is this: a life that is structured only
in that way is led not in the world, but only in an environment. For a
creature whose life has only that sort of shape, the milieu it lives in can
be no more than a succession of problems and opportunities, consti
tuted as such by those biological imperatives.
When we acquire conceptual powers, our lives come to embrace
not just coping with problems and exploiting opportunities, consti
tuted as such by immediate biological imperatives, but exercising
spontaneity, deciding what to think and do. A naturalism of second
nature allows us to put it like that; we can take in our stride something
that is problematic in the context of a different sort of naturalism, that
these exercises of freedom are elements in our lives, our careers as
living and therefore natural beings. Of course it had better not be that
our being in charge of our lives marks a transcendence of biology; that
looks like a version of the rampant platonist fantasy. But we do not
fall into rampant platonism if we say the shape of our lives is no
longer determined by immediate biological forces. To acquire the
4. See Truth and Method, pp. 438-56, especially pp. 443-5.

5. I say "at most indirectly", and it is not clear that there is always even an indirect
connection. That depends on such questions as whether playing can be completely :�c·
counted for in terms of, for instance, the honing of skills that are normally called on for
survival.
spontaneity of the understanding is to become able, as Gadamer puts

it, to "rise above the pressure of what impinges on us from the world"
( Truth and Method, p� 444)-that succession of problems and oppor
tunities constituted as such by biological imperatives-into a " free,
distanced orientation" (p. 445 ) . And the fact that the orientation is
free, that it is above the pressure of biological need, characterizes it as
an orientation to the world. For a perceiver with capacities of sponta
neity, the environment is more than a succession of proble�s and op
portunities; it is the bit of objective reality that is within her perceptual
and practical reach. It is that for her because she can conceive it in
ways that display it as that.6
When I say that for a creature with a merely animal life the milieu
in which it lives can be no more than a succession of problems and
opportunities, I am not saying that it conceives its environment in
those terms. That would be to try to attribute to mere animals a full
fledged subjectivity, involving a conceptually mediated orientation
that ought, as such, to count as an orientation to the world, even
though we restrict the concepts in question to concepts things satisfy
in virtue of how they relate to biological imperatives, which is to ac
knowledge that the orientation lacks the freedom and distance that
would be required for it to be an orientation to the world at all. The
point of the distinction between living merely in an environment and
living in the world is precisely that we need not credit mere animals
with a full-fledged subjectivity, an orientation to the world, at all, not
even one that is restricted in that way. This is not to imply that fea
tures of the environment are nothing to a perceiving animal. On the
contrary, they can be problems or opportunities for it, as I have been
saying. The point is just that we must distinguish that from saying that
the animal conceives the features as problems or opportunities.
This talk of what features of the environment are for an animal
expresses an analogue to the notion of subjectivity, close enough to
ensure that there is no Cartesian automatism in our picture. Exactly
not: we need to appeal to an animal's sensitivity to features of its en
vironment if we are to understand its alert and self-moving life, the
6. Gadamer's topic, in the passage I am exploiting, is the role of language in disclosing

the world to us; it is language, he claims, that makes the "free, distanced orientation" pos
sible. I postpone any discussion of the connection between language and the spontaneity of
the understanding until the sketchy remarks at the end of this lecture; meanwhile I adapt
Gadamer's remarks to my purposes.
precise way i n which it copes competently with its environment. But

by putting the notion of sensitivity to this or that in the context of the
idea of inhabiting an environment, we ensure that we need not try to
credit a mere animal with an orientation to the world, even the world
as conceptualized in purely behaviour-related ways. To register how
far we are from the Kantian structure, we might say that what is in
question here is proto-subjectivity rather than subjectivity.
In a merely animal mode of life, living is nothing but responding to
a succession of biological needs. When Gadamer describes the con
trast as a " free, distanced orientation", the note of emancipation from
the need to produce behaviour can suggest the idea of the theoretical.
And certainly one thing that is absent from any sane conception of a
merely animal mode of life is a disinterestedly contemplative attitude,
towards the world at large or towards something in particular in it.
But the point is not just that with spontaneity the activities of life
come to include theorizing as well as acting. The lack of freedom that
is characteristic of merely animal life is not enslavement to the practi
cal as opposed to the theoretical, but enslavement to immediate bio
logical imperatives. Emancipation into the " free, distanced orienta
tion" brings intentional bodily action on to the scene no less than
theoretical activity. The picture of full-fledged subjectivity that is in
play here is not a picture of that dubiously intelligible kind of thing,
an observer and thinker that does not act in the world it observes and
thinks about.
Gadamer's account of how a merely animal life, lived in an environ
ment, differs from a properly human life, lived in the world, coincides
strikingly with some of what Marx says in his 1 844 manuscript on
alienated labour.7 ( Gadamer does not note the parallel.) This conver
gence should help to exorcize the idea of the passive observer.8 For
Marx, of course, a properly human life is nothing if not active: it in
volves the productive making over of " nature, the sensuous exterior
world" (p. 1 35). If productive activity is properly human, it can in
principle range freely over the world. This contrasts with merely ani
mal life. As in Gadamer's description, merely animal life is a matter of
dealing with a series of problems and opportunities that the environ-
7. I shall cite from the translation of David Mclellan, in Karl Marx: Early Texts (Basil
Blackwell, Oxford, 1 972), pp. 1 33-45.
8 . The convergence is surely not a coincidence. It reflects a Hegelian influence on both
texts.
ment throws up, constituted as such by biologically given needs and

drives. Marx complains memorably of a dehumanization of humanity
in wage slavery. The part of human life that should be most expressive
of humanity, namely, productive activity, is reduced to the condition
of merely animal life, the meeting of merely biological needs. And al
though it is freedom that gives its di stinctively human .character to
human life, wage slavery restricts freedom to the merely animal as
pects of what are thus only incidentally human lives. "Man {the
worker) only feels himself freely active in his animal functions of eat
ing, drinking, and procreating, at most also in his dwelling and dress,
and feels himself an animal in his human functions" {pp. 1 37-8).
Marx sums up his vision of what a properly human life would be in
a striking image: without alienation, "the whole of nature" is "the
inorganic body of man" {p. 1 39).9 We can point up the convergence
with Gadamer by glossing the image like this: the world is where a
human being lives, where she is at home. Contrast the relation of an
environment to an animal life. An environment is essentially alien to a
creature that lives in it; it is the source of "the pressure of what im
pinges on {the animal} from the world" . Not that a merely animal life
is a constant struggle, whereas a distinctively human life is peculiarly
easy. In Marx as in Gadamer, the point is not that a properly human
life is easy but rather that it is distinctively free. And that is the same
fact as that it is lived in the world, as opposed to consisting in coping
with an environment. Q
Of course we must not understand the contrast between possessing
the world and merely inhabiting an environment in terms of the ab
surd notion that when one comes to possess the world one stops hav
ing an environment-as if being human exempted one from having to
be somewhere in particular. And of course it is already open to a mere
animal to leave its present environment, in the straightforward sense
of going somewhere else. That is one response to the pressures im
posed on an animal by its present environm�nt: insufficient food, no
sexual partners, various sorts of threat. Coming to possess the world
is in part acquiring the capacity to conceptualize the facts that under
lie this already available behavioural possibility, so that one conceives
the present environment as the region of the world within one's pres-
9. He adds "in so far as it is nor itself a human body". Of course my ordinary (organic)
body is parr of nature; the striking thought is that the rest of nature is, in a different way,
my body also.
ent sensory and practical reach: as where one happens to be, in con
trast with other places where one might be.
And of course there is more than that to coming to possess the
world. For instance, possessing the world shows also in the pointless
knowable detail that one's current environment typically makes avail
able to one. Consider the richness of a normal adult human being's
visual field, which is far beyond anything that could matter for a ca
pacity to cope with merely animal needs. Marx says man is unique in
producing "according to the laws of beauty " (p. 1 40), and the point
he is making in that remark shows up also here, in a distinctive feature
of our consciousness. Our very experience, in the aspect of its nature
that constitutes it as experience of the world, partakes of a salient
condition of art, its freedom from the need to be useful.
5. As I noted in my third lecture (§3), the potential embarrassment I

have been discussing does not stop with the denial that mere animals
have "outer experience" . They cannot have " inner experience" either,
on the conception of " inner experience" I have recommended. This
generates a parallel worry that I am obliterating their sentience. But if
the considerations that can be summed up in the label " proto-subjec
tivity" can disarm the worry as it arises in the case of "outer experi
ence", they should be no less capable of disarming it as it arises in the
case of "inner experience" .
As Gada mer describes merely animal lives, they are made up of cop
ing with the "pressures " imposed on them by the environment. I have
been insisting that if we refuse to find an orientiation to the world in
such a life, that does not commit us to denying that it includes a proto
subjective perceptual sensitivity to features of the environment. And
such a life has just as much room for, say, pain or fear. Perceptual
sensitivity to the environment need not amount to awareness of the
outer world; I have been defending the claim that awareness of the
outer world can be in place only concomitantly with full-fledged sub
jectivity. Somewhat similarly, feelings of pain or fear need not amount
tq awareness of an inner world. So we can hold that an animal has no
inner world without representing it as insensate and affectless.
Sensations, emotional states, and the like are there for our subjec
tivity in an inner world. To talk in those terms is to employ the idea of
objects of experience, present in a region of reality. In my second lec
ture (§5), I suggested that we should understand this application of
the idea of objects of experience as a limiting case, because here the

objects of awareness do not exist independently of the awareness. So
the idea of the inner woPid is a limiting case of the idea of a region of
reality. ln these actualizations of our sensibility, as in others, concep
tual capacities are passively drawn into operation, in this case in a
first-person and present-tense mode. But we can recognize the relevant
conceptual capacities as operative here only l'>ecause these operations
of them embody an understanding that they are not restricted to the
first-person and present-tense mode: the very same circumstances of
which these motions of our conceptual powers constitute awareness
are also thinkable otherwise than in the first-person present. That is
what entitles us to apply the structure of awareness and object: since
the circumstances are understood to be essentially such as to admit of
that alternative angle on them, we can conceive the first-person angle
on them as a case of awareness of something, even though the object
of this awareness is really nothing over and above the awareness itself.
Now it would be absurd to try to fit this complexity of structure,
which is needed to sustain the idea that the inner world is a region of
reality, into a description of a merely animal way of living. And it
would not meet the point if we tried saying that sensations and emo
tional states are present to the proto-subjectivity of a mere animal. It
would be hopeless to claim that sensations and emotional states are
there for a mere animal in the way problems and opportunities
thrown up by the environment are there for it. That would be to sug
gest that when we refuse to credit mere animals with an inner world,
we could compensate by crediting them with an inner something else
(we could hardly say " inner environment"), somehow analogous to
the outer environment they are alive to in the way that is central to
their proto-subjectivity. "Inner environment" does not make sense,
and it does not seem plausible that we could force sense into the sug
gestion by careful choice of a different noun. But in any case nothing
I have said about the inner world prevents us from acknowledging
that mere animals can feel pain and fear.
Only for a full-fledged subjectivity can what it is to feel pain or fear
amount to a limiting-case awareness of a degenerately substantial
inner state of affairs. The limiting case of the structure of awareness
and object is in place only because of how the awareness is structured
by understanding. But nothing in the concepts of pain or fear implies
that they can get a grip only where there is understanding, and thus
Lecture VI. Ra tio na l and Otl1er Animals 121
full-fledged subjectivity. There is no reason to suppose that they can

be applied in a non-first-person way only to something capable of ap
plying them to itself in a first-person way.
6. At this point I want to stress again something I said in my third

lecture (§4). I am rejecting a picture of a mere animal's perceptual
sensitivity to its environment: a picture in which the senses yield con
tent that is less than conceptual but already such as to represent the
world. What I am rejecting is a picture of what perceptual states and
occurrences are for an animal. I have said nothing about how things
look when someone tackles scientific questions about how an animal's
perceptual machinery works. And it is hard to see how those ques
tions could be addressed without exploiting an idea of content that
represents the world but cannot be conceptual in the demanding sense
I have been using, since no animal's perceptual machinery (not even
ours) possesses the spontaneity of understanding. I do not mean to be
objecting to anything in cognitive science.
What I do mean to be rejecting is a certain philosophical outlook,
one that could be expressed like this, if its proponents would consent
to use my terms: delineating the contours of a subjectivity and deline
ating the contours of a proto-subjectivity are two tasks of much the
same kind; they differ only in that they involve two different modes of
orientation to the world, and so two different sorts of content. Ac
cording to this outlook, both tasks call for saying how the world
strikes a perceiver. (Among other things: we should also need to char
acterize other aspects of subjectivity, or proto-subjectivity, such as
sensations or emotions. ) It is just that in one kind of case the content
involved in the world's striking the perceiver in a certain way is non
conceptual.
For a vivid example of this philosophical outlook, consider the im
portance Thomas Nagel attributes to the question "What is it like to
be a bat?"10 Think first about a different question: what would it be
like if we had a sensory capacity for echo-location? That question
challenges our imagination in a striking way. We have to project our
imagination into an alternative possible world in which our subjectiv
·
ity is in part differently constituted, and the question is challenging
10. "What Is It Like ro Be a Bar?H, in Nagel's Mortal Questions (Cambridge University

Press, Cambridge, 1979), pp. 165-80.
because we have no sensory basis for this imaginative extrapolation.

Now compare the question Nagel actually asks: what is it like for bats
to do their echo-locating? Nagel's idea is that this presents just the
same challenge to the imagination, but now in a form that makes our
failure to meet it not a failure to conceptualize a mere possibility, but
a failure to get our minds around part of the layout of the actual
world. In my terms, this is to treat what is only a proto-subjectivity as
if it were a full-fledged subjectivity. Nagel's picture is that bats have a
full-fledged subjectivity whose shape is beyond the reach of our con
cepts.
I do not believe the question what it is like to be a bat should seem
any more intractable than the question what it is like to be a dog or a
cat, where there are no senses that we do not share. To answer such
questions, we need an account of the biological imperatives that struc
ture the lives of the creatures in question, and an account of the sen
sory capacities that enable them to respond to their environment in
ways that are appropriate in the light of those biological imperatives.
Saying that bats can locate prey or cave walls by sonar can be part of
such an answer in the case of bats, in much the way saying that the
vision of cats is sensitive to green and blue but not red can be part of
an answer in the case of cats. It is true that when the question is what
it is like to be a bat or a cat, the answer must try to characterize some
thing on the lines of the creature's point of view. But the accounts that
I envisage do that, to the fullest extent that makes sense, without war
ranting the thought that in the case of bats at least there are facts that
elude our understanding. These accounts capture the character of the
proto-subjectivity of the creatures in question, the distinctive ways in
which they are alive to their environments.
We are familiar "from within" with what it is like to see colours. It
is tempting to think that this equip� us to comprehend a fully subjec
tive fact about what the colour vision of cats is like, which we report
when we say they can see green and blue but not red. There must then
be parallel fully su bjective facts about bat echo-location, but they de
feat our understanding. But this is just another form of the Myth of
the Given. The idea is that mere animals already enjoy perceptual ex
perience in which the world strikes them as being a certain way, and
the only difference our understanding makes for us is that we can im
pose conceptual form on the already world-representing but less than
conceptual content that, like them, we receive in experience. Then the
problem about bats is that our imagination cannot extend to how the
conversion into conceptual form would go, in the case of the content
yielded by the echo-locating capacity. So the picture is that mere ani
mals only receive the Given, whereas we not only receive it but are
also able to put it into conceptual shape. To think like that is to set
one's foot on a familiar philosophical treadmill. 1 1
7 . How has i t come about that there are animals that possess the
spontaneity of understanding? That is a perfectly good question.
There was a time when there were no rational animals. Suppose we
had a credible account of how forces that are intelligibly operative in
nature might have led to the evolution of animals with conceptual
powers. That would definitively avert a form of rampant platonism:
the idea that our species acquired what makes it special, the capacity
to resonate to meaning, in a gift from outside nature. If we took that
seriously, we would have to suppose that when succeeding genera
tions are initiated into responsiveness to meaning, what happens is
that upbringing actualizes a potential for the development of an extra
natural ingredient, a potential implanted in the species in the sup
posed extra-natural evolutionary event.
But this request for an evolutionary story need not look very press
ing. Evolutionary speculation is not a context in which rampant
platonism is somehow particularly tempting. Reflection about the
Bildung of individual human beings should be enough to distinguish
the naturalized platonism I have recommended from rampant platon
ism. And in this reflection we can regard the culture a human being is
initiated into as a going concern; there is no particular reason why we
should need to uncover or speculate about its history, let alone the
origins of culture as such. Human infants are mere animals, distinctive
only in their potential, and nothing occult happens to a human being
in ordinary upbringing. If we locate a variety of platonism in the con
text of an account of Bildung that insists on those facts, we thereby
ensure that it is not a rampant platonism. Mere ignorance about how
1 1 . Nagel could have made many of the points he wanted without leaving the domain of
subjectivity properly so called (by my lights). Perhaps Martians have an echo-locating ca
pacity, which ligures in the rational basis of their world-view in the way our senses do in the
basis of ours. I have no need to deny that there might be concepts anchored in sensory
capacities so alien to ours that the concepts would be unintelligible to us. What I am object
ing to is only the way this point gets focused, in the case of bats, on a supposed non-concep
tual content that we cannot convert inro conceptual shape.
... � .. '
. . -
human culture might have come on the scene in the first place is hardly
a plausible starting point for an argument that initiation into it must
actualize an extra-natural• potential in human beingsY
And in any case, if we do speculate about how animals might have
evolved into a way of living that includes initiating their young into a
culture, we must be clear that that is what we are doing. It would be
one thing to give an evolutionary account of the fact that normal
human maturation includes the acquisition o ( a second nature, which
involves responsiveness to meaning; it would be quite another thing to
give a constitutive account of what responsiveness to meaning is. I
have been granting that it is reasonable to look for an evolutionary
story. This is not a concession to the sort of constructive philosophical
& account of meaning that I discussed in my last lecture (§3): something
whose point would be to make the relevant sort of intelligibility safe
for a naturalism without second nature. That is a misbegotten idea,
and there is no comfort for it here.
8. Michael Dummett has claimed that the fundamental tenet of ana

lytic philosophy is that philosophical questions about thought are to
be approached through language.13 In these lectures, I have concerned
myself with thought; I have tried to describe a way of conceiving how
thought bears on the world that would be immune to some familiar
philosophical anxieties. And so far I have scarcely mentioned lan
guage. So it might seem that I have enrolled myself as an opponent of
analytic philosophy in Dummett's sense.
But any such impression would be quite superficial.
I have followed Kant in taking thought to be an exercise of the un
derstanding: "the mind's power of producing representations from it
self, the spontaneity of knowledge" . 14 The power of spontaneity com
prises a network of conceptual capacities linked by putatively rational
connections, with the connections essentially subject to critical reflec-
1 2. It is true, however, that the good questions we can raise in the evolutionary context
come as dose as good questions can to the philosophical questions I want to exorcize.
13. See "Can Analytical Philosophy be Systematic, and Ought lr to Be?", in his Truth
mrd Other Enigmas (Duckworth, London, 1 978), pp. 437-5 8. At p. 442 Dummett writes:
" for frege, as for all subsequent analytical philosophers, the philosophy of language is the
foundation of all other philosophy beca use it is only by the ana lysis of language that we can
ana lyse thought."
1 4. Critique of Pure Reason. A51/B7 5.
tion. I have claimed that experience must stand in rational relations to

judgement if we are to be able to understand the very possibility of
empirical content; and I have claimed that we can make sense of ratio
nal relations between experience and judgement only in the context of
an equation between the space of concepts and the space of reasons.
Thought can bear on empirical reality only because to be a thinker at
all is to be at home in the space of reasons. And being at home in the
space of reasons involves not just a collection of propensities to shift
one's psychological stance in response to this or that, but the standing
potential for a reflective stance at which the question arises whether
one ought to find this or that persuasive.
Now it is not even clearly intelligible to suppose a creature might be
born at home in the space of reasons. Human beings are not: they are
born mere animals, and they are transformed into thinkers and inten
tional agents in the course of coming to maturity. This transformation
risks looking mysterious. But we can take it in our stride if, in our
conception of the Bildung that is a central element in the normal mat
uration of human beings, we give pride of place to the learning of
language. In being initiated into a language, a human being is intro
duced into something that already embodies putatively rational link
ages between concepts, putatively constitutive of the layout of the
space of reasons, before she comes on the scene. This is a picture of
initiation into the space of reasons as an already going concern; there
is no problem about how something describable in those terms could
emancipate a human individual from a merely animal mode of living
into being a full-fledged subject, open to the world. A mere animal,
moved only by the sorts of things that move mere animals and exploit
ing only the sorts of contrivances that are open to mere animals, could
not single-handedly emancipate itself into possession of understand
ing. Human beings mature into being at home in the space of reasons
or, what comes to the same thing, living their lives in the world; we
can make sense of that by noting that the language into which a
human being is first initiated stands over against her as a prior embod
iment of mindedness, of the possibility of an orientiation to the world.
This way of accepting the basic tenet of analytic philosophy is some
distance from any Dummett considers. Dummett focuses on two
"principal functions" of language: as "instrument of communication"
and as "vehicle of thought" . His conclusion is that we should take
neither of them to be primary.15 But that is because he thinks those

functions of language are both fundamental. In the picture I am sug
gesting, they are secondary. The feature of language that really mat
ters is rather this: that a natural language, the sort of language into
which human beings are first initiated, serves as a repository of tradi
tion, a store of historically accumulated wisdom about what is a rea
son for what. The tradition is subject to reflective modification by
each generation that inherits it. Indeed, a standing obligation to en
gage in critical reflection is itself part of the inheritance. (See Lecture
I, §5; Lecture II, §7.) But if an individual human being is to realize her
potential of taking her place in that succession, which is the same
thing·as acquiring a mind, the capacity to think and act intentionally,
at all, the first thing that needs to happen is for her to be initiated into
a tradition as it stands. 16
15. See "Language and Communication", in Alexander George, ed., Reflections on
Chomsky (Basil Blackwell, Oxford, 1989), pp. 1 92-2 1 2. When I delivered this lecture, I
wrongly attributed to Dummett the view that language's function as instrument of commu
nication is primary. My point was just to lead up to language's function as repository of
tradition, and I have felt free to recast what I say so as to arrive there without falsifying
Dummett's position. I am grateful to Christopher Peacocke for setting me straight.
16. The concept of tradition is central in Gadamer's thinking about understanding; see
Truth and Method, passim.
A F T E R W O R D
P A R T I
Davidson in Context
1. In the lectures, I use Davidson's coherentism exclusively as a foil

to the view of experience that I recommend. Here I want to place
Davidson's thinking about the epistemology of empirical thought in a
historical context, constituted by a strand in the recent development
of the American pragmatist tradition. I hope this will bring out some
thing the lectures obscure: the extent to which I can count Davidson
as an ally rather than an opponent.
2. The two dogmas of empiricism that W. V. Quine attacked in his

celebrated essay of that title were, first, that there is a "fundamental
cleavage" (p. 20) between the analytic, in the sense of statements true
by virtue of meaning alone, and the synthetic, in the sense of state
ments whose truth is dependent not only on meaning but also on the
world; and, second, that "empirical significance" can be parcelled out
statement by statement among the body of statements that express
our view of the empirical world.
What we ought to say instead of the second dogma, according to
Quine, is this: "The unit of empirical significance is the whole of sci
ence" (p. 42: this is the source for the way I have put the second
dogma). An alternative formulation is this: "our statements about the
external world face the tribunal of experience not individually but as
a corporate body" (p. 41 ) . If these are two formulations of the same
thought, Quine is implicitly glossing empirical significance in terms of
being subject to the tribunal of experience. That makes it look as if
Quine's conception of "empirical significance" corresponds to the
Kantian conception of empirical content, or bearing on the empirical
world, that I defend in the lectures. According to that conception, the

fact that something, say a belief, or-more congenially to Quine-a
whole world-view, bears on the world, in the sense of constituting a
stand taken as to how things are, depends on its being vulnerable to
the world for a verdict on its acceptability; and the verdict can be
delivered only through experience.
The rejected first dogma claims that the truth of a synthetic state
ment depends on two factors, meaning and the world; an analytic
statement is one for which the "world" factor is null. Now Quine's
positive picture retains this duality of factors on which truth depends.
He says (p. 36): "It is obvious that truth in general depends on both
language and extralinguistic fact." His claim is not that there are not
those two factors, but just that we cannot separate them out statement
by statement. In the context of an apparently uncontentious empiri
cism, the "world" factor is just the answerability to experience that is
summed up as "empirical significance". So Quine can reformulate the
"obvious" dependence of truth on both language and extralinguistic
fact, in an expression of the thesis that rejects the second dogma, by
saying (p. 41): "Taken collectively, science has its double dependence
upon language and experience; but this duality is not significantly
traceable into the statements of science taken one by one."
This validates an impression that is anyway given by the structure
of Quine's essay: that rejecting the second dogma is the fundamental
point. Quine's positive thinking in the essay is encapsulated in the the
sis that the unit of empirical significance is the whole of science. With
the duality of factors retained, the first dogma figures as something
that can be correct only if the second dogma is, so that rejecting the
second can suffice to reject both. The first dogma is the thesis that
there are true statements that are analytic in the sense that for them
the "world" factor-the factor of dependence on experience, "empir
ical significance"-is null. If "empirical significance" cannot be
shared out among individual statements anyway, that undermines the
very idea of a statement with no "empirical significance" . "No empir
ical significance of its own" could only be a special case of "some
empirical significance of its own" . If it makes no sense to suppose a
particular statement has its own positive quantity of "empirical
significance", a determinate share of the "empirical significance" of
the total world-view in whose expression the statement figures, then it
Afterword, Part I 131
equally makes n o sense to suppose there might be statements for

which the quantity is zero. 1
3. A s I said, Quine's positive picture, his "empiricism without the

dogmas" (p. 42), retains a counterpart to the duality that figures in
spelling out the idea of analyticity. Truth-which we must now think
of as primarily possessed by a whole world-view-depends partly on
" language" and partly on "experience" . " Language" here labels an
endogenous factor in the shaping of systems of empirical belief, dis
tinguishable-though only for whole systems-from the exogenous
factor indicated by "experience " . When we acknowledge this exoge
nous factor, we register that belief is vulnerable, by way of "the tri
bunal of experience ", to the world it aims to be true of. So " lan
guage" figures, in Quine's holistic context, as a counterpart to
"meaning" as " meaning" figured in the now debunked contrast be
tween "true by virtue of meaning alone" (true in a way that involves
no vulnerability to the world) and "true by virtue of both meaning and
the way the world is" .
"Empirical significance" registers the exogenous factor i n this con
trast, the answerability to something outside the system. And the
nearest thing in Quine's positive thinking to the old notion of meaning
is "language", which stands on the other side of the retained duality,
as the endogenous factor. " Empirical significance" is not meaning as
it figured in the idea that there might be statements true by virtue of
meaning alone; and it is not a functional descendant of meaning, so
conceived, in the novel environment of Quine's holism, but a func
tional descendant of what stood precisely in contrast to meaning in
the old version of the duality.
The fact that Quine's "empirical significance" is lone side of a holis
tic counterpart to the old duality means that, in spite of the conver
gence between Quine's talk of facing the tribunal of experience and
my talk of a rational vulnerability to intuitions, we cannot gloss
Quine's "empirical significance" in terms of bearing on the empirical
world in the sense in which I use that phrase: what stand one takes
on how things are in the world when one adopts a belief or a world
view. For Quine, the two factors are distinguishable, even though
I . The structurally parallel idea, the idea of statements for which the "meaning" factor is
null, is clearly self-defeating.
only for whole systems, and that means that the "empirical
significance" of a world-view cannot amount to its empirical content
in the sense of how, in• adopting the world-view, one takes things to
be in the empirical world. That requires the other factor, the endo-
genous one, also.
So far, this might be merely a terminological oddity about Quine's
use of the phrase "empirical significance". It is Quine's own point that
"empirical significance" does not amount t� content, in the sense of
what stand one takes on how things are in the empirical world. In the
thesis that translation is indeterminate, which is meant to elaborate
the moral of "Two Dogmas", his aim is to stress "the extent of man's
conceptual sovereignty" in the formation of world-views:2 that is-to
put it in a way that brings Quine into explicit contact with Kant-the
extent to which the content of world-views is a product of spontaneity
operating freely, uncontrolled by the deliverances of receptivity. And
from Quine's point of view, it is a merit of the notion of "empirical
significance" that it stands on the wrong side of the descendant dual
ity to be a descendant of the old notion of meaning. Quine is no friend
to the old notion of meaning, and the descendant notion, the notion
of "language" as the endogenous factor, bound up as it is with "man's
conceptual sovereignty", retains in Quine's thinking some of the in
tellectual dubiousness of its ancestor. In contrast, "empirical signifi
cance" is an intellectually respectable notion, because it is explicable
entirely in terms of the law-governed operations of receptivity, un
tainted by the freedom of spontaneity. To put it in a more Quinean
way, "empirical significance" can be investigated scientifically. "The
extent of man's conceptual sovereignty", the extent to which the con
tent of a world-view goes beyond its "empirical significance", is just
the extent to which such a notion of content lies outside the reach of
science, and therefore outside the reach of first-rate intellectual en
deavour.
It is not just a verbal point that "empirical significance" is on the
wrong side of the duality to be a descendant of the notion of meaning.
We have to discount the rhetoric that makes it look, at first sight, as if
Quine's notion corresponds to the Kantian notion of empirical con
tent. Quine speaks of facing the tribunal of experience, which seems
2. Word and Object (MIT Press, Cambridge, Mass., 1 960), p. 5. For the indeterminacy
thesis, see chap. 2 of that work.
Afterword, Part I 1 33
to imply a vulnerability to rational criticism grounded in experience.

But he conceives experience as "the stimulation of . . . sensory recep
tors".3 And such a conception of experience makes no room for expe
rience to stand in rational relations to beliefs or world-views. The cash
value of the talk of facing the tribunal of experience can only be that
different irritations of sensory nerve endings are disposed to have dif
ferent impacts on the system of statements a subject accepts, not that
different courses of experience have different rational implications
about what system of statements a subject ought to accept. In spite of
the juridical rhetoric, Quine conceives experience so that it could not
figure in the order of justification, as opposed to the order of law-gov
erned happenings. This is all of a piece with the idea that "empirical
significance" is a topic for natural science.
At one point in "Two Dogmas" (p. 43), Quine writes: "Certain
statements . . . seem peculiarly germane to sense experience-and in a
selective way: some statements to some experiences, others to oth
ers . . . . But in this relation of 'germaneness' I envisage nothing more
than a loose association reflecting the relative likelihood, in practice,
of our choosing one statement rather than another for revision in the
event of recalcitrant experience." The only connection he counte
nances between experience and the acceptance of statements is a
brutely causal linkage that subjects are conditioned into when they
learn a language. It is not that it is right to revise one's belief system
thus and so in the light of such-and-such an experience, but just that
that revision is what would probably happen if one's experience took
that course.4 Quine conceives experiences so that they can only be out
side the space of reasons, the order of justification.
It may still seem that Quine is subject at most to criticism of his
rhetoric. But his talk of facing the tribunal of experience is not just a
slip, which we could easily excise; it has roots that go deep in Quine's
3. This is from p. 75 of " Epistemology Naturalized", in Ontological Relativity and

Other Essays (Columbia Universiry Press, New York, 1969).
4. This must go for the concept of recalcitrant experience, which is central to Quine's
celebrated image of "a man-made fa bric which impinges on experience only along the
edges" (p. 42). It is tempting to spell out the idea like this: an experience is recalcitrant if
one cannot rationally take it at face value while continuing to believe everything one be·
lieves; it is rationally incumbent on one either to recast one's world-view (and the Duhemian
point from which Quine starts is that there will be more than one way to do this) or to
discount the experience. But for Quine what it is for an experience to be recalcitrant can
only be that a subject who has the experience will probably change her beliefs.
thinking. We cannot simply register that "empirical significance" is

not genuinely a kind of significance if experience is not in the order of
justification, and leave the substance of Quine's thinking intact.
If experience is not in the order of justification, it cannot be some
thing that world-views transcend or go beyond. But Quine needs that
for his talk of "the extent of man's conceptual sovereignty". What a
world-view might transcend; so that adoption of it might be an exer
cise of spontaneity or "conceptual sovereignty", is evidence that tells
less than conclusively in favour of it. But if experience plays only a
causal role in the formation of a world-view, not a justificatory role,
then it does not serve as evidence at all.
And if experience does not stand to world-views as evidence to the
ory, that puts in question the capacity of Quine's picture to accommo
date world-views at all. It is true that Quine wants the idea of a world
view co stand exposed as intellectually second-rate. But he does not
want to abandon it altogether. That would be to abandon the point he
wants to make by talking of "the extent of man's conceptual sover
eignty". There would be nothing for the thesis of the indeterminacy of
translation to be about if instead of talking about arriving at world
views we had to talk exclusively about acquiring propensities to feel
comfortable with certain vocalizations. Now the idea of an interac
tion between spontaneity-"conceptual sovereignty"-and receptiv
ity, which is Kantian so far as it goes, can so much as seem to make
room for the idea of adopting a world-view only if the deliverances of
receptivity are understood to belong with the adopted world-view in
the order of justification. If we try to suppose that exercises of "con
ceptual sovereignty" are only causally affected by the course of expe
rience, and not rationally answerable to it, there is nothing left of the
idea that what "conceptual sovereignty" produces is something that is
about the empirical world, a stance correctly or incorrectly adopted
according to how things are in the empirical world. And if we lose
that, there is nothing left of the idea that what is operative is "concep
tual sovereignty". The notion of a world-view, formed in an exercise
of "conceptual sovereignty", is not just the notion of a perturbation
produced jointly by impacts from the world and by some force oper
ating from within the subject, in a way that is partly (but only partly)
determined by those impacts.5
5. Understandably, Quine's language often tries to have it both ways. A characteristic
formulation is this (from p. 75 of "Epistemology Naturalized"): "The stimulation of his
If we clean up Quine's formulations by eliminating the juridical

rhetoric, we deprive him of the very idea of "conceptual sovereignty" ,
and the effect i s t o threaten the idea that we are i n touch with the
empirical world at all. It is not that this reading makes Quine suggest
we may be wildly wrong about the world, like an old-fashioned philo
sophical sceptic. But without the "tribunal" rhetoric and its compan
ion idea of "conceptual sovereignty", which have emerged as strictly
illicit by Quine's own lights, he puts in question the very idea that we
have the world in view at all, that anything that we do constitutes
taking a stand, right or wrong, even wildly wrong, on how things are
in it.6
'
4. The awkward position of experience in Quine's thinking has im-
plications for an otherwise attractive reading of what he is doing
when he rejects the first of the two dogmas. In the lectures, I derive
from Sellars, and trace to Kant, a rejection of the idea that something
is Given in experience, from outside the activity of shaping world
views. The attractive reading has Quine making a counterpart point,
rejecting the idea that something is Given from within the very struc
ture of the understanding. 7
Sellars says: "empirical knowledge, like its sophisticated extension,
science, is rational, not because it has a foundation but because it is a
self-correcting enterprise which can put any claim in jeopardy, though
not all at once."8 We must think of empirical rationality in a dynamic
way, in terms of a continuing adjustment to the impact of experience.
To reject the idea of an exogenous Given is to follow this prescrip
tion in part. It is to refuse to conceive experience's demands on a sys-
sensory receptors is all the evidence anybody has had to go on, ultimately, in arriving ;lt his
picture of the world." This sentence begins with a formulation that would lit onI)· 'om<··
thing outside the order of j ustification, but continues ( " has had to go on . . . in arrivin�: at
his picture of the world") in a way that would make sense only of something within the
order of justification. What one goes on in arriving at one's picture of the world is not the
stimulation of one's sensory receptors, experience as Quine officially conceives it, but how
things appear to one, which belongs in a quite different conception of experience.
6. For a much fuller discussion of Quine's thinking on these lines, see chap. 6 of Barry
Stroud, The Significance of Philosophical Scepticism (Clarendon Press, Oxford, 1 984 ), to
which I am much indebted.
7. This is how Rorty reads Quine; see chap. 4 of Philosophy and the Mirror of Nuture.
(I capitalize "Given", as throughout, to contrast the problematic conception with one that
is innocuous; see Lecture I, §4.)
8. "Empiricism and the Philosophy of Mind", p. JQO.
136 .\1 I � D A N D W 0 R L D
rem of beliefs as imposed from outside the activity of adjusting the

system, by something constituted independently of the current state of
the evolving system, or a state into which the system might evolve.
The required adjustmeflts to the system depend on what we take ex
perience to reveal to us, and we can capture that only in terms of the
concepts and conceptions that figure in the evolving system. What we
take experience to tell us is already part of the system, not an external
constraint on it.
That is to say that nothing is Given from outside the evolving
system of belief. The counterpart claim-the claim that nothing is
Given from inside the understanding, the intellectual capacity that
is operative in the continuing activity of shaping the system-is in
Sellars too; it is implicit in the remark about the rationality of sci
ence that I have just quoted. It is true that when Sellars debunks the
Myth of the Given in detail, he focuses on the supposed external
constraint, but he begins "Empiricism and the Philosophy of
Mind " by saying (pp. 253-4) that the idea of something Given in
experience is a specific application of a conception that is much more
general. It can easily seem that rejecting an endogenous Given re
quires us to say what Sellars says in the remark I have quoted: that
any of our beliefs, including beliefs about structures that must be in
stantiated in intellectually respectable belief systems-the beliefs that
implicitly or explicitly govern adjustments in belief systems in re
sponse to experience-are open to revision. And this has a distinctly
Quinean ring.
Bur there is one thing wrong with reading Quine like this, as claim
ing that nothing is Given from inside the evolving system either, and
that is the word "either". We cannot without qualification take it that
Quine fills out the attack on the Given with a counterpart to Sellars's
focus on the idea of something Given from outside. The trouble for
this reading comes from the uneasy status of Quine's juridical rheto
ric. From one perspective Quine leaves the external Given in place: on
his official conception, experience would have to lack the putative ra
tional bearing on belief that is distinctive of the Myth of the Given,
but the "tribunal" rhetoric, which is not mere decoration, implies a
rational link between experience and belief. So Quine's thinking looks
like an awkward combination: he tries to reject an endogenous Given
without definitively rejecting an exogenous Given. In Quine's picture,
"man's conceptual sovereignty" has no internally generated limits to
its freedom of play, but it operates within limits set from outside its
domain.
5. Davidson's recommendation of coherentism, which I have used as

a target of criticism in these lectures, is a response to a genuine and
important insight, which we can identify with the point I have made
about Quine: Quine's play with (in effect) spontaneity and receptivity
impossibly tries to have it both ways-to exploit the idea of experi
ence as a tribunal that stands in judgement over beliefs, while conceiv
ing experience so that it has to stand outside the order of justification.
Davidson is also, admirably by my lights, hostile to any philosophy
that generates problems about how we can be in touch with the em
pirical world at all, as I have claimed Quine's does.
Davidson argues that the duality of endogenous and exogenous fac
tors, as it persists into Quine's "empiricism without the dogmas", is
"itself a dogma of empiricism, the third dogma" .9 He attacks this per
sisting dualism, "the dualism of conceptual scheme [Quine's "lan
guage"] and empirical content [Quine's "empirical significance"]" , 1 0
from both sides.
In "On the Very Idea of a Conceptual Scheme", the attack is from
the side of the endogenous factor. A vivid expression of Quine's idea
of "conceptual sovereignty", a freedom only partly constrained by
the exogenous factor, is the thesis that there might be world-views
that are mutually unintelligible-that different exercises of "con
ceptual sovereignty" might diverge that far. But Davidson argues
that the idea of mutually unintelligible world-views does not make
sense.
In "A Coherence Theory of Truth and Knowledge", the attack is
from the other side. There Davidson argues that experience cannot
constitute "a basis for knowledge outside the scope of our beliefs"
(p. 310). The argument is reminiscent of Sellars. In the terms I have
used, Davidson's point is that experience cannot be both inside the
space of reasons, as it would need to be if it were to constitute "a basis
for knowledge", and outside it, as Davidson claims it would need to
be if it were to be "outside the scope of our beliefs".
9 . " O n the Very Idea of a Conceptual Scheme", p . 1 89. Davidson continues: "The third,
and perhaps the last, for if we give it up it is not clear that there is anything distinctive left
to call empiricism."
10. "On the Very Idea of a Conceptual Scheme", p. 1 89.
Now Davidson's aim in these thematically connected writings is to

exorcize a style of thinking whose effect-even when this is not the
intention-is to make a mystery out of thought's bearing on the em
pirical world. He ends "On the Very Idea of a Conceptual Scheme"
like this (p. 1 98): "In giving up the dualism of scheme and world [that
is: world conceived as the source of demands imposed on our thinking
in experience, from outside thoughtJ, we do not give up the world, but
re-establish unmediated touch with the familiar objects whose antics
make our sentences and opinions true or false. " Focusing on this re
mark, we can see that the image I recommend in the lectures, in which
thought is unbounded, should be fully congenial to Davidson. The
rejected image, with a boundary that encloses thought and sets it off
from the world, would give pictorial expression to the idea that there
are philosophical problems about the relation between thought as
such and its objects. Rejecting the image is refusing to let our "unme
diated touch" with the familiar world be threatened by a set of philo
sophical assumptions that give only the illusory appearance of being
compulsory. Viewed in this light, Davidson's response to Quine's con
tinuing dualism of scheme and world looks like something that ought
to be a model for me. 1 1
And indeed it is, up to a point. I gladly acknowledge that Davidson
has the essential point, that philosophy must not be allowed to make
a mystery out of thought's bearing on its objects. And he is right that
Quine cannot have it both ways: experience as Quine conceives it
cannot be a tribunal. But the aspiration of eliminating mystery is one
thing; realizing it is another. In my view Davidson resolves the ten
sion he finds in Quine in the wrong direction, and the result is pre
cisely to leave us with the philosophical problems he wants to eli
minate.
Davidson sees that Quine has no coherent way to make something
out of, in effect, the idea that systems of empirical belief result from
the co-operation of spontaneity and receptivity. He goes wrong by my
lights in concluding that there is nothing to be made of the idea: that
it can only reflect an unworkable dualism. As I said (§3 above), the
idea of an interaction between spontaneity and receptivity can so
much as seem to make it intelligible that what results is a belief, or a
1 1 . Christopher Hookway and, in a different way, Aryeh Frankfurter urged this on me.
system o f beliefs, about the empirical world-something correctly or

incorrectly adopted according to how things are in the empirical
world-only if spontaneity's constructions are rationally vulnerable
to the deliverances of receptivity. Quine's official view of experience
disrupts Quine's attempt at a version of that picture. And Davidson
shares that view of experience: for Davidson, receptivity can impinge
on the space of reasons only from outside, which is to say that nothing
can be rationally vulnerable to its deliverances.12 Davidson differs
from Quine only in that he is explicit about this, and clear-sightedly
draws the consequence: we cannot make sense of thought's bearing on
the world in terms of an interaction between spontaneity and recep
tivity. If we go on using the Kantian terms, we have to say that
the operations of spontaneity are rationally unconstrained from out
side themselves. That is indeed a way of formulating Davidson's
coherentism.
Of course Davidson's thinking does not obliterate all forms of the
idea that empirical thinking is rationally vulnerable to the course of
experience. "The course of experience" might be interpreted to mean
the succession of circumstances that consist in its appearing to one
that things are thus and so, and appearings do impinge rationally on
the activity of shaping one's world-view. That is something we can say
without confusion, by Davidson's lights, because its appearing to one
that things are thus and so is, as such, already within the space of
concepts, and hence qualified to stand in rational relations to other
inhabitants of that space. But what we cannot say without confusion,
according to Davidson, is that sensory impressions, impacts of the
world on our senses, impose rational demands on our empirical think
ing. Or if we can say that, it is only by dint of packing some complex
ity into " impose". Perhaps a sensory impression causes it to appear to
a subject that things are thus and so, and the appearing has implica
tions for what the subject ought to think. But sensory impressions
12. The poinr does not turn on the detail of Quine's conception of experience, as stimu
lation of sensory surfaces . There can be less resolut.ly ami-mentalistic conceptions of expc·
rience that nevenh.less match Quine's conception at a more a bstract level, in that they take
.experiences to be deliverances of receptiviry. Davidson's general thought is that if experi
ence is understood as what receptiviry provides us with, then, whatever the details of the
conception, experience is eo ipso understood in a way that removes it from the space of
reasons.
themselves, as Davidson conceives them, cannot stand in rational re

lations to what a subject is to think. 1 3
I have not put this thought in Davidson's own terms. Davidson
does not talk about appearings, and he writes as if only beliefs can fill
the role that I have suggested appearings can fill, for instance when he
expresses his coherentism by the remark I quote in the lectures: "no
thing can count as a reason for holding a belief except another belief"
(p. 3 1 0). 14 If Davidson wants to acknowtedge a grounding role for
appearings, this focus on beliefs seems unhappy, at least as a matter of
terminology. Its appearing to me that things are thus and so is not
obviously to be equated with my believing something. Certainly not
with my believing that things are thus and so. No doubt when it ap
pears to me that things are thus and so, I usually (at least) believe that
it appears to me that things are thus and so, but it is not obvious that
the appearing is the belief; and whether it is or not, we can innocu
ously credit the appearing itself with rational implications for what I
ought to think. But for my immediate purposes, there is no substantial
issue here. Davidson's terminology fits what looks like an overly sim
ple formulation that he chooses for his coherentist position. He could
have made the same substantial point if he had said: nothing can
count as a reason for holding a belief except something else that is also
in the space of concepts-for instance, a circumstance consisting in its
appearing to a subject that things are thus and so. (There is a more
substantial problem with Davidson's formulation, which I shall come
to in due course.)
The echo of Sellars that I mentioned earlier sounds here.15 Sellars
devotes part of "Empiricism and the Philosophy of Mind" to defend
ing a notion of sensory impressions.1 6 The point of the defence is to
1 3. At least in "A Coherence Theory of Truth and Knowledge", Davidson seems to

reserve the word "experience" for sensory impressions as distinct from appearings. See, for
instance, p. 3 1 3, where he formulates a view he opposes like this: "whatever there is to
meaning must be traced back somehow to experience, the given, or patterns of sensory
stimulation, something intermediate between belief and the usual objects our beliefs are
about." ( " Belief" here serves as a code for "things in the space of concepts"; see the next
paragraph in my text. ) But how we should use the word "experience" is not the point here.
14. See " A Coherence Theory of Truth and Knowledge", p. 3 1 1 : "The trouble we have
been running into is that the justification seems to depend on the awareness, which is just
another belief."
1 5 . In fact I have formulated what is innocuous by Davidson's lights in terms that are
more Sellarsian than Davidsonian.
1 6. The defence occupies most of the concluding part of the essay, from §45 .
distinguish impressions from bits o f the Given, and Sellars effects this
by carefully refusing to attribute any direct epistemological
significance to impressions. They have an indirect epistemological
significance, in that without them there could not be such directly
significant circumstances as seeing that things are thus and so, or hav
ing it look to one as if things are thus and so. But it is only in that
indirect way that impressions enter into the rational responsiveness of
empirical thinking to the course of experience. We can have an inno
cent interpretation of the idea that empirical thinking is rationally re
sponsive to the course of experience, but only by understanding "the
course of experience" to mean the succession of appearings, not the
succession of impressions.
Impressions are, by definition we might say, receptivity in opera
tion. So the picture that is common to Sellars and Davidson is this.
Receptivity figures in the explanatory background of circumstances
that belong together with evolving world-views in the order of
justification. But receptivity itself cannot rationally interact with
spontaneity, in the way that Quine's rhetoric implies, though his offi
cial conception of receptivity precludes such interaction.
Against this, I claim that although Quine's half-hearted attempt to
picture world-views as products of a rational interaction between
spontaneity and receptivity is unacceptable, as Davidson sees, that is
no reason to discard the very idea of such an interaction. The trouble
lies not in the idea itself, but in the half-heartedness-in the fact that
while the rhetoric depicts the interaction as rational, Quine conceives
receptivity in such a way that it cannot impinge rationally on any
thing. We can have a whole-hearted version of the idea if we can see
our way to saying that the impressions of the world on our senses, the
deliverances of our receptivity, are-as such-the appearings (or at
least some of them) that, as Davidson and Sellars agree, can inno
cently be taken to belong together with our world-views in the space
of reasons, since they are already in the space of concepts. That way,
we can hold on to the attractive thought that Quine only half
heartedly embraces. There really is a prospect of finding empirical
content, as possessed by exercises of spontaneity, unmysterious if we
can think of it on the lines that Davidson and Sellars disallow, and
that Quine is officially committed t'o disallowing. We ought to have
no problem about how an exercise of "conceptual sovereignty" can
bear on the empirical world-can constitute taking a stand on how
things are, a posture correctly or incorrectly adopted according to the

way the world is arranged-if "conceptual sovereignty" is rationally
answerable to how the world impresses itself on the subject in experi
ence.
It is not enough to say, with permission from Sellars and Davidson,
that the exercise of "conceptual sovereignty" in shaping world-views
is rationally responsible to the su(:cession of appearings to the subject.
Not that that is wrong. But if we follow Sellars and Davidson in dis
tancing the appearings from the impressions, saying what they will let
us say does not entitle us to find no philosophical mystery in thought's
bearing on the world.17
According to Davidson and Sellars, we can place appearings within
the space of concepts, so that they can coherently be taken to stand in
rational relations to beliefs, only by distinguishing them from the
world's impacts on our senses. Appearings are just more of the same
kind of things beliefs are: possessors of empirical content, bearing on
the empirical world. And now we cannot make the question "How
can beliefs (say) have empirical content?" look any less pressing by
talking about a rational interplay between appearings and beliefs. The
question is really "How can anything have empirical content?", and it
is no good just helping ourselves to the fact that appearings do.
Contrast how things look if we manage to conceive the rational
answerability of beliefs to appearings as a rational answerability to
receptivity itself. Now we cannot be accused of merely shifting the
mystery of empirical content from beliefs to appearings. Now there is
no mystery. If the rational answerability is to receptivity itself, not just
to something linked with receptivity only in that receptivity figures in
its explanatory background, then in being subject to the tribunal of
experience, exercises of "conceptual sovereignty" are rationally an
swerable to the world itself. (Recall the image of experience as open
ness to the world.) There cannot be a problem about the idea of a
stand as to how things are, correctly or incorrectly adopted according
17. "Entitle" matters here. Rorty (whom I shall come to shortly) is very good at debunk
ing the moves that are available if we suppose there are such problems for philosophy.
Perhaps the uselessness of the moves indicates, as it were from the outside, that this concep
tion of philosophy's task must be mistaken. But this external approach can easily leave the
philosophical questions still looking as if they ought to be good ones, and then the result is
continuing philosophical discomfort, not an exorcism of philosophy. Exorcism requires a
different kind of move, which Rorty is much less good at.
to the layout of the world, if shaping a world-view is rationally an

swerable, by way of experiential openness, to the world itself.
This conception of experience contrasts with one that pictures expe
riences as emissaries from the world. Davidson remarks ( " A Coherence
Theory", p. 312) that if we picture experiences as emissaries, puta
tively informing us about the world, we have the problem that they
"may be lying . . . we can't swear intermediaries to truthfulness" . But
the real trouble with conceiving experiences as intermediaries is that
we cannot make sense of experiences, so conceived, as purporting to
tell us anything, whether truthfully or not. When we take receptivity
itself to impinge rationally on belief, we equip ourselves to understand
experience as openness to the world. And now the problem of making
it intelligible that experience is endowed with content lapses, and the
question of truthfulness takes on a different look. We achieve an intel
lectual right to shrug our shoulders at sceptical questions, if they are
asked with the usual philosophical animus, namely, to point up a sup
posed problem about whether our thought is in touch with its pur
ported topics. Of course we are fallible in experience, and when expe
rience misleads us there is a sense in which it intervenes between us and
the world; but it is a crucial mistake to let that seem to deprive us of the
very idea of openness-fallible openness-to the world, as if we had to
replace that idea with the idea of emissaries that either tell the truth or
lie. It is only because we can understand the notion of appearings con
stituted by the world's making itself manifest to us that we can make
sense of the empirical content, the bearing on the world, embodied in
the idea of a misleading appearance. When we are not misled by expe
rience, we are directly confronted b.y a worldly state of affairs itself, not
waited on by an intermediary that happens to tell the truth.18
From this angle we can see that there is more than an excess of
simplicity in Davidson's formulation "nothing can count as a reason
for holding a belief except another belief". I suggested this emenda
tion: nothing can count as a reason for holding a belief except some
thing else that is also in the space of concepts. In fact the emended
wording is fine by my lights. It need not give expression to what, in the
lectures, I call "unconstrained coherentism", the thesis that there are
no external rational constraints on exercises of spontaneity. The
emended wording allows that exercises of spontaneity can be ration-
18. See Lecture VI, SJ.

1 ·1-1 \ 1 I :-.; D .\ :-.; D \l" 0 R l D
ally constrained by facts, when the facts make themselves manifest in

experience; that is a constraint from outside exercises of spontane·
ity-from outside the activity of thinking, as I put it in the lectures,
though not from out�de what is thinkable, so not from outside the
space of concepts (Lecture II, §3). But this idea of external constraint
is genuinely available only if we can contrive to accept a rational en
gagement with spontaneity on the part of receptivity itself, and David
son thinks that is impossible. So when Da"¢'idson says that only beliefs
can stand in rational relations to beliefs, that is not just the termino
logical awkwardness that I discounted earlier-the easily discarded
implication that an appearing is a belief. Davidson's formulation
reflects something deeper: he cannot countenance external rational re
quirements on exercises of spontaneity, so his coherentism is genu
inely unconstrained. Even if we do not take the word "belief" very
seriously, the restriction to beliefs conveys something on these lines:
only subjective things belong together with evolving world-views in
the space of reasons. My claim is that this is disastrous: it ensures that
we cannot refuse to find a mystery in the bearing of belief, or anything
else, for instance appearing, on the empirical world.
The conception of impressions that is common to Sellars and Da
vidson does not completely remove impressions from the domain of
epistemology, even apart from their indirect relation to what one
should believe. The way impressions causally mediate between the
world and beliefs is itself a potential topic for beliefs, and these beliefs
can stand in grounding relations to other beliefs. Consider a belief
that credits an observable property to an object. In the context of a
rationally held theory about how impressions figure in causal interac
tions between subjects and the world, such a belief might be rationally
grounded in a belief about an impression. One might be justified in
believing that the object has the property by the fact that one has an
impression of a type that is, according to one's well-grounded theory,
caused in suitable circumstances (for instance the prevailing illumina·
tion) by an object's possessing that property.19
19. Compare Sellars's discussion o f the authority of observation, "Empiricism and the
Philosophy of Mind" pp. 296-8. This passage comes before Sellars has undertaken to reha
bilitate the notion of impressions; it makes the authority of an observational judgement that
something is green turn on the subject's knowledge that her own report "This is green" is
reliably correlated, in the right conditions, with something's being green. But once the im·
pression of green is in view, it can figure in a parallel grounding, in a position corresponding
to that of the report in the kind of grounding Sellars envisages.
But that is quite distinct from saying, as my picture allows and the
picture common to Sellars and Davidson does not, that the belief that
an object has an observable property can be grounded in an impres
sion itself: the fact's impressing itself on the subject. In my picture
impressions are, so to speak, transparent. In the picture common to
Sellars and Davidson they are opaque: if one knows enough about
one's causal connections with the world, one can argue from them to
conclusions about the world, but they do not themselves disclose the
world to one. They have an epistemological significance like that of
bodily feelings in diagnosing organic ailments. And my claim is that
that undermines Davidson's aim of eliminating mystery. If we cannot ·
conceive impressions as transparent, we distance the world too far
from our perceptual lives to be able to keep mystery out of the idea
that our conceptual lives, including appearings, involve empirical con
tent.
In the style of thinking I have been attacking, impressions need not
be separated from appearings as causes and effects. Another version
of the picture might allow one form of the claim that appearings (at
least some of them) are impressions. What would make this still a
version of the picture that I am attacking would be an insistence that
something's being an appearing must be at a conceptual remove from
something's (perhaps, in this version of the picture, the same thing's)
being an impression. The identification of one and the same thing as
both an impression and an appearing would straddle the boundary
between two radically different modes of conceptualization; we
would have to insist that it is not by virtue of being the impression it
is that an item is the appearing it is. This version of the picture might
be more congenial to Davidson than the version I have been working
with, according to which impressions belong in the explanatory back
ground of appearings; that seems to be Sellars's line.
As my counting it a version of the same picture implies, I do not
believe identifying appearings with impressions in this way makes any
difference to the main point. It will still be true, in the context of the
identification understood like this, that impressions as such are
opaque. If an item that is an impression is credited with empirical con
tent, because it is said to be also an appearing, it is not supposed to be
by virtue of being the impression it · is that it possesses that content.
This is just another way of refusing to countenance a rational engage
ment between spontaneity and receptivity as such, and I think it still
leaves us short of an entitlement to find empirical content unmyster

ious.20 What that requires is that we see our way to accepting that an
impression as such can be an appearing, a circumstance constituted by
the world's making an appearance to the subject.
In the picture I recommend, although the world is not external to
the space of concepts, it is external to exercises of spontaneity. Al
though we are to erase the boundary that symbolized a gulf between
thought and the world, the picture still has an in-out dimension. Link
ages between what is further in and what is further out stand for the
availability of rational groundings, and the world-which is as far out
as possible-is ultimate in the order of justification. What I have been
urging, against Davidson and Sellars, is that we must find a place for
impressions, the deliverances of receptivity, along this in-out dimen
sion. They must figure in the order of justification. Of course there are
other dimensions along which we can trace connections between men
tal items and the world, and we can interpret phrases such as "the
impact of the world on the senses" so as to apply to items that stand
between minds and the world only along one of these other dimen
sions. But we must not suppose, with Sellars and Davidson, that that
is the only sort of sense we can give to the idea of deliverances of
receptivity.
6. A clearer target for my criticism than Davidson himself would be

Rorty's reading of Davidson, in "Pragmatism, Davidson, and Truth" .
There Rorty singles out for commendation precisely the aspects of
Davidson's thinking that I have objected to. Playing them up yields a
reading that makes no room for other aspects of Davidson's thinking.
And the effect is to provide an object lesson in how not to rid our
selves of the illusory intellectual obligations of traditional philosophy.
In presupposing that this is the aim, I assume Rorty is right to dis
count "the fact that Davidson . . . does not present himself as repudi
ating the skeptic's question, but as answering it" (p. 342).21 I also as
sume that philosophical concerns about the possibility of knowledge
express at root the same anxiety as philosophical concerns about how
20. The point here is of a piece with something I suggest in Lecture IV, §4.
2 1 . In M Afterthoughts, 1 987" (in Alan Malachowski, ed., Reading Rorty [Blackwell,
Oxford, 1 990]), Davidson attributes to Rorty the claim that Ml should not pretend that I am
answering the skeptic when I am really telling him to get lost", and says Ml pretty much
concur with him " .
content i s possible, an anxiety about a felt distance between mind and

world. Davidson and Rorty usually focus on concerns of the former
sort, whereas I focus on concerns of the latter sort; I take it that the
underlying thought is the same, that we ought to exorcize the feeling
of distance rather than trying to bridge the felt gap.
Rorty is very strong in his conviction that the supposed obligations
of traditional philosophy are illusory, and I have every sympathy with
that. My objection to Davidson is not that he does not answer the
question how anything can have empirical content, but rather that he
does not ensure, as he wants to, that the question lacks urgency. But
Rorty gives expression to the conviction in a way that spoils the point.
What he praises Davidson for recommending is a way of thinking
within which, on inspection, it turns out to be unintelligible how em
pirical content can be anything but a mystery. It is true that Rorty
resists the blandishments of traditional philosophy, but the effect of
the framework he assumes is that he can do that only by plugging his
ears, like Odysseus sailing past the Sirens.
Rorty puts the point of Davidson's coherentism, which he endorses,
like this: we must hold apart the view of beliefs "seen from the outside
as the field linguist sees them (as causal interactions with the environ
ment)" and the view of beliefs seen "from the inside as the pre-episte
mological native sees them (as rules for action)" (p. 345 ). We must
"abjure the possibility of a third way of seeing them-one which
somehow combines the outside view and the inside view, the descrip
tive and the normative attitudes" (p. 345).22 The outside view, the
field linguist's view, is descriptive; it links beliefs with objects and cir
cumstances in the believer's environment, in a structure whose consti
tutive relations are causal. (In the outside view, beliefs are " seen . . . as
causal interactions with the environment".) The inside view is norma
tive: it is "the . . . point of view of the earnest seeker after truth"
(p. 347), a point of view in which beliefs are linked with what is taken
to give them their rational credentials, that is, located in the space of
reasons.
Rorty thinks different uses of the notion of truth can be shared out
between these points of view. He commends Davidson for a "contri-
22. The abjured third view is what Rorty earlier describes a � " a confused attempt to ne
inside and outside the language-game at the same time" (p. 342). Compare also the criticism
of Hilary Putnam for wanting "a synoptic vision which . . . will somehow bring the outside
and the inside points of view together" (p. 34 7).
bution to pragmatism" that "consists in pointing out that ["true"] has

a disquotational use in addition to the normative uses seized on by
james" (p. 342). The contrast with "normative", which, as we have
seen, Rorty associates with the inside point of view, licenses us to read
this as assigning the disquotational use of "true" to the descriptive
view, the outside view.
That anyway fits the fact that Rorry identifies the outside view with
the field linguist's view. Davidson's field linguist aims to construct the
ories of truth for this or that language in the style of Tarski, and such
theories use "true" disquotationally, in an extended sense. We have
disquotation strictly so called in the theorems of a Tarskian theory of
truth formulated in a minimal extension of the language for which it
is a theory: what Davidson calls "neutral snow-bound trivialities",23
in a reminder of the notorious example '"Snow is white' is true (in
English) if and only if snow is white". Here disquoting is what one
needs to do in order to get from the left-hand side of a theorem to
what should figure on the right-hand side. But that idea easily gener
alizes to the idea of cancelling semantic ascent; that need not be ef
fected by a move in the language from which the ascent was made, as
with disquotation strictly so called.24 Davidson's field linguist aims at
theories of truth whose theorems are disquotational in the generalized
sense: they are inter-language counterparts to what figure, in the intra
language case, as the "neutral snow-bound trivialities". And Rorry's
idea is that such theories are constructed and understood from the
outside point of view, the point of view whose concern is descriptive
as opposed to the normative concern of the "earnest seeker after
truth".
Now this holding apart of the two points of view is quite unsatisfy
ing, both in itself and as a reading of Davidson.
We can appreciate how unsatisfying it is in itself by considering a
move Rorty makes against Hilary Putnam. Objecting to a position
like the one Rorty finds and applauds in Davidson, Putnam once
wrote:
If the cause-effect-description [of our linguistic behavior qua production
of noises] is complete from a philosophical as well as from a behavioral-
23. From p. 51 of "True to the Facts", in Inquiries into Truth atrd Interpretation,
pp. 37-54; quoted by Rorry at p. 343.
24. For the generalization, see pp. 10-13 of Quine's Philosophy of Logic (Prentice-Hall,
Englewood Cliffs, N.J., 1 970).
scientific point of view; if all there is to say about language is that it

consists in the production of noises (and subvocalizations) according to
a certain causal pattern; if the causal story is not to be and need not be
supplemented by a normative story . then there is no way in which the
. .
noises we utter . . . are more than mere "expressions of our subjectiv

ity".25
Rorty responds as follows, in a passage I have already quoted from:

The line I have italicized suggests that disquotationalist theorists of truth
think that there is only one story to be told about people: a behavioristic
one. But why on earth should such theorists not allow for, and indeed
insist upon, supplementing such stories with "a normative story " ? Why
should we take the existence of the outside point of view of the field
linguist as a recommendation never to assume the inside point of view of
the earnest seeker after truth? Putnam, I think, still takes a "philosophi
cal account of X" to be a synoptic vision which will somehow synthesize
every other possible view, will somehow bring the outside and the inside
points of view together.
Until the last sentence, this strikes me as refusing to listen to

Pumam's concern. Admittedly the concern is not well expressed in the
passage Rorty exploits. But it takes only minimal charity to see what
Putnam is driving at. Putnam does not suppose that wanting to tell a
story about causal relations between human vocalizations and the en
vironment is, in itself, enough to debar someone from so much as hav
ing a story to tell that represents the vocalizers as expressing thoughts
and making assertions, trying to get things right.26 What he objects to,
as the last sentence I have quoted from Rorty in effect acknowledges,
is precisely the thesis Rorty finds and applauds in Davidson, that the
two stories must be held apart. But the acknowledgement comes too
late. By saddling Putnam with a different target for his complaint,
Rorty contrives to make it look legitimate for him to cap a series of
supposedly well-placed rebukes to Putnam, to the effect that the one
story does not exclude the other, a point that is in fact irrelevant, with
an unargued reaffirmation of the thesis Putnam really means to ques
tion, that the two stories have to be kept separate.
Putnam's worry cannot be squelched by merely insisting that both
25. "On Truth", in Leigh S. Cauman et al., eds. How Many Questions? (Hackett, india·
napolis, 1 983), p. 44; quoted by Rorry at p. 347. (The interpolation and emphasis are
Rorty's.)
26. See the passage Rorty quotes from Putnam's Realism and Reason, at pp. 345-6.
stories are there to be told; the worry is precisely about the thesis that
they cannot be told together. That implies that if we occupy a stand
point from which our beliefs are in view along with their objects and
our causal engagements with the objects, then we cannot, from that
standpoint, bring the beliefs under the norms of inquiry. And
Putnam's worry about this is well placed: the result is to make it a
mystery how what we are talking about can be beliefs, stances with
respect to how things are in the world, at all. It does not help to insist,
as Rorty does, that there is another standpoint from which beliefs are
seen as subject to the norms of inquiry. If the view from this second
standpoint is not allowed to embrace the causal interactions between
believers and the objects of their beliefs-since those interactions are
the preserve of the outside view, which has to be held separate-then
it simply becomes mysterious how we can be entitled to conceive what
organizes the subject matter of the second standpoint as the norms of
mqmry.
The point comes out vividly in the way Rorty deals with disquota
tion. There is an obvious connection between disquotation, whether
in the strict or in the extended sense, and a straightforward notion of
getting things right. It is because "La neige est blanche" is true in
French if and only if snow is white that, since snow is indeed white, I
shall be getting things right if I express a belief in French by saying " La
neige est blanche". Rorty's remarks about disquotation give the field
linguist, the occupant of the outside point of view, responsibility for
the question whether beliefs achieve truth in the sense of disquotabil
ity. The question whether a belief achieves disquotability is supposed
to be descriptive as opposed to normative, and Rorty's picture keeps
it apart from any question we address in our capacity as "earnest
seekers after truth"-in our efforts to be responsive to what we
should like to think of as the norms of inquiry. But this severs what
we want to think of as responsiveness to the norms of inquiry from
any connection with that unproblematic notion of getting things right.
And the effect is to make it unintelligible how it can be norms of in
quiry that are in question. Norms of inquiry are normative for the
process of inquiry precisely because disquotability is the norm for its
results.
Amazingly enough, Rorty seems to think it is merely routine to sep
arate what we want to think of as norms of inquiry from the straight
forward notion of getting things right that is connected with the no-
tion o f disquotability. A t p . 336, h e says, without ceremony, that it

"seems paradoxical" to suggest that "There are rocks" is implied by
"At the ideal end of inquiry, we shall be justified in asserting that there
are rocks", because " there seems no obvious reason why the progress
of the language-game we are playing should have anything in particu
lar to do with the way the rest of the world is". But that is an extraor
dinary thing to say. It is the whole point of the idea of norms of in
quiry that following them ought to improve our chances of being right
about "the way the rest of the world is". If following what pass for
norms of inquiry turns out not to improve our chances of being right
about the world, that j ust shows we need to modify our conception of
the norms of inquiry. Rorty implies that to say that sort of thing is to
succumb to the attractions of traditional philosophy. But the world as
I invoke it here, in this expression of the thought that Rorty's separa
tion of viewpoints is intolerable, is not the world that is well lost, as
Rorty sees, when we follow Davidson in rejecting "the dualism of
scheme and world". 27 It is the perfectly ordinary world in which there
are rocks, snow is white, and so forth: the worldtthat is populated hy
"the familiar objects whose antics make our sentences and opinions
true or false", as Davidson puts it. It is that ordinary world on which
our thinking bears in a way that Rorty's separation of viewpoints
leaves looking mysterious, precisely because it separates relatedness to
the world from the normative surroundings that are needed to make
sense of the idea of bearing-rational bearing-on anything. It is loss
of that ordinary world that Rorty threatens us with, when he insulates
norms of inquiry from disquotability.
Once one has adopted a style of thinking that has that effect, it is
too late to refuse to listen to expressions of philosophical discomfort.
We should indeed want philosophical problems about how thought
can be in touch with the world to stand revealed as illusory, but Rorty
deprives himself of the right to take that attitude. His own thinking
makes those problems urgent, so that his refusal to address them can
only be an act of will, a deliberate plugging of the ears. In one sense
the refusal is well justified, since Rorty has such an excellent appreci
ation of how hopeless our predicament is if we try to address the ques-
27. I echo the engaging title of Rorty's "The World Well Lost" , in his Consequences of
Pragmatism (University of Minnesota Press, Minneapolis, 1982), pp. 3-1 8. Tha t paper
stands to Davidson's "On the Very Idea of a Conceptual Scheme" much as " Pragmatism,
Davidson, and Truth" stands to "A Coherence Theory of Truth and Knowledge".
tions. But in another sense the refusal is arbitrary, since Rorty's own
thinking, so far from being shaped so that the questions cannot arise,
positively exacerbates theic apparent pressingness.
Rorty's assignment of truth as disquotability to the outside point of
view is also quite unsatisfying as a reading of Davidson.
It is true that the data available to a Davidsonian field linguist,
when she begins on radical interpretation of a l�nguage, are restricted
to vocal or otherwise putatively linguistic behaviour, with its causal
connections to the environment. As long as the language, if that is
what it is, has not been interpreted, the linguist has no handle on
what, if anything, its speakers count as a reason for what, though she
can observe which environmental circumstances are likely to prompt
them to which vocalizations or other putatively linguistic actions.
While the interpreter is in this position, it cannot yet stand quite firm
for her that what is in question is linguistic behaviour at all; that de
pends on the behaviour's turning out to be interpretable-that is, ca
pable of being intelligibly placed in the space of reasons.
But that is how it is only at the outset of radical interpretation. The
field linguist's aim is not merely to codify those causally connected
data, or to construct a theory that postulates further connections of
the same kind, so as to make the data intelligible in the way a theory
in the natural sciences makes intelligible the data it is based on. Ex
actly not: Davidson's field linguist aims to work into an appreciation,
as from within, of the norms that constitute the language she investi
gates: the specific sense of when it is right to say what according to
which that language-game is played. That is what she aims to capture
in a theory for the language that is disquotational in the extended
sense. She begins as an occupant of the outside standpoint, but if she
succeeds in her aim, she ends up equipped to give expression in her
own terms to part of how things look from the inside standpoint her
su bjects occupied all along. When Rorty suggests that the results of
the field linguist's endeavours employ a notion of truth unconnected
with norms, and hence separated-by the supposed gulf between the
two standpoints-from, for instance, a conception in which the truth
is seen as what ought to be believed ("the normative uses seized upon
by James" ), he obliterates the significance of the transition from start
ing predicament to achieved interpretation.
The outside standpoint as Rorty conceives it is a standpoint from
sideways on. (For that image, see Lecture II, §4.) Davidson's radical
interpreter starts with a sideways-on view of the relation betw�en her

subjects and the world. But she finishes with a theory whose point is
exactly that it is not from sideways on: a theory that enables her to
capture some of her subjects' relations to the world from their own
point of view, though in her terms rather than theirs. It is just the
beaury of the notion of disquotation in the extended sense that it is
available for this capturing of the inside viewpoint. When Rorry sep
arates disquotation from the standpoint of the speakers of the lan
guage, as "earnest seekers after truth", he precisely misses what
makes a disquotational notion of truth suitable for summarizing the
results of interpretation.28
7. When Rorty insists that the two points of view must be kept sepa
rate, that is an expression of a dualism of nature and reason. In this ver
sion of the dualism, nature figures as the subject matter of the outside
view, and the space of reasons figures as the normative organization that
things have when they are seen from the inside view; what is dualistic is
insisting that the two modes of organization cannot be combined.
In the lectures, I point to the dualism of nature and reason as the
source of the merely apparent difficulties confronted by traditional
philosophy. I find the dualism operative in Davidson's thinking: it ac
counts for his attitude to the idea that spontaneity interacts rationally
with receptivity. So I am not in a position to dissent outright from
Rorty's reading of Davidson. But in my reading, Davidson's vulnera
bility to the dualism is a defect; it is out of line with his better thinking
on interpretation, and it ensures failure in the aim of exorcizing tradi
tional philosophical anxieties. In contrast, Rorty centres his reading of
28. Rorty is not alone in supposing that the sideways-on character of the radical
interpreter's starting orientation-what makes the interpretation radical-persists into the
results of radical interpretation. See Charles Taylor, "Theories of Meaning", in his Human
Agency and Language: Philosophical Papers, 1 (Cambridge University Press, Cambridge,
1 985), pp. 248-92, especially pp. 273-82. Taylor takes it that Davidson's thinking ex
cludes Gadamer's idea of a fusion of horiwns. (See Lecture II, S4.) Like Rorty, Taylor
thinks Davidson's approach to interpretation is inextricably committed to an outside point
of view. And Cora Diamond seems to suggest something similar, at pp. 1 1 2-3 of "What
Nonsense Might Be", in her The Realistic Spirit: Wittgenstein, Philosophy, and the Mind
(MIT Press, Cambridge, Mass., 1 99 1 ), pp. 95- 1 1 4. I think such readings miss the distance
between Davidson and Quine. (Davidson may be partly to blame for this, because of the
way he systematically understates that distance. I say something about this at p. 73 of my
"In Defence of Modesty", in Barry Taylor, ed., Michael Dummett: Contributions to Philo
sophy [Martinus Nijhoff, Dordrecht, 1 987], pp. 59-80.)
Davidson on the dualism, and he applauds it as what it exactly is not,

a way to escape from the obsessions of tra ditional philosophy.29
It is ironic that I can put things like this. Rorty begins "Pragmatism,
Davidson, and Truth" with an admiring description of pragmatism,
into which he wants to enrol Davidson, as "a movement which has
specialized in debunking dualisms and in dissolving traditional prob
lems created by those dualisms" _{p. 333). But Rorty's own thinking is
organized around the dualism of reason and nature, and that means
he can be at best partly successful in being a pragmatist in his own
sense. No wonder his attempt to dissolve traditional problems has the
aspect of refusing to listen to questions that still stubbornly look as if
they ought to be good ones, rather than supplying a way of thinking
within which the questions genuinely do not arise.
Of course Rorty does not cast his view about nature and reason as
a dualism. He speaks, for instance, of "patiently explaining that
norms are one thing and descriptions another" (p. 347). That sounds
like calmly drawing a distinction; it is not the obsessive mode of utter
ance characteristic of a philosopher insisting on a dualism. But I have
been urging that if we try to think as Rorty says we must, we are stuck
with the philosophical anxieties that he wants to avoid. Cultivating a
non-obsessive tone of voice is not enough to ensure that philosophical
obsessions are out of place.
I cited Rorty's suggestion that Putnam wants "a synoptic vision
which will somehow synthesize every other possible view, will some
how bring the outside and the inside points of view together". Rorty
means to accuse Putnam of the grandiose aspirations of traditional
philosophy, which he thinks we should discard: thought is to be
brought into alignment with its objects, minds with reality. My sug
gestion has been that bringing the outside and the inside points of
view together (not " somehow", which suggests a mystery) is exactly
the sort of dualism-debunking and problem-dissolving move for
which Rorty himself admires pragmatism. So what I recommend in
29. I do not mean to suggest that we can easily isolate the dualism's role in motivating
Davidson's coherentism. It is also at work elsewhere in his thinking: notably in the thesis
that causal relations can hold between occupants of the space of reasons only because they
can be identified with elements in the realm of law. (Compare Rorty's analogue to this
thesis: it is only in the inside view that items are placed in the space of reasons, and causal
relations are not present in that view at all.) This Davidsonian thesis is at issue in Lecture
IV, S4.
the lectures could be represented as a pragmatislh in Rorty's sense,

even though, in trying to give expression to it, I borrow from thinkers
such as Kant, whom Rorty finds utterly suspect.30 And I am claiming
that Rorty's own pragmatism is half-baked, according to standards
set by his own account of what pragmatism is.
8. Davidson's objection to the third dogma of empiricism is this:

even as it tries to make out that sensory impressions are our avenue of
access to the empirical world, empiricism conceives impressions in
such a way that they could only close us off from the world, disrupt
ing our " unmediated touch" with ordinary objects. Now Rorty gener
alizes that thought into rejecting a whole array of candidate interme
diaries between us and the world, on the ground that to accept them
is just to saddle ourselves with pointless anxieties about our hold on
the world. He speaks of "such tertia as, in Davidson's words, 'a con
ceptual scheme, a way of viewing things, a perspective' (or a transcen
dental constitution of consciousness, or a language, or a cultural tra
dition) " (p. 344 ).
In the lectures, I explain how the notion of impressions can be in
nocuous. We can take it that spontaneity is rationally vulnerable to
receptivity without the unwelcome effect that receptivity seems to get
in the way between us and the world, if we reject the framework that
is the real source of the problems of traditional empiricism, namely,
the dualism of reason and nature. In the context of a full-blown prag
matism, impressions can come into their own as precisely a mode of
openness to the world. And something similar goes for at least some
of Rorty's other "tertia ". Conceptual schemes or perspectives need
not be one side of the exploded dualism of scheme and world. Thus
innocently conceived, schemes or perspectives can be seen as embod
ied in languages or cultural traditions. So languages and traditions can
figure not as "tertia " that would threaten to make our grip on the
world philosophically problematic, but as constitutive of our un
problematic openness to the worldY (This Gadamerian conception of
30. This might give one pause about Rorty's attitude to such thinkers. Moves in the
language of traditional philosophy can be aimed at having the right not to worry about its
· problems, rather than at solving those problems. I think Rorty is insufficiently alive to this
possibiliry.
3 1 . The idea of a transcendental constitution of consciousness sounds harder to reha bi 1-
itate, but perhaps even that would not be impossible; see the previous footnote .
•
tradition figures at the end of Lecture VI.) Rorry suggests (p. 344) that
" intentionalistic notions'� just as such breed unhealthy philosophical
worries ("inserting imaginary barriers between you and the world");
in the context of a pragmatism less half-baked than Rorty achieves,
that suggestion can stand exposed as absurd. 32
9. I think if we follow Davidson in rejecting ;he third dogma of em

piricism, that has a devastating effect on the familiar cornerstones of
Quine's philosophy. Here I evidently diverge from Davidson himself:
with respect to the distinction between the analytic and the synthetic,
and to the indeterminacy of meaning, he declares himself " Quine's
faithful student" ("A Coherence Theory of Truth and Knowledge",
pp. 3 1 2-3).
Quine's thesis that translation is indeterminate gives vivid expres
sion to his notion of "conceptual sovereignty" . The indeterminacy is
supposed to mark how far the products of "conceptual sovereignty"
fall short of being determined by the scientifically treatable facts about
"empirical significance". Now the point Quine is aiming to make here
is inextricably bound up with the persisting dualism of the endoge
nous and exogenous factors ( "conceptual sovereignty" and "empiri
cal significance" ), the very thing Davidson rejects as the third dogma.
Quine's point is precisely to insist on how far the exogenous factor
falls short of determining meaning in the intuitive sense, bearing on
the world. When we discard the third dogma, we discard the very
framework within which this point seems to make the sort of sense
Quine wants. Now it should stand revealed as unsurprising that
meaning is undetermined by "empirical significance"-so called. That
just reflects the fact that " empirical significance" cannot really be a
kind of significance at all, since, conceived as dualistically set over
against "conceptual sovereignty", it is excluded from having anything
to do with the order of justification.
That meaning is indeterminate with respect to " empirical
significance" has no tendency to show, what would indeed be interest
ing, that meaning is indeterminate, period. That would require that
we have an ineliminable freedom of play when we look for a kind of
32. No doubt Rorty is encouraged here by the example of Quine. But to the extent that
Quinean suspicion of the intentional is more than mere scientism, its basis is thorou11hly
undercut by the debunking of the third dogma. See S9 below.
understanding that takes us outside the ambit of "empirical

significance": a kind of understanding that involves seeing how the
phenomena of our subjects' lives can be organized in the order of
justification, the space of reasons. If meaning is indeterminate in this
interesting sense, that is not something one could learn at Quine's
feet. 33
As for the distinction between the analytic and the synthetic, here
Quine is on the track of a real insight. But once the insight is properly
formulated, it undercuts that way of trying to give it expression. The
insight is the one that comes to explicitness only in Davidson's correc
tion to Quine: it is the rejection of the third dogma, "the dualism of
scheme and world". The suspect notion of the analytic is the notion of
truths that are such by virtue of being constitutive of conceptual
schemes in the suspect sense, the sense in which schemes are conceived
as dualistically set over against the world. The world plays no role in
accounting for the truth of these statements, though it is supposed to
help account for the truth of others. But once we have discarded the
dualism of endogenous and exogenous factors, that can no longer
seem a good interpretation for the idea of statements that are true by
virtue of their meaning. Meaning is not to be identified with the en
dogenous factor. If we embrace the picture I recommend in the lec
tures, in which the conceptual realm is unbounded on the outside, we
make it unintelligible that meaning's impact on determining what we
are to believe is endogenous as opposed to exogenous. (Not that it is
exogenous instead, or even jointly endogenous and exogenous; the
need to make this kind of determination simply lapses.) That is to say:
when we reject the dualism of scheme and world, we cannot take
meaning to constitute the stuff of schemes, on the dualistic conception
of schemes. But that does not deprive us of the very idea of meaning.
So if I am right that Quine's insight is really a glimpse of the unaccept
ability of the dualism, perhaps we can rehabilitate the idea of state
ments that are true by virtue of their meaning, without flouting the
real insight.
If, as I suggested (§ 8), the notion of a conceptual scheme need not
belong to the dualism, meaning can co�stitute the stuff of schemes in
33. I say more about this at pp. 245-6 of " Anti-Realism and the Epistemology of Under·
standing", in Herman Parret and Jacques Bouvercssc, cds., Meaning and Understanding (De
Gruyter, Berlin, 1981), pp. 225-48.
an innocent sense. We can reject the two factors without threatening

the idea that there are limits to what makes sense: that our minded
ness, as Jonathan Lear puts it,34 has a necessary structure. The idea of
a structure that must be found in any intelligible conceptual scheme
need not involve picturing the scheme as one side of a scheme-world
dualism. And analytic truths (in an interesting sense, not just
definitionally guaranteed truisms such as "A vixen is a female fox " )
might b e just those that delinea�e such a necessary structure. 35
This need not involve backtracking from what is right about the
Sellarsian thought that I considered in §4 above. Sellars claims that
nothing is Given, and argues in detail against the idea of an exogenous
Given. My topic there was a reading of Quine's attack on analyticity,
according to which the claim is that nothing is endogenously Given
either. The point was that there is a problem with "either": it suggests,
falsely, that Quine is with Sellars in firmly rejecting the exogenous
Given.
The Sellarsian thought in full is that nothing is either exogenously
or endogenously Given. This need not tell against rehabilitating an
idea of analyticity in the way that I have suggested. It would be wrong
to conceive what is necessary in any intelligible conceptual scheme as
fixed because it is Given, either exogenously or endogenously. That
very distinction has lapsed. Sellars claims, in a Hegelian spirit, that all
forms of Givenness are to be overcome; Davidson rejects the dualism
of the endogenous and exogenous factors. Perhaps these are just two
different expressions for the same insight. In that case Sellars's
thought, general though it is ("all forms of Givenness"), did not re
quire him to claim that absolutely everything we think is up for revi
sion. Immunity to revision, come what may, is a mark of Givenness
only if it is understood in terms of the two factors, and it need not be.
The idea that any intelligible conceptual scheme has a necessary
structure needs care. If we find ourselves inclined to take the thought
that the structure is necessary as a reassurance that our thinking must
be on the right track, I think we have gone astray, in the direction of
supplying a solution, rather than a dissolution, of traditional philo
sophical problems. As far as I can see, such reassurance is the point of
34. See "Leaving the World Alone", Journal of Philosophy 79 ( 1 982), 382-403.
35. Perhaps this is the category in which we should place at least some of the "hinge
propositions" to which Wittgenstein attributes a special significance in On Certainty (Basil
Blackwell, Oxford, 1969).
Afterword, l'art 1
the vestigial transcendental idealism that Lear finds in the later

Wittgenstein.3 6 A full-blown transcendental idealism reassures by af
firming, to put it crudely, that we cannot be fundamentally wrong
about the world we think about, since it is constituted by us. Lear's
vestigial transcendental idealism affords a version of the same reassur
ance by affirming that the "we" who figure in the idea of " our concep
tual scheme" " disappear"; the effect is that there cannot be a general
worry about how we suppose things to be, a worry that this is just our
line (as if there might be another). I think there is something right
about the idea of the disappearing "we", but (though this is admit
tedly a fine distinction) the disappearance of the "we" should not take
on the aspect of a reassurance, but should rather figure as part of the
reason why a reassurance should never have seemed to be needed.
One way to discover the limits of what makes sense is by an activity
characteristic of the later Wittgenstein: working out towards them in
thought-experiments, and noting the extent to which one finds oneself
losing one's grip. I borrow that image from Bernard Williams, in a
paper called "Wittgenstein and Idealism" . 37 Lear is following Wil
liams when he takes it that the role of "how we go on" in Wittgenstein
reveals his view as a form of transcendental idealism. But it is difficult
to make this fit. " How we go on" summarily introduces what Lear
describes as our mindedness, and we might begin by trying to find in
Wittgenstein a thought on these lines: world and mind, or minded
ness, are transcendentally made for each other. What makes it appro
priate to call the Kantian version of such a thought " idealism" is, to
put it crudely again, that the constituting of this harmony between
world and mind is supposed to be a transcendental operation of mind:
not, of course, the empirical mind, which is in constituted harmony
with the world, but an off-stage transcendental mind. But nothing
matches that in Wittgenstein. "How we go on" is just our minded
ness, which is ex hypothesi in constituted harmony with our world; it
is not something that constitutes the harmony, as it were from out
side. And now we should be struck by the thought that there is no
thing in Wittgenstein's picture to do the constituting of the harmony.
The appearance of a piece of vestigially transcendental philosophy is
hard to sustain.
. 36. See "Leaving the World Alone" and Lear's contribution to the symposium "The
Disappearing 'We'", Proceedings of the Aristotelian Society, supp. vol. 58 ( 1 984), 219-42.
37. In his Moral Luck (Cambridge University Press, Cambridge, 1 982), pp. 1 44-63 .
•
In §2, I cast Quine's rejection of analyticity as a corollary of his

claim that "empirical sianificance" cannot be apportioned between
statements: if there is no such thing as the "empirical significance" of
an individual statement, it does not make sense to suppose that some
statements may have null "empirical significance" . Now once we ap
preciate the peculiarity of Quine's notion of "�mpirical significance",
we can see a gap between the premise that "empirical significance"
cannot be apportioned among individual statements and the conclu
sion that we cannot make sense of analyticity. An analytic statement
should be a statement with no vulnerability to experience, and it is
true that we cannot make sense of that if we cannot make sense of the
idea that an individual statement might have its own vulnerability to
experience. But here we must gloss " vulnerability to experience" in
terms of rational answerability. When Quine speaks of facing the tri
bunal of experience, that sounds as if it expresses the idea of rational
answerability, but the rhetoric is hollow; that is the point Davidson
fastens on. So if "empirical significance" in Quine's sense cannot be
apportioned between statements, that does nothing to show that ra
tional answerability to experience cannot be apportioned between
statements.
And in fact, once we understand experience so that it really can be
a tribunal, we commit ourselves to conceiving it in such a way that
rational answerability to it can be apportioned between statements.
Consider an experience with a content partly capturable by "Here's a
black swan". Such an experience poses a rational problem for the
statement, or belief, that there are no black swans; there is a germane
ness relation between them that is not restricted, as in Quine's picture,
to the likelihood that the belief will be abandoned.
Quine's argument for the indeterminacy of translation exploits a
familiar Duhemian point, which we can put like this: vulnerability to
experience cannot be shared out among the statements of a theory. If
the invocation of vulnerability is not just empty rhetoric, but alludes
to a rational relation, this claim can indeed figure in an argument for
an indeterminacy of meaning. The argument can work only if the lan
guage in which we capture experience can be held separate from the
language of the theory, so that the relevant experience does not, as it
were, already speak the language of the theory. In that case we can put
the Duhemian point by saying that individual statements of the theory
are indeterminate in meaning with respect to the observational state-
ments on which the theory is grounded. Now it may be that in some

of the contexts in which we think of theory as based on experience,
the language of theory can indeed be held separate from the language
of observation; so we can argue on those lines that the observational
significance of individual theoretical statements is indeterminate. But
we cannot extract a general indeterminacy of meaning from these con
siderations. We might hope to do so, but only by falling into the con
fusion of the third dogma of empiricism. That way we embrace, at the
general level, a limiting case of the separation of languages. We push
all meaning into theory, and we leave experience not speaking any
language at all, not even metaphorically. But this defeats the hope,
because it eliminates the rational relation that figures in the cogent
Duhemian argument for indeterminacy. The cogent argument can at
best be local. So if we see our way through the third dogma, we cut the
Duhemian point down to size.
P A R T I I
Postscrip t to Lecture III
1 . In the lectures, I claim that we can coherently credit experiences

with rational relations to judgement and belief, but only if we take it
that spontaneity is already implicated in receptivity; that is, only if we
take it that experiences have conceptual content. Evans's thinking
stands as an obstacle to this: he undertakes to conceive experiences as
a rational basis for judgements, even while he excludes experiences
from the conceptual sphere. In Lecture Ill (§4), I argue that Evans's
position is unacceptable, because it is a version of the Myth of the
Given.
This can seem difficult to sustain. The very idea of representational
content brings with it a notion of correctness and incorrectness: some
thing with a certain content is correct, in the relevant sense, j ust in
case things are as it represents them to be. I can see no good reason not
to call this correctness "truth" . But even if, for some reason, we re
serve that title for correctness in this sense when it is possessed by
things with conceptual content, it seems a routine thought that there
can be rational connections between the world's being as a possessor
of one bit of content represents it and the world's being as a possessor
of another bit of content represents it, independently of what kind of
content is in question.
Christopher Peacocke aims to exploit this routine though.t in A
Study of Concepts, when he argues for a position like Evans's in that
it takes certain judgements and beliefs to be rationally grounded in
non-conceptual content possessed by experiences. At p. 80 he defends
the claim that the non-conceptual content he attributes to experiences
can afford " not merely reasons but good reasons" for judgements and
beliefs. The defence, for one representative case, involving an experi
ential basis for judging that something is square, goes like this: "If the
thinker's perceptual systems are functioning properly, so that the non
conceptual representational content of his experience is correct, then
when such experiences occur, the object thought about will really be
square." Peacocke comments: " In this description of why the linkages
are rational linkages, I make essential use of the fact that the non
conceptual content employed in the possession condition [the condi
tion for a subject to possess the concept square] has a correctness con
dition that concerns the world. The account of the rationality of this
particular linkage turns on the point that when the correctness condi
tion of the relevant nonconceptual contents is fulfilled, the object will
really be square. "
But this falls short of establishing what Peacocke needs: namely,
that non-conceptual content attributable to experiences can intelligi
bly constitute a subject's reasons for believing something.
There is a familiar style of explanation for circumstances that in
volve a subject (such as the circumstance of someone's believing
something): an explanation of the sort I mean shows how it is that
the explanandum is as it should be from the standpoint of rationality
( for instance true, if the explanandum is a belief). Now this is not eo
ipso to give the subject's reasons for whatever the explanation ex
plains. The subject may not even have reasons. Consider, for in
stance, the bodily adjustments that a skilled cyclist makes in
rounding curves. A satisfying explanation might show how it is that
the movements are as they should be from the standpoint of rational
ity: suited to the end of staying balanced while making progress on
the desired trajectory. But this is not to give the cyclist's reasons for
making those movements. The connection between a movement and
the goal is the sort of thing that could be a reason for making the
movement, but a skilled cyclist makes such movements without need
ing reasons for doing so. Why would it not be similar with experience
and judgement, if experiences had the non-conceptual content that
Peacocke says they have ?
The routine thought shows that the linkages between experience,
conceived as Peacocke conceives it, and belief are rational linkages,
but only in the sense in which the linkages between considerations
about what is required to keep one's balance and the cyclist's adjust
ments are rational linkages. This does not establish that something
with non-conceptual content can intelligibly figure on the in-out di

mension I mentioned earli�r in this Afterword (Part I, §5). That is: the
routine thought does not entitle Evans to his talk of judgement and
belief as " based upon" experience, or Peacocke to his counterpart talk
of beliefs being formed for reasons constituted by experience's being
as it is. 1
How could we ensure that a story like Peacocke's displayed experi
ence not just as part of the reason why, but as yielding reasons for
which a subject forms her beliefs?
One way would be to have it that the subject accepts the story, and
uses it in deciding what to believe, or at least would be disposed to cite
it if challenged. Suppose we are told that someone argues to the con
clusion that an object she can see is square, from the premise that in
seeing it she enjoys an experience with a certain non-conceptual con
tent. Its having that content involves a correctness condition ration
ally linked to believing that the object is square in the way that figures
in Peacocke's exploitation of the routine point. The correctness condi
tion is that the object is indeed square. In this version, the story clearly
depicts someone who forms the belief, with its conceptual content, for
a reason supplied by an experience, with its non-conceptual content.
But this is not what Peacocke needs. This subject would have to be in
command of Peacocke's more or less abstruse conceptual apparatus
for talking about non-conceptual content: the concepts of scenario
content, protopropositional content, and so forth. But Peacocke
a wants the supposed rational grounding in experience to figure in an
account of conceptual capacities exploited in the observational judge
ments of quite ordinary subjects, not j ust those who are philosophi
cally speaking in the know.2
1 . See, for instance, p. 7: "The thinker must . . . be disposed to form the belief for the
reason that the object is so presented." Here it is not something with non-conceptual con
tent that is said to lie further out than a belief on my in-out dimension. But when, in a
context in which non-conceptual content is in play, Peacocke writes of "perceptual experi
ences that give good reasons for judging . . . (certain conceptual) contents" (p. 66), he must
mean that the judging is done for those reasons, as in the formulation on p. 7.
2. Similarly with " for the reason that" at p. 7, where it is not non-conceptual content
that is supposed to figure further out than belief on my in-out dimension. A parallel defence
of "for the reason that" here would require the subject to have the concept of sensational
properties of regions of the visual field. Peacocke is sketching a candidate account of what
it is to possess the concept red, and he would not dream of suggesting that anyone who
possesses that concept must have the concept of sensational properties of regions of the
visual field.
Afterword, Part II 1 65
And it seems right not to require possession of the theory, even

independently of the fact that the rational connection with experi
ence is supposed to figure in accounts of what it is for anyone to
possess this or that observational concept. If theory were needed to
mediate the rational connection of experience to judgement and be
lief, that would undercut the very point of placing experience be
tween belief and the world on the in-out dimension. If we restrict the
role of experience in empirical thought to its being something from
which we can argue to a conclusion about the world, given that we
know the relevant theory, then we cannot conceive experience as it
self constituting access to the world. Something whose role is to yield
premises for arguments about the world could only be opaque, as I
put it in a parallel context when I was discussing the conception of
impressions that is common to Sellars and Davidson (in this
Afterword, Part I, §5).
Once we are clear how far the routine point is from what Peacocke
needs, it becomes possible to see how unattractive Peacocke's position
really is. In the reflective tradition we belong to, there is a time
honoured connection between reason and discourse. We can trace it
back at least as far as Plato: if we try to translate "reason" and " dis
course" into Plato's Greek, we can find only one word, logos, for
both. Now Peacocke cannot respect this connection. He has to sever
the tie between reasons for which a subject thinks as she does and
reasons she can give for thinking that way. Reasons that the subject
can give, in so far as they are articulable, must be within the space of
concepts.
I do not mean to suggest any special degree of articulateness; that
would be just as unsatisfactory as requiring command of Peacocke's
theory. But suppose one asks an ordinary subject why she holds some
observational belief, say that an object within her field of view is
square. An unsurprising reply might be "Because it looks that way".
That is easily recognized as giving a reason for holding the belief. Just
because she gives expression to it in discourse, there is no problem
about the reason's being a reason for which . . . , and not just part of
·
the reason why. . . .

In that simplest kind of case, what the subject says counts as giving
a reason for her belief because the way the object looks is the way she
believes it to be. In other cases the connection between reason and
belief is less simple. A minimally articulate reply to the request for a
" uu M l N lJ A N D W U K L lJ
reason might have to be less specific, perhaps " Because of the way it
looks" . But this makes no difference to the essential point. Here, too,
the reason is articulable (even if only in the form " It looks like that"),
s o it must b e no less conceptual than what it i s a reason for.
The routine point is really no more than that there can be rational
relations between its being the case that P and its being the case that
Q (in a limiting case what replaces " Q " can simply be what replaces
"P"). It does not follow that something whose content is given by the
fact that it has the correctness condition that P can eo ipso be
someone's reason for, say, j udging that Q, independently of whether
the content is conceptual or not. We can bring into view the rational
relations between the contents-its being the case that P and its being
the case that Q-only by comprehending the putatively grounding
content in conceptual terms, even if our theory is that the item that has
that content does not do its representing in a conceptual way. A the
ory like Peacocke's does not credit ordinary subjects with this compre
hensive view of the two contents, and I think that leaves it unintelligi
ble how an item with the non-conc.::p tual content that P can be
someone's reason for judging that Q.3
2. Why is Peacocke convinced that the rational grounding of belief

and judgement in experience has to set up links between the concep
tual realm and something outside it?
The conviction is largely sustained by a requirement of non-circu
larity that he imposes on accounts of what it is to possess this or that
concept. Avoiding circularity, in the relevant sense, does not require
that the concept in question not be used within the account at all, but
just that it not be used in a specification of the content of a conceptual
state. An account that violates this requirement "will not have eluci
dated what it sets out to elucidate" (p. 9).
Now if we set out to give an account of what it is to possess an
observational concept, we shall have to exploit the way judgements
3. Evans appropriates "seem" as "our most general term for the deliverances of the in·
formational system", that is, items with non-conceptual content (The Varieties of Refer
"
ence, p. 154; see also p. 1 80). It is certainly an easy thought that seemings can be our rea
sons for judgements and beliefs. But I think that is only because we understand the content
of seemings as conceptual; Evans's appropriation just undermines the intelligibility of
"seem". For the c.ontent of seemings (specifically lookings) as conceptual, see Sellars, "Em
piricism and the Philosophy of Mind", pp. 267-77.
I\{terwura, run H
and beliefs that employ it can be rationally grounded in experience.4

So if we hold that the content of the relevant experiential appearings
already involves the concept in question, we cannot satisfy Peacocke's
requirement of non-circularity. When we set out to give an account of
what it is to possess, for instance, the concept red, we shall find our
selves saying things like this: to possess the concept red one must be
disposed-if one takes it that the lighting conditions and so forth are
of the right sort-to make j udgements in whose content that concept
is applied predicatively to an object presented to one in visual experi
ence, when the object looks red to one, and for that reason. 5 But this
use of "looks red" presupposes, on the part of an audience to whom
the account might be addressed, not just the concept red, which would
be innocent, but the concept of possession of the concept red, which is
implicit in the idea of being able to have things look red to one. And
that is just what the account was supposed to be an account of. It is
obvious how this motivates the thought that the grounding experience
must be characterizable otherwise than in terms of conceptual con
tent.
But this just shifts the question. Why should we suppose that it is
always possible to give accounts of what it is to possess concepts, in
conformity with Peacocke's non-circularity requirement? Notice that
what is in question is accounts of what it is to possess concepts. Per
haps we could give a condition satisfied by all and only those who
possess a certain concept, without presupposing the concept of pos
session of that concept. For all I know, for instance, there may be a
neurophysiologically specifiable condition that 'distinguishes those
who possess the concept red from those who do not. But that specula
tion does not promise an account of what it is to possess the concept.
The neurophysiological condition would not bear on the question
4. Of course the applicability of an observational concept is not restricted to cases in

which one has the grounding type of experience. (A subject must understand that something
can be red without looking red to her.) This point is implicitly covered by the wording in the
text. Suppose someone took a predication of Mred" to be warranted only when she had an
experience like those we descrihe in terms of something's look ing red to us. Such a per"'"
would not qualify as possessing the concept red; that is not how judgements and hclicb
exploiting that concept are rationally grounded in experience. (Things could not even look
red to such a person; that is why I needed the circumlocution Man experience like
those . . . " . )
5. Compare the first clause of Peacocke's sketch a t pp. 7-8.
what someone thinks when she thinks that something is red. And
Peacocke wants his accoutits to bear on such questions. That is why
an account of an observational concept must place employments of
the concept in the space of reasons, even though the non-circularity
requirement compels Peacocke to hold that the experiences that con
stitute the rational basis of employments of such concepts are outside
the space of concepts.
What is at issue here is whether it is in general possible to give side
ways-on accounts of concepts, in the sense that figures in Lecture II,
§4, and in my discussion of Rorty earlier in this Afterword (Part I,
§6). The non-circularity requirement is in effect an insistence on side
ways-on accounts. In Lecture II, I deny, in effect, that sideways-on
accounts of concepts are in general possible. I cannot see that
Peacocke gives any reason to suppose that is wrong. In fact this seems
to work the wrong way round for him. In § 1 above, I urged that it is
difficult to see how experiences, conceived as Peacocke conceives
them, could constitute a believer's reasons for believing something.
This suggests I was right to deny that sideways-on accounts are possi
ble; the problem for the motivated thought tends to undermine the
• motivating thought.
Peacocke's proposed accounts are indeed offered as accounts of
what it is to possess this or that concept, and they talk of beliefs or
judgements, which figure in the accounts as having contents that em
ploy the concepts in question. But that does not mean that the ac
counts are not from sideways on in my sense. What is true is that the
accounts explicitly represent themselves as being about thinkers, users
of the concepts in question. But they do not say-indeed they carefully
refrain from saying-what the thinkers think when they use the con
cepts in question. Avoidance of circularity requires the accounts to
come at what the thinkers think only from the outside, identifying it
as something that one thinks when . . . , where what follows "when"
is a condition external to possession of the concept. The accounts em
body the claim that there is an inside view, but they are not given from
it. Peacocke is responsive to the suspicion that this externality threat
ens the project of capturing content. He thinks he can meet the threat
by linking the external condition to the thinking not j ust with "when"
but also with " for the reason that". However, I have urged, in § 1
above, that the required externality undermines the very intelligibility,
here, of "for the reason that". So I see no reason to give up, or qualify,
Afterword, Part II 1 69
the claim I made in Lecture II. I see no reason to suppose, and plenty
of reason not to suppose, that it is always possible to give accounts of
concepts in conformity with the non-circularity requirement.6
Non-circular accounts might be available, so far as these consider
ations go, in cases where what follows "when and for the reason
that . . . " can be a mention of conceptual states whose content in
volves concepts other than the one of which an account is being given;
that is, cases in which a concept can be captured in terms of how em
ployments of it are rationally grounded in employments of other con
cepts. But that is of course exactly not how it is with observational
concepts. These non-circular accounts would be given not from a side
ways-,on orientation to the whole conceptual realm, but only from
outside the conceptual capacities of which they are accounts; whereas
Peacocke's proposed accounts of observational concepts would be
given from outside the conceptual realm altogether.
Is my scepticism about accounts that do everything Peacocke wants
some kind of obscurantism? Peacocke makes a suggestion to that ef
fect at pp. 3 5-6:
Theories are developing in the literature of what it is to possess certain
specific concepts: the first person, logical notions, and many others.
While there is much that is still not understood and not all of what has
been said is right, it is hard to accept that the goal of this work is com
pletely misconceived. On the contrary, there are often phenomena
specific to the concept treated that are explained by these accounts.
McDowell would not let us say that these accounts are theories of what
it is to possess these concepts. But I cannot see what else they can be, and
we can hardly just dismiss them.
A lot depends here on what "the goal of this work" really is. We are
not restricted to just two options: either accepting that the presuppo
sitions of this undismissable work are made explicit in Peacocke's de
siderata for a theory of concepts, avoidance of circularity and all; or
6. What I was driving at in "In Defence of Modesty" was this denial that the sideways-on
perspective can capture concepts. I cannot recognize my side of the debate in Peacocke's
representation, at pp. 33-0. As Peacocke puts what he takes to be my point, it ought to be
satisfiable by a sideways-on story, so long as the story announces itself as being about
thoughts. My point was that one cannot fix what is thought from the outside, identifying it
only as something that one thinks when . . . ; and I do not believe it helps to add " for the
reason that. . . (Dummett makes much of the claim that the connections are rational in
w.
his response to me: see pp. 260-2 of " Reply to McDowell", in Taylor, ed., Michael
Dummett: Contributions to Philosophy, pp. 253-68 . )
refusing to acknowledge that there might be anything to be said

about, for instance, the specific rational connections between first-per
son thinking and experience.
That case is worth considering in a little more detail. As Peacocke
remarks ( p. 72), it does not individuate the concept of the first person
to say that it "is that concept m such that judgements a bout whether
Fm display a certain sensitivity to experiences that represent Fm as the
case " . But we do not have that- problem if we identify the relevant
judgements, for a given subject, as judgements that display a certain
sensitivity to experiences that represent F(herself) as the case. Of
course this violates the non-circularity requirement. But we can leave
the violation standing and still have plenty of room for substantive
inquiry into the character of the "certain sensitivity" . No doubt the
amended formulation is much too simple. But the possibility of locat
ing its substantive promise in the "certain sensitivity", leaving the cir
cularity undisturbed, is suggestive. Peacocke is here aiming to co-opt
Evans's discussion of the first person as an example of work directed
towards an account of the sort Peacocke envisages. But I think Evans
is much better read as primarily concerned with the "certain sensitiv
ity", and not particularly interested in avoiding circularity.
3. Evans claims that we do not have enough concepts of, say, colours
for it to be possible that the content of our visual experience is concep
tual. In Lecture III, §5, I urge, against that, that we can express all the
concepts we need, in order to capture the finest detail of our colour
experience, by utterances of "that shade ". We do not have all these
concepts in advance, but we do have whichever we need, exactly when
we need them.
An utterance of " that shade" depends for its meaning on the iden
tity of a sample shade. We might lay down the rule that something
counts as having that shade j ust in case it is indiscriminable in colour
from the indicated sample. (Of course we can actually say something
in these terms only in the presence of a sample.)
Now there is a familiar pitfall here. It may be tempting to lay down
a second rule as well: that something counts as having a shade if it is
indiscriminable in colour from something else that counts as having
that shade. ·But if we say that, we run afoul of a sorites paradox: we
undermine the idea that utterances of "that shade" can express a de
terminate meaning at all, because indiscriminability in colour is not
A{terwura, (urt 11
transitive. With a n artfully arranged series o f samples, we could work

out from the original sample to one that would have to count as hav
ing the shade in question, by repeated applications of the second rule,
even though it is discriminable in colour from the original sample,
contrary to the first rule. The two rules do not cohere.
So we should restrict ourselves to the first rule, and resist the temp
tation to endorse the second. Something that counts as having a shade
does not thereby count as a sample of the shade, bringing into the
shade's extension anything indiscriminable from it; that is what leads
to the familiar collapse. The status of a sample, a determinant of the
extension of the concept expressed by "that shade" in the relevant
use, must be reserved for the original sample, or at least for things
whose promotion to that status would not contribute to slippage in
the concept's extension.7
In the lecture I allow that a shade concept can be given overt expres
sion later in the duration of the recognitional capacity that constitutes
possessing it, if experience happens to provide something to anchor
the reference of a later utterance of the demonstrative "that shade" . In
view of the point I have just made, we need to treat that thought with
some care. If there are things that would be indiscriminable in colour
from the new anchor but would have been discriminable in colour
from the original sample, then even if one is not wrong in recognizing
the new anchor as an exemplifier of the shade in question (since it
would be indiscriminable from the original sample), one would be
wrong if one supposed it would do as a sample of the original shade.
And it is hard to suppose that when it anchors the reference of an
utterance of "that shade", it can be serving as anything but a sample
of whatever shade is in question. This case brings out the possibility
that a putatively recognition-based conceptual capacity can lapse.
One can suppose one is empioying a recognitional capacity, but
wrongly because one's propensities to arrive at what one takes to be
recognitions have been skewed; and one way that can happen is that
one has promoted an unsuitable example to the status of a secondary
sample. In this kind of case, the subsequent utterance of "that shade"
does not express the original shade concept. The subject has lost the
original shade concept without realizing it.
7. I see no need for the apparatus Peacocke introduces in response to the sorites threat,
pp. 83-4.
1 72 .\ ! ! :-; D A :-.; D \\ 0 R L D
4. What makes it recognizably a concept that is made available by

the presence to experience "f the original sample is a capacity of mem
ory; so I claim in the lectures. A person can retain a capacity to recog
nize things as having that shade, and while this recognitional capacity
persists, possibly for quite a short period, the subject can embrace just
that shade in thought. (It is redundant, in my �sage, to say "concep
tual thought" . ) The most direct manifestation of the conceptual ca
pacity after its onset is a judgement in which the recognitional capac
ity that constitutes it is directly employed: something seen later is
judged to have the shade in question. But this kind of memory-based
capacity to embrace a precise shade in thought can also be exercised
in thinking that is not geared to present experience. Consider, for in
stance, someone who remembers the colour of a rose that she no
'
longer has in view, and thinks "I should like the walls of my room to
be painted that shade".
The kind of thought I mean here needs to be distinguished from one
that might be expressed by something like "I should like the walls of
my room to be painted the shade of the rose that I saw on such-and
such an occasion " . One can have a thought of this second kind with
out remembering the shade itself, as we can naturally put it. Perhaps
one remembers that it was a shade of pale apricot, and that it struck
one as ideal for the walls of one's room. In the case I mean, one has
the shade itself in mind, as we can naturally put it: it does not figure
in one's thinking only as what fits some specification. One way of cap
turing this is co say that one sees the shade in one's mind's eye.
In the presence of the original sample, a subject who has the con
cept of a shade is enabled to classify items as possessing the relevant
shade or not, by direct inspection for colour match. We can retain, for
a while at least, a capacity to classify items as possessing the relevant
shade or not, in a way that corresponds with the verdicts we would
have given on the strength of direct inspection for colour match, even
a fter the original sample is no longer in view. (In fact this capacity of
memory is drawn on even while the original sample is still in view, if
a candidate for being credited with the shade cannot be juxtaposed
with the sample in a single overview. ) We can be confident, and with
good grounds, that the verdicts, or potential verdicts, of the memory
involving capacity correspond with the verdicts direct comparison
would have yielded, and that fact is at least part of what underlies
how natural it is to say things like "I see it in my mind's eye". It is as
.-\/1<'1'11'• •rd. 1'.11"1 II 1 7. 1
i f a sample i s still available for comparison with any candidates for

being credited with the shade.
The idiom is a fine way to make the distinction between having the
shade itself in mind and bringing one's thinking to bear on it by way
of a specification. But it can elicit a familiar and dubious philosophical
thought, according to which the idiom's obvious rightness points to a
mechanism by which the retained classificatory capacity operates. The
idea is that the subject attributes the shade by comparing things she
sees with a retained inner sample of the shade; it is like checking things
for match in colour with, say, a paint manufacturer's sample, except
that here the sample is before an inner eye. Wittgenstein warns against
ideas like this (see, for instance, Philosophical Investigations, §604 ) .
"I see it in my mind's eye" is a natural way to affirm one's capacity to

embrace the shade itself in one's thinking, a capacity that would be
overtly displayed, given a suitable opportunity, in verdicts as to
whether things exemplify the shade. The idiom does not allude to psy
chological machinery that would be operative in one's production of
those verdicts. 8
5. What brings into being the recognitional capacities that constitute

abilities to embrace determinate shades in one's thinking? In the lec
tures I consider only confrontations with exemplifiers of the shades in
question. But recall Hume's question whether a subject could form, in
pure thought, the idea of a shade that has been missing from her col
our experience.9 That introduces what may be a different possibility. I
am already committed to claiming that we have all concepts of the
relevant kind potentially, just by virtue of having the concept of a
shade. But Hume's question is about possession of such a concept in
actuality: the suggestion is that a subject might equip herself to see the
missing shade in her mind's eye, by an exercise of imagination alone.
8. "I see it in my mind's eyeH embodies a picture. The proper attitude to it is like the one
Wittgenstein expresses, in Philosophical Investigations, §427, to the picture embodied in
saying things like "While I was speaking to him I did not know what was going on in his
head. H Wittgenstein says: "The picture should be taken seriously. We should really like to
see into his head. And yet we only mean what elsewhere we should mean by saying: we
should like to know what he is thinking. H Taking this picture literally, and even supposing
it is somehow intellectually obligatory to do so, is depressingly common in contemporary
philosophy of mind.
9. A Treatise of Human Nature, ed. L. A. Selby-Bigge and P. H. Nidditch (Clarendon
Press, Oxford, 1978), 1 . 1 . 1 .
Perhaps this is indeed possible; I have no thesis that the possibility

would threaten, in the way it seems to threaten Hume's empiricism
about concepts. The main thing to insist on is that if it is possible, this
is just another possibility for the genesis of the sort of recognitional
capacity we can claim by saying "I see the shade in my mind's eye" .
There i s n o extra support here for the idea that exercises o f the asso
ciated recognitional capacity are based on comparison with an inner
sample.10
10. If there i s a possibility here, the point i s that imagination can fi l l gaps in a repertoire
of concepts of shades. But the repertoire must be mainly initiated by experience, in the way
I discuss in the lecture. Shade concepts in general must still be constitutively dependent on
intuitions, in the way that, according to my speculation in the lecture, underlies the fact that
Evans does not even consider them as a possible range of concepts.
P A R T I I I
Postscrip t to Lecture V
1 . According to Crispin Wright, Wittgenstein commits himself to

elaborating a certain conception of meaning, though his " quietism"
prevents him from discharging the commitment. In Lecture V, §3, I
suggest that the consequences Wright himself attributes to that con
ception of meaning are intolerable. I leave the suggestion undefended,
and turn instead to urging that we must miss Wittgenstein's point if
we read him in a way that makes his " quietism" an awkward fact
about a set of texts that actually set substantive tasks for philosophy,
and even give hints as to how to execute them.
This does not strike me as a real gap. If we form a proper under
standing of Wittgenstein's " quietism", the doctrines Wright finds him
committed to can collapse of their own weight. It is only under pres
sure from philosophy that anyone could dream of thinking there is
any problem about the idea of how things are anyway, independently
of communal ratification.
Wright thinks Wittgenstein reveals just such a philosophical com
pulsion. According to Wright, Wittgenstein poses a genuinely urgent
question about how meaning is possible: a question of a sort philoso
phy characteristically addresses, in a kind of activity whose very point
is bound up with our not being able to rule out in advance the possi
bility of affront to what passes for common sense. And Wright thinks
a conception of meaning can be legitimized, in the face of the problem
Wittgenstein supposedly elaborates, only if we are willing to rethink
what passes for common sense about the idea of how things are any
way. So Wittgenstein's "quietism" looks like an embarrassing fa ilure
to acknowledge the character of his own �hilosophical achievement.
But it is a mistake co think Wittgenstein points up a good question

about how meaning is possible.
The " quietism " , the av� idance of any substantive philosophy, is
really the point. Questions such as " How is meaning possible? " ex
press a sense of spookiness, and Wittgenstein's point is that we
should not indulge the sense of spookiness, b�ut rather exorcize it.
The question looks like an urgent one from the standpoint of a
• world-view that is inhospitable to meaning: a standpoint from
which it looks like a task for philosophy to shoehorn into the world
something as close as we can get to our previous conception of
meaning. But philosophy's task is rather to dislodge the assump
tions that make it look difficult to find a place for meaning in the
world. Then we can take in our stride meaning's role in shaping our
lives. We do not need a constructive legitimizing of its place in our
conception of ourselves.
Wittgenstein aims to cast suspicion on an aura of mystery that cer
tain thoughts about meaning acquire in an uncongenial environment.
The thoughts are thoughts like this: the meaning o f, say, an instruc
tion that specifies an arithmetical series, for instance the instruction
"Add 2", " determines the steps in advance" (compare Philosophical
Investigations, § 1 90), in such a way that-to bring the thought into
direct connection with Wright's concerns-the fact that such-and
such a move is the correct one at a certain point in the expansion of
the series does not depend on ratification by the relevant community,
those who count as understanding the instruction. Such a thought
can seem uncanny, as if it credited meaning with magical powers.
Wright's mistake is to take it that Wittgenstein means to cast suspi
cion on such thoughts themselves. But Wittgenstein's target is the at
mosphere of uncanniness. The thought itself is all right. 1
The contrast between rampant and naturalized platonism helps to
bring out this possibility. The relevant thoughts are platonistic, and if
we can envisage only a rampant platonism the aura of uncanniness is
inescapable; our only recourse is a philosophical construction in
which we pull in our horns, about objectivity or whatever. But the
I. Compare Philosophical Investigations, S 195. An interlocutor voice says: "But I don't

mean that what I do now (in grasping a sense) determines the future use causally and as a
matter of experience, but that in a queer way, the use itself is in some sense present." The
responding voice says: "But of course it is, ' i n some sense'! Really the only thing wrong with
what you say is the expression 'in a queer way'. The rest is all right . "
Afterword, Part Ill 1 77
problem is not with the platonistic thoughts themselves. Within a nat

uralized platonism they do not have the aura of uncanniness. So an
other way of putting Wright's mistake is to say that he is blind to the
possibility of a naturalized platonism.2
2. In the lecture I credit Wittgenstein with the aspiration of seeing

through the apparent need for ordinary philosophy. This needs to be
taken with care. I do not mean to suggest that Wittgenstein seriously
contemplates a state of affairs in which ordinary philosophy no longer
takes place. The intellectual roots of the anxieties that ordinary phi
losophy addresses are too deep for that. This point comes out dramat
ically in the multiplicity of voices in Wittgenstein's later writing� its
dialogical character. The voices that need to be calmed down, recalled
to sobriety, are not alien voices; they give expression to impulses he
finds, or at least can imagine finding, in himself. When he writes, "The
real discovery is the one that makes me capable of stopping doing
philosophy when I want to" (Philosophical Investigations, §133), we
should not take him to be envisaging a post-philosophical culture
(an idea that is central to Rorty's thinking). He is not even envisag
ing a future for himself in which he is definitively cured of the philo
sophical impulse. The impulse finds peace only occasionally and tem
porarily.3
But I do not think it follows that there can be no role, in a style of
thinking that is genuinely Wittgensteinian in spirit, for the sort of dia
gnostic move I propose in the lectures. What I suggest is that our philo
sophical anxieties are due to the intelligible grip on our thinking of a
modern naturalism, and we can work at loosening that grip. It is a
way of making this suggestion vivid to picture a frame of mind in
which we have definitively shrugged off the influences on our thinking
that lead to philosophical anxieties, even if we do not suppose we
2. I elaborate a reading on these lines, with more detailed reference to Wittgenstein than
seemed appropriate in these lectures, in � Meaning and Intentionality in Wittgenstein's Later
Phi!osophyft, in Peter A. French, Theodore E. Uehling, Jr., an-1 Howard K. Wettstein, eds.,
Midwest Studies in Philosophy, vo!. 17: The Wittgenstein Legacy (University of Notre
Dame Press, Notre Dame, 1992), pp. 42-52. For kindred thoughts, see Cora Diamond, The
Realistic Spirit. At p. 6, she describes one of her targets like this: "Wittgenstein's criticism
of . • . mythology or fantasy-in particular, his criticism of the mythology attached to logi
cal necessity-is read as if it were rejection of the mythology as a false notion of how things
are."
3. I have been persuaded here by James Conant and Lisa Van Alstyne.
could ever have such a frame of mind as a permanent and stable pos
session. Even so, this identification of a source for our apparent diffi
culties can be one of our resources for overcoming recurrences of the
philosophical impulse: recurrences that we know there will be.
3. When I describe the relaxed platonism made possible by a natural

ism of second nature, I say things like this: the structure of the space
of reasons is not constituted in splendid isolation from anything
merely human. Wittgenstein's " quietism", properly understood, is a
good context in which to stress that remarks like that should not in
vite the question "So what does constitute the structure of the space
of reasons? " If we take ourselves to be addressing that question, my
invocation of second nature, sketchy and unsystematic as it is, will
seem at best a promissory note towards a proper response. But that
would miss my point. I think the response we should aim at being
entitled to, if someone raises a question like "What constitutes the
structure of the space of reasons? " , is something like a shrug of the
shoulders. It is a thought well expressed by Rorty that questions like
that should not be taken to be in order without further ado, j ust be
cause it is standard for them to be asked in philosophy as we have
been educated into it. Their sheer traditional status cannot by itself
oblige us to take such questions seriously. Rather, there is an assumed
background that is supposed to make them urgent. When I invoke
second nature, that is meant to dislodge the background that makes
such questions look pressing, the dualism of reason and nature. It is
not meant to be a move-which could be at best a first move-in con
structing a response to that question.
4. In Lecture V, §5, I discuss Kant's suggestion that the continuity of

self-consciousness embodies only a formal conception of persistence.
It is worth comparing the suggestion Wittgenstein makes in a well
known passage (pp. 66-7 of The Blue Book ): that "the use [of "I"]
as subject" does not refer. We can formulate what Wittgenstein
suggests there, in a way that brings out the resemblance to the Kantian
thought, like this: the reference-plus-predication structure in, say, " I
have a toothache" is merely formal. The considerations that induce
Wittgenstein to make this suggestion seem closely similar to those
that are operative in the Paralogisms. And our verdict should be the
same. The motivation for the suggestion-to undermine the basis for
Afterword, Part III 1 79
a Cartesian conception of the ego-is admirable, but once we under

stand the wider context in which the use of "I" as subject functions,
the suggestion can be detached from its motivation. There is no obsta
cle to supposing that my use of "I" as subject refers to the human
being I am.4
5. It may be helpful to j uxtapose my remarks on reference to partic

ulars, in Lecture V, §6, with what I say in Lecture II, especially § 3 .
In Lecture II, I exploit Wittgenstein's "truism" t o discourage the
idea of a gap between thought as such and the world. An objector
might say something like this: "You can make it look as if your drift
is not idealistic, so long as you consider the world only as something
whose elements are things that are the case. In that context, you can
exploit the claim that it is no more than a truism that when one's
thought is true, what one thinks is what is the case. But as soon as we
try to accommodate the sense in which the world is populated by
things, by objects (and there had better be such a sense), it will emerge
that your image of erasing an outer boundary around the realm of
thought must be idealistic in tenor, perhaps in an extended sense.
Even if the image allows for a kind of direct contact between minds
and facts, it obliterates a certain possibility that we should not be will
ing to renounce, a possibility of direct contact between minds and ob
jects, which must surely be external to the realm of thought. This pos
sibility is what has been brought to our attention in the recoil from the
generalized Theory of Descriptions. "
Given the identity between what one thinks (when one's thought is
true) and what is the case, to conceive the world as everything that is
the case (as in Tractatus Logico-Philosophicus, § 1) is to incorporate
the world into what figures in Frege as the realm of sense. The realm
of sense (Sinn) contains thoughts in the sense of what can be thought
(thinkables) as opposed to acts or episodes of thinking. The identity
displays facts, things that are the case, as thoughts in that sense-the
thinkables that are the case. But objects belong in the realm of refer
ence (Bedeutung), not the realm of sense. The objection is that
Wittgenstein's "truism" yields an alignment of minds with the realm
of sense, not with the realm of reference.
4. See Evans's discussion of rhe Wirrgensrein passage in The Varieties of Referr�r<"c,

pp. 2 1 7-20.
I can indeed formulate a main point of my lectures in terms of the

Fregean notion of sense, l V<e this: it is in the context of that notion
that we should reflect about the relation of thought to reality, in order
to immunize ourselves against the familiar philosophical anxieties.
This is just another way to put the thought I express in the lectures in
terms of Sellars's image of the logical space of reasons. Frege's notion
of sense operates in the space of reasons: the whole point of the notion
of sense is captured by the principle that thoughts, potential senses of
whole utterances, di ffer if a single subject can simultaneously take ra
tionally conflicting stances towards them (say, any two of acceptance,
rejection, and neutrality) without thereby standing convicted of irra
tionality. If failing to distinguish senses would leave us liable to have
to attribute to a rational and unconfused subject, at the same time,
rationally opposed stances with the same content, then we must dis
tinguish senses, so as to rnake possible a description of the subject's
total position that has different contents for the stances, and so does
not raise a question about the position's rationality.5
The objector I have imagined thinks that if we take this Fregean
view, that thought and reality meet in the realm of sense, then we can
purport to accommodate thought's bearing on objects-a relation be
tween minds and inhabitants of the realm of reference-only by em
bracing some version of the generalized Theory of Descriptions. We
lose the insights of those who have insisted on a more direct relation
than that allows between minds and objects.
Now what I say at the end of Lecture V can be seen as pre-empting
any such objection. On a proper understanding of the Fregean appa
ratus, my exploitation of Wittgenstein's "truism" in Lecture II, which
can indeed be reformulated by saying thought and reality meet in the
realm of sense, already allows for what is right in the recoil from the
generalized Theory of Descriptions. If the relevant senses are rightly
understood, the role of sense, in a picture that leaves the relation of
thought to the world of facts unproblematic, already ensures that
there is no mystery about how it can be that the relevant thoughts
bear on the relevant particulars, inhabitants of the realm of reference,
in the non-specificatory ways that proponents of the recoil rightly in
sist on.
5. See Evans, The Varieties of Reference, pp. 1 8-19.
P A R T I V
Postscript to Lecture VI
1 . My talk of Aristotelian innocence may raise some suspicions. It

may seem to fly in the face of an obvious fact: Aristotle explicitly dis
cusses positions that are, in some respects, strikingly like modern nat
uralism.
It is true that the ancient atomists (to single out perhaps the best
case for the objection) already have a somewhat modern-seeming con
ception of nature, in the sense of what is comprehended by the most
fundamental understanding of things. They take nature, in that sense,
to be empty of meaning and value. And it is true that Aristotle sticks
to his different conception of nature in self-conscious opposition to
conceptions like that. But in such ancient anticipations of the disen
chanted conception of nature, the thesis that nature is empty of mean
ing and value lacks a certain status it has in modern thinking. It does
not figure as another way to formulate a rightly entrenched view of
the kind of understanding aimed at by properly scientific investiga
tion: a view that is not open to dispute, but part of what one must take
for granted if one is to count as an educated person.
As I insist in the text, we must recognize that when the realm of
properly scientific understanding came to be generally seen as disen
chanted, that marked an intellectual advance. That is why it is so dif
ficult for us to get out from under the philosophical anxieties I have
been concerned with. It can easily go unnoticed that what it is right for
us to conceive as disenchanted need not be identified with nature.
Aristotle's innocence consists in his not being subject to that intellec
tual pressure. Certainly he is aware that it is possible to see nature,
identified as the topic of the most fundamental understanding, in a
way that disenchants it. But for him that reflects a merely optional
and, from his intellectual standpoint, not very well supported-view
of the most fundamental understanding. He does not have to resist a
temptation to let the label " nature" affix itself to something that he is
anyway intellectually committed to countenancing as what is compre
hended by scientific understanding. Aristotle has no inkling of a per
fectly correct thought that we can formulate like this: if we identify
nature as the topic of scientific understanding, we must see it as disen
chanted.1
2. In Lecture VI, §4, I refuse to credit non-human animals with

orientations towards the world. Understandably, that raises some
hackles.
It may help to stress that what I am committed to denying in the
case of mere animals is precisely, and only, something correlative with
possession of spontaneity. In the text of the lecture, I try to minimize
any appearance that I am debunking animal mentality, by disowning
a reductive conception of biological imperatives, which are what
shape the lives of mere animals. And perhaps this talk of biological
imperatives already suggests a harder line than I need, even without
the reductive conception I disown. The point is just that dumb ani
mals do not have Kantian freedom. That is perfectly compatible with
acknowledging that they can be, in their ways, clever, resourceful, in
quisitive, friendly, and so forth. I do not suggest that they are some
how " out of it" . Indeed my whole point, in appropriating Gadamer's
notion of an environment, is to provide language for saying quite the
reverse, even while I deny that, lacking spontaneity as they do, dumb
animals can possess the world. And it is important that the freedom I
claim they lack is precisely Kantian spontaneity, the freedom that con
sists in potentially reflective responsiveness to putative norms of rea
son. No one without a philosophical axe to grind can watch, say, a
dog or a cat at play and seriously consider bringing its activities under
the head of something like automatism. But we can deny Kantian
spontaneity while leaving plenty of room for the self-movingness that
is plain to the unprejudiced eye in such a scene.
The risks of falling into philosophical anxiety that I am concerned
with in the lectures arise in relation to a conception of orientation
1 . This section responds to a question raised by M. F. Burnyeat.

Afterword, Part I V 1 83
towards the world that is correlated with spontaneity in the Kantian

sense. They arise precisely because of what is peculiar about sponta
neity in the Kantian sense, when it is viewed from the standpoint of
the familiar modern conception of nature as the disenchanted realm
of law.
If someone wants to work out a conception of orientation towards
the world that is detached from spontaneity in the Kantian sense, with
a view to making the language of world-directedness available for
talking about the mentality of brutes, that is, so far, perfectly all right
by my lights. I have no wish, apart from this context, to say anything
at all about mere animals, and certainly no wish to play down the
respects in which their lives are like ours. It is part of what I want to
insist on that we are animals too, not beings with a foothold outside
the animal kingdom. And in some respects, the lives of mature human
beings simply match the lives of mere animals; it would be absurd to
suppose that Bildung effects a transfiguration, so to speak, of every
thing that happens in a human life.2
What I do want to resist is the suggestion that such a neutral con
ception of orientation towards the world will do for all purposes.
That would amount to rejecting all talk of a spontaneity that is pecu
liar in the Kantian way. This may be motivated by the conviction that
if we let an acknowledged peculiarity of that kind lead to an outbreak
of philosophy, our intellectual position is hopeless, and I have made it
clear that I sympathize with this motivation. But in the lectures I aim
to show how we can acknowledge that we have a spontaneity that is
peculiar in the Kantian way, and that that fact is crucial to a special
way in which we are in touch with the world, without thereby landing
ourselves in that philosophical predicament. Wet'do not need to stop
this kind of philosophy before it begins, by refusing to allow that there
is any basis for the fuss it makes about spontaneity (to echo something
I say in Lecture IV, §4). When we acknowledge the peculiarity of
spontaneity, we should be aware of how we thereby risk falling into
unprofitable philosophical anxiety. But the risk need not be realized.
We can understand and exorcize the philosophical impulse, not j ust
repress it.
We can understand how it can seem that philosophical problems
2. Even those aspects of mature human life that are shaped hy Bildung show una�simi
lated residues from their evolution out of mere nature (first nature). That is a way of pun in�
a central thought of Freud.
are posed by a notion of orientation towards the world that is corre

lated with the Kantian peculiarity of spontaneity. It seems so because
of the pressure to suppose that spontaneity would have to be non-nat
ural. But I have tried to show how this appearance of a philosophical
task can be revealed as illusory. There is no call for elaborate and,
outside the study, incredible philosophical appanrtus to respond to
the question how it is possible for empirical thinking, understood as
an exercise of "conceptual sovereignty ", to bear on the world, let
alone to yield knowledge of the world. This is a recipe for a potentially
satisfying exorcism of philosophical anxiety, because it fully acknowl
edges the thought that generates the anxiety. And we can go this far
with the philosophical impulse only if we draw a line between posses
sors and non-possessors of spontaneity, in the way I do throughout
the lectures.
3. When I invoke tradition at the end of Lecture VI, I mean to do no

more than open the door to the topic, about which, no doubt, much
more needs to be said. I am not going to attempt a proper discussion
here. But it may help to clarify the note I mean to sound if I make
some remarks about the way my stance sets me in opposition not only
to Dummett, as I register in the lecture, but also to Davidson.
The Gadamerian note that the lectures end on is that understanding
is placing what is understood within a horizon constituted by tradi
tion, and I suggest that the first thing to say about language is that it
serves as a repository of tradition. Initiation into a language is initia
tion into a going conception of the layout of the space of reasons.
That promises to make it intelligible how, beginning as mere animals,
human beings mature into being at home in the space of reasons. On
this view, a shared language is the primary medium of understanding.
It stands over against all parties to communication in it, with a kind
of independence of each of them that belongs with its meriting a kind
of respect.3 We can understand communication across boundaries be
tween traditions by moving out from this basic case, where the hori
zon is pretty much given (not Given! ) by the tradition embodied in the
3. Not just the respect due to an effective instrument, which is enough to account for
recoiling when people misuse words like "disinterested" or "careen". The respect I mean is
the respect that is due to something to which we owe our being what we are. (Of course
what we do with our language can change it; for instance, what was misuse can cease to be
misuse. But that does not undermine the sense in which the language is independent of us.)
Afterword, Part I V 1 85
language, to the sort of case where horizons need to be fused, which

rnay take hard work.
Davidson does not concede this kind of importance to the idea of a
shared language. On the contrary, for Davidson communicative inter
action stands in no need of anything to play the role of a medium in
the sense I have gestured at. (Of course in a different sense there has
to be a medium: speech, smoke signals, or whatever.) In Davidson's
conception, the parties to communication are self-standing individu
als. They do not need a language-a specific repository of tradition as
to the shape of the space of reasons-to constitute them as potential
parties to communication, or indeed participants in any other activity
that requires conceptual capacities. In Davidson's view, the idea of a
language shared between parties to communication is at best short
hand for an explicable degree of correspondence in idiolects. Such
correspondence can make hypotheses for interpretation easier to
come by between some pairs of people, but mutual understanding be
tween people whom we think of as sharing a language is in principle
no different from the most radical interpretation. The "shared lan
guage" is no more than an aid in a cognitive performance that could
be undertaken without it; the capacity for mutual understanding
needs no philosophically interesting background.4
I cannot properly argue against this vision here, but I shall allow
myself a suggestion. Davidson conceives the kind of understanding of
persons that is in question here in just the way that I do in the lectures:
that is, in terms of placing what they think and do (including what
they say) in the space of reasons. It was as much from Davidson as
from Sellars that I learned to think that way. Where I use the Sellars
ian image of the space of reasons, Davidson talks of "the constitutive
ideal of rationality"; but the idea is clearly the same. (I exploit this
correspondence in Lecture IV, §4.) So Davidson's mutual interpreters
must come to their cognitive task already equipped with a sense of the
layout of the space of reasons, a substantive conception of what "the
constitutive ideal of rationality" requires. Now I think we should be
suspicious of the thought that we can simply credit human individuals
4. See "A Nice Derangement of Epitaphs", in LePore, ed., Truth and Interpretation:
Perspectives on the Philosophy of Donald Davidson, pp. 43�6. The germ of the thought
is already present in Davidson's claim that "all understanding of the speech of another
involves radical interpretation ": p. 125 of "Radical Interpretation", in Inquiries into Truth
and Interpretation, pp. 1 25-39.
with this equipment, without benefit of anything like my appeal to

initiation into a shared language and thereby into a tradition. I think
the idea that this cognitive equipment needs no such background is
just another outcropping of Givenness. If we want to attack the Myth
of the endogenous Given, in a counterpart to Sellars's Kant-inspired
attack on the Myth of the exogenous Given, this would be a much
better target than the idea of analyticity, or unrevisability come what
may (see Part I of this Afterword, § §4, 9). To put it in Hegelian terms,
Givenness is not unrevisability as such, but a supposed unrevisability
that reflects absence of mediation from our picture; and Davidson re
nounces the only available mediation for the capacity of human be
ings to understand one another.
In recent work, Davidson has undertaken to build the concept of
objectivity out of a "triangulation" between these self-standing sub
jects, pairwise engaged in mutual intcrprctation. 5 This comes into
conflict with the Kantian thesis of interdependence that I consider in
Lecture V, §5, and reconsider in Lecture VI, §4. By my lights, if sub
jects are already in place, it is too late to set about catering for the
constitution of the concept of objectivity. We must take subjectivity
and the concept of objectivity to emerge together, out of initiation into
the space of reasons.
4. I shall end by saying something to avert a perhaps unreal risk: that

by invoking tradition, I may seem to commit myself to a hidebound
conservatism about the possibilities for intelligibility. At the close of
the lectures, I repeat something I stress at various earlier points: that
being at home in the space of reasons includes a standing obligation to
be ready to rethink the credentials of the putatively rational linkages
that constitute the space of reasons as one conceives it at any time.
That leaves exactly as much room for innovation as there is. If a bit
of, say, vocal behaviour is to constitute making a novel remark, as
opposed to mere babble, it must be capable of being understood by
people who would not have thought of saying that themselves. One
kind of originality calls on those who understand it to alter their prior
5. See "Meaning, Truth, and Evidence", in Robert B. Barrett and Roger F. Gibson, eds.,
Perspectives on Quine (Basil Blackwell, Oxford, 1 990), pp. 68-79. Davidson had sketched
this exploitation of "triangulation" at the end of "Rational Animals", in Ernest LePore and
Brian McLaughlin, eds., Actions and Events: Perspectives on the Philosophy of Donald
Davidson (Basil Blackwell, Oxford, 1 985), pp. 473-80.
Afterword, Part IV 1 87
conception of the very topography of intelligibility. A remark with

this kind of originality is not just a move hitherto unimagined but still
within the possibilities as they were already comprehended, at least in
general terms. (That is how it is with even the most radical innovation
in chess.) Rather, the remark changes a hearer's conception of the
structure that determines the possibilities for making sense. But even
in this kind of case, it can only be a matter of warping a prior concep
tion of the topography of intelligibility. An utterance could not make
a place for itself in a comprehending mind from scratch, reshaping
wholesale its audience's conception of the possibilities. Even a
thought that transforms a tradition must be robted in the tradition
that it transforms. The speech that expresses it must be able to be
intelligibly addressed to people squarely placed within the tradition as
it stands.
Index
Agency, 89-91 Davidson, Donald, 3-4, 14-18, 24, 25,

Allison, Henry E., 43n 34, 35, 62-3, 64, 67-9, 70, 72, 74-6,
Analytic-synthetic distinction, 129-37, 78, 129, 137-46, 147, 148, 1 52-6, 1 65,
157-6 1 1 84-6
Animals, 49-50, 63-5, 69-70, 1 1 4-23, Dennen, Daniel, 55n
182-4 Descriptions, the Theory of, 105-7, 1 79-80
Anscombe, G. E. M., 104n Diamond, Cora, 153n, 177n
Aristotle, 78-84, 85, 91, 108-10, 1 8 1-2 Donnellan, Keith S., 105n
Ayer, A. J., 36n Dualism: of scheme and content, 3-7, 137-
9, 1 56-6 1 ; of reason (or norm) and na
Belief, 48, 60, 140 ture, 85-6, 93-5, 108-1 1 , 153-5. See
Bildung, 84, 87-8, 95, 1 23-4, 125, 183. also Naturalism; Nature
See also Nature Duhem, Pierre, 133n, 1 60-1
Bodily movement. See Agency Dummen, Michael, 124-6, 159n, 1 84
Brandom, Robert, Sn, 97n
Burge, Tyler, 105n Embodiment, 91, 1 02-4
Burnyeat, M. F., 84n, 1 82n Environment (vs. world), 1 1 5-19
Evans, Gareth, 32n, 37n, 47-65, 67-9, 70,
Cartesian philosophy of mind, 90n, 100- 72, 76, 100-1n, 102, 1 06-7, 1 14, 162,
4, 1 14, 1 1 6, 1 78-9 166n, 170, 179n, 1 80n
Cassam, Quassim, lOOn Experience, 5, 9-13, 25-6, 31-4, 51-5,
Causation, 14-18, 68, 71n, 75 60-3, 66-9, 97-8, 1 1 1-13, 1 1 9, 138-
Cognitive science, 55, 121 46, 162-6. See also Impressions; Intu
Cohen, Zvi, 8n itions
Coherentism, 14-1 8, 68, 1 37-46, 147-56
Colours, 12, 29-33, 56-60, 170-3 Fichte, J. G., 43n
Conant, James, 1 77n Frankfurter, Aryeh, 138n
Concepts and the conceptual, 4-5, 7-13, Frege, Gonlob, 106-7, 179-80
46-7, 56-60, 105-7; observational con Freud, Sigmund, 1 83n
cepts, 6-7, 166-9
Content, 3-4, 28-9; empirical, 5-6, 14- Gada mer, Hans-Georg, 36n, 8 1 n, 1 1 5-19,
18, 42-3, 129-30; non-conceptual, 47- 126n, 155, 182, 1 84
65, 121-3, 162-70 Geach, P. T., 7, 20, 105n
1 90 Index
Gibson, J. J., ! On 1 08; of second nature, 85-6, 91-2, 94-

Given, the, 4, 5-9, 14-23, 39, 5 1-5, 135- 5, 1 09-10, 178
7, 1 8 5-6. See also Dualism, of scheme Nature, 65, 70-86, 87-8, 96-7, 108-10,
and content 1 8 1-2; second nature, 84-6, 87-8,
Grounding. See Rational relations 9 1 -2, 103-4, 109-10. See also
Bildung
Hegel, G. W. F., 43n, 44-5, 83, 1 1 1 , Neurath, Otto, 8 1
1 1 7n, 1 8 6
Hookway, Christopher, 138n Peacocke, Christopher, 5 6 n , 5 8 n , 1 62-70,
Hume, David, 97, 173-4 171n
Peirce, C . S . , 40n
idealism, 26-7, 39-40, 41-5; transcenden Pippin, Robert B., 44n, 1 1 1 n
tal, 4 1-4, 95-9, 1 58-9; Absolute, 44-5 Plato, 77n, 1 1 0, 165
Impressions, 9-10, 34, 1 39-46. See also Platonism, 77-8, 83-4, 88, 91-5, 109-10,
Experience; Intuitions; Receptivity 1 23-4, 1 76-7, 178
Indeterminacy of translation, 132, 134, Pragmatism, 129, 1 54-5
1 5 6-7, 160-1 Private Language Argument, 1 8-23
Inner sense and its objects, 1 8-23, 30, 33, Putnam, Hilary, 147n, 148-50, 154
36-8, 49-50, 1 1 9-23
Intelligibility, 70-2, 109 Quine, W. V., 13n, 1 29-37, 138, 139,
Interpretation, 16-17, 35, 1 52-3, 1 85 1 4 1 , 148n, 1 56-61
Intuitions, 4, 9, 59-60. See also Dualism,
of scheme and content; Given, the; Im Rational relations, 5, 7-13, 14-18, 42-3,
pressions; Receptivity 52-3, 79, 146, 1 62-6. See also Reasons,
space of
James, William, 148, 1 5 2 Reasons, space of, 5-13, 70-86, 1 80
Judgement, 1 1 , 29-34, 47, 60-3 Receptivity, 4, 9-13, 28-9, 29-34, 51-5,
Justification. See Rational relations 141, 142-3. See also Dualism, of
scheme and content; Experience; Given,
Kant, Immanuel, 3-5, 18, 41-4, 46, 5 1-5, the; Impressions; Intuitions
60, 66, 71n, 74n, 82-3, 85, 87, 89, 95- Reference, 1 04-7, 1 79-80
107, 1 10-1 1 , 1 14, 1 15, 124, 132, 134, Reflection, 12-13, 40, 47, 81-2, 126
135, 155, 1 78-9, 1 8 2-4 Rorty, Richard, 17, 35n, 71n, 85-6, 95,
Knowledge, 5-6, 15-18, SOn, 146-7. See 135n, 142n, 146-56, 168, 177
also Rational relations Russell, Bertrand, 71n, 1 05-7, 1 79-80
Kripke, Saul A., 92-3n, 105n Ryle, Gilbert, 90n
Language, 1 1 6n, 1 24-6, 1 84-6 Searle, John, 106, 107

Lear, Jonathan, 1 5 8-9 Secondary qualities, 1 1-12, 29-33, 38
Life, 78, 95, 1 03-4, 1 1 5 Sedivy, Sonia, 60n
Locke, John, 100 Self-consciousness, 47n, 99-104, 178-9
Lockwood, Michael, 97n Sellars, Wilfrid, 5, 12n, 3 1 nn, 70, 7 l n, 74,
SOn, 135-7, 140-6, 1 58, 165, 166n,
Macbeth, Danielle, 38n 1 85-6
Macintyre, Alasdair, 79n Sense. See Frege, Gottlob; Reference
Marcus, Ruth Barcan, 105n Sensibility. See Receptivity
Marx, Karl, 1 1 7- 1 9 Sideways-on views of understanding, 34-
6, 4 1-2, 82-3, 1 52-3, 1 68-9
Nagel, Thomas, 40n, 1 21-3 Spontaneity, 4-5, 7-13, 29-34, 43, 46-7,
Naturalism: bald, 67, 73, 76-7, 88-9, 5 1-5, 60-3, 71-86, 97-8
Index 191
Strawson, P. F., 2 1 n, 37n, 43n, 54n, 99n, Understanding. See Spontaneiry

lOOn, 1 02, 1 03n, 1 04n, 1 07n, l l l n
Stroud, Barry, 42n, 135n Van Alsryne, Lisa, 1 77n
Subjectivity, 99-104, 1 1 4-23; proto-sub
jectivity, 1 1 7, 1 1 8-23 Warrant. See Rational relations
Weber, Max, 70
Williams, Bernard, 79n, 96, 1 5 9
Tarski, Alfred, 148 Wittgenstein, Ludwig, 7 , 1 8-23, 27-8, 36-
Taylor, Charles, 71n, 1 53n 7, 57, 86, 92-5, 1 04n, 1 1 0, 1 5 H-9, I ?J,
Tradition, 98-9, 1 26, 1 84-7 175-80
Truth, 1 47-53 Wright, Crispin, 92-3nn, 17 5-7

Mind and World - John McDowell

Uploaded by

Copyright:

Available Formats

Mind and World - John McDowell

Uploaded by

Document Information

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

Mind and World - John McDowell

Uploaded by

Copyright:

Available Formats

MIND AND

With a New Introduction

HARVARD UNIVERSITY PRESS

First Harvard University Press paperback edition, 1996

Library of Congress Cataloging-in-Publication Data

Lecture II. The Unboundedness of the Conceptual 24

Lecture III. Non-conceptual Content 46

Lecture N. Reason and Nature 66

Lecture V. Action, Meaning, and the Self 87

Lecture VI. Rational and Other Animals 108

Part I. Davidson in Context 129

Part II. Postscript to Lecture III 162

Part III. Postscript to Lecture V 175

Part IV. Postscript to Lecture VI 181

Third, I have tried not to erase an unguardedness that seemed

I have many substantial debts to acknowledge.

I made my first sketches of the kind of formulation I have arrived at

1. This book first appeared without an Introduction. Since then,

2. A good way into the picture I offer is to consider the plausibility

so--must be a posture or stance that is correctly or incorrectly

1. "Two Dogmas of Empiricism", in W. V. Quine, From a Logical Point of View (Har­

3. It may seem surprising that I am associating empiricism with a

nature is the logical space in which the natural sciences function, as we

5 . I have been suggesting that we can trace some distinctive a nxieties

6. One option would be to renounce empiricism, at least with expe­

7. Sellars and Davidson think we are forced to renounce empiricism,

8. My alternative holds on to the thought rejected by bald natural­

9. The position I call "bald naturalism" figures in this book only as

posed problems together in a question on these lines: "How is empir­

responsiveness to reasons can fit into the natural world. (Indeed I

10. It should be clear that reflection about perceptual experience has

Concepts and Intuitions

1. The overall topic I am going to consider in these lectures is the

2. This Kantian background explains why what stands over against

mand of concepts, in terms of spontaneity. A schematic but suggestive

reedy responsive to experience. The supposed requirement is reflected

6. Mental Acts (Routledge and Kegan Paul, London, �957), SS6-l l .

What we wanted was a reassurance that when we use our concepts

4. The original Kantian thought was that empirical knowledge re­

in this sense are already possessed by impressions themselves, im­

5. I said (§4) that when we enjoy experience conceptual capacities

a faculty of spontaneity. That may make it seem that I am not really

that someone who has them is responsive to rational relations, which

ing willingness t o refashion concepts and conceptions i f that is what

6. It need not be a mere superficial oversight if someone fails to see a

1 1 . I intend the imagery of this paragraph to be reminiscent of the well-known closing

belong to spontaneity could be inextricably implicated in an operation

allay the worry. I think we should be suspicious of his bland

nected. By rejecting that, Davidson undermines his right to the idea

7. The Myth of the Given expresses a craving for rational constraint

general point by considering the way it shows up in those familiar

tier. The work of abstraction is the private ostensive definition that

rational responsiveness to a bare presence, into a composite concept

1 8. Philosophical Investigations, trans. by G. E. M. Anscom be (Basil Blackwell, Oxford,

to insist that they are indeed impressions, products of receptivity. So

8. In this lecture, I have claimed that we are prone to fall into an

1 . In my first lecture I talked about a tendency to oscillate between a

positions. My aim will be to suggest that our diffic11lties in this area

2. What I want to consider in this lecture is an objection on the score

anyway are becomes available to exert the required rational control,

1. "Two Dogmas of Empiricism", in W. V. Quine, From a Logical Point of View (Har

6. One option would be to renounce empiricism, at least with expe

8. My alternative holds on to the thought rejected by bald natural

posed problems together in a question on these lines: "How is empir

4. The original Kantian thought was that empirical knowledge re

in this sense are already possessed by impressions themselves, im

4. Wittgenstein's aphorism can be reworked like that for any concep

idea: that, when w e work at making someone else intelligible, we ex

that an object o f " inner experience" i s not independent o f the experi

7 . I have been urging that, i n judgements of experience, concep

9. In these first two lectures, I have introduced the tendency to oscil

tricably implicated in deliverances of receptivity. We must not sup

ing picture within a peculiar version of the sideways-on view I men

depict anything recognizable as empirical judgement; we have obliter

bility, o f capacities that are genuinely elements in a faculty of spon

without concepts, they are blind. And actually there is an interpreta

ence that ex hypothesi affords a suitable sample-one can give linguis

be given linguistic expression again, if the course of experience is fa

tlement to it can always be raised. We can sum up what Evans is sug

with the testimony system, when we do understand the linguistic per

2. Now I want to start uncovering the presumably deep-rooted men

ity by placing things in a logical space that is sui generis, by compari

1 5. According to the kind of reading I am disputing, Aristotle attributes this founda