0199661790

The Critical Imagination
oxford philosophical monographs

Editorial Committee
anita avramides r. s. crisp
william child antony eagle stephen mulhall
other titles in this series include
Nietzsche and Metaphysics

Peter Poellner
Understanding Pictures
Dominic Lopes
Things That Happen Because They Should
A Teleological Approach to Action
Rowland Stout
The Ontology of Mind
Events, Processes, and States
Helen Steward
Wittgenstein, Finitism, and the Foundations of Mathematics
Mathieu Marion
Semantic Powers
Meaning and the Means of Knowing in Classical Indian Philosophy
Jonardon Ganeri
Hegels Idea of Freedom
Alan Patten
Metaphor and Moral Experience
A. E. Denham
Kants Empirical Realism
Paul Abela
Against Equality of Opportunity
Matt Cavanagh
The Grounds of Ethical Judgement
New Transcendental Arguments in Moral Philosophy
Christian Illies
Of Liberty and Necessity
The Free Will Debate in Eighteenth-Century British Philosophy
James A. Harris
Plato and Aristotle in Agreement?
Platonists on Aristotle from Antiochus to Porphyry
George E. Karamanolis
Aquinas on Friendship
Daniel Schwartz
The Brute Within
Appetitive Desire in Plato and Aristotle
Hendrik Lorenz
The Critical
Imagination
James Grant
1
3
Great Clarendon Street, Oxford, OX2 6DP,
United Kingdom
Oxford University Press is a department of the University of Oxford.
It furthers the Universitys objective of excellence in research, scholarship,
and education by publishing worldwide. Oxford is a registered trade mark of
Oxford University Press in the UK and in certain other countries
# James Grant 2013
The moral rights of the author have been asserted
First Edition published in 2013
Impression: 1
All rights reserved. No part of this publication may be reproduced, stored in
a retrieval system, or transmitted, in any form or by any means, without the
prior permission in writing of Oxford University Press, or as expressly permitted
by law, by licence or under terms agreed with the appropriate reprographics
rights organization. Enquiries concerning reproduction outside the scope of the
above should be sent to the Rights Department, Oxford University Press, at the
address above
You must not circulate this work in any other form
and you must impose this same condition on any acquirer
British Library Cataloguing in Publication Data
Data available
ISBN 9780199661794
Printed by the MPG Printgroup, UK
For Terrence Comeau
This page intentionally left blank
Acknowledgements
Many people have provided feedback on this book. For their comments,
I would like to thank Mart Abrusn, Paloma Atencia-Linares, George
Botterill, Emily Brady, Emily Caddick, Davide Cargnello, Dan Cavedon-
Taylor, Rafael De Clercq, David Davies, Martin Davies, Heather Demar-
est, Dorothy Edgington, Paul Faulkner, Miranda Fricker, Stacie Friend,
Michael Garnett, Berys Gaut, Jonathan Gilmore, Dominic Gregory, Ian
Ground, Alex Grzankowski, Samuel Guttenplan, Andy Hamilton, Louise
Hanson, John Hawthorne, Allan Hazlett, Rob Hopkins, Jennifer
Hornsby, Keith Hossack, Andrew Huddleston, Dan Isaacson, Susan
James, Rosanna Keefe, Claire Kirwin, Andrew Klevan, Deborah Knight,
Stephen Laurence, Stephen Leighton, Sam Liao, Paul Lodge, Dominic
McIver Lopes, Sabina Lovibond, Ofra Magidor, Derek Matravers,
Andrew McGonigal, Jennifer McMahon, Aaron Meskin, Peter Millican,
Margaret Moore, Daniel Morgan, Victoria Moul, Bradley Murray, Chris
Norbury, Yuuki Ohta, Toby Ord, Alex Paseau, Ian Phillips, Jennifer
Saul, Elisabeth Schellekens-Damann, Severin Schroeder, Vid Simoniti,
Maarten Steenhagen, Josef Stern, Robert Stern, Scott Sturgeon, Christo-
pher Timpson, Kate Tunstall, Ralph Walker, Milly Zimeta, and the
readers for Oxford University Press. I would also like to thank audiences
at the American Society for Aesthetics 2010 Pacic Division Annual
Meeting; Birkbeck, University of London; the British Society of Aesthet-
ics 2010 and 2011 Annual Conferences; Hertford College, Oxford; the
London Aesthetics Forum; Oxfords 2009 Philosophy Graduate Confer-
ence, Doctoral Thesis Seminar, Ockham Society, and Topics in Aesthet-
ics Research Seminar; The Queens College, Oxford; and the University
of Shefeld. I am grateful to Peter Momtchiloff and Eleanor Collins of
Oxford University Press for their help in preparing the book for publica-
tion. I thank the Social Sciences and Humanities Research Council of
Canada for funding my research.
Some previously published material appears here with substantial revi-
sions. I am grateful to the publishers for permission to reprint it. Material
from the following papers is reprinted by permission of Oxford University
Press, on behalf of the British Society of Aesthetics: James Grant, The
viii ACKNOWLEDGEMENTS
Dispensability of Metaphor, British Journal of Aesthetics 50/3 (2010):

255272 <http://dx.doi.org/10.1093/aesthj/ayq018>; James Grant,
Metaphor and Criticism, British Journal of Aesthetics 51/3 (2011):
237257 <http://dx.doi.org/10.1093/aesthj/ayr016>. Material from
James Grant, The Value of Imaginativeness, Australasian Journal of Philoso-
phy 90/2 (2012): 275289 <http://dx.doi.org/10.1080/00048402.2011.
574143> is # the Australasian Association of Philosophy and is reprinted
by permission of Taylor & Francis Ltd <http://www.tandfonline.com>
on behalf of the Australasian Association of Philosophy.
I owe special thanks to my sister, Elizabeth Grant, who provided two of
the books illustrations. My parents, William and Joan Grant, have given
me invaluable support throughout the writing of this book, as they have in
everything else. My ance, Anna Kemp, has made this book much better
than it would otherwise have been. I would need to be a writer of her
calibre to express how much better she has made its authors life. Peter
Lamarque and Stephen Mulhall read the manuscript in its entirety, and
provided many valuable suggestions. Malcolm Budd has been unfailingly
generous with his time and considerable philosophical intelligence. I am
very grateful for his perceptive comments on several drafts. Lastly, I owe a
special debt of gratitude to my doctoral supervisor, John Hyman. Johns
acuity, intellectual rigour, and independence of mind have saved me from
many pitfalls. Time and again his example has reminded me what good
philosophy looks like. I hope this book shows I learned something.
Contents
List of Figures xi
Introduction 1
1. The Aims of Criticism 5
2. Criticism and Appreciation 29
3. Criticism and Imagination 53
4. Metaphor and Likeness 87
5. The Dispensability of Metaphor 125
6. Metaphor and Criticism 149
Conclusion 173
Bibliography 179
Index 189
List of Figures
1.1 C. Elizabeth Grant. Drawing after Antoine Babuty Desgodets, Second

Ordre du Colise, Rome. 2012. 26
2.1 C. Elizabeth Grant. Palazzo del T, Mantua. 2012. 47
6.1 Attic red-gure pyxis decorated with women and erotes or cupids.
Close to the Meidias Painter. End of 5th century bc. 158
6.2 Raphael. The Miraculous Draught of Fishes. 15151516. 165
Introduction
The highest Criticism, writes Oscar Wilde, is in its way more creative
than creation. Wilde is an advocate of a view with a long history in
aesthetics. According to this tradition, art criticism offers as much scope for
imaginativeness as creating art does. The critics freedom to imaginatively
interpret, experience, and describe works of art is a signicant part of what
makes criticism rewarding to write and to read. Engaging actively and
intelligently with artworks involves exercising one of the aptitudes most
closely associated with making them.
This attractive idea has taken many forms. Some think the interpret-
ation of literature offers good evidence for it. Literary works seem to admit
of many different interpretations. Some hold that this gives critics great
scope to interpret works of literature imaginatively, just as theatre directors
have great scope to adapt plays imaginatively. Others think a developed
sensitivity to art involves the disposition to look at artworks imaginatively.
A sensitivity to architecture might be manifested by seeing a row of
columns spaced close together as tense and forbidding, or columns spaced
farther apart as stately, serene, meditative.1 So too, many have thought
that critics commonly need to be imaginative to adequately describe the
appearance of artworks or the effects they have on us. Vivid metaphors and
other gurative descriptionsof music as shimmering, poems as tightly-
knit, arches as soaring, and so forthare commonly employed when
critics try to say what is of aesthetic interest in a work. Such descriptions
have received much attention in analytic philosophy of art. They have
been one of the main sources, within contemporary analytic philosophy,
of the view of criticism as a notably imaginative pursuit.
1
John Summerson, The Classical Language of Architecture (London: Thames and Hudson,
1980), p. 26.
2 INTRODUCTION
This view of criticism is of interest for several reasons. I will mention

two. In the rst place, it would help explain why it is rewarding to give
artworks the kind of sustained attention critics give them. Why engaging
with art is rewarding has been a question of perennial interest, and it has
received a variety of answers. Aristotle attributed our enjoyment of art-
works to our propensity to enjoy reasoning and understanding. Both,
he thought, are involved in recognizing what a representational artwork
represents. A common view in eighteenth-century aesthetics was that
beauty pleases because it engages our cognitive powers in certain ways.
On both views, many of the pleasures art provides are explained by
reference to the pleasures of understanding. Engaging with art involves,
or is importantly like, acquiring knowledge from perception. On the view
I am considering, by contrast, engaging with art is importantly like creating
art. Criticism is rewarding because it involves exercising imaginativeness.
Second, clarifying the role of imaginativeness in criticism will help to
explain what it is, exactly, critics do. Criticism is, to many people, a rather
mystifying practice. Even after reading much of it, many nd it unclear
how a critic reaches her conclusions, what constraints there are on the
claims she makes, and what the aims of criticism are. Indeed, these are
matters of dispute among critics themselves. This division is reected in
the different attitudes people have to the claim that criticism may be, or
ought to be, highly imaginative. Some worry that those who stress the
creative element in criticism open the way to overly speculative, impres-
sionistic criticism, losing sight of criticisms genuine aims and the standards
to which it must be sensitive. This raises the question of what those aims
and standards are. In determining the role of imaginativeness in criticism,
we shall shed light on this question. We shall also shed light, thereby, on
what characteristics, aptitudes, and intellectual virtues make someone a
good critic.
This book has two main parts. The aim of Chapters 13 is to say what the
role of imaginativeness in criticism is. In Chapters 46, I examine metaphor
and its use in criticism. Many inuential writers, as I have said, take the
prevalence of metaphor in criticism to indicate the role imaginativeness
plays in criticism and aesthetic experience. It is therefore appropriate to
consider, in the light of my account of the role of imaginativeness in
criticism, why critics do use metaphor so frequently. This account will not
only explain why metaphor is commonly used in critics descriptions. It will
also provide reasons why metaphor is such an effective device for
INTRODUCTION 3
communication in general. The language of criticism is not alone in at-

tracting philosophical attention on account of being frequently metaphor-
ical. As we shall see, the widespread use of metaphor in the sciences,
metaphysics, and theology has also been thought to reveal something
important about these areas. My conclusions have implications for these
debates as well.
In the course of my discussion, I shall answer several questions about
metaphor, imaginativeness, and criticism that are important in their own
right. Those interested in these questions can read the chapters in which
I answer them on their own.
In Chapters 1 and 2, I argue for a new view about what the aims of
criticism are. Philosophers have disagreed on whether the point of criti-
cizing an artwork is to evaluate it, to explain it, to modify our response to
it, or to achieve something else besides. Clearly, we need to know what
critics are trying to achieve if we are to say what role imaginativeness and
metaphor play in criticism. In the rst chapter, I take issue with ve
inuential views on the aims of criticism, which have been suggested by
Arnold Isenberg, Monroe Beardsley, Frank Sibley, Arthur Danto, and
Nol Carroll, among others. None of these views is ultimately successful.
A good account of the aims of criticism, however, ought to accommodate
the truth in each of them.
I provide my own account in the second chapter. To get clear on what
the aims of criticism are, I argue, we need to get clear on what it is to
appreciate art. At the beginning of this chapter, I present a new account of
what art appreciation is. This discussion of appreciation enables me to give
an account of the aims of criticism that avoids the problems affecting other
views. I identify a constitutive aim of criticism and a non-constitutive aim
of criticism, both of which are connected to appreciation. I then explain
what characteristics make someone a good critic, given that the aims I have
identied are aims of criticism. I conclude by comparing this conception of
the critic with Humes conception of the true judge of art.
This account of critics and criticism puts us in a position to say what the
role of imaginativeness in criticism is. This is the subject of Chapter 3.
Although the belief that imaginativeness has a signicant role in criticism
has a long history, many philosophers today are unlikely to be familiar with
this history. To familiarize my readers with the debate in which I am
participating, I give a short overview of the history of this idea at the
beginning of this chapter, outlining the views of Hume, Kant, Oscar
4 INTRODUCTION
Wilde, and Roger Scruton, among others. I then develop my own

contribution to this debate in two stages. First, I defend a new account
of what imaginativeness is. Second, based on my accounts of criticisms
aims, of the characteristics of good critics, and of imaginativeness, I explain
what the role of imaginativeness in criticism is.
In Chapter 4, I present an argument in the philosophy of language.
I argue for a claim about metaphor that goes against much current thinking
in the philosophy of language. I argue that likeness always plays a certain
role in determining what is communicated with metaphor and in enabling
us to grasp it. My claim resembles, but is not, the claim that metaphors are
comparisons or compressed similes. That view is rejected by most philoso-
phers working on metaphor today, and hostility to it has helped prevent
philosophers from recognizing the truth of the claim I defend here. I reply
to twelve objections, presented by Donald Davidson, John Searle, Richard
Moran, William Lycan, Robert Fogelin, Samuel Guttenplan, and others.
My view enables us to explain several facts about our use of metaphor.
I use it for this purpose in the rest of the book.
Chapter 5 is an attack on the belief that metaphor is needed in order to
think, express, communicate, or discover certain things. Many contem-
porary philosophers believe this. Some have appealed to this claim to
explain metaphors prevalence in critics descriptions of artworks and of
our responses to them. However, I argue that recent arguments for
metaphors indispensability, by Stephen Yablo, Richard Boyd, Elisabeth
Camp, and Berys Gaut, are unconvincing. Moreover, the view that
metaphor is indispensable for these purposes is commonly confused with
other, more plausible claims. We ought not to explain the prevalence of
metaphor in criticism by appealing to it.
I explain in the nal chapter why metaphor is so effective, and therefore
so prevalent, in criticism of the arts. Building upon the conclusions about
criticism reached in the rst half of the book, and upon the view of
metaphor defended in Chapters 45, I argue that there are two import-
antly different kinds of metaphor used by critics. This distinction has not,
to my knowledge, been previously recognized; but drawing it is essential
to explaining why metaphor is used so often in criticism. The explanation
of why metaphor is such an effective way of achieving the aims of criticism
is different for each of these kinds of metaphor. In the course of providing
my explanation, I also explain why, as philosophers since Aristotle have
observed, metaphor has a particularly close connection with perception
and with imagination.
1
The Aims of Criticism
To characterize the role imaginativeness plays in criticism of the arts, we

need to know what the aims of criticism are. If we know what criticism is
an attempt to achieve, we shall be better placed to say how imaginativeness
enables critics to achieve it.
Philosophers and other theorists of the arts have often made claims
about what the aims of criticism are. There are at least two things we
might be interested in when we ask this question. First, we might want to
know what the constitutive aim or aims of criticism are. Being written
with this aim or these aims would be part of what makes a piece of writing
an instance of art criticism. If criticism has a constitutive aim, it would be
both interesting in itself to know what it is, and useful for saying what role
endowments such as imaginativeness play in criticism.
Second, we might want to know what makes a piece of criticism a piece
of good criticism. Perhaps there are aims such that achieving them (or
being such as to achieve them) makes a piece of criticism that has them
good criticism. That is to say, it makes it good as criticism: a passage can be
an example of good writing but bad or indifferent criticism. Aims of this
kind may or may not also turn out to be constitutive aims, as we shall see.
If there are such aims, knowing what they are, too, would enable us to say
what the role of imaginativeness in criticism is.
In what follows, I will identify both a constitutive and a non-
constitutive aim of criticism. Saying how imaginativeness enables a critic
to achieve these aims will show what its role in criticism is. In this chapter,
I will consider ve answers to the question of what the aims of criticism
are. I will argue that none of them is an aim of all criticism. However, each
proposal is instructive for what it reveals about the variety of things critics
do and attempt to achieve. An account of the aims of criticism should
make sense of the fact that critics do and attempt to achieve these things.
In Chapter 2, I will provide such an account.
6 THE AIMS OF CRITICISM
1. Helping readers choose

Monroe Beardsley argues that the primary aim of criticism is to help the
critics readership decide which artworks to choose to experience. He says
that critics, in the strict sense, are those who set themselves up, or are set
up by others, to make public judgments for the purpose of guiding the
choices of others who are less qualied than they, perhaps by the lack of
talent or time.1 He calls this the Consumers union model of (professional)
criticism, and argues that it is essentially correct as an account of art
criticism, capturing its primary character, on which its other features
depend.2 Critical activities such as explaining and interpreting are under-
taken for this purpose.3
Beardsleys model is best suited to much journalistic criticism, particu-
larly reviews. Many reviews by theatre critics, lm critics, music critics,
etc., are written with the intention of helping readers to decide whether to
see, watch, listen to (etc.) the work. Virginia Woolf makes this aim explicit
in her essay on The Faerie Queene, introducing her piece as some general
observations made by one who has gone through the experience, and
wishes to urge others, who may be hiding their yawns and their polite
boredom, to the same experience.4
However, Beardsleys account is a poor model of much academic
criticism. It is implausible that guiding the choices of others is the aim of
much academic criticism. Such criticism is normally written for readers
who have already chosen to experience the work. It is usually presumed
that they are reading the criticism because they are studying the work.
Good academic criticism is certainly capable of providing information that
would be useful to audiences trying to decide whether to experience the
work. But it is implausible that being useful in this way is what makes it
good criticism, or that it is defective if it is not useful for this purpose.
1
Monroe C. Beardsley, What Are Critics For?, in The Aesthetic Point of View: Selected
Essays, ed. Michael J. Wreen and Donald M. Callen (Ithaca: Cornell University Press, 1982),
p. 149.
2
Ibid., pp. 156157.
3
Ibid., p. 160. Stein Haugom Olsen expresses a similar view of the aim of what he calls
criticism in the judgmental sense. Criticizing, in this sense, is pointing out good or bad
qualities in a literary work in order to make a recommendation of some sort either to the artist
or to the reading public (Stein Haugom Olsen, Criticism of Literature and Criticism of
Culture, Ratio 22/4 (2009): 439440).
4
Virginia Woolf, The Faery Queen, in Hugh Maclean and Anne Lake Prescott, eds,
Edmund Spensers Poetry, 3rd edn (New York: Norton, 1993), p. 672.
2. PERCEPTION 7
Furthermore, not even all journalistic criticism has this aim. Some
reviews (for example, of a theatre production that has already nished)
may simply be written to inform the reader of what was of interest in the
work and why it was of interest. It may be of great interest to know how a
certain director adapted a certain play, in what respects it was successful or
unsuccessful, etc. For works that the reader can no longer experience, the
aim of criticizing them cannot be to help the reader to decide whether to
choose to experience them. But it is clearly not pointless to write criticism
of such works.
2. Perception
Many hold that critics describe works in order to cause their readers to
perceive features of the work. Stuart Hampshire, Frank Sibley, Michael
Baxandall, and many others have held this view.5 The main argument
offered in support of it is presented by Arnold Isenberg. Isenbergs paper
has exercised great inuence.6 It is widely regarded as the classic defence of
particularism in aesthetics. And his view that the aim of critical description
is to cause perception has been not only supported but taken for granted
by many aestheticians.
5
See Stuart Hampshire, Logic and Appreciation, in William Elton, ed., Aesthetics and
Language (Oxford: Blackwell, 1959), p. 165; Frank Sibley, Aesthetic Concepts, in Approach to
Aesthetics: Collected Essays on Philosophical Aesthetics, ed. John Benson, Betty Redfern, and
Jeremy Roxbee Cox (Oxford: Oxford University Press, 2001), pp. 1520; Frank Sibley,
Aesthetic and Non-Aesthetic, in Approach to Aesthetics, p. 38; Michael Baxandall, The
Language of Art History, New Literary History 10/3 (1979): 455; James Shelley, The
Character and Role of Principles in the Evaluation of Art, British Journal of Aesthetics 42/1
(2002): 51.
6
See, e.g., Mary Mothersill, Critical Reasons, The Philosophical Quarterly 11/42 (1961):
7478; John Casey, The Language of Criticism (London: Methuen, 1966), pp. 172173; Mary
Mothersill, Beauty Restored (Oxford: Oxford University Press, 1984), p. 354; Ted Cohen, On
Consistency in Ones Personal Aesthetics, in Jerrold Levinson, ed., Aesthetics and Ethics: Essays
at the Intersection (Cambridge: Cambridge University Press, 1998), pp. 112113; Joel
J. Kupperman, Value . . . and What Follows (Oxford: Oxford University Press, 1999), p. 53;
Dominic McIver Lopes, Sight and Sensibility: Evaluating Pictures (Oxford: Oxford University
Press, 2005), pp. 105106; Robert Hopkins, Critical Reasoning and Critical Perception, in
Matthew Kieran and Dominic McIver Lopes, eds, Knowing Art: Essays in Aesthetics and
Epistemology (Dordrecht: Springer, 2007), pp. 137153; James Shelley, Critical Compatibi-
lism, in Matthew Kieran and Dominic McIver Lopes, eds, Knowing Art: Essays in Aesthetics
and Epistemology (Dordrecht: Springer, 2007), pp. 125136; Keith Lehrer, Art, Self and
Knowledge (Oxford: Oxford University Press, 2012), pp. 21, 29.
Isenberg begins his argument by examining

a theory of criticism . . . which divides the critical process into three parts.
There is the value judgment or verdict (V): This picture or poem is good.
There is a particular statement or reason (R): because it has such-and-such a
quality. And there is a general statement or norm (N): and any work
which has that quality is pro tanto good.7
Isenberg agrees that critics make value judgements. He also agrees that
when we speak of justifying or giving reasons for our critical judg-
ments, we refer to something which . . . does go on in the world.8 And he
agrees that the value judgement is in some sense conditional upon R.9
His objection to the theory of criticism he describes is that, though V is
in some sense conditional upon R, the truth of R never adds the slightest
weight to V .10 What we describe as giving reasons in support of our value
judgements is not a case of making statements whose truth supports our
judgements.
He uses this claim to argue that getting the reader to perceive is the aim
of critical description. The overarching structure of his argument is as
follows:
(A) The truth of R offers no support for the value judgement.

(B) If the truth of R offers no support for the value judgement, the
best explanation of Rs function is that using it is a way of getting
the reader to perceive.
Therefore,
(C) The best explanation of Rs function is that using it is a way of

getting the reader to perceive.
Isenberg devotes most effort to defending (A). This defence involves
denying that critics do rely on norms of the kind on which, according to
the simple theory he rejects, they rely.11
7
Arnold Isenberg, Critical Communication, Philosophical Review 58/4 (1949), p. 330.
8
Ibid., p. 333 n. 3.
9
Ibid., p. 331.
10
Ibid., p. 338.
11
James Shelley has recently emphasized that Isenberg is not arguing that there are no true
critical norms, but that critics do not appeal to any critical norms (Shelley, Critical Compa-
tibilism, pp. 128129). This is true. However, Isenberg appears to be arguing that critics do
not appeal to any critical norms by arguing that the only norms to which they could possibly
be appealing are untrue.
2. PERCEPTION 9
His argument for (A) appeals to the following passage by the art critic
Ludwig Goldscheider. Discussing El Grecos The Burial of Count Orgaz,12
Goldscheider writes:
Like the contour of a violently rising and falling wave is the outline of the four
illuminated gures in the foreground: steeply upwards and downwards about
the grey monk on the left, in mutually inclined curves about the yellow of the
two saints, and again steeply upwards and downwards about . . . the priest on
the right. The depth of the wave indicates the optical center; the double curve
of the saints yellow garments is carried by the greyish white of the shroud down
still farther; in this lowest depth rests the bluish-grey armor of the knight.13
Isenberg comments:
This passagewhich, we may suppose, was written to justify a favorable
judgment on the paintingconveys to us the idea of a certain quality
which, if we believe the critic, we should expect to nd in a certain painting
by El Greco. And we do nd it: we can verify its presence by perception. . . .
But the same quality (a steeply rising and falling curve, etc.) would be found
in any of a hundred lines one could draw on the board in three minutes. It
could not be the critics purpose to inform us of the presence of a quality as
banal and obvious as this.14
The point, it seems, is that there is no true norm to the effect that any work
which has a steeply rising and falling curve, etc., is pro tanto good. So the
argument appears to be:
(A1) If the truth of R supports the value judgement, then there are
true norms to the effect that any work with the property attrib-
uted by R is pro tanto good.
(A2) But there are no true norms to this effect.
Therefore,
(A) The truth of R offers no support for the value judgement.
That result raises a problem: as long as we have no alternative interpret-
ation of the import and function of R, we must assume either that R is
perfectly arbitrary or that it presupposes and depends on some general
claim.15 The above argument seems to show that R does not presuppose
12
For an image of this work, see <http://en.wikipedia.org/wiki/The_Burial_of_the_
Count_of_Orgaz> accessed 14 June 2012.
13
Quoted in Isenberg, Critical Communication, p. 335.
14
Ibid., pp. 335336.
15
Ibid., p. 335.
and depend on some general claim, so it seems that R is perfectly

arbitrary. This leads Isenberg to his alternative interpretation of the
function of R.
He continues the discussion of Goldscheider in this way:
It seems reasonable to suppose that the critic is thinking of another quality, no idea
of which is transmitted to us by his language, which he sees and which by his use of
language he gets us to see. This quality is, of course, a wavelike contour; but it is not
the quality designated by the expression wavelike contour. Any object which
has this quality will have a wavelike contour; but it is not true that any object
which has a wavelike contour will have this quality. . . . Now the critic . . . gives
us directions for perceiving, and does this by means of the idea he imparts to us,
which narrows down the eld of possible visual orientations and guides us in the
discrimination of details, the organization of parts, the grouping of discrete objects
into patterns. . . . It is a function of criticism to bring about communication at the
level of the senses, that is, to induce a sameness of vision, of experienced content.16
Isenbergs positive account thus has two components. There is the infer-
ence to the best explanation of the aim of criticism in providing R:
The aim of criticism in providing R is to get the reader to perceive
certain properties.
Isenberg defends the view that this is the best alternative explanation of Rs
function by claiming that reading criticism, otherwise than in the pres-
ence, or with direct recollection, of the objects discussed is a blank and
senseless employment.17
However, his view has a second component independent of this one.
Isenberg also claims that
The properties criticism aims to get the reader to perceive are not the
properties attributed by the critic in providing R.
One might have thought the aim is to cause the reader to see the property
the critic attributes (even though the truth of the claim that the object has
that property does not support the value judgement). But this is not
Isenbergs view. He is explicit that criticism does not actually designate
16
Ibid., p. 336.
17
Ibid., p. 337. Compare Baxandall: The work of art we discourse on is to some extent
present or available, if only in reproduction or in the memory or even more marginally as a
visualization derived from knowledge of other objects of the same class (The Language of
Art History, p. 455).
2. PERCEPTION 11
the qualities to which it somehow directs our attention.18 Presumably, the

argument for this additional claim is that, as he says, the property attributed
by the critic in providing R is banal and obvious. Not only can it not be
the critics purpose to inform us of the presence of such a banal and
obvious property: it also cannot be her purpose to make us see it.
I will take issue with both components of Isenbergs view. However,
there is some truth in what Isenberg says. In fact, it is obvious that critics
sometimes try to draw our attention to, or get us to perceive, various
properties of a work. Critics include reproductions of paintings in their
work to get us to look them, and quote lines of poetry to get us to read
them. Sometimes critics explicitly instruct us to look at certain features.
Nevertheless, Isenbergs account is awed.
The rst problem is that, when critics try to get us to see a certain quality,
they do normally designate this quality. It is worth examining how Isenberg
is led to deny this. Other philosophers have also held this. Mary Mothersill,
for instance, says that Isenberg has shown that, in a certain sense, critics
do not . . . mean what they say .19 But even in the passage of criticism
Isenberg uses to support his point, the critic designates the quality he gets
us to see.
The argument for this is simple. The quality Goldscheider gets us to see is
the outline of the four illuminated gures in the foreground. He designates
this quality with the denite description the outline of the four illuminated
gures in the foreground. Therefore, he designates the quality he gets us
to see.
Isenberg claims that the quality Goldscheider gets us to see is not the
quality designated by the expression wavelike contour. This may well be
true. But even if it is true, it does not establish the conclusion that criticism
does not actually designate the qualities to which it somehow directs our
attention. Goldscheider designates the quality to which he directs our
attention. It is just that he does not designate it with the expression
wavelike contour. He designates it with a different expressionnamely,
the outline of the four illuminated gures in the foreground.
18
Isenberg, Critical Communication, p. 339. Elsewhere he says that the grounds to
which [the critic] is really appealing are not the same as those which he explicitly states or
designates (ibid., pp. 343344).
19
Mothersill, Critical Reasons, p. 77. Compare Baxandall, The Language of Art
History, pp. 455456, and Lehrer, Art, Self and Knowledge, p. 21.
Goldscheider does not use the expression wavelike contour in the

passage Isenberg quotes: he uses the predicate x is like the contour of a
violently rising and falling wave. I assume Isenbergs claim is a concise way
of making the point that the quality he gets us to see is not designated by
that predicate. And that claim is plausible. I assume that, in Isenbergs
usage, what is designated by an expression is what is denoted or referred to
with (or by) that expression. If anything is designated, in this sense, by the
predicate x is like the contour of a violently rising and falling wave, then
it is the quality being like the contour of a violently rising and falling wave. So
Isenberg is clearly right that the quality Goldscheider gets us to see is not
designated by this predicate. The outline of the gures in the foreground
(the quality he gets us to see) is not identical to the quality, being like the
contour of a violently rising and falling wave. But this is an uninteresting result.
For Goldscheider does designate the quality he gets us to seewith a
different expression.
What is certainly true is that Goldscheider characterizes the quality he
gets us to see with the predicate x is like the contour of a violently rising
and falling wave. He characterizes, but does not designate, the outline of
the gures with this predicate.20 What is also true is that Goldscheider gets
us to see the outline, and gets us to see that the outline has the quality of
being like the contour of a violently rising and falling wave. He gets us to
see one quality (the outline), and he gets us to see that it has a further
quality (being like the contour of a violent wave). Isenberg does not make
the distinction between designation and characterization, or the distinc-
tion between seeing and seeing-that. But if we make these distinctions,
then another thesis suggests itself.
Isenberg might mean that Goldscheider does not characterize the out-
line as having the quality he gets us to see that it has. Isenberg believes that
it could not be the critics purpose to inform us of the presence of a quality
as banal and obvious as being like the contour of a violently rising and falling
wave. This is the quality he thinks could be possessed by any of a hundred
lines one could draw on the board in three minutes. And this is the quality
Goldscheider characterizes the outline as having. So perhaps Isenberg
20
Again, assuming that what is designated by an expression is what is denoted or referred
to with (or by) that expression. If Isenberg thinks what is designated by an expression is what is
characterized with or by that expression, then (contra Isenberg) Goldscheider does designate
the outline with x is like the contour of a violently rising and falling wave.
2. PERCEPTION 13
means that Goldscheiders purpose is not to get us to see that the outline
has this quality. Rather, his aim is to get us to see that the outline has some
other quality, which he does not characterize it as having.
This claim is implausible for different reasons. First, even if it is not the
critics purpose merely to inform us that the outline is wavelike, it does not
follow that it is not his purpose to get us to see that the outline is wavelike.
It may be interesting for the critics readers to see that this is the case, but
not especially interesting merely to be informed that it is the case.
Second, even if one can easily draw hundreds of wavelike lines on a
blackboard, it does not follow that the fact that the outline of the gures in
El Grecos painting is like a violent wave is banal or obvious. How easily
we can draw a wavelike line has nothing to do with whether this fact about
the El Greco is interesting or banal. If El Greco had given one of the
gures a square head, that certainly would be an interesting fact, even
though squares are easily reproduced on the blackboard, too.
So we should not accept Isenbergs claim that critical communication
differs from ordinary communication in the ways he says it doeseven in
those cases in which the critic is trying to get her readership to see.
Goldscheider gets us to see the outline of the gures, and he gets us to
see that this outline is like a violent wave. He designates the quality he gets
us to see, and he characterizes it as having a quality he gets us to see that
it has.
Further problems attach to the other component of Isenbergs conclu-
sion: the view that getting us to perceive properties is the critics aim. The
rst point to highlight is that Isenbergs argument only supports the
conclusion that this is the point of making R-type statements. As
I have noted, his argument is presented as an examination of the function
of R (rather than that of V or N), and it only supports such a conclusion as
regards R-type statements.
This point is worth emphasizing. The upshot of it is that, for all Isenberg
has shown, critics do support their value judgementsonly not with
R-type statements. They may, for example, support some of their evalu-
ations with other evaluations. As I will explain further below, Isenberg
regards R-type statements as descriptive, non-evaluative claims. So even if
Isenbergs argument about R-type claims were awless, it would not show
that one value judgement cannot support another.
Discussions of Isenberg tend to overlook the fact that the scope of his
thesis is limited in this way. For example, Daniel Kaufman and Nol
Carroll have recently objected to Isenberg on the grounds that such claims
as Roger van der Weydens Descent from the Cross displays Christs human-
ity well or Harold Lloyds Safety Last contains many successful pratfalls
support, respectively, such value judgements as Rogers Descent from the
Cross is good21 and Safety Last is good (pro tanto).22
Kaufman and Carroll are wrong to claim that these are counterexamples
to the claim Isenberg makes. These are cases of supporting one evaluation
with another: to say that Christs humanity is displayed well and that the
pratfalls are successful is to evaluate. Isenbergs problem, however, is
whether evaluations can be supported by descriptive, non-evaluative
claims.
It is true, however, that it is possible to support one value judgement
with another. It would be absurd to hold that the truth of the claim, Safety
Last contains many successful pratfalls, cannot add the slightest weight to
the claim that it is a good slapstick comedy. So not only does Isenbergs
thesis not imply that value judgements cannot be supported by other value
judgements: it is clear that they can indeed be so supported, and that they
often are in criticism.
There is another little-noted consequence of the fact that Isenbergs
thesis is restricted to R-type statements: Isenberg needs to tell us what an
R-type statement is. If his thesis about their function is true, they cannot
really be those statements whose function is to support value judgements,
even though this may appear to be their function. What, then, are they,
according to Isenberg?
Isenberg says that R is a statement describing the content of an art
work.23 He contrasts descriptive statements with evaluative statements,
as many philosophers have done. He acknowledges that this is an
idealization:
V and R, it should be said, are often combined in sentences which are at
once normative and descriptive. If we have been told that the colors of a
certain painting are garish, it would be astonishing to nd that they were all
very pale and unsaturated; and to this extent the critical comment conveys
21
Daniel A. Kaufman, Critical Justication and Critical Laws, British Journal of Aesthetics
43/4 (2003): 399.
22
Nol Carroll, On Criticism (New York: Routledge, 2009), p. 167.
23
Isenberg, Critical Communication, p. 331.
2. PERCEPTION 15
information. . . . We shall be concerned exclusively with the descriptive

function of R.24
As I mentioned above, Isenberg also says that the critics value judgement
is in some sense conditional upon R.
This is all he tells us by way of characterizing R. And this creates a
number of problems. Suppose Isenberg means that an R-type claim is any
descriptive, non-evaluative critical claim made about a work (including
the descriptive, non-evaluative claim made with sentences which are at
once normative and descriptive) upon which the value judgement is in
some sense conditional. If that is so, then it is not at all plausible that the
function of every such claim is to get the reader to perceive properties.
First, various kinds of review provide counterexamples. Reviews of
works the critics readers cannot perceive, such as a theatrical production
that has completed its run, are counterexamples. So are negative reviews
discouraging the reader from perceiving the work. The aim of making the
descriptive claims in such reviews cannot be to cause the reader to perceive
features of the work. Reviews like these sometimes do reproduce the
work or parts of it, to enable the reader to perceive certain properties. But
we do not necessarily fault the review if it does not do this, even if it
contains many non-evaluative descriptions on which a value judgement is
somehow conditional. Indeed, it is not always possible to enable the reader
to perceive every single property ascribed with such descriptions in a
review. Reproducing enough of a novel, lm, or live performance to
allow the reader to perceive every property ascribed in the review is often
impractical or impossible. But the review may be none the worse for that.
Second, there are several kinds of descriptive critical statement on which
a value judgement is sometimes conditional, but which do not always have
causing perception or directing perceptual attention as their aim. For
example, critics sometimes try to persuade a reader that something is
true in the world of the work: that Hamlet only feigns madness; that the
governess in The Turn of the Screw did see ghosts; or that Ugolino in the
Inferno ate his children. Beardsley has called this the elucidation of a
representational artwork.25 A value judgement may well be conditional
24
Ibid. The example of garish indicates that, when Isenberg says R describes the content
of an artwork, he cannot mean that it always gives the representational content of the work.
25
See Monroe C. Beardsley, Aesthetics: Problems in the Philosophy of Criticism, 2nd edn
(Indianapolis: Hackett, 1981), pp. 242247, 278280.
on the truth of such a claim. A critic of The Turn of the Screw may regard the
story as better if the governess is suffering from the effects of repression
than it would be if she really saw ghosts. But the aim of an elucidation is
often to cause the reader to believe that something is true in the world of
the work, not to guide perception.
Other examples are easily found. The aim of some description may be
to get the readership to interpret the work in a certain way, to get them to
believe that the work belongs to a certain artistic category or genre, or to
inform them of something about the works historical or cultural context
(e.g., that it had a certain religious or political function). A value judge-
ment can be conditional on the truth of any such claim, even though the
aim is not to guide perception.
So if an R-type claim is any descriptive, non-evaluative claim on which
the value judgement is somehow conditional, Isenbergs thesis about the
function of such claims is false. But if an R-type claim is not just any claim
of this kind, then it is not clear which claims are R-type claims. Isenberg
offers no more clarication of this than what I have quoted. So Isenbergs
position is problematic. What he writes is unclear, and there do not appear
to be clarications of what he means that render his claims plausible.
It is certainly true that the point of some art-critical statements is to get
the critics readership to perceive properties, or to perceive that the work
has certain properties. In fact, this is not only true, but obvious. Isenbergs
thesis is the more ambitious claim that every critical statement of a certain
very general kind has this function. Without a characterization of the kind
of statement he is talking about that makes his thesis plausible, we are
left with the more modest claim that some critical statements have this
function.
3. Evaluation
Carroll argues that the aim of criticism is to provide a sound justication
for an evaluation of a work. As he puts it, criticism, properly so-called, is
not merely a matter of evaluating an artworkof giving it a thumbs-up or
thumbs-down. Critics are expected to supply reasonsindeed, good
reasonsin support of their evaluations.26 Other critical operations,
26
Carroll, On Criticism, p. 13. See also pp. 15n, 18n, 19, 4347. Although he does not say
so explicitly, Carroll presumably holds that the aim is to provide a sound justication for a
correct, plausible, or convincing evaluation.
3. EVALUATION 17
such as description, classication, contextualization, elucidation, inter-

pretation, analysis . . . typically function as grounds for evaluation.27
His argument for this is that evaluation is the feature that sets criticism
off from comparable discourses28 about art:
For example, certain forms of historical discourse about art will mobilize
description, elucidation, contextualization, classication, interpretation, and/
or analysis. For example, an economic historian of art might describe and
analyze Rembrandts tendency to have large swaths of black in his pictures in
order to explain that in this way Rembrandt was able to undertake, for the
purpose of maximizing his prot margin, a very large number of commis-
sions, since those empty, unarticulated, black spaces of canvas could be
painted very quickly. . . . The notion that, additionally, criticism engages in
evaluation provides us with a ready differentia or rationale, which suggestion
is also amply supported in everyday speech.29
Thus, evaluation is an essential feature of criticism such that if a piece of

discourse lacks explicit or implicit evaluation, it would not qualify as
criticism.30
Carroll holds that reasoned evaluation is possible because, contra Isen-
berg, there are indeed principles of critical evaluation. But they do not take
the form: Every artwork with such-and-such a characteristic is good ( pro
tanto). Genuine critical principles are not claims about all artworks what-
soever. Rather, they are claims about artworks of a given category.31
According to Carroll, an artwork of a given category that fulls the artistic
purposes or function of artworks of that category is, normally, good ( pro
tanto). If this is so, he argues, then there are many true critical principles.
For example, the principle he uses to support his judgement of Lloyds
Safety Last is: Given the purpose or function of slapstick comedy, slapstick
comedies that contain many successful pratfalls, all other things being
equal, are good (pro tanto).32
27 28
Ibid., p. 84. Ibid., p. 18n.
29 30
Ibid., pp. 1617. Ibid., pp. 4344.
31
Compare Kaufman, Critical Justication and Critical Laws, and Daniel A. Kaufman,
Normative Criticism and the Objective Value of Artworks, The Journal of Aesthetics and Art
Criticism 60/2 (2002): 151166. See Richard Shusterman, Wittgenstein and Critical
Reasoning, Philosophy and Phenomenological Research 47/1 (1986): 9495 for a brief discussion
of philosophers who have supported this idea (e.g., Aristotle, Collingwood, and Weitz) and of
critics who have applied it (e.g., Addison, Johnson, and Coleridge).
32
Carroll, On Criticism, p. 167. I take it that, by other things being equal, Carroll means
normally. He writes: The Isenbergian may grumble about my Harold Lloyd example,
To illustrate such category-relative evaluation, Carroll uses a review by

the dance critic Joan Acocella. Acocella is reviewing a production of Mark
Morriss Mozart Dances, a work of modern abstract choreography. She
writes: Why is he so popular? One reason, I think, is that he gives people
the modern pleasure of seeing abstract work without leaving them scratch-
ing their heads over what it was about. Though he may not have a story on
the surface, he always has one underneath, in the form of movement
motifs.33 She then describes these movement motifs and the story they
suggest. Carroll says of her review:
Acocella enables her readers to understand her grounds for maintaining that
Morris has subtly articulated the outline of a story. This, in turn, she maintains,
gives the viewer a way into a dance of the sort that is often confusing to
audiences, presupposing, as she does, that a narrative, typically, enhances
accessibility. . . . She is not supposing that a suggested narrative is a good-making
feature of every artwork. Rather, she is restricting her claim to works of modern
abstract choreography and saying that, all things being equal, it is a good-making
feature in such works.34
As I have said, Carroll is right to hold both that it is possible to provide

support for a critical evaluation and that critics do this. However, like
Beardsley and Isenberg, he overstates the prevalence of the aim he identies.
First, some good criticism merely provides evaluations of artworks
without supporting them. For example, many entries in The Penguin
Guide to Recorded Classical Music simply rate recordings on a scale of one
star to ve stars without further comment.35 This is bad criticism if the
evaluation is wrong, or if the critic herself is not justied in giving it the
rating she gives it. But it is not necessarily bad criticism if she does not
provide a justication for the verdict. The readership being addressed may
be one that is only interested in the verdict of a good critic, rather than a
justication for it. They may look to the critic, as Beardsley held, to guide
suggesting that it can be the case that under some strange conditions a particular pratfall might
not contribute to the goodness of a slapstick comedy. But that is why the ceteris paribus clause
has been added to our formulation (ibid., p. 168). The mention of strange conditions
suggests that, in Carrolls usage, all other things being equal means normally.
33
Joan Acocella, Mozart Moves, The New Yorker, 20 August 2007, <http://www.
newyorker.com/arts/critics/dancing/2007/08/20/070820crda_dancing_acocella> accessed
14 June 2012.
34
Carroll, On Criticism, pp. 154155, 168169.
35
See Ivan March et al., The Penguin Guide to Recorded Classical Music 2010 (London:
Penguin, 2009).
3. EVALUATION 19
their choices, and for this purpose they may only be interested in her
verdict.
Second, Carroll gives us no good reason to believe that the point of
other critical operations, such as interpretation and classication, is always
to support an evaluation. Some of the counterexamples to Isenbergs view
also serve as counterexamples to Carrolls view. For example, it seems that
a convincing or plausible elucidation, in Beardsleys sense,36 of a contro-
versial question about the world of a work could be good criticism even if
it is not used to support an evaluation. If a critic came up with a convincing
answer to the question of why Hamlet procrastinated, that in itself would
be excellent criticism. It would not be awed if the elucidation was not
presented in support of an evaluation.
Carroll responds to the objection that much good criticism (e.g., of
canonical works) appears not to include evaluation:
Since it is through operations like interpretation, description, analysis,
classication, contextualization, etc. that one grounds ones evaluations,
when it comes to criticizing canonical works, if, for example, these routines
bring to readers a view of the unity, complexity, sophistication, and
wisdom of the work, then it may not be necessary to round off ones
critical remarks with overt commendation. The recommendation may be
implicit. But this only shows that the evaluative moment in criticism need
not be explicit.37
This reply is weak because at most it establishes that, when there is no

explicit evaluation, there may be an implicit evaluation. This is true. But
that does not answer the objection that it seems possible to produce good
criticism that contains neither an explicit nor an implicit evaluation.
It seems, therefore, that Carroll must rely on his argument that a piece of
criticism must contain evaluation, because otherwise criticism would be
indistinguishable from similar discourses about art, which can also contain
the other operations he mentions. But this argument is also unsuccessful. If
the claim is that no piece of writing about an artwork lacking evaluation
of the work could be criticism, then the elucidation of Hamlet I suggested
seems simply to be a counterexample. Such an elucidation would gener-
ally be acknowledged to be criticism. The fact that such a piece of writing
36
Carroll himself uses elucidation in a different sense. See Carroll, On Criticism,
pp. 108110.
37
Ibid., p. 21.
appears to be criticism is a datum that any account of criticism must respect

or explain away. Carrolls theory does not do this. Carroll is right that the
challenge that confronts the skeptic regarding the claim that criticism is
essentially concerned with evaluation is to propose another distinctiona
more effective and more persuasive dividing line between criticism and
comparable modes of discoursethan the one I am advancing.38 I will
answer this challenge in Chapter 2. But we do not have to answer it to
show that the claim that all criticism contains implicit or explicit evaluation
is mistaken.
So it seems there can be good criticism containing evaluation but
no support for it, and good criticism containing no evaluation. There is
a third reason to doubt that providing support for an evaluation is always
the aim of criticism. Even when criticism contains evaluations, and even
when what the critic says could be used to support those evaluations, it is
not always plausible that the point of the criticism is to support those
evaluations.
For example, often the evaluations implicit or explicit in criticism
of well-known works are already very well supported. This makes it
implausible that the point of criticizing those works again is to provide
further support for those evaluations. The readership being addressed may
not need to be persuaded, and may also be justied in believing, that King
Lear is a masterpiece. Their experience of the work may justify them in
believing this, and past criticism may have provided ample support for this
evaluation. It may therefore be unnecessary to provide further support for
it. But it is not pointless to criticize such works further.
Carroll might reply that, although this may be true of judgements of
the overall value of a well-known work (e.g., the claim that the work is a
masterpiece), those are often not the kinds of evaluation critics argue for.
He writes that the category-relative evaluation of an artwork is a pro tanto
evaluation.39 For instance, the evaluation in his Harold Lloyd example is
not a judgement of the comedys overall value. It takes the form: This
work is good ( pro tanto).40 So Carroll might reply that, although it is often
unnecessary to provide support for judgements of the overall value of
well-known works, it is often necessary to provide support for various pro
tanto evaluations of such works.
38 39 40
Ibid., p. 17. Ibid., p. 180. Ibid., p. 167.
3. EVALUATION 21
What, however, does good pro tanto mean? This expression is fre-
quently used in philosophy without explanation. Some writers say they
use good pro tanto to refer to somethings being good in some respect or
good in a way.41 This clearly will not help Carroll. A critics readers also
frequently have plenty of evidence that King Lear is good in a way
indeed, that it is good in many ways. The point of criticizing such works
cannot be to provide evidence for this extremely modest claim.
The obvious alternative is to construe good pro tanto as meaning good
in that way or good in that respect. The dictionary meaning of pro tanto
is: to such an extent, to that extent. Good to that extent or good to
such an extent seem to be equivalent to good in that respect or good in
that way. If this is the form that a pro tanto evaluation takes, then the claim
that something is pro tanto good, unlike the claim that there is some respect
or way in which a certain thing is good, can only be understood in a
context that makes it clear what way or respect is being referred to. Pro
tanto good is comparable to a phrase containing a demonstrative referring
to a respect or way in which the work is being said to be good. For
instance, the claim that Safety Last is pro tanto good, as it occurs in Carrolls
example, amounts to the claim that its many successful pratfalls are a good
thing about it. So perhaps the claim Carroll would make is this: when the
overall value of the work is already well established, the aim of criticism is
to support the claim that the work is good (or bad) in such-and-such a
respect or way.
It is certainly true that, even when we have ample justication for the
belief that a work is a masterpiece, and for the belief that there are many
ways in which it is good, it remains of great interest to learn what is good
or bad about it. We can continue to learn such things long after we have
established a works overall value. It is also true that much criticism
involves telling us what is good or bad about a work.
What is doubtful is that the aim of criticism is always to provide support
for the claim that such-and-such is a good or bad thing about a work.
Much good criticism simply asserts or implies that this or that is a good
thing about the work. This can be of great interest. But the criticism often
does not supply reasons to back up the claim that this or that is a good
thing about the work.
41
Christine Tappolet, Through Thick and Thin: Good and Its Determinates, Dialectica
58/2 (2004): 210.
For example, Samuel Johnson, discussing Othello, writes: The scenes

from the beginning to the end are busy, varied by happy interchanges and
regularly promoting the progression of the story; and the narrative in the
end, though it tells but what is known already, yet is necessary to produce
the death of Othello.42 William Hazlitt writes: Macbeth and Lear, Othello
and Hamlet, are usually reckoned Shakespeares four principal tragedies.
Lear stands rst for the profound intensity of the passion; Macbeth for the
wildness of the imagination and the rapidity of the action; Othello for
the progressive interest and powerful alternations of feeling; Hamlet for the
rened development of thought and sentiment.43 Here we have several
claims about what is good about some acknowledged masterpieces. John-
son claims that it is a good thing about Othello that the scenes are busy and
varied by happy interchanges, and that they regularly promote the pro-
gression of the story. Hazlitt claims that the profound intensity of the
passion is a good thing about Lear, and that its rened development of
thought and sentiment is a good thing about Hamlet.
But Johnson and Hazlitt do not supply us with any reasons in support of
their claims that these are good things about these plays. Moreover, it is
clear why they do not: it is unnecessary to provide support for the claim
that these are good things about the plays. It would certainly be necessary
to provide support for the claim that it is a bad thing about Othello that its
scenes are varied by happy interchanges and regularly promote the pro-
gression of the story. But it seems no more necessary to support the pro
tanto evaluation that this is a good thing about Othello than it would be to
support the all-things-considered evaluation that Othello is a great tragedy.
An important difference between these evaluations is that a reader may not
have entertained the thought that this is one good thing about Othello,
whereas she is almost certain to have entertained the thought that it is a
great tragedy. But it is no more necessary to support the one claim than it is
to support the other.
So if this is what a pro tanto evaluation is, then we should agree with
Carroll that much good criticism contains many such evaluations. But
even in such criticism, it is implausible that the aim is always to provide
42
Samuel Johnson, Selections from the Notes to the Edition of Shakespeares Plays, in
Samuel Johnson on Shakespeare, ed. H. R. Woudhuysen (London: Penguin, 1989), p. 247.
43
William Hazlitt, Macbeth, in Selected Writings, ed. Jon Cook (Oxford: Oxford
University Press, 1991), p. 336.
4. EXPLANATION 23
reasoned pro tanto evaluations. Often, it will indeed be necessary to provide

justications of ones pro tanto evaluations. But it may be enough that ones
criticism provide numerous correct and insightful pro tanto evaluations of
the work. Johnson and Hazlitts criticism is awed if what they say are
good things about the plays are not good things about them. But their
criticism does not seem to be awed on account of not including justica-
tions of their pro tanto evaluations.
4. Explanation
It is often said that critics explain why a work has the value or the aesthetic
properties it has. For example, Arthur Danto holds that criticism is a kind
of education, and that the aim of criticism is to provide explanations of a
certain kind: Education is not training people to say, Mitchells Hemlock is
better than Mardens Cold Mountain. It is rather explaining how and why
each of them is good in its own way.44 Sibley discusses explanation in
criticism at greater length than many. He writes:
A critic frequently tries, as one of his central occupations, to say why a picture
is unbalanced, or what gives a complex work its grace, unity, or serenity. . . .
He may mention a concentration of blues and greys as responsible for the
unity of tone, certain wavy lines as giving a restless quality, a change in key as
giving a sombre or indecisive character.45
Sibley describes explanation as one of the central activities of critics,46

though he does not claim that providing explanations is the aim of all
criticism.47
There are, indeed, several kinds of fact that critics explain, apart from
those mentioned by Danto and Sibley. In addition to facts about a works
value and aesthetic properties, critics also explain facts about the world of a
work. Sometimes these facts are explained by appealing to or postulating
other facts about the world of the work. I have already mentioned the
44
Arthur C. Danto, The Fly in the Fly Bottle: The Explanation and Critical Judgment of
Works of Art, in Unnatural Wonders: Essays from the Gap Between Art and Life (New York:
Farrar, Straus and Giroux, 2005), p. 361. See also Monroe C. Beardsley, Critical Evaluation,
in The Aesthetic Point of View, p. 321.
45
Sibley, Aesthetic and Non-Aesthetic, p. 36.
46
Ibid., p. 37.
47
Indeed, Sibley regards helping people to see and judge for themselves that things have
aesthetic qualities as a more important critical task (ibid., p. 38).
example of attempting to explain why Hamlet procrastinated, which

might proceed by postulating facts about Hamlets psychology. Some-
times, by contrast, facts about the world of the work are explained by
citing other kinds of fact. Peter Lamarque holds that an incident in a ction
can sometimes be explained by citing a fact about the contribution the
episode makes to the completed artistic structure of the work, as when a
critic explains why Tess killed Alec dUrberville by saying that it signies
an end to her journey.48 Similarly, a commonplace about Henry IV, Part
One is that Hotspur is impetuous and hot-headed because this makes him a
foil to the calculating Prince Hal.
Critics also sometimes explain why a work leaves a certain impression or
elicits a certain response. Thomas De Quincey does this in his essay, On
the Knocking at the Gate in Macbeth. De Quincey writes that, ever since
his childhood,
the knocking at the gate, which succeeds to the murder of Duncan, produced
to my feelings an effect for which I never could account: the effect wasthat
it reected back upon the murder a peculiar awfulness and a depth of
solemnity: yet, however obstinately I endeavoured with my understanding
to comprehend this, for many years I never could see why it should produce
such an effect.49
His answer is that the knocking at the gate marks the re-establishment of
the goings-on of the world in which we live after the murder, and this
makes us profoundly sensible of the awful parenthesis that had suspended
them.50 De Quincey argues that if the reader reects on cases in which
ordinary goings-on are dramatically interrupted and then resumed, he will
be aware that at no moment was his sense of the complete suspension and
pause in ordinary human concerns so full and affecting as at that moment
when the suspension ceases, and the goings-on of human life are suddenly
resumed.51 This is so, he suggests, when one sees a woman revive after
fainting, or hears the rattling wheels of the carriage break the silence of the
funeral procession of a great national hero; and the same kind of effect, he
argues, occurs in Macbeth.52
48
Peter Lamarque, The Philosophy of Literature (Oxford: Blackwell, 2009), p. 207.
49
Thomas De Quincey, On the Knocking at the Gate in Macbeth, in Confessions of an
English Opium-Eater, and Other Writings, ed. Grevel Lindop (Oxford: Oxford University Press,
1985), p. 81.
50 51 52
Ibid., p. 85. Ibid., p. 84. Ibid.
4. EXPLANATION 25
Not only are there several kinds of fact that critics explain. There are
also several kinds of explanation that critics provide. Bas van Fraassen
claimed that an explanation is an answer to a why-question.53 Explan-
ations of why something is the case loom large in aestheticians examples of
the kinds of explanation critics provide. But as several philosophers have
pointed out, not every explanation is an explanation of why something is
the case.54 And the most important explanations in a piece of criticism are
not always explanations of why something is the case.
Some explanations provide answers to how-questions. For example,
John Summerson explains how the Romans solved the problem of
combining the Ionic order of Greek temple architecture with the arch-
and-vault system needed for major Roman buildings like the Colosseum
(see Figure 1.1):
Here you have a grammatical construction which is a pretty complete thing.
It is controlled by an Ionic order which obeys nothing but its own traditional
aesthetic rules. The shape and size of the piers behind the columns and of the
arch, on the other hand, have come about through the exigencies of con-
venience and construction. The two disciplines have got to meet each other
harmoniously and I think we may agree that they do. The pedestal moulding
of the order ranges with the sill height of the arched gallery. The impost of
the arch strikes the columns a little above half their height and the arch sits
comfortably between the columns and the architrave above. If this arrange-
ment is satisfactory it has been achieved by a very careful balancing of needs,
the aesthetic dictatorship of the Ionic order and the practical needs of the
building as a thing of use.55
So too, some explanations provide answers to what-questions. You can

explain what a word means, what a gesture signies, or what an image
symbolizes. Such explanations, unlike the previous ones we have con-
sidered, do not take facts as their explananda. For example, when you
explain what a word means, what you explain is its meaning, not the fact
that it has a meaning, or any other fact. Clearly, explanations of this kind,
too, are prevalent in criticism.
53
Bas C. van Fraassen, The Scientic Image (Oxford: Oxford University Press, 1980),
p. 134.
54
See, for example, Sylvain Bromberger, Why-Questions, in R. Colodny, ed., Mind and
Cosmos: Essays in Contemporary Science and Philosophy (Pittsburgh: University of Pittsburgh
Press, 1966), pp. 8990.
55
1980), pp. 2122.
Figure 1.1. C. Elizabeth Grant. Drawing after Antoine Babuty Desgodets, Second
Ordre du Colise, Rome. Pen on paper. After Antoine Babuty Desgodets, Les
dices antiques de Rome: dessins et mesurs trs exactement (Paris: Jean Baptiste
Coignard, 1682), p. 263. # 2012 C. Elizabeth Grant. With permission.
It would appear, then, that explanation plays a signicant role in

criticism. This is all the more obvious when we remind ourselves of the
various kinds of explanation that there are and the various explananda we
want critics to explain. And it is unsurprising that explanation has such a
signicant role in criticism. Good explanations provide us with under-
standing, and this is one of the benets of good criticism.
Nevertheless, it is not the aim of all criticism to provide explanations.
First, several of the examples mentioned in section 3 tell against this
claim, such as criticism that merely provides unsupported verdicts. Second,
although some criticism explains why a work elicits certain responses,
some criticism tells us what responses the work elicits, or can elicit, from
5. AIDING APPRECIATION 27
someone who appreciates it, without providing an explanation of why it

does so. Third, I noted in section 3 that critics often tell us what is good or
bad about a work. Telling us what is good or bad about a work may or
may not be part of an explanation. You can certainly explain why a work
is good by telling us what is good about it. But you can also tell us what
is good about it without explaining why it is good: you can tell us what is
good about it without believing, claiming, or implying that it is good, and
therefore without explaining why it is.
5. Aiding appreciation
As I have stressed, critics sometimes do guide perception, provide evalu-
ations, support them, elucidate, interpret, explain, and describe apprecia-
tive responses. This is not meant to be an exhaustive list. It might be that all
of these, and perhaps more, are the aims of criticism, and that there is no
further aim. Clearly, however, it would be desirable to nd an aim for the
sake of which critics do all of these things.
An attractive suggestion is that the aim of criticism is to enable the
critics readers to appreciate the work better than they would be likely to if
they experienced the work (or a suitable reproduction, performance,
token, etc.) without having read the criticism. This is meant to include
the case in which the readers would probably not appreciate the work at all
without the criticism. Plausibly, all of the above critical activities are ways
of enabling a person to appreciate a work better.
This, I shall suggest, is close to the truth of the matter. In Chapter 2,
I will show how all of the critical activities we have identied can aid
appreciation. However, there are also counterexamples to the claim that
aiding appreciation is the aim of all criticism.
Criticism of works that the critic knows cannot be experienced any-
more, and of which she knows there are no appropriate reproductions,
performances, etc., provide counter examples. Such criticism can be
addressed to readers who never have been and never will be in a position
to appreciate the work. The aim of criticizing them cannot be to aid
appreciation of such works.
Furthermore, when a work can still be appreciated, the information
provided by good criticism may not enable the readership to appreciate it
better than they would be likely to without the criticism. Reviews that
describe the work in such a way as to provide information that is useful to

someone trying to decide whether to experience the work provide
examples. It might be very useful to know whether a lm is terrifying,
funny, challenging, clichd, thrilling, etc., when trying to decide whether
to watch it. However, such information might be useless for the purpose
of gaining a better appreciation of the work than one would be likely to
acquire by experiencing it without the criticism. For those who have seen
or are watching the lm, being told that it received four stars will probably
not enable them to appreciate it any better. Similarly, anyone who sees the
lm might be unlikely to fail to see that it is terrifying, clichd, disturbing,
etc., and so this information will not enable them to appreciate it any
better than they would be likely to without the criticism.
Nevertheless, there clearly are close links between criticism and appre-
ciation. And we ought not to abandon the attempt to identify an aim
that unies the various critical activities. In Chapter 2, I shall suggest that
such an aim can be found, and that it does have something to do with
appreciation. To say what it is, however, we shall rst have to say more
about appreciation.
2
Criticism and Appreciation
In this chapter, I will argue for a view about what the aims of criticism are.
These aims, I hold, must be understood in terms of the notion of appreci-
ation. So I will begin by providing an account of art appreciation.
This account of appreciation is meant to be a modest one. As I explain
below, there are many more specic claims one could make about appre-
ciation that are compatible with it. The facts about appreciation that I will
attempt to establish are meant to provide an illuminating account of
criticism.
I will then identify a constitutive and a non-constitutive aim of criti-
cism. The constitutive aim is shared by all criticism, whereas the non-
constitutive aim is not. But achieving the non-constitutive aim makes
criticism that has this aim good criticism. Both aims relate to appreciation.
This account enables us to say what makes someone a good critic.
I identify nine endowments that make critics good at achieving these
aims. I conclude by comparing my discussion with Humes well-known
account of the characteristics of the true judge of art. This will enable us, in
the third chapter, to explain what the role of imaginativeness in criticism is.
1. Appreciation
Appreciating a work is clearly not just a matter of knowing facts about it.
You can be extremely well-informed about aspects of a work relevant to
the appreciation of it without appreciating the work yourself. You could
know that the work is beautiful, that it is a satire on the Church, that it
achieves a perfect harmony between form and content, that it expresses a
horrifying vision of a future dystopia, etc., without appreciating it.
A necessary condition of appreciating a work is to be or to have been
aware of the works features by appropriate means. For example, for many
30 CRITICISM AND APPRECIATION
objects of appreciation, some form of perceptual acquaintance with the

work, or with tokens, reproductions, representations, or performances of
it, is an appropriate means of awareness of its features. Forms of perceptual
acquaintance such as seeing, looking at, watching, listening to, and hearing
are paradigmatic examples of appropriate means of awareness of a works
features. In some cases, it may be that more than one form of perceptual
acquaintance can be an appropriate means of awareness of the same works
features. Perhaps one can appreciate sculpture either through sight alone
or touch alone.1
Perceptual acquaintance with something other than the work itself can
be an appropriate means of awareness of some works features. It would be
implausible to claim that no one who has seen high-quality reproductions
of a drawing, but not the original, is in a position to appreciate the
drawing. If musical works are abstract, and if abstracta cannot be perceived,
then such works can be appreciated by perceiving performances of them.
The formulation, awareness of a works features by appropriate means,
also allows for the possibility that there are appropriate means of awareness
of its features other than perceptual acquaintance with something. There
are at least two reasons why one might think that this is, in fact, the case.
First, one might deny that reading a work is a form of perceptual acquaint-
ance with the work. But obviously one can appreciate a literary work by
reading it, and one can appreciate a musical work by reading its score.
Second, it is plausible that accurately imagining perceiving a (token,
performance, etc., of a) work can sometimes give one some appreciation
of it. Perhaps this is how Beethoven appreciated the works he composed
when deaf. We might want to allow that a person could appreciate
Malevichs Black Square, or certain works of conceptual art, if she imagined
perceiving these works accurately enough.
So awareness of a works features by appropriate means is a necessary
condition of appreciating it. The second point to emphasize is that one
appreciates a work by responding appropriately to it, to its parts, to its
features, or to what is represented.
We can, I suggest, think of appreciation in the following way. For
various parts of a work, features of a work, and represented persons, items,
and events, there are various responses to them that appreciation can
1
See Robert Hopkins, Sculpture, in Jerrold Levinson, ed., The Oxford Handbook of
Aesthetics (Oxford: Oxford University Press, 2003), pp. 575577.
1. APPRECIATION 31
involve. You appreciate the work in virtue of responding in some of these

ways to these parts, features, and represented elements. So, for example,
pitying Oedipus is plausibly a response that appreciation of Oedipus Rex
can involve. Being amused at his fate is almost certainly not.
Let us call the responses that appreciation of a work can involve appro-
priate responses. There are three points to clarify about appropriate
responses, as I understand them. First, I do not mean that having any one
appropriate response would be sufcient for appreciating the work. Second,
I do not mean that every response appreciation of a work can involve is a
response appreciation must involve. Some appropriate responses are such
that you can appreciate the work without having them, but you may
appreciate the work better if you do respond in those ways. Third, there
are many properties of a work (e.g., a paintings being 8.51 inches high) to
which there is not an appropriate response.
I suggest we can distinguish ve kinds of response appreciation can
involve. As we shall see, some appropriate responses can be of more than
one of these kinds. But it is useful to distinguish the kinds to which
appropriate responses can belong.
First, there are perceptual responses. Appreciating a Chinese jade can
involve looking at the smoothness of the surface, the translucence of the
stone, or the intricate patterns carved into it; seeing the serene expression
on the face of the gures on it; or seeing that it has been made with great
skill or that a piece of white lychee-esh jade has been used. If, as Frank
Sibley holds, we perceive aesthetic properties like exquisiteness, neness,
grace, and elegance, then appreciation will normally involve perceiving its
aesthetic properties as well.2
Second, there are cognitive responses. Plausibly, appreciation of a work
can always involve acquiring the knowledge, by appropriate means, of
some facts about the work. Appreciating The Waste Land can involve
recognizing the many allusions it contains. Appreciating versions of
well-known stories, such as Shakespeares history plays or the Greek
tragedians versions of Greek myths, can involve recognizing how they
change, add to, or emphasize certain aspects of the story. Grasping the
themes of a work, too, can be an important part of appreciating it.
2
See Frank Sibley, Aesthetic and Non-Aesthetic, in Approach to Aesthetics: Collected Papers
on Philosophical Aesthetics, ed. John Benson, Betty Redfern, and Jeremy Roxbee Cox (Oxford:
Oxford University Press, 2001), p. 34.
Acquiring knowledge by appropriate means is not the only kind of

cognitive response appreciation can involve. Appreciation can also involve
conrming what you know, by appropriate means. You may have learned
via testimony that the design of a certain artefact is very well suited to its
function, but you may never have seen that this is the case. Appreciation
can involve seeing that this is so, even though your seeing that this is so is
not a case of acquiring the knowledge that this is so, for you already have
this knowledge. And as this example indicates, some appropriate percep-
tual responses are a sub-set of appropriate cognitive responsesnamely,
perceiving that something is so.3
Third, it is worth distinguishing another class of responses. M. R. Bennett
and P. M. S. Hacker distinguish the general cognitive concepts of know-
ledge and memory from the cogitative concepts of belief, thought and
imagination.4 Appreciation can certainly involve cogitative responses. For
example, it is unclear whether Ugolino, in the Inferno, ate his own children.
In Borgess view, Dante did not want us to believe it, but he wanted us
to suspect it.5 Coming to suspect that Ugolino ate his own children is a
cogitative response to the poem, and it is plausible that appreciation of the
Inferno can involve having this response. An appropriate response to a descrip-
tive passage in a written work or an evocative piece of music may be to
imagine what is described or evoked. It is also sometimes said that, when
reading a play, one should imagine how it might be staged. Finally, it
is plausible that appreciation can involve acquiring a justied belief in
certain interpretations of a work, even if one does not know that this
interpretation is right.
Fourth, one can also appreciate at least some works by having appropri-
ate affective or emotional responses: delighting in the colours and sheen of
3
Both acquiring and conrming knowledge should be distinguished from possessing
knowledge. Knowledge you already possess (e.g., historical, sociological, or anthropological
knowledge) can enable you to appreciate a work better than you otherwise could. But
possessing knowledge is not a response, and a fortiori not a response that appreciation can
involve. Furthermore, appreciation that is enabled or deepened by the knowledge that p may
not involve acquiring or conrming the knowledge that p. Knowing what Shakespeares
contemporaries believed about English history can enhance ones appreciation of the history
plays. Such knowledge may aid appreciation even if acquiring or conrming it is not a
response appreciation can involve.
4
M. R. Bennett and P. M. S. Hacker, Philosophical Foundations of Neuroscience (Oxford:
Blackwell, 2003), p. 172.
5
Jorge Luis Borges, The False Problem of Ugolino, in The Total Library: Non-Fiction
19221986, ed. Eliot Weinburger, trans. Esther Allen (London: Penguin, 2001), p. 278.
1. APPRECIATION 33
a polished gem, being awestruck by the Alhambra, and cringing at a poorly

played piece of music are all examples of appropriate affective or emotional
responses.
Fifth, there are many appropriate responses that do not obviously t
into any of the last four categories. It seems that appreciation can involve
being engrossed by a story, taking an interest in the expression of a sitter in
a portrait, being fascinated by the movements of water in a fountain, and
so forth. These responses involve desiring or being disposed to continue
experiencing the object. They are not merely affective or emotional
responses. Indeed, it is doubtful that some, such as taking an interest in
something, are affective or emotional responses at all. As they involve
desire, we might describe such responses as conativea term dened by
the OED as pertaining to, or of the nature of, the faculty of volition and
desire. Perhaps some conative responses are also affective responses. But as
it is not clear that they all are, it is worth distinguishing this fth kind of
response.6
That appreciation can involve affective, emotional, and conative re-
sponses is a familiar point. However, it is worth observing that, for many
works to which affective, emotional, or conative responses are appropri-
ate, one can appreciate them by acquiring or conrming for oneself, by
appropriate means, the knowledge or justied belief that those responses
are appropriate, without having those responses oneself. This might be the
case if you are in too foul a mood to take pleasure in the works beauty. It
might also be the case if the work originated in a culture whose sensibilities
are very different from your own. Both factors, it seems clear, can prevent
6
This account of conative responses will remind some of Kants account of pleasure. He
writes:
The consciousness of the causality of a representation in respect of the state of the
subject as one tending to preserve a continuance of that state, may here be said to denote in
a general way what is called pleasure; whereas displeasure is that representation which
contains the ground for converting the state of the representations into their opposite
(for hindering or removing them). (Immanuel Kant, Critique of Judgement, ed. Nicholas
Walker, trans. James Creed Meredith and Nicholas Walker (Oxford: Oxford
University Press, 2007), Ak. 5: 220).
I am not claiming that pleasure is the consciousness of a representations causality. But I am
stressing that appreciation can involve being disposed to continue experiencing the object.
Compare Malcolm Budds view that a positive aesthetic response to an artwork or a natural
object involves the disposition to continue to attend to it (Malcolm Budd, The Aesthetic
Appreciation of Nature: Essays on the Aesthetics of Nature (Oxford: Oxford University Press,
2002), p. 14).
you from having an appropriate affective response yourself, but do not

necessarily prevent you from appreciating the work. Indeed, it is not clear
that they would necessarily prevent you from appreciating the work
as well as someone capable of having those responses. It is important,
however, that one have acquired or conrmed ones understanding
of what the appropriate affective, emotional, or conative responses are
by appropriate means. For example, learning by testimony that a certain
response to a feature of the work is appropriate, while nding it unintelli-
gible on the basis of your awareness of the work how that could be an
appropriate response to it, would (probably) not be a response appreciation
can involve.
In short, some appropriate responses are meta-responses. One kind of
appropriate cognitive or cogitative response can be to acquire or conrm
by appropriate means the knowledge or justied belief that some emo-
tional, affective, or conative response is an appropriate response to a feature,
part, or represented element.
There are two nal points to note about the responses appreciation can
involve. First, it might sound odd or misleading to say, of some responses,
that you appreciate the work by responding in those ways, or that appreci-
ating the work involves having those responses. This might be so when the
appropriate response is not a favourable one. It might sound odd to say that
one appreciates a boring work by being bored by it, that appreciation of a
certain terrible piece of music involves cringing at it, and so on. There is a
common use of appreciate in which saying one appreciates something
implies that one has evaluated it positively or responded to it favourably.
We normally say that someone has failed to appreciate the work when we
mean to imply that she has not had a favourable response to it that would
have been appropriate.
In such cases, it will be more natural to say, of something about the
work rather than the work itself, that it is appreciated by responding in
these ways. For instance, if an experienced theatre critic is bored by a play
the audience enjoys, it would be more natural to say that she is the only
one who appreciates how badly written it is. She is the only one who
showed an appreciation of the plays aws. Appreciate is also used in these
two ways in non-artistic contexts. Someone might fail to appreciate the
danger she is in. In saying this, we do not imply that she fails to respond
favourably to the danger she is in. But if the danger is posed by an
oncoming train, it would be odd to express the same point by saying she
1. APPRECIATION 35
fails to appreciate the oncoming train. So when I talk of the responses that
appreciation of a work can involve, these include some responses that it
might be more natural to describe as responses that appreciation of some-
thing about the work can involve.
The last point to make about appropriate responses concerns the
relation between them and appropriate awareness. Appreciating a work
requires that one must have been aware of the works features by appro-
priate means. But it does not require that one must respond appropriately
only when one is aware of its features by appropriate meansonly when
one is looking at the painting (or a suitable substitute), listening to the
music, reading the poem, etc. One can also respond appropriately when
reecting upon the work after experiencing it. Ones appreciation of a lm
is often deepened by discussing it with others after watching it.
Apart from appropriate awareness and appropriate responses, there is a
third aspect of appreciation. Often, appreciation does not simply involve
responding appropriately. It involves responding appropriately for appro-
priate reasons. There are certain reasons to pity Oedipus, and appreciating
the play can involve pitying him for those reasons, but cannot involve
pitying him for no reason or for other reasons. So too, appreciation of the
Inferno can involve coming to suspect, on appropriate grounds, that Ugolino
ate his children: for Ugolino appears to allude darkly to such an act.
Appreciating it cannot involve simply coming to suspect him of this crime
for no reason. When appreciation of a work can involve responding for a
certain reason, I will call that reason an appropriate reason for that response.
There are at least two relevant conceptions of responding for a reason to
distinguish here. One might think of a reason merely as a fact about the
work that explains, or partly explains, your response. So, for example,
Stephen Jay Gould has argued that the way in which Mickey Mouse is
drawn has changed over the years, making him look more childlike.7 He
has acquired, for example, a larger eye size as a percentage of head length,
and a larger head length as a percentage of body length. Gould argues that
Mickey Mouse is more attractive to us because he looks more like a child.
If Gould is right, the fact that Mickey looks more like a child explains our
response. But the claim is not that our awareness of this fact explains
our response. The fact that he looks more childlike, and our awareness of
7
Stephen Jay Gould, A Biological Homage to Mickey Mouse, in The Pandas Thumb:
More Reections in Natural History (New York: Norton, 1980), pp. 93107.
the features in virtue of which he looks more childlike, explain our

response. But our awareness of this fact about these features may not
explain our response. We may not even be aware of this fact.
Alternatively, one might think of a reason as a fact about the work your
awareness of which explains, or partly explains, your response. This is the
kind of reason I am talking about when I say appreciation can involve
responding appropriately for appropriate reasons. For example, it is our
awareness that the speaker has been caught by the monster that explains
our amusement at the end of the poem Slithergadee:
The Slithergadee has crawled out of the sea.
He may catch all the others, but he wont catch me.
No you wont catch me, old Slithergadee,
You may catch all the others, but you wo8
Similarly, a person might be aware that the Purgatorio and Paradiso have
33 cantos each, while the Inferno has 34, and aware that it is appropriate for
parts of the poem about the saved to have such a salient relation to the
number of the persons of the Trinity, and for a part about the damned to
lack this relation. Her awareness of these facts can partly explain her
admiration for the design of the Divine Comedy.
These are the kinds of fact that can be what I am calling appropriate
reasons. If you responded appropriately for appropriate reasons, then you
became aware of the facts in question. By contrast, if peoples awareness of
the fact that Mickey looks more childlike does not explain their response to
him, it is not what I am calling a reason for which they respond as they do.9
It can often be unclear whether our awareness of a fact, or only the fact
itself and not our awareness of it, explains our response. Awareness comes
in degrees, and it can be unclear whether one was aware of a certain fact.
For example, suppose you nd a certain typeface in a newly published
book strange. Suppose someone points out that the typeface is more
characteristic of books published in the 1970s than of books published
today. This is not new information about that sort of typeface: you already
8
Shel Silverstein, Slithergadee, in Uncle Shelbys Zoo (New York: Simon and Schuster,
1964), quoted in Kendall L. Walton, Mimesis as Make-Believe: On the Foundations of the
Representational Arts (Cambridge, MA: Harvard University Press, 1990), p. 162.
9
When I say that our awareness of a fact sometimes explains our response, I do not mean
that the fact itself does not explain our response. The contrast is between cases in which our
awareness of the fact explains our response, and cases in which it does not.
1. APPRECIATION 37
knew that this sort of typeface is characteristic of 1970s books. But you do
not recall consciously noting that the typeface of this book is in a 70s style.
In such a case, it may be unclear to you whether you were already dimly
aware that the book uses a typeface typical of the 1970s. It may therefore
be unclear to you whether your awareness of this fact explains your
response to the typeface. This kind of experience is not uncommon
when reading criticism, as we shall see.
At least sometimes, an appreciator must be aware of the fact that is an
appropriate reason by certain means rather than others. Suppose you are told
that a tree in a landscape painting looks like a lonely person, but you cannot
see this for yourself. The fact that it looks like a lonely person may be an
appropriate reason to respond to the painting with a certain melancholy. But
if, improbably, you responded with melancholy for this reason, without
having seen that the tree looks like a lonely person, you would not thereby
be appreciating the painting. You are responding appropriately for an appro-
priate reason, but you are not aware of the reason by appropriate means.
Finally, for some features, it seems true that appreciation involves
responding to them in a certain way, but false that appreciation involves
responding to them in that way for certain reasons. This often seems true of
aesthetic features, if these are indeed features objects have. For example,
appreciation can often involve admiring an objects beauty. But it would
rarely involve admiring an objects beauty for some reason. Appreciation
can involve admiring a paintings colours because they are subtle, harmoni-
ous, expressive, etc., or being amused by a turn of phrase because it is witty.
But rarely, if ever, would it involve admiring beauty for some reason.10
This account of appreciation, as I said initially, is a minimal one. Many
further claims about appreciation are compatible with it. For example, I do
not provide a general account of what makes a response or a reason
appropriate. I also do not claim or deny that, as Kendall Walton holds,
appreciation of representational works of art is primarily a matter of
participation in games of make-believe, or that appreciation not involv-
ing participation is nevertheless to be understood in terms of it.11 My
intention has been to provide enough detail to enable us to specify the
aims of criticism by reference to appreciation.
10
For a discussion of qualities that can be admired aesthetically for themselves, see Frank
Sibley, Aesthetics and the Looks of Things, in Approach to Aesthetics, pp. 2432.
11
Walton, Mimesis as Make-Believe, pp. 213, 275.
2. Appreciation and the aims of criticism

There are clearly connections between appreciation and criticism. One of
them is this. Appreciation, I said, requires awareness of the works features
by appropriate means. Someone who criticizes a work represents herself as
having been aware of its features by some means needed in order to
appreciate it. This is a pragmatic implication of criticizing a work. Some-
one who criticizes a literary work implies that she has read it or heard it
read; someone who criticizes a painting or sculpture implies that she has
seen it or a suitable reproduction or representation of it; a theatre critic
implies that she has seen a performance of the production she criticizes; a
music critic implies that she has heard a performance or read the score of
the piece she criticizes; and so forth.
This is a striking fact about criticism. No matter how much you know
about a work, it is illegitimate to criticize it if you have not been aware of
its features by means required for appreciation. If a lm critic had not seen
the lms she reviewed, she would be guilty of a kind of dishonestyeven
if everything she said about the lm was true, and even if she had good
reasons for believing it. Reviewers can lose their jobs for criticizing works
they have not seen.
This enables us to answer Carrolls challenge, mentioned in Chapter 1,
to identify a feature that distinguishes criticism from comparable forms of
discourse about art. You can engage in many comparable forms of dis-
course about art without implying that you have been aware of the works
features by means required to appreciate it. When you tell someone that
you have it on good authority that the lm is excellent, you are not
criticizing. This is clear in any case. But we can now explain why: you
are not representing yourself as having seen the lm, which is a necessary
condition of criticizing it. One way in which art criticism differs from art
history is that the former but not the latter is necessarily governed by this
requirement. Art historians can still write about the lost paintings of Zeuxis
and Parrhasios, and there is not necessarily anything untoward about the
art history they produce. But no one can legitimately criticize such works
anymore, because to do so is to imply that one has been aware of their
features by means required for appreciation.
A critic also implies that her belief in at least some of what she commu-
nicates is based on her awareness of the works features by appropriate
means. A critic who has seen a building must not tell her reader that the
2. APPRECIATION AND THE AIMS OF CRITICISM 39
spacing of its columns creates a solemn tempo if her belief in this is not
based on awareness of the buildings features by appropriate means. It is
not enough that she saw the building: her belief in what she says must be
based on her experience.12
Let us call the requirement that a critic must have been aware of the
works features by means required for appreciation, and that her belief in at
least some of what she communicates must be based on this awareness, the
acquaintance requirement. The critic says what she does with a certain
authority, and this authority must derive from her awareness of the works
features by means required to appreciate it.
The second connection between criticism and appreciation relates
to the aims of criticism. One appreciates a work, I said, by responding
appropriately to it, its parts, features, or represented elementsand often,
by responding appropriately for appropriate reasons. One criticizes an
artwork only if one aims to communicate:
(a) what parts, features, or represented elements appreciation can
involve responding to; or
(b) what responses appreciation of it can involve; or
(c) what appropriate reasons for these responses there are.
A necessary condition of criticizing, I suggest, is that one aim to give ones
reader to understand that such-and-such is an appropriate response, or an
appropriate reason, or a part, feature, or represented element to which
appreciation can involve responding.
Note that I do not claim that critics must say that a fact of one of these
kinds obtains. Giving ones reader to understand this is enough. Often,
critics simply ascribe properties to the work, or describe the responses it
elicits. What makes this criticism is that the point of this is not just to inform
the reader that the work has these properties or elicits these responses,
but to convey that these properties are objects of appreciative responses, or
that these responses are appreciative responses, or that the fact the work has
these properties is a reason to respond appreciatively. A reader who recog-
nizes what she is reading as criticism understands that this is the point.
Communicating facts of this kind (that is, (a), (b), or (c)) is, I suggest, a
constitutive aim of criticism. One is not engaging in criticism if one does
not have this aim. For example, footnotes in an edition of Shakespeare that
12
This does not mean that it cannot also be based on testimony.
explain the meaning of unfamiliar words are not necessarily critical

remarks. One reason why is that such footnotes do not necessarily give
us to understand that these words or their meanings are to be responded to
appreciatively. Similarly, a psychologist might describe the responses edu-
cated readers have to a work. This is not necessarily criticism, because there
need be no suggestion that these are responses appreciation can involve.
This constitutive aim thus provides a second way of differentiating criti-
cism from many comparable forms of discourse about art.
The claim that this is a constitutive aim of criticism also enables us to
explain several other things.
First, it enables us to explain the acquaintance requirement. If a critic
aims to tell us what responses appreciation can involve, what reasons for
such responses are appropriate, or what the objects of appropriate responses
are, then her criticism has a distinctive authority if it is based on awareness of
the works features by means required for appreciation. Criticism is a source
of information or insight about such matters that is supposed to have this
authority.13
This also enables us to explain why some things the critic says need not
be based on such awareness. The critics assertions of various historical,
sociological, and anthropological facts do not necessarily have to be based
on her awareness of the works features by appropriate means. These
assertions can be based on testimony. But when she tells us that such-
and-such is an appropriate response, reason, or object of an appropriate
response, the acquaintance requirement applies.
Second, the claim that this is a constitutive aim of criticism allows us to
explain why critics do several things I said in Chapter 1 they do. Critics
evaluate for many reasons. For example, they tell us what is good or bad
about a work because appreciation can involve responding to what is good
and bad about it. They might tell us that a work has a certain value because
appreciating the work can involve recognizing or conrming, by appro-
priate means, that it has that value. Appreciation can also involve respond-
ing in a certain way to a work, or to something about it, because it has a
certain value. Critics interpret and elucidate because appreciation can
involve having cognitive or cogitative responses to the work that have
13
This is not to deny that one can learn such things about appreciation of a work via
testimony. It is just that appropriate awareness is a source of greater authority about such
matters.
2. APPRECIATION AND THE AIMS OF CRITICISM 41
the content of interpretations and elucidations as their content. They tell

us of affective, emotional, and conative responses a work elicits because
appreciation can involve having such responses or recognizing that such
responses are appropriate.
Critics provide explanations for several reasons. For example, appreci-
ating a work can often involve grasping or seeing that p because q: that the
painting has a certain unity of tone because it has that concentration of
blues and greys; that Tess kills Alec because she is a certain type of folk
heroine; or that the knocking at the gate in Macbeth reects upon the
murder a peculiar awfulness because it re-establishes the normal goings-on
of the world. Critics explanations often reveal appropriate reasons for
appropriate responses. Summerson presents it as a reason for admiring the
design of the Colosseum that its architects combine the Ionic order with
arch-and-vault architecture by having the pedestal moulding of the order
range with the sill height of the gallery; by having the impost of the arch
strike the columns just above half their height; by having the arch sit
comfortably between the columns and the architrave above; and so on.
What about the other candidate aims of criticism I considered, namely,
helping readers choose what works to experience, aiding appreciation, and
guiding perception? I said that critics sometimes, but do not always, try to
achieve these things. We can now see why.
One of the reasons why one might tell the reader what responses
appreciation can involve, what appropriate reasons for them there are, or
what objects of those responses there are, is to help her decide whether to
experience the work. Criticism necessarily aims to provide information of
a certain kind, and one can provide information of this kind for many
purposes. Helping readers choose what works to experience is one of these
purposes, but not the only one.
The same is true of aiding appreciation. A very common reason why a
critic tells us these things about the appreciation of a work is to help us
appreciate the work better. Again, this is not the only reason why one
might provide the information about appreciation that the critic provides.
The information might be of interest to the critics readership even if it
cannot aid appreciation (due to the work no longer being available for
appreciation, for instance). But it is certainly a very common reason why
critics criticize.
In fact, I suggest that aiding appreciation is not just a very common
purpose to which criticism is put. Aiding appreciation is itself an aim of
criticismthough not a constitutive aim. I said in Chapter 1 that an aim is

an aim of criticism if achieving, or being such as to achieve, that aim makes
criticism that has it good as criticism. Much good criticism, it seems clear, is
good criticism because it is such as to enable its readership to appreciate the
work better than they would be likely to if they were aware of the works
features by appropriate means, but had not read the criticism. This is not
just one purpose to which criticism can be put. Works are sometimes
criticized to help establish when they were made or who the artist was.
If the criticism does help establish this, that may be a good thing. The
criticism furthers art-historical scholarship. But achieving these aims is not
the kind of thing that makes it good criticism. Being such as to aid the
readerships appreciation, by contrast, makes it good as criticism.
This aim, in turn, allows us to explain why critics try to guide percep-
tion or cause us to have other responses to the work. They do this because
this can enable us to appreciate the work better than we otherwise would.
Note, nally, that a piece of criticism could achieve the constitutive aim
I have identied without this making it good criticism. For example, the
criticism might communicate very uninteresting facts of the relevant kinds
about appreciation. Having the constitutive aim enables it to qualify as
criticism, and it would be a defect if the criticism did not achieve even this
aim. But achieving this aim in this way clearly does not make it good
criticism. This is one reason why it is worth distinguishing between
constitutive and non-constitutive aims of criticism.
In the remainder of this chapter, I wish to focus on these two aims: the
constitutive aim of criticism I have identied and the non-constitutive aim
of aiding appreciation. I wish to determine what endowments make one a
good critic. I will do this by determining what endowments make a critic
good at achieving these aims.
3. Better appreciation
If my account of appreciation is right, there are three basic changes one can
effect to enable someone to appreciate a work better.14 If O is an object of an
14
For ease of exposition, I will focus on the diachronic case of enabling someone to
appreciate a work better than she did before. But the aim of criticism I identied is to enable
someone to appreciate a work better than she would be likely to if she were aware of the
works features by appropriate means without having read the criticism.
3. BETTER APPRECIATION 43
appropriate response (a part, feature, or represented element), R an appropriate

response to O, and P an appropriate reason for R, we can say the following.
One can enable someone to appreciate a work better by enabling her:
(1) To become aware by appropriate means of O, or to acquire or
conrm by appropriate means the knowledge or justied belief
that the work represents O; or
(2) To have R to O on the basis of appropriate awareness of the
works features; or
(3) To have R to O for P on the basis of appropriate awareness of the
works features.
Note that it would also be possible to effect these changes without
enabling a person to appreciate the work better. For instance, the person
might become aware of O by appropriate means, but not be able to
respond to it appropriately or grasp what an appropriate response to it
would be. Whether effecting these changes enables a given individual to
appreciate a work better depends on what else is true of her.
There are three points to note about how critics effect these changes.
First, critics effect these changes by achieving the constitutive aim of
criticism I identied. A person could effect these changes by other means.
Giving someone a drug might enable her to respond in appropriate ways in
which she did not respond before. Giving someone glasses could enable
her to become aware of features of a painting of which she hadnt been
aware. Educating someone in the appreciation of other works of art might
enable her to appreciate a given work better. But none of this would be
criticism of the given work. Criticism effects the changes mentioned above
by communicating to the reader what responses appreciation can involve,
what the objects of those responses are, or what appropriate reasons there
are for those responses.
Second, criticism can effect these changes either directly or indirectly. For
example, telling someone what features appreciation involves responding
to, can, obviously, enable her to become aware of those features. This
is a change the criticism effects directly. The critics readership may also be
sensitive or savvy enough to respond appropriately for appropriate reasons
once made aware of this feature, and thereby appreciate the work better.
This is a change the criticism effects indirectly. In such cases, it is unnecessary
to tell readers, in addition, what responses appreciation can involve having,
or what facts about the work are reasons to have it.
The same can be true of criticism that merely tells the readership what
responses appreciation can involve. This can have the direct effect of
enabling her to have those responses. It can also have indirect effects. It
can, for instance, make her aware of features to which that response is an
appropriate response. Told that appreciating a work can involve having a
certain response, we naturally seek what it would be an appropriate
response to. Walter Pater recognized this when he wrote: To see the
object as in itself it really is, has been justly said to be the aim of all true
criticism whatever; and in aesthetic criticism the rst step towards seeing
ones object as it really is, is to know ones own impression as it really is, to
discriminate it, to realise it distinctly.15
Third, critics not only effect these direct and indirect changes by making
their readers aware of facts, features, etc. of which they had not been aware
before. Sometimes, critics effect such changes by making readers more
aware of facts, features, etc. of which they had been only dimly aware
before. For awareness comes in degrees. A reader may not have responded
appropriately to a feature because she was not aware enough of the feature.
Alternatively, she may have responded only weakly or intermittently to
the feature. Making her more aware of it can make her response stronger
or more sustained. Likewise, she may not have responded for a certain
reason because, though aware of the fact that is a reason, she was not aware
enough of it to respond for that reason. Critics can aid appreciation by
increasing, and not only by creating, awareness.
A good example of this is Samuel Johnsons description of Miltons
style:
Through all his greater works there prevails an uniform peculiarity of Diction,
a mode and cast of expression which bears little resemblance to that of any
former writer, and which is so far removed from common use, that an
unlearned reader, when he rst opens his book, nds himself surprised by a
new language. . . . Both in prose and verse, he had formed his style by a
perverse and pedantick principle. He was desirous to use English words with
a foreign idiom. . . . The disposition of his words is, I think, frequently Italian;
perhaps sometimes combined with other tongues.16
15
Walter Pater, Studies in the History of the Renaissance, ed. Matthew Beaumont (Oxford:
16
Samuel Johnson, Milton, in The Lives of the Poets: A Selection, ed. Roger Lonsdale
(Oxford: Oxford University Press, 2009), p. 112.
4. ENDOWMENTS OF GOOD CRITICS 45
Many readers of Milton are well aware that his style is peculiar. However,
many readers, I suspect, are only dimly aware that he uses words with a
foreign idiom, even if they are familiar with the syntax of other languages.
They may not become aware enough of it for their response to be much
affected by their awareness of this fact. Johnsons criticism aids their
appreciation of Milton by making them more aware of it.
Another example is the following observation by Frank Kermode:
now and again Shakespeare uses a word neither the original nor the modern
audience had ever heard before, which yet remains intelligible to both, as
when Goneril (King Lear, I.iv.249) advises her father A little to disquantity
his train. The dictionary records no earlier use of this word, and it did not
catch on, but to the modern ear it has a disturbingly bureaucratic ring, rather
like the euphemisms produced by government departments, and it must have
surely struck the rst audience also as a cold and ofcial-sounding word for a
daughter to use in conversation with her father.17
It is likely that many readers of the play are only dimly aware of the ofcial-
sounding ring of disquantity. They would be made more aware of this by
Kermodes remark.18 A readers increased awareness of this can enable her
to respond to the line in new ways: she might, for example, now nd the
line chilling. Alternatively, she might respond for a reason for which she did
not respond before. She might now admire the line because it has an
ofcial-sounding ring, for this makes it a very apt line to give Goneril.
4. Endowments of good critics

These, then, are the basic changes a critic can effect to aid appreciation,
and some of the principal ways in which critics effect them. The endow-
ments that make a critic good at aiding appreciation will therefore be
endowments that make critics good at effecting the above changes by
achieving the constitutive aim of criticism. Because of the acquaintance
requirement, they will include endowments that enable a critic to base
what she says on her awareness of a works features by appropriate means.
First, to be good at effecting any of changes (1)(3) by criticizing, one
must be good at communicating to ones readership. One must express
17
Frank Kermode, Shakespeares Language (London: Penguin, 2000), p. 5.
18
Critics sometimes overstate how aware we or the audience already are of the facts and
features they point out, or how responsive we already are to them.
oneself in such a way that the readership can grasp, without undue
difculty, what responses, objects of responses, and reasons one has in
mind. An impressive display of this skill is Paters description of the
landscapes in the backgrounds of Leonardos paintings:
In him rst, appears the taste for what is bizarre or recherch in landscape;
hollow places full of the green shadow of bituminous rocks, ridged reefs of
trap-rock which cut the water into quaint sheets of light . . . all solemn effects
of moving water . . . Through his strange veil of sight things reach him so; in
no ordinary night or day, but as in faint light of eclipse, or in some brief
interval of falling rain at daybreak, or through deep water.19
In short, the critic must be articulate.

Articulacy, however, is not enough. You might convey that appreciation
can involve admiring a certain feature, but your readers may be unable to
gure out why it merits admiration. To appreciate the work better, they may
need to be told what the appropriate reasons for admiring it are. Again, it may
not be enough simply to claim that a certain fact is an appropriate reason for a
certain response. To aid the readers appreciation, it may be necessary to
persuade her that the fact obtains, or that it is a reason, so that she can see for
herself that it obtains and respond for this reason. (This is one function of
providing evidence for a critical evaluation.) A critic must therefore be a good
judge of what the readership being addressed needs to be told in order to
enable them to appreciate the work better.
Furthermore, a critic needs to be a good judge of how to communicate
this to her readership. For example, a critic may need to choose her words not
only to inform, but to guide the readers perception effectively. She must
consider not only what information she conveys, but other effects of her
words that can aid appreciation. We shall see examples of this in Chapter 6.
These endowments help make one good at effecting changes of all three
kinds by criticizing. Now consider what makes one good at effecting
particular kinds of change.
Take changes (1) and (3). Given the acquaintance requirement, a critic
must become aware, by appropriate means, of what she wants her reader
to become aware of by appropriate means. This is so whether she wants
the reader to become aware of a fact that is a reason for an appropriate
response, or of a part, feature, or represented element that is an object of an
19
Pater, Studies in the History of the Renaissance, p. 63.
Figure 2.1. C. Elizabeth Grant. Palazzo del T, Mantua. Pen on paper. # 2012
C. Elizabeth Grant. With permission.
appropriate response. I suggest we can identify three endowments that

make the critic good at acquiring such awareness. The critic can be more
knowledgeable, more observant, or more perceptive than the readership.
Being more knowledgeable, in some respect, than someone else can
enable you to become aware of what she is unaware of. In a discussion of
Giulio Romanos design for the Palazzo del T (Figure 2.1), John Sum-
merson writes:
Here is a very strange performance. You recognize, of course, the Doric order.
And the major columns are approximately on the triumphal arch pattern. But
the pediment rests not on columns but on brackets emerging from the wall,
and the keystone of the arch pushes violently up into the pediment. Every-
thing is a bit uneasy, a bit wrong. Do you notice that in the entablature some
of the stones have slipped? . . . It is irrational, impressionistic.20
It might be obvious to someone with enough knowledge of the rules and

history of classical architecture, and of the date of this buildings design,
that it is a very strange performance. For it might be obvious to such a
20
1980), p. 46.
person that the position of the keystone inside the pediment, and the
dropped stones of the entablature, constitute dramatic departures from
architectural norms. The fact that the building is strange in virtue of such
features is a reason to take an interest in them. A viewer of the building less
knowledgeable than Summerson might be unaware of this fact.
Sometimes you become aware of what another person misses, not because
you are more knowledgeable, but because you are more observant. It was
very observant of Helen Vendler to have seen what she calls the Couplet
Tie in Shakespeares sonnets. She pointed out that in nearly every sonnet, a
thematically important word in the quatrains is repeated in the couplet. For
example, words like old (sonnet 2), single (sonnet 8), and time (sonnets
12, 15) are repeated in the procreation sonnets urging the addressee to
marry so that he does not die without producing an heir. Vendler claims that
Shakespeare clearly depended on this device not only to point up the
thematic intensities of a sonnet, but also to show how the same words take
on different emotional import as the poem progresses.21 Those who failed
to notice the Couplet Ties prior to Vendler did not fail through lack of
knowledge, but because they were, in this regard, less observant.
Greater perceptiveness, too, enables critics to become aware of what
their readers are liable to miss. For example, it was perceptive of Kermode
to see that a line in King Lear is an allusion to a common saying:
The Renaissance, like St. Paul, found much value in folly, and Erasmus, who
wrote a famous book about it, also recorded the adage Kings and fools are
born, not made, which Shakespeare may have recalled when he has Lear ask,
Dost thou call me fool, boy? and receives the reply All thy other titles thou
hast given away, that thou wast born with (I.iv.14850, Q only).22
It is partly Kermodes knowledge of the adage that enables him to recognize

the Fools line as an allusion to it. But it would not be obvious to anyone who
knows this adage and reads the line that this is an allusion to it. Kermode
displays both knowledgeability and perceptiveness in recognizing the allusion.
To be good at effecting changes (1) and (3), it is not enough that a critic
be good at becoming aware of what is in fact an object of an appropriate
response, or of what is in fact an appropriate reason. She must also be good
at recognizing, or acquiring the justied belief, that it is an appropriate
21
Helen Vendler, The Art of Shakespeares Sonnets (Cambridge, MA: Harvard University
Press, 1997), p. 28.
22
Kermode, Shakespeares Language, p. 187.
reason or an object of an appropriate response. As I will put it, she must be

a good judge of the appreciative relevance of what she becomes aware of.
What endowments allow critics to effect change (2)that is, enabling a
reader to have an appropriate response she is unlikely to have? Clearly, to
effect this change, the critic must be a good judge of the appreciative
relevance of responses. She must be good at recognizing that a given
response is one appreciation can involve. She must also be good at becom-
ing aware, by appropriate means, of such responses in the rst place. We can
distinguish two basic endowments that can make someone good at becom-
ing aware of appropriate responses.
First, the critic can normally become aware of an appropriate response if
she has that response herself. Joan Mirs report of his response to Cour-
bets Stormy Sea provides a good example of an appropriate response many
of his readers are unlikely to have.23 He writes: One feels physically drawn
to it, as by an undertow. It is fatal. Even if this painting had been behind
our backs, we would have felt it.24 Many viewers are unlikely to respond
to the painting as intensely as Mir.
Second, being good at thinking of appropriate responses can make a
critic good at becoming aware of them. Provided she thinks of them on
the basis of her awareness of the works features by appropriate means, it is
not necessary for her to respond in that way herself. A critic might judge,
for example, that melancholy would be an appropriate response to a work,
even if she does not feel melancholy herself.
It seems, then, that the following endowments make a critic good at
enabling her readership to appreciate the work better:
(i) Articulacy.
(ii) Good judgement regarding what the readership needs to be told
to be enabled to appreciate the work better.
(iii) Good judgement regarding how to communicate so as to enable
the readership to appreciate the work better.
(iv) Knowledgeability.
(v) Being observant.
(vi) Perceptiveness.
23
For an image of this work, see <http://www.musee-orsay.fr/index.php?id=851&L=1&tx_
commentaire_pi1[showUid]=8986> accessed 14 June 2012.
24
Quoted in Michael Fried, Courbets Realism (Chicago: University of Chicago Press,
1990), p. 215.
(vii) A disposition to have appropriate responses the readership is

unlikely to have.
(viii) Being good at thinking of appropriate responses the readership
is unlikely to have.
(ix) Good judgement regarding the appreciative relevance of re-
sponses, objects of responses, and reasons for responses.
As we shall see in Chapter 3, this account by no means excludes a role for
imaginativeness in criticism. Rather, I will argue that this account allows us
to understand just what its role is. Before proceeding, however, let us
compare my account with the most prominent philosophical account of
the endowments of true judges of art.
5. Humean judges
The best-known account of what makes one a good judge of art is Humes
Of the Standard of Taste. Humes view is often described as an account of
the true critic. Humes concern, however, is only to say what makes
someone a good judge of a works beauty. Being a good judge of a works
beauty is part, but not all, of what makes someone a good critic in the
current sense.25 Comparing my account with Humes is instructive for what
it reveals about how the endowments of a good critic, in todays sense, differ
from the endowments of a good judge, or at least a good Humean judge, of
a works beauty. Hume identies ve characteristics: good sense, delicacy,
practice (experience contemplating other works), the ability to compare
the work one is judging with many others, and freedom from prejudice.
First, communication skills gure in my account, but not in Humes.
The rst three endowments I identifyarticulacy, good judgement about
what ones readership needs to be told, and good judgement about how
to communicate itare communication skills. These mark a difference
between what makes someone good at criticizing art and what only makes
25
In the rst edition of Johnsons Dictionary, published two years before Humes essay,
critick is dened as: A man skilled in the art of judging of literature; a man able to distinguish
the faults and beauties of writing. Note that this suggests not only that a critic, in the sense
then current, was only a judge. It also implies he was a judge only of literature. This would
explain why all of Humes examples are from literature. Even Humes reference to relishing a
ne stroke may be a literary example: Johnson denes stroke as A touch; a masterly or
eminent effort, and illustrates this sense with a quotation that discusses strokes of poetry.
5. HUMEAN JUDGES 51
someone good at judging art. To be a good judge, it is not clear that one
needs to be articulate; and it is clearly not necessary to have good judge-
ment about what and how to communicate to others to help them better
appreciate works.
Second, both a good judge and a good critic need what Hume calls
delicacy. Hume explains the need for delicacy by observing that qualities
naturally tted to produce the sentiment of beauty may be found in a small
degree, or may be mixed and confounded with each other. . . . Where the
organs are so ne, as to allow nothing to escape them; and at the same time
so exact as to perceive every ingredient in the composition: This we call
delicacy of taste.26 Any of knowledgeability, being observant, or percep-
tiveness can endow a critic with delicacy. In Humes example from Don
Quixote, Sanchos kinsmen, who detect the taste of leather and metal in a
wine, are more observant than their companions. Summersons knowledge
of architecture enables him to notice the slipped stones of the entablature,
and nd them strange, more readily than people less knowledgeable.27
Similarly, both a good critic and a good judge need good sense. For
Hume, a judge with good sense appears to be someone generally intelli-
gent: he says that her good sense enables her to overcome prejudice,
apprehend the relations between a works parts, determine how well a
work fulls its functions, and follow chains of reasoning in the work.28 My
claim that critics must be good judges of the appreciative relevance of
responses, objects of responses, and reasons for responses is, in effect, a
claim about what critics ought to have good sense about.
The reasons why a good critic needs delicacy and good sense, however,
are not limited to the reason why, according to Hume, a good judge needs
these qualities. In Humes view, judges need these characteristics to be able
to have appropriate sentiments. Prejudice, similarly, prevents appropriate
sentiments.
26
David Hume, Of the Standard of Taste, in Essays: Moral, Political, and Literary, ed.
Eugene F. Miller (Indianapolis: Liberty Fund, 1985), p. 235.
27
This explains why, as Hume says, practice in a particular art, and the frequent survey or
contemplation of a particular species of beauty is something that tends further to encrease
and improve delicacy (ibid., p. 237). One reason why practice can do this is that practice is a
way of acquiring knowledge. Knowing, for example, what aspects of an architectural or
literary style are conventional can enable you to notice those features of a work that constitute
subtle departures from convention.
28
Ibid., p. 240.
In my view, a disposition to have appropriate sentiments can certainly

help make one a good critic. But delicacy and good sense are not qualities
of good critics only because they can enable critics to have appropriate
sentiments. They can also enable a critic to provide insightful interpret-
ations or sensitive descriptions. This helps make one a good critic whether
or not it leads one to have appropriate sentiments.
Finally, being good at thinking of appropriate responses gures in my
account of the good critic, but not in Humes account of the true judge.
He is concerned with what allows a person to have appropriate sentiments.
One reason why I hold that being good at thinking of appropriate
responses helps make one a good critic is that this can compensate for a
critics inability to have those responses to culturally or historically distant
works. Hume discusses the judgement of culturally and historically distant
works. In judging oratory, he says, a critic of a different age or nation . . .
must place himself in the same situation as the audience, in order to form a
true judgment of the oration.29 But the point of this, it seems, is to enable
the judge to have an appropriate sentiment, not to enable her to think of
one. Regarding works containing speculative errors, he writes: There
needs but a certain turn of thought or imagination to make us enter into all
the opinions, which then prevailed, and relish the sentiments or conclu-
sions derived from them.30 By contrast, in the case of immoral works,
a very violent effort is requisite to change our judgment of manners, and
excite sentiments of approbation or blame, love or hatred, different from
those to which the mind from long custom has been familiarized. And where
a man is condent of the rectitude of that moral standard, by which he
judges, he is justly jealous of it, and will not pervert the sentiments of his heart
for a moment, in complaisance to any writer whatsoever.31
This suggests, though it does not show decisively, that the point of placing
ourselves in the situation of the original audience, when we are supposed
to do it, is only to enable a sentiment to be excited in us.
It may be, then, that Hume would not regard being good at thinking of
appropriate responses as a trait that makes one a true judge.32 By contrast,
being good at thinking of responses, and of much else besides, is a very
important characteristic of a good critic. This endowment will play an
important role in the next chapter.
29 30 31
Ibid., p. 239. Ibid., pp. 246247. Ibid., p. 247.
32
For discussion of a similar issue in Humes moral philosophy, see Barry Stroud, Hume
(London: Routledge and Kegan Paul, 1977), pp. 188192.
3
Criticism and Imagination
We are almost in a position to say what role imaginativeness plays in

criticism. We have an account of the aims of criticism, and we have an
account of what makes someone good at achieving these aims. In this
chapter, I will provide a new account of what imaginativeness is. On the
basis of my conclusions, I will then argue for a view about what the role of
imaginativeness in criticism is.
Before doing this, however, I want to say more about why one might
associate criticism and imaginativeness in the rst place. After all, it is
typically the production of artworks, not the reception of them, that we
associate with imaginativeness. Great artists are among the most salient
examples of imaginative people. Great critics, on the other hand, are
associated with judiciousness, sensitivity, and perceptiveness, but not,
normally, with imaginativeness.
Many who have reected on criticism, however, have thought that this
common-sense view of the critic is incomplete. They have held that
intelligently appreciating and criticizing artworks also involves, at least in
central or important cases, responding imaginatively to them.
They have arrived at this conclusion by several different routes. The
interpretation of artworks, especially literary works, has been thought to
support this view. The kinds of interpretation some literary critics provide,
and the fact that literary works seem to admit of multiple interpretations,
suggest to many that imaginativeness plays a crucial role in interpretation.
On this view, we cannot explain what makes someone good at interpret-
ation only in terms of qualities like sensitivity and judiciousness. Others
reach similar conclusions about the importance of imaginativeness in
criticism by thinking about aesthetic experience. According to them, a
distinctive kind of perceptual response, characterized by imaginativeness, is
involved in aesthetic experience, at least in certain important cases. For
example, some hold that when we perceive architecture or instrumental
54 CRITICISM AND IMAGINATION
music as expressive of emotion we perceive imaginatively. Likewise, many

see in the critics use of language an indication of imaginativenesss role in
criticism. Descriptions of artworks and their effects are often couched in
imaginative gurative language. Often, this not only serves a decorative
function, but also makes the critics writing more effective as criticism.
Many great works of criticism, such as the writings of John Ruskin and
Walter Pater, are striking examples of imaginative writing, and are read as
literature in their own right. Considerations like these have led many to
conclude that having the communication skills of a good critic involves, or
even requires, having the ability to use language imaginatively.
Such views have been inuential. Beliefs about the role of imaginative-
ness in interpretation have shaped some philosophers accounts of literary
meaning; their evaluations of certain kinds of literature; and even their
views on the ontology of literature. Views on the imaginativeness of
aesthetic experience have inuenced views on the nature and reality of
aesthetic properties. Famous instances of imaginative criticism have shaped
many theorists views on what the point of criticism is. Many have
appealed to claims about the critics imaginativeness to mount defences
of the value of criticism.
The persistence of this conception of criticism, and the variety of forms
it has taken, are striking. However, the history of this idea, and the reasons
why people have endorsed it, are likely to be unfamiliar. It is not widely
recognized that, for example, contemporary philosophers who appeal to
metaphor to argue for a signicant role for imaginativeness in criticism are
endorsing, on independent grounds, a view held by several of the most
important aestheticians of the last three hundred years. I will therefore
begin this chapter with a brief history of the view that appreciation and
criticism involve responding imaginatively to art. My aim is not to give a
comprehensive history of this idea, but to familiarize readers with the
debate to which I will be contributing. As this discussion will make
clear, my approach to this topic, as well as the conclusions I reach, differ
signicantly from those of others.
1. The pleasures of the imagination

Throughout the eighteenth century, philosophers of art took a great
interest in the question of what characteristics enable a person to appreciate
1. THE PLEASURES OF THE IMAGINATION 55
and judge art. Humes account of the characteristics of true judges in Of

the Standard of Taste is today the best known of these discussions, but
contributions to this debate were also made by Joseph Addison, the Abb
Du Bos, Edmund Burke, Alexander Gerard, and James Beattie, among
others.1 Among these endowments, imagination, or at least what philoso-
phers of the time called imagination, was commonly regarded as having a
crucial role. Kants view that aesthetic responses to free beauty are a matter
of the understanding and what he called imagination being set in a
harmonious free play is today the most familiar of these theories. Indeed,
Paul Guyer goes so far as to say that the central idea to emerge in
eighteenth-century aesthetics is the idea of the freedom of the imagin-
ation, and it was the attraction of this idea that provided much of the
impetus behind the explosion of aesthetic theory in the period.2 Whether
or not this is true, it is certainly true that philosophers of the time
repeatedly invoke imagination in their theories of our responses to art-
works. One of the main reasons why the imagination was so often
appealed to was to explain why artworks give us pleasure.
Addisons very inuential essays of 1712, On the Pleasures of the
Imagination, are about pleasures that arise from visible Objects, either
when we have them actually in our view, or when we call up their
Ideas into our Minds by Paintings, Statues, Descriptions, or any the like
Occasion.3 Pleasures arising in the latter way are obviously the ones more
closely related to what we would today describe as imagination, and
Addison calls them the Secondary Pleasures of the Imagination.4 Repre-
sentational artworks provide such pleasures, he says, by prompting that
Action of the Mind, which compares the Ideas arising from the Original
1
See Joseph Addison, No. 409 of The Spectator, in Joseph Addison et al., The Spectator, ed.
Donald F. Bond (Oxford: Oxford University Press, 1965), vol. 3, pp. 527531; Jean-Baptiste
Du Bos, Rexions Critiques sur la Posie et sur la Peinture (Paris: Pierre-Jean Mariette, 1733),
vol. 2, ss. 2129 (available at <http://books.google.co.uk/books?id=crYPAAAAQAAJ>
accessed 14 June 2012); Edmund Burke, A Philosophical Enquiry into the Origin of Our Ideas
of the Sublime and Beautiful, ed. Adam Phillips (Oxford: Oxford University Press, 1990),
pp. 1126; Alexander Gerard, An Essay on Taste, 3rd edn (Delmar, NY: Scholars Facsimiles
and Reprints, 1978); James Beattie, Of Taste, and its Improvement, in Selected Philosophical
Writings, ed. James A. Harris (Exeter: Imprint Academic, 2004), pp. 161182.
2
Paul Guyer, The Origins of Modern Aesthetics: 17111735, in Values of Beauty:
Historical Essays in Aesthetics (Cambridge: Cambridge University Press, 2005), p. 5.
3
Addison, No. 411 of The Spectator, vol. 3, pp. 536537.
4
Ibid., p. 537.
Objects, with the Ideas we receive from the Statue, Picture, Description,
or Sound that represents them.5 We cannot explain why the activity of
comparison is pleasurable, Addison says, but many examples, such as our
pleasure in mimicry, rhyme, and puns, establish that it is.6 Addison uses this
principle to provide a solution to the classic problem of how artistic
representations of what is unpleasant can cause us pleasure. When some-
thing disagreeable is described, he says, it is not the image of the disagree-
able thing that provides pleasure. Rather, if the description is apt, we
experience pleasure because the mind compares the Ideas that arise from
Words, with the Ideas that arise from the Objects themselves.7 Similarly,
the representation of something dangerous, as in a tragedy, causes the
Pleasure we receive from the Sense of our own Safety, which results from
the secret Comparison which we make between our selves and the Person
who suffers.8
Hume also appeals to imagination to explain our pleasure in beautiful
objects. In the Treatise, he argues that the imagination has a set of passions
belonging to it, upon which our sentiments of beauty much depend.9 In
particular, most kinds of beauty are derivd from sympathetic passions.10
For example,
A man, who shows us any house or building, takes particular care . . . to point
out the convenience of the apartments, the advantages of their situation, and
the little room lost in the stairs, anti-chambers and passages; and indeed tis
evident, the chief part of the beauty consists in these particulars. . . . As this is a
beauty of interest, not of form, so to speak, it must delight us merely by
communication, and by our sympathizing with the proprietor of the lodging.
We enter into his interest by the force of imagination, and feel the same
satisfaction, that the objects naturally occasion in him.11
Likewise, someone who knows that a plain, overgrown with furze and
broom has less agricultural value than a hill coverd with vines or olive-
trees will always nd the latter more beautiful than the former:
5
6
Ibid., p. 560.
7
8
Ibid., p. 568.
9
David Hume, A Treatise of Human Nature, ed. David Fate Norton and Mary J. Norton
(Oxford: Oxford University Press, 2000), III.iii.1, p. 373.
10
Ibid., II.v.5, p. 235.
11
Ibid.
1. THE PLEASURES OF THE IMAGINATION 57
this is a beauty merely of imagination, and has no foundation in what appears

to the senses. Fertility and value have a plain reference to use; and that to
riches, joy, and plenty; in which tho we have no hope of partaking, yet we
enter into them by the vivacity of the fancy, and share them, in some
measure, with the proprietor.12
Kant, as is well known, also thinks pleasure in beauty derives from an

activity of the imagination, namely, the harmonious free play of imagination
and understanding. However, in Kants discussion of ne art, we also nd an
account of the effect on the imagination produced by artworks possessing a
special merit he calls spirit. Artworks can lack spirit even if we nd nothing
to censure in them as far as taste goes.13 What is distinctive of artworks with
spirit is that they furnish their readers, audiences, or viewers with aesthetic
ideas. Aesthetic ideas are mental images with two distinguishing features.
First, they suggest rational ideas to us. Rational ideas are concepts of that
which we cannot experience or imagine, such as God, freedom, and im-
mortality. Second, aesthetic ideas stimulate the imagination into producing
many other images related to the rational ideas they suggest. In Kants words,
an aesthetic idea encourages the imagination to spread its ight over a whole
host of kindred representations that provoke more thought than admits of
expression in a concept determined by words.14 This, indeed, is the proper
task of an aesthetic idea.15 Thus an aesthetic idea allows a concept to be
supplemented in thought by much that is indenable in words, and so with
language, as a mere thing of the letter, [it] combines spirit.16
To illustrate these points, Kant gives the example of an artwork using an
eagle to symbolize the god Jupiter. According to Guyers reading of this
passage, the aesthetic idea furnished by such a work is the image of Jupiter
himself. The idea of Jupiter, in turn, suggests the rational idea of the
12
Ibid. See also ibid., III.iii.1, pp. 368369, 373374. For a recent study of Humes views
on these matters, to which I am indebted, see Paul Guyer, The Standard of Taste and the
Most Ardent Desire of Society , in Values of Beauty: Historical Essays in Aesthetics (Cam-
bridge: Cambridge University Press, 2005), pp. 4657. Sibley suggested a similar explanation
of why we aesthetically admire certain qualities for themselves. See Frank Sibley, Aesthetics
and the Looks of Things, in Approach to Aesthetics: Collected Papers on Philosophical Aesthetics,
ed. John Benson, Betty Redfern, and Jeremy Roxbee Cox (Oxford: Oxford University Press,
2001), p. 31.
13
Immanuel Kant, Critique of Judgement, ed. Nicholas Walker, trans. James Creed Mere-
dith and Nicholas Walker (Oxford: Oxford University Press, 2007), Ak. 5: 313.
14
Ibid., Ak. 5: 315.
15
Ibid.
16
Ibid., Ak. 5: 316.
sublimity and majesty of creation,17 and stimulates the imagination into

ranging over a host of representations kindred to the idea of natures
sublimity and majesty.18
Part of the signicance for Kant of this imaginative activity is that it gives
us a feeling of freedom from psychological laws of the association of ideas.
In these responses to artworks, the imagination . . . displays a creative
activity, and it sets the faculty of intellectual ideas (reason) into move-
ment.19 The imagination is following principles which have a . . . seat in
reason, and
by this means we come to feel our freedom from the law of association
(which attaches to the empirical employment of the imagination), with the
result that the [perceptible] material [supplied by the artwork] can be
borrowed by us from nature in accordance with that law, but be worked
up by us into something elsenamely, what surpasses nature.20
Kants account of our responses to such artworks is similar to his account of

the sublime, according to which our feeling of pleasure in the sublime,
too, is due to our awareness of aspects of ourselves that transcend nature.21
Many other philosophers in this period, including Burke, Lessing,
Moses Mendelssohn, Gerard, Beattie, and Archibald Alison, also assign
the imagination an important role in criticism and appreciation.22 Alison
lays particular stress on the role of imagination in responses to beauty and
sublimity. In Alisons view, when something excites the emotions of
beauty and sublimity, the simple perception of the object, we frequently
17
Ibid., Ak. 5: 315.
18
See Paul Guyer, Kant and the Claims of Taste, 2nd edn (Cambridge: Cambridge
19
Kant, Critique of Judgement, Ak. 5: 315.
20
Ibid., Ak. 5: 314.
21
See ibid., ss. 2529.
22
See Burke, A Philosophical Enquiry into the Origin of Our Ideas of the Sublime and Beautiful,
pp. 1626; Gotthald Ephraim Lessing, Laocon: An Essay on the Limits of Painting and Poetry, ed.
and trans. Edward Allen McCormick (Indianapolis: Bobbs-Merrill, 1962), ch. 3; Moses
Mendelssohn, On the Sublime and Naive in the Fine Sciences, in Philosophical Writings,
ed. and trans. Daniel O. Dahlstrom (Cambridge: Cambridge University Press, 1997),
pp. 195196; Gerard, An Essay on Taste, Part III, s. 1, entitled How far Taste depends on
the Imagination; Beattie, Of Taste, and its Improvement, pp. 160167; Archibald Alison,
Essays on the Nature and Principles of Taste, 2nd edn (Boston: Cummings and Hilliard, 1812),
Essay I (available at <http://books.google.co.uk/books?id=_PIRAAAAYAAJ> accessed 14
June 2012).
2. THE CRITIC AS ARTIST 59
nd, is insufcient to excite these emotions.23 The emotions produced in

a person who experiences something beautiful or sublime can be ex-
plained only by a train of thought being immediately awakened in his
imagination:
The landscapes of Claude Lorrain, the music of Handel, the poetry of Milton,
excite feeble emotions in our minds, when our attention is conned to the
qualities they present to our senses. . . .
It is then, only, we feel the sublimity or beauty of their productions, when
our imaginations are kindled by their power, when we lose ourselves amid
the number of images that pass before our minds, and when we waken at last
from this play of fancy, as from the charm of a romantic dream.24
This emphasis in both Kant and Alison on the multitude of thoughts,

images, and associations produced in aesthetic experience marks a differ-
ence between their accounts of imaginative response and those of earlier
writers. Addison and Hume, for example, do not stress this. This concep-
tion of imaginative responses to artworks would become more common in
the work of later writers.
2. The critic as artist

In the nineteenth century, the imaginative powers of the critic were
frequently compared to those of the artist. The primary purpose of the
comparison was often to defend the practice of criticism, commonly
against attacks by artists. Thus Charles Baudelaire replied to artists com-
plaints about criticism by asserting that the best criticism is the criticism
that is entertaining and poetic, and that the best accounts of a picture may
well be a sonnet or an elegy.25 One of the more popular ways of
defending criticism, it seems, was to argue that great critics display qualities
highly valued in artists.
Matthew Arnolds celebrated essay, The Function of Criticism at the
Present Time, begins by discussing a biography of Wordsworth in which
Wordsworth is reported as holding the critical power very low, innitely
23
Alison, Essays on the Nature and Principles of Taste, p. 18.
24
Ibid., pp. 1819.
25
Charles Baudelaire, The Salon of 1846, in Selected Writings on Art and Literature, ed. and
trans. P. E. Charvet (London: Penguin, 1992), p. 50.
lower than the inventive, and as having said that the time spent criticizing
literature would be better spent trying to compose it.26 Arnold writes:
The critical power is of lower rank than the creative. True; but in assenting
to this proposition, one or two things are to be kept in mind. It is
undeniable that the exercise of a creative power, that a free creative
activity, is the highest function of man; it is proved to be so by mans
nding in it his true happiness. But it is undeniable, also, that men may
have the sense of exercising this free creative activity in other ways than in
producing great works of literature or art; if it were not so, all but a very
few men would be shut out from the true happiness of all men. They may
have it in well-doing, they may have it in learning, they may have it even
in criticising.27
While granting that judging is often spoken of as the critics one business,
and so in some sense it is, Arnold goes on to say that mere judgment and
application of principles is, in itself, not the most satisfactory work to the
critic; like mathematics, it is tautological, and cannot well give us, like fresh
learning, the sense of creative activity.28 Arnold concludes his essay by
reiterating that to have the sense of creative activity is the great happiness
and the great proof of being alive, and it is not denied to criticism to have
it; but then criticism must be sincere, simple, exible, ardent, ever
widening its knowledge. Then it may have, in no contemptible measure,
a joyful sense of creative activity.29 Nevertheless, Arnold tempers his
conclusion by adding: Still, in full measure, the sense of creative activity
belongs only to genuine creation; in literature we must never forget
that.30
Oscar Wilde regarded these claims as too moderate. His dialogue, The
Critic as Artist, is in large part a reply to Arnolds essay. In reply to the
assertion that the creative faculty is higher than the critical, Wilde has
one of the interlocutors answer that that ne spirit of choice and delicate
instinct of selection by which the artist realizes life for us . . . is really the
critical faculty in one of its most characteristic moods, and no one who
26
Quoted in Matthew Arnold, The Function of Criticism at the Present Time, in
Lectures and Essays in Criticism, ed. R. H. Super (Ann Arbor: The University of Michigan
Press, 1962), p. 259.
27
Ibid., p. 260.
28
Ibid., p. 283.
29
Ibid., p. 285.
30
Ibid.
2. THE CRITIC AS ARTIST 61
does not possess this critical faculty can create anything at all in art.31 To
the objection that a great work of art will be a thing so complete and
perfect that there will be nothing left for the critic to do, he replies that:
Criticism is itself an art. And just as artistic creation implies the working of the
critical faculty . . . so Criticism is really creative in the highest sense of the
word. Criticism is, in fact, both creative and independent. . . . Criticism is no
more to be judged by any low standard of imitation or resemblance than is
the work of poet or sculptor. The critic occupies the same relation to the
work of art that he criticizes as the artist does to the visible world of form and
colour, or the unseen world of passion and of thought.32
This leads to an attack on Arnolds dictum that the proper aim of

Criticism is to see the object as in itself it really is. This, Wilde writes, is
a very serious error:
the highest Criticism, being the purest form of personal impression, is in its
way more creative than creation, as it has least reference to any standard
external to itself. . . . No ignoble considerations of probability, that cowardly
concession to the tedious repetitions of domestic or public life, affect it
ever. . . . Who cares whether Mr Ruskins views on Turner are sound or
not? What does it matter? That mighty and majestic prose of his, so fervid and
so ery-coloured in its noble eloquence, so rich in its elaborate symphonic
music, so sure and certain, at its best, in subtle choice of word and epithet, is
at least as great a work of art as any of those wonderful sunsets that bleach or
rot on their corrupted canvases in Englands Gallery. . . . Some resemblance,
no doubt, the creative work of the critic will have to the work that has stirred
him to creation, but it will be such resemblance as exists, not between Nature
and the mirror that the painter of landscape or gure may be supposed to
hold up to her, but between Nature and the work of the decorative artist.33
Wilde thinks that, along with Ruskins criticism, Paters famous descrip-
tion of the Mona Lisa embodies this ideal. Pater writes that, as depicted by
Leonardo, Mona Lisa
31
Oscar Wilde, The Critic as Artist, in The Major Works, ed. Isobel Murray (Oxford:
Oxford University Press, 2000), p. 253. Despite ranking the critical faculty below the creative,
Arnold did agree that the sound exercise of the one is necessary for the sound exercise of
the other: the critical power . . . tends, at last, to make an intellectual situation of which the
creative power can protably avail itself. It tends to establish an order of ideas . . . to make the
best ideas prevail. . . . Out of this stir and growth come the creative epochs of literature
(Arnold, The Function of Criticism at the Present Time, p. 261).
32
Wilde, The Critic as Artist, p. 260.
33
Ibid., pp. 261262, 266.
is expressive of what in the ways of a thousand years man had come to desire.
Hers is the head upon which all the ends of the world are come, and the
eyelids are a little weary. It is a beauty wrought out from within upon the esh,
the deposit, little cell by cell, of strange thoughts and fantastic reveries and
exquisite passions. Set it for a moment beside one of those white Greek
goddesses or beautiful women of antiquity, and how would they be troubled
by this beauty, into which the soul with all its maladies has passed? All the
thoughts and experience of the world have etched and moulded there in that
which they have of power to rene and make expressive the outward form, the
animalism of Greece, the lust of Rome, the reverie of the middle age with its
spiritual ambition and imaginative loves, the return of the Pagan world, the sins
of the Borgias. She is older than the rocks among which she sits; like the
vampire, she has been dead many times, and learned the secrets of the grave; and
has been a diver in deep seas, and keeps their fallen day about her; and trafcked
for strange webs with Eastern merchants; and, as Leda, was the mother of Helen
of Troy, and, as Saint Anne, the mother of Mary. . . . The fancy of a perpetual
life, sweeping together ten thousand experiences, is an old one; and modern
thought has conceived the idea of humanity as wrought upon by, and summing
up in itself, all modes of thought and life. Certainly Lady Lisa might stand as the
embodiment of the old fancy, the symbol of the modern idea.34
A great critic, in Wildes view, will be attracted to such works as make

him brood and dream and fancy, to works that possess the subtle quality of
suggestion, and seem to tell one that even from them there is an escape
into a wider world.35
3. Criticism and metaphor

In contemporary philosophy, those who believe imagination has a signi-
cant role in criticism hold this view for various reasons. One of the main
reasons is that vivid, novel metaphors are often used when critics describe
34
Walter Pater, Studies in the History of the Renaissance, ed. Matthew Beaumont (Oxford:
Oxford University Press, 2010), pp. 7071.
35
Wilde, The Critic as Artist, p. 265. Compare Harold Bloom: criticism is either a genre
of literature or it is nothing. It has no hope for survival unless it is a genre of literature
(Antonio Weiss, Harold Bloom: The Art of Criticism No. 1, The Paris Review 118, Spring
1991, <http://www.theparisreview.org/interviews/2225/the-art-of-criticism-no-1-harold-
bloom> accessed 14 June 2012). See also Geoffrey H. Hartman, How Creative Should
Literary Criticism Be?, The New York Times Book Review, 5 April 1981, <http://www.
nytimes.com/1981/04/05/books/how-creative-should-literary-criticism-be.html> accessed
14 June 2012; James Elkins, What Happened to Art Criticism? (Chicago: Prickly Paradigm Press,
2003), ch. 2.
3. CRITICISM AND METAPHOR 63
artworks and our responses to them. This aspect of criticism will be the
focus of the second half of this book.
The use of metaphor in criticism is not itself a recent phenomenon.
Famously, the term katharsis, which originally meant cleansing or puri-
cation, is used non-literally by Aristotle to characterize the emotional
effect of good tragedy on an audience. Horace coined the phrase purple
patch to describe an excessively ornate descriptive passage in a piece of
writing.36 Connoisseurs of Chinese jades have for centuries distinguished
between the colours of lychee-esh, spinach, and mutton-fat jade,
among many other kinds.37 Examples of this sort could be multiplied.
However, philosophical reection on the signicance of metaphors
prevalence in criticism is more recent. Two factors, in particular, have
brought the use of art-critical metaphors to philosophers attention.
The rst is the observation that many so-called aesthetic descriptions
are metaphorical. In several very inuential papers, Frank Sibley distin-
guished between what he called aesthetic terms, such as unied, balanced,
integrated, lifeless, serene, sombre, dynamic,38 and non-aesthetic terms, such as
red, noisy, brackish, clammy, square.39 Sibley claims that the difference
between them is that it requires the exercise of taste, perceptiveness, or
sensitivity, of aesthetic discrimination or appreciation to judge that an
aesthetic term applies to something, but not to judge that a non-aesthetic
term applies.40 Sibley also pointed out that when we employ words as
aesthetic terms we are often making and using metaphors.41 We might
describe a passage of music as chattering, carbonated, or gritty, a painters
colouring as vitreous, farinaceous, or effervescent, or a writers style as
glutinous, or abrasive.42 Some have suggested that literal aesthetic descrip-
tions are actually in the minority.
36
Horace, Ars Poetica, in The Satires and Epistles of Horace and Persius, ed. and trans. Niall
Rudd (Harmondsworth: Penguin, 1979), p. 190, l. 15.
37
See Craig Clunas, Jade Carvers and Their Customers in Ming China, The Bulletin of the
Friends of Jade 6 (1989): 36; Angus Forsyth and Brian McElney, Jades from China (Bath: The
Museum of East Asian Art, 1994), pp. 304, 354355.
38
Frank Sibley, Aesthetic Concepts, in Approach to Aesthetics, p. 1. See also Frank Sibley,
Aesthetic and Non-Aesthetic, in Approach to Aesthetics, pp. 3351.
39
Sibley, Aesthetic Concepts, p. 2.
40
Ibid., p. 1.
41
Ibid., p. 2.
42
Ibid., p. 2 n. 2.
The second factor is the belief that metaphors employing psychological,

motor, and spatial concepts are common, or even indispensable, when we
describe important aspects of artworks, especially music. It is widely
believed that descriptions of a works expressive character are often meta-
phorical, as when we say that a piece of music is sad, jaunty, or pensive.
Similarly, some hold that descriptions of melodies as rising and falling, of
musical movements, of chords as hollow and open, and so forth, are
metaphors capturing something essential about music.
Many philosophers have been struck by these facts, or apparent facts,
about art-critical metaphors. Why are so many aesthetic descriptions
metaphorical? Whether aesthetic descriptions are those we need taste to
apply, or those that attribute aesthetic properties, or something else, the
fact that so many are metaphorical calls for explanation. There is no
antecedently obvious reason why such descriptions should tend to be
metaphors. Many feel that the prevalence of metaphor in criticism must
reect something important about the nature of our responses to art; about
artistic expression; or about aesthetic properties.
Numerous views about what metaphor reveals about these matters have
been advanced and disputed.43 In particular, many have concluded that
43
See, for example, Roger Scruton, Art and Imagination (London: Methuen, 1974);
Nelson Goodman, Languages of Art: An Approach to a Theory of Symbols, 2nd edn (Indianapolis:
Hackett, 1976), ch. 2; Monroe C. Beardsley, What Is an Aesthetic Quality?, in The Aesthetic
Point of View: Selected Essays, ed. Michael J. Wreen and Donald M. Callen (Ithaca: Cornell
University Press, 1982), pp. 106110; Malcolm Budd, Music and the Emotions (London:
Routledge and Kegan Paul, 1985); Peter Kivy, Sound Sentiment: An Essay on the Musical
Emotions (Philadelphia: Temple University Press, 1989), pp. 5455; Richard Wollheim,
Correspondence, Projective Properties, and Expression in the Arts, in The Mind and Its
Depths (Cambridge, MA: Harvard University Press, 1993), pp. 147148; Stephen Davies,
Musical Meaning and Expression (Ithaca: Cornell University Press, 1994), pp. 137166; Roger
Scruton, The Aesthetics of Music (Oxford: Oxford University Press, 1997); Roger Scruton,
Understanding Music, in The Aesthetic Understanding: Essays in the Philosophy of Art, 2nd edn
(South Bend, IN: St Augustines Press, 1998), pp. 89115; Derek Matravers, Art and Emotion
(Oxford: Oxford University Press, 1998), ch. 6; Nick Zangwill, The Metaphysics of Beauty
(Ithaca: Cornell University Press, 2001), ch. 10; R. A. Sharpe, Philosophy of Music: An
Introduction (Chesham: Acumen, 2004), pp. 102108; Roger Scruton, Musical Movement:
A Reply to Budd, British Journal of Aesthetics 44/2 (2004): 184187; Paul Boghossian,
Explaining Musical Experience, in Kathleen Stock, ed., Philosophers on Music: Experience,
Meaning, and Work (Oxford: Oxford University Press, 2007), p. 123; Malcolm Budd,
Understanding Music, The Characterization of Aesthetic Qualities by Essential Metaphors
and Quasi-Metaphors, and Aesthetic Realism and Emotional Qualities of Music, in Aesthetic
Essays (Oxford: Oxford University Press, 2008), pp. 122153, 171184; Christopher Pea-
cocke, The Perception of Music: Sources of Signicance, British Journal of Aesthetics 49/3
(2009): 257275; Paul F. Snowdon, Peacocke on Musical Experience and Hearing
4. MY APPROACH 65
the prevalence of metaphor in critical language reects the major role

imagination plays in aesthetic experience. Roger Scruton is the best-
known exponent of this view. Scruton holds that metaphors express an
experience of imaginatively perceiving one thing as another. The frequent
use of metaphor in aesthetic description therefore indicates that aesthetic
experience is itself an experience of imaginative perceiving-as. Scruton
also argues that, in order to hear sound as music at all, we need to employ
concepts of movement and space metaphorically in auditory perception.44
Linguistic metaphors employing concepts of space and movement occur
in descriptions of music because musical experience itself is founded in
metaphor, arising when unreal movement is heard in imaginary space.45
As we shall see in Chapter 6, Scruton views his conclusions about the role
of imagination and metaphor in responses to art as lending support to an
anti-realist view of aesthetic properties. Debate about each of these matters
continues.46
4. My approach
My own approach to this topic in the remainder of this book differs from
that of others in several respects.
Metaphorically-As, British Journal of Aesthetics 49/3 (2009): 277281; Christopher Peacocke,

Experiencing Metaphorically-As in Music Perception: Clarications and Commitments,
British Journal of Aesthetics 49/3 (2009): 299306; Christopher Peacocke, Music and Experi-
encing Metaphorically-As: Further Delineation, British Journal of Aesthetics 50/2 (2010): 189
191; Stephen Davies, Music and Metaphor, in Musical Understandings and Other Essays on the
Philosophy of Music (Oxford: Oxford University Press, 2011), pp. 2133. This is not an
exhaustive list.
44
Scruton, Understanding Music; The Aesthetics of Music, chs 3 and 6; and Musical
Movement: A Reply to Budd.
45
Scruton, The Aesthetics of Music, p. 239.
46
For discussion of Scrutons views on these matters, see, in addition to several of the
references given already: Philip Pettit, The Possibility of Aesthetic Realism, in Eva Schaper,
ed., Pleasure, Preference and Value: Studies in Philosophical Aesthetics (Cambridge: Cambridge
University Press, 1983), pp. 1738; Berys Gaut, Metaphor and the Understanding of Art,
Proceedings of the Aristotelian Society 97 (1997): 223241; Paul Boghossian, On Hearing the
Music in the Sound: Scruton on Musical Expression, The Journal of Aesthetics and Art Criticism
60/1 (2002): 4955; Brandon Cooke, Imagining Art, British Journal of Aesthetics 47/1 (2007):
2945; Rafael De Clercq, Melody and Metaphorical Movement, British Journal of Aesthetics
47/2 (2007): 156168; Malcolm Budd, Musical Movement and Aesthetic Metaphors, in
Aesthetic Essays (Oxford: Oxford University Press, 2008), pp. 154170.
First, we ought to distinguish between (i) doing or producing some-

thing imaginative, and (ii) doing or producing something by using the
imagination. Many locutions we use blur this distinction. We can char-
acterize someone as having done something imaginative by saying that
she used her imagination, put a great deal of imagination into it, or
showed she has an excellent imagination. Conversely, we can use the
word imaginative in the sense given by the OED as of, relating to, or
concerned in the exercise of imagination as a mental faculty. This is the
sense in which we speak of imaginative literature to distinguish it from
non-ction. It is not, however, the sense in which the imaginative is
opposed to the unimaginative. Rather, it is the sense in which the
imaginative is opposed to what does not relate to imagining. Moreover,
we do not use imaginative, in this sense, to attribute the property of
imaginativeness.
In what follows, I will use imaginative only in those senses in which
the imaginative is opposed to the unimaginative, and in which we use it
to attribute imaginativeness. It is not the case that whenever one has
imagined, one has done or produced something imaginative, as opposed
to unimaginative.47 There can be imaginative imagining and unimagina-
tive imagining. Many works of imaginative literature are unimaginative.
Moreover, I will argue below that one need not imagine to do or produce
something imaginative.
My focus, then, is on the role of imaginativeness, not imagination, in
criticism. It is often unclear, in other accounts, whether the claims being
made are about imagination or imaginativeness, or both. Much of the
interest of the claims that have been made about the place of imagination
in criticism is due to what these claims suggest about the role of imagina-
tiveness in criticism. Kants view that an aesthetic idea encourages the
imagination to spread its ight over a whole host of kindred representa-
tions suggests that there is great scope for imaginativeness in engaging
with artworks that furnish aesthetic ideas. The conception of the critic as
artist, which so attracted Wilde and many others, plainly involved
47
Compare Berys Gauts claim that not all imagining involves creative acts (Berys Gaut,
Creativity and Imagination, in Berys Gaut and Paisley Livingston, eds., The Creation of Art:
New Essays in Philosophical Aesthetics (Cambridge: Cambridge University Press, 2003), p. 155).
Gauts point is about creativity, but he regards imaginative, in one use, as a synonym for
creative (ibid., p. 151).
4. MY APPROACH 67
regarding great criticism as manifesting imaginativeness. Paters description

of the Mona Lisa is exceptionally imaginative. Scruton claims that imagin-
ing must go beyond the obvious and involves thinking of a description as
appropriate to an object;48 but as we shall see, these points are, in fact,
closer to the truth about imaginativeness. And what is striking about the
metaphors used in good criticism is that so many are imaginative or express
imaginative responses to artworks. It is not, or not merely, that they
indicate that criticism involves imagining.
To make these points is not to deny that there are important and
interesting questions about how imagining, whether imaginative or
unimaginative, is involved in our engagement with art. Rather, it is to
separate out a line of inquiry that is commonly bound up with such
questions, and which is interesting in its own right.
Second, my focus is on the role of imaginativeness in criticism, not
in aesthetic experience or appreciation. Other philosophers draw con-
clusions regarding the role of imaginativeness in criticism, in appreci-
ation, and in aesthetic experience from the prevalence of metaphor
in criticism. I do not deny that one can draw conclusions about these
other matters from considerations about art-critical metaphors. But
one must distinguish these questions carefully. The prevalence of im-
aginative metaphors in criticism suggests, in the rst instance, that
imaginativeness has a role of some kind in criticism. Moreover, as we
shall see, not all effective, imaginative art-critical metaphors indicate
anything about the role of imaginativeness in appreciation or aesthetic
experience.
Third, I will remain neutral on the question of the reality or mind-
independence of aesthetic properties. The account of metaphor I will
adopt is consistent with both realism and anti-realism about aesthetic
properties. And although I have said that appreciative responses to
artworks can include imaginings, I do not claim (or deny) that these
imaginings put the reality or mind-independence of aesthetic properties
in doubt.
To say what the role of imaginativeness in criticism is, we need an
account of what imaginativeness is.
48
Scruton, Art and Imagination, pp. 97100.
5. The imaginativeness of acts, omissions,

and products
I am interested in what it is for a person to be imaginative. To explain this,
however, I need to explain what it is for a persons acts (including her
mental acts), her omissions, and the products of her acts to be imaginative.
Acts, such as impersonations, retorts, acts of kindness, and theatrical
performances, and the products of a persons acts, such as stories, haircuts,
paintings, and plans, can be imaginative in a different sense than persons
can be. Omissions can also be imaginative. It might be imaginative of a
novelist, in writing dialogue, to leave certain things unsaid by the charac-
ters. The imaginativeness of persons is to be understood in terms of the
imaginativeness that acts, omissions, and products can have.49
The rst point to note is that an act, omission, or product that is both an
f and a g can be an imaginative f but not an imaginative g. For example,
dancing might be an imaginative way of getting someones attention in a
crowded train station. But it would not be an imaginative way of respond-
ing to a request to demonstrate what dancers do. To take another example,
after the novel We Need to Talk About Kevin had become a best-seller, there
appeared a popular-science book entitled We Need to Talk About Kelvin.
That is an imaginative title for a popular-science book. But it would not be
an imaginative subject-line for an email in which you tell someone you
need to talk about your troubled friend Kelvin. In these examples, what is
imaginative is imaginative as an act or product of a certain typeas a way
of getting someones attention and as a title for a popular-science book. In
what follows, I will say what it is for a product or act to be imaginative as a
product or act of a certain type. Whenever I use the schema an imagina-
tive f , I am talking about what is imaginative as an f.
Second, imaginativeness is relative to persons and contexts. Using a
stream-of-consciousness style was, at one time, an imaginative way of
writing a novel. By the 1930s, however, it no longer was. Similarly, a
solution to a problem might be an imaginative solution for a child to think
of, but not an imaginative solution for an adult to think of. Ascriptions of
49
Perhaps there can also be imaginative emotions and attitudes. Perhaps it can be
imaginative, in certain situations, to feel affection for someone you normally dislike, or to
adopt a tranquil attitude in the midst of a struggle. I will not discuss emotions and attitudes,
but the account I provide could be extended, with slight alterations, to imaginative emotions
and attitudes, if there are any.
5. THE IMAGINATIVENESS OF ACTS , OMISSIONS , AND PRODUCTS 69
imaginativeness to something do not always specify a (type of ) person or

context, relative to which the thing is imaginative. But when they do not,
they rarely, if ever, imply that what is described as imaginative would be
imaginative relative to any person or context.50
Third, something can be an imaginative f to f but not an imaginative f
to c. Marshalling her forces in a certain pattern could be a highly imagina-
tive tactic for a general to think of at one time, without being an
imaginative tactic for her to adopt at that time. Perhaps the tactic would
be an imaginative tactic to adopt only if a certain kind of situation arises
on the battleeld. It would be suicidal, and not at all imaginative, to adopt
it beforehand. But it could still be an imaginative tactic to think of
beforehand.51
This example reveals a fourth point. Marshalling her forces in that
pattern might be an imaginative tactic for the general to think of if she
has thought it up as a response to one kind of situation. But it may not be
an imaginative tactic for her to think of if she thought it up as a response to
another kind of situation. Whether something is an imaginative f to think
of depends not only on the context in which it is thought of. It can depend
on the context for which it is thought of.52
These facts will help us provide an account of what imaginativeness is.
Preferably, an account of imaginativeness should also enable us to explain
these facts. I now turn to the construction of such an account.
It is natural to think that an imaginative f must be an original f.
Imaginativeness and originality are closely associated, and one might
50
We should be aware of a possible ambiguity here. If we say, That solution was an
imaginative solution for a child to think of , we are relativizing the imaginativeness of the
solution to children. However, we can use the construction, X was an imaginative f for N to
f, without relativizing the imaginativeness of X to N. Suppose you say, Painting Christ that
way was an imaginative approach for Salvador Dal to take. You do not necessarily mean
that, even for Salvador Dal, this approach was imaginative. That would imply that it was even
more imaginative than what Dal normally does. You may mean only that it was imaginative
of Dal to take that approach in his painting. There are parallel cases of this kind. Compare
saying, That was a megalomaniacal thing for Napoleon to do. We do not necessarily mean
that, even for Napoleon, that was a megalomaniacal thing to do. We may mean only that it
was megalomaniacal of Napoleon to do that. If we are not alert to this ambiguity, it can seem
as though we explicitly relativize imaginativeness to persons more often than we actually do.
51
This is not to imply, however, that an imaginative f can always be described as an
imaginative f to f.
52
This is not to imply that there always is a context for which an imaginative f is thought
of.
think originality is a necessary condition of imaginativeness. However, if

being original means being new or very different in some salient, valuable,
or signicant respect, this proposal will not work. Another popular-
science writer could independently think of the title We Need to Talk
About Kelvin after the rst one did. Her book would have an imaginative
title too. But her books title would not be new or very different in any
salient, valuable, or signicant respect.
This example suggests another possibility. What seems to matter here
is that the second writer came up with her title independently of the rst. If
she had gotten the idea from the earlier books title, her title would not
have been imaginative. So perhaps an imaginative f cannot be a derivative f.
Being new or unprecedented in signicant ways is not necessary, but
perhaps being non-derivative is. Indeed, it is plausible that original is
sometimes used to mean just not derivative.53
It is certainly true that many imaginative acts and products are not
derivative. And this is not accidental. Below, I will explain why imagina-
tiveness and non-derivativeness are so often connected. Nevertheless, it is
not true that an imaginative f cannot be a derivative f.54 Suppose that a
poet, knowledgeable about the history of literature, chooses to write a
poem in a form not used for centuries, as a result of reading poems written
in that form. This form turns out to be strikingly effective and appropriate.
Using that form could be an imaginative way of writing a poem today. But
it is a derivative way of writing a poem today. So being a derivative f is
consistent with being an imaginative f.
A better suggestion is that imaginativeness is necessarily connected, not
with the new or the underivative, but with the unobvious. Using the
archaic poetic form is an imaginative way of writing a poem today, despite
being derivative, partly because, nowadays, using that form is not an
obvious way of writing a poem. If it had been an obvious way of writing
a poem, it would not have been an imaginative way of writing a poem.
This appears to cover other examples as well. We Need to Talk About Kelvin
is not an obvious title to think of; dancing is not an obvious way of getting
53
Sibley claims this. See Frank Sibley, Originality and Value, in Approach to Aesthetics,
p. 121.
54
In earlier work, I claimed that non-derivativeness is a necessary condition of imagina-
tiveness. See James Grant, The Value of Imaginativeness, Australasian Journal of Philosophy
90/2 (2012): 277278.
someones attention in a crowded train station; and using a stream-of-

consciousness style was an unobvious way of writing a novel to adopt.
I suggest that the unobvious, rather than the new or the underivative, is
the notion we want.
There are, however, at least three kinds of obviousness. Only one of
these kinds is the kind I mean to oppose to imaginativeness.
The rst kind is exemplied by what is perceptually salient. A colossal
sculpture, for instance, might be an obvious feature of the landscape. The
second kind is the kind we have in mind when we say, It is obvious that p.
We might say that it is obvious that she is happy about something, or that
the artwork is a forgery, or that it will be hours before we get home. Here
we are not talking about how perceptually salient something is. We are
talking about how evident it is that something is true. Indeed, It is obvious
that p is at least nearly synonymous with It is evident that p. We do not
always use the construction It is obvious that p when we have this kind of
obviousness in mind. A fact can be described simply as obvious, and we
might describe an artwork as an obvious forgery, meaning that it is
obvious that it is a forgery.
The third kind of obviousness is the kind we can attribute with the
construction, Such-and-such is an obvious f to f. We can describe
something as an obvious move to make, an obvious strategy to adopt, or
an obvious description to come up with. Here, we are not making a point
about how perceptually salient something is, or how evident it is that
something is true. There is perhaps no succinct phrase that captures this
kind of obviousness as well as perceptually salient and evident capture
the other two kinds. The OED, however, supplies a rough equivalent:
roughly (but only roughly), what is obvious in this way is such as common
sense might suggest. We do not always use the construction an obvious
f to f when attributing this kind of obviousness. We might describe the
use of a certain poetic form simply as obvious or as an obvious way of
writing a poem, meaning that it is an obvious way of writing a poem to
think of.
My claim is that imaginativeness is opposed to this third kind of
obviousness. Each example that led us to connect imaginativeness with
the unobvious is an unobvious f to f. In fact, each of these examples is,
specically, an unobvious f to think of: an unobvious title to think of, an
unobvious way of getting someones attention to think of, an unobvious
way of writing a poem to think of, and an unobvious way of writing a

novel to think of.
Being an unobvious f to think of is, I suggest, the notion we need.
Before I develop this proposal further, however, it will help to clarify two
things about thinking-of.
First, when we use think of as I am using it, we mean think up or
come up with. Sometimes, think of means think about, bring to
mind, or consider. You might be asked to think of an elephant, or to
think of your summer vacation, or to think of those less fortunate than
you. But when we say that you thought of a solution to your problem, and
mean that you came up with a solution to your problem, we do not mean
that you thought about a solution, and we do not mean only that you
brought one to mind (if this is even part of what we mean). You can bring
to mind a solution you did not come up with, if you have learned about it;
and after you have thought of a solution yourself, you may later, perhaps
while reminiscing proudly, think of the solution you thought ofi.e., you
may bring to mind the solution you came up with. But recalling a solution
is not coming up with it again.
Second, saying that a person thinks of something does not imply that
she says or visualizes something to herself, or that she rst has a plan and
then executes it. Not all cases of thinking of something conform to the
model of having a ash of insight and then acting on it. The comedian
who delivers an off-the-cuff reply to a heckler can still be said to have
thought of a reply, without its being true that she rst thought of the reply
and then gave it.
Let us now consider what role thinking-of should play in our account of
imaginativeness. In view of what I have said so far, it would be natural
simply to claim that an imaginative f must be an unobvious f to think of.
This proposal certainly covers many of the examples I have considered.
However, it will not cover all cases. There are two kinds of counter-
example.
First, suppose a rugby player makes an imaginative pass. It is doubtful
that such a player can always be said to have thought of a pass. Accord-
ingly, it is doubtful that an imaginative pass must be an unobvious pass to
think of. This is so even though we should allow that thinking-of can
occur in situations, such as rugby matches, where we do not have time to
deliberate. It is mistaken, nevertheless, to say that a rugby player who
makes an imaginative pass must have thought of a pass.
We should deal with this kind of case in the following way. A person
who makes an imaginative pass does not necessarily think of a pass, but she
does think of something else: she thinks of an unobvious way of putting
her teammates in a better position to score, or an unobvious way of
keeping the ball from the other team, or something else besides. The
pass is, of course, the way of keeping possession that she thought of. But it
still might not be true to say that she thought of a pass. Therefore,
sometimes an imaginative f is an unobvious g to think of, though not an
unobvious f to think of.
The second kind of counterexample is different. Suppose someone
writes an imaginative play or carves an imaginative sculpture. Such a
person cannot necessarily be said to have thought of a sculpture or to
have thought of a play. In this respect, these cases are like the imaginative
pass. But in these cases, it is doubtful that there is always some f, such that
an imaginative play or sculpture is an unobvious f to think of. For example,
it is doubtful that an imaginative play is, in any given case, either an
unobvious comedy to think of, or an unobvious tragedy to think of, and
so forth. These cases need to be handled differently.
In these cases, we should consider what makes a play or sculpture
imaginative. A play might be imaginative because of the intricate way in
which the events of the plot are arranged, as in many Elizabethan comedies.
Arranging the events of the plot in this way was an unobvious way of
writing a play to think of. Similarly, a sculpture might be imaginative
because the stone was left rough-hewn in certain areas, and leaving the
stone rough in those areas was an unobvious way of making a sculpture of
that kind to think of. In each case, the imaginative product has certain
properties, such that giving it or leaving it with those properties was an
unobvious f to think of.
So the basic idea that imaginativeness is to be explained using the notion
of being an unobvious f to think of seems right. An imaginative f must
either be an unobvious f to think of, or be an unobvious g to think of, or
have properties such that giving it or leaving it with those properties is an
unobvious g to think of.
These cases suggest something further. It also seems that an imaginative f
must either be an imaginative f to think of, or be an imaginative g to
think of, or have properties such that giving it or leaving it with those
properties is an imaginative g to think of. The imaginative pass is not only
an unobvious, but an imaginative, way of keeping possession to think of.
Leaving the sculpture rough-hewn is not only an unobvious, but an

imaginative, way of making such a sculpture to think of. In short, what
it is to be an imaginative f can be explained in terms of the notion of being
something it is imaginative to think of. I will therefore focus on the latter
notion in what follows. So far, we have established that something is an
imaginative f to think of only if it is an unobvious f to think of.
This is not a sufcient condition of being an imaginative f to think of.
A recipe for chocolate-chip cookies that included generous amounts of
balsamic vinegar would not be an obvious recipe to think of. But it would
not necessarily be an imaginative recipe to think of. You cannot come up
with an imaginative recipe merely by thinking of an unobvious recipe.
The imaginative is not just opposed to the obvious. It is also opposed, in
John Passmores words, to the gimmicky, the merely fanciful, the point-
lessly innovative, the kitsch.55
One suggestion is that an imaginative f to think of must be valuable.
A number of philosophers have recently tied imaginativeness to being
valuable. Berys Gaut, for example, claims that, for something to be
creative, it must have considerable value.56 This suggestion accounts for
the recipe. A recipe for chocolate-chip cookies with balsamic vinegar is
unlikely to be imaginative because it is unlikely to be a good chocolate-
chip cookie recipe. This explanation appeals not merely to the recipes
likely value, but to its likely value as a chocolate-chip cookie recipe. So we
might suppose that, for something to be an imaginative f to think of, it
must have substantial value as an f.57
There are, however, counterexamples to this claim. For there are
striking examples of imaginative failures. The ying machines envisaged
by Leonardo would not work. His designs for them were imaginative
ying-machine designs to think of. But they do not have substantial value
as ying-machine designs, because the machines envisaged would not be
capable of ight.
55
John Passmore, Serious Art: A Study of the Concept in all the Major Arts (London: Duck-
worth, 1991), p. 98.
56
Gaut, Creativity and Imagination, p. 150.
57
Plausibly, imaginativeness itself often helps to make what possesses it (for instance,
artworks) valuable of its kind. If so, then this claim must be understood as the claim that an
imaginative f to think of must have substantial value as an f in virtue of something other than
its own imaginativeness. An imaginative cookie recipe, for example, might satisfy this
condition by having value as a recipe in virtue of being a recipe for cookies that taste good.
However, we can accommodate this case with slight adjustments. It is

signicant that Leonardos designs are very intelligent failures. It was
plausible for him to believe that the machines he designed had at least a
reasonable chance of being capable of ight. At the time, this was not a
foolish or crazy belief. This contrasts with the recipe. Normally, it would
be implausible to believe that a recipe for chocolate-chip cookies with
balsamic vinegar has even a reasonably good chance of being a good
chocolate-chip cookie recipe. Therefore, it seems that, for something to
be an imaginative f to think of, it must be plausible to believe that it has a
reasonable chance of having substantial value as an f.58
This proposal will work for some cases. But it will not work for all.
Gettiers counterexamples to the tripartite analysis of knowledge were
imaginative counterexamples to think of. But it seems irrelevant that it
was plausible for him to believe they had value as counterexamples. What
matters is that it was plausible for him to believe that these were counter-
examples. Certainly, a counterexample can be imaginative when, and
partly because, it is plausible to believe it has substantial value as a counter-
example. Being vivid and memorable, for instance, can enhance some-
things value as a counterexample. But Gettiers cases were imaginative
counterexamples to think of at least largely because they are plausibly
counterexamples, and were not obvious counterexamples to think of. It
seems mistaken to require that it must also be plausible to believe they
have substantial value as counterexamples.
This shows that, sometimes, thinking of an f is in itself a kind of success.
Thinking of a solution to a diplomatic crisis, and thinking of a way of
forcing checkmate, are other examples. In such cases, thinking of an f with
substantial value as an f, when that is possible, would be a further success.
By contrast, in other cases, only thinking of an f with value as an f is a
success. Thinking of an objection to a claim, unlike thinking of a counter-
example to a claim, can fail to be any sort of success. The objection
thought of might be terrible. Success in thinking of an objection is
thinking of a good objection.
58
The claim here is not that this belief must be justied. If a belief cannot be justied by
false beliefs, then, in the case of some imaginative acts and products, the belief in question is
not justied. Some of Leonardos beliefs about aerodynamics could have been false, and if his
belief that his ying-machine designs had a reasonable chance of being good designs was based
on them alone, then it was not justied, if this view about justication is right. But it was still
plausible for him to believe this.
Imaginativeness is related to success in thinking-of. For something to be

an imaginative f to think of,
It must be plausible to believe that it has a reasonable chance of being

an f or of having substantial value as an f,
depending on whether thinking of an f, or thinking of an f with value as an
f, constitutes success. It will be convenient to have a term for what is
thought of when thinking-of is a success. I will use the term achievement
for this. We can now re-state the above condition concisely. Something is
an imaginative f to think of only if
(1) It is an unobvious f to think of,
and
(2) It is plausible to believe that it is reasonably likely to be an
achievement.
Gettiers counterexamples satisfy condition (2) because it was plausible for
him to believe they were counterexamples. Leonardos designs satisfy
condition (2) because it was plausible for him to believe that they would
be good designs.
Satisfying conditions (1) and (2) is not, for every f, a sufcient condition
of being an imaginative f to think of. I discuss this further in a note.59 But
these conditions do give the important features of imaginativeness for my
59
Suppose that, in houses in a certain culture, the bedroom is always on the ground oor on
the right-hand side. There is no particular reason for this. A design placing the bedroom on the
top oor on the left-hand side might then be an unobvious house design to think of. It might
also be plausible to believe that the design is a good house design. Conditions (1) and (2) are
satised. But such a design would not necessarily be imaginative. If there was no particular
reason to place the bedroom where it is normally placed, a design placing it elsewhere might not
have much value as a design in virtue of doing so. It might be a good design solely in virtue of
other features. What makes it unobvious might be entirely unrelated to what makes it good.
Contrast this with Frank Gehrys imaginative design for the Bilbao Guggenheim. One
reason why this design is an unobvious design to think of is that it gives the buildings exterior
that unusual shape. Giving the building this shape also contributes signicantly to the designs
value as a design: the shape makes the building dynamic, expressive, spectacular from a
distance without being overwhelming from street level, and so forth. Here, some properties
that make it an unobvious design to think of coincide with some of the properties that it was
plausible to believe would contribute signicantly to its value.
Let us call cases like designs coincidence cases. Fs are coincidence cases just when, in order
for something to be an imaginative f to think of, it must have properties such that (i) these
properties make it an unobvious f to think of, and (ii) it is plausible to believe that these same
properties have a reasonable chance of contributing signicantly to its value as an f. I will
ignore coincidence cases in what follows.
purposes. Moreover, this account enables us to explain several facts about

imaginativeness.
First, it explains why something that is both an f and a g can be an
imaginative f but not an imaginative g. One reason why this is possible is
that something can be an unobvious f to think of but an obvious g to think
of. Dancing can be an imaginative way of getting someones attention in a
crowded train station, but not an imaginative way of responding to a
request to demonstrate what dancers do, because it is an unobvious way of
getting someones attention in a crowded train station to think of, but an
obvious way of responding to such a request to think of.
Second, this account explains why imaginativeness is relative to persons
and to contexts. The reason is that obviousness and plausibility are relative
to persons and contexts. A solution can be an imaginative solution for a
child to think of, but not an imaginative solution for an adult to think of,
because it can be an unobvious solution for a child to think of, but an
obvious solution for an adult to think of. One reason why using a certain
poetic form can be an imaginative way of writing a poem to think of
today, but not in the Middle Ages, is that it can be an unobvious way
of writing a poem to think of today, but would have been an obvious
way of doing so to think of in the Middle Ages. A military tactic might be
an imaginative tactic to think of for one situation, but not for another,
because it is plausible to believe it is reasonably likely to be successful in the
one situation, but it is not plausible to believe it has much chance of success
in the other situation.60
60
In saying this, I do not mean to imply that an imaginative f to think of is an unobvious f
to think of in every particular context in which it is an imaginative f to think of. An example
will illustrate what I mean. Suppose you are a diplomat dealing with an international crisis. As
you are a gifted diplomat, you consider many facts about the situation that other diplomats in
your situation would not consider. After taking these facts into account, you think of a
solution. This solution might be an obvious solution to think of if those facts are taken into
account, but not an obvious solution to think of otherwise. But it might still be an imaginative
solution to think of, even after those facts have been taken into account.
What matters here is that the solution is an unobvious solution to think of in a certain kind
of crisis, and you thought of the solution in a crisis of that kind. It does not matter that it is an
obvious solution to think of if one takes those facts into account. In assessing the imagina-
tiveness of your solution, certain context-types (e.g., a diplomatic crisis of this kind) have a
kind of relevance that other context-types (e.g., a diplomatic crisis of this kind in which those
facts are taken into account) do not have. Why this is so is a further question, which we need
not answer. The point is that, although the relativity of obviousness and plausibility explains
the relativity of imaginativeness, it is false that something is an imaginative f to think of in a
particular context only if it is an unobvious f to think of in that particular context.
Third, we can explain why something can be an imaginative f to f but

not an imaginative f to c. I said above that marshalling ones forces in a
certain pattern can fail to be an imaginative tactic to adopt in a context,
even if it is an imaginative tactic to think of in that context. One reason
why this is possible is that it can be implausible to believe that the tactic
would be an effective tactic to adopt in that context, but plausible to
believe, in that context, that it would be an effective tactic to adopt in
some other context for which it is thought of.
Fourth, my account enables us to explain why imaginativeness is asso-
ciated with originality. If something is very different from what came
before, that is often why it is an unobvious f to think of. If this difference
also makes it valuable as an f, then it will not be a coincidence if it is
plausible to believe that it is an achievement. That is why there is a non-
accidental connection between being original, in the sense of being
very different from what came before in a valuable respect, and being
imaginative.
Fifth, we can explain why imaginativeness is associated with non-
derivativeness. The fact that something is derivative is evidence that it
was an obvious f to think of. A derivative act or product has to derive from
something the agent or producer is already familiar with. Familiarity with
something can, though it need not, make it an obvious f to think of.
The familiarity of European novelists of the 1930s with the stream-of-
consciousness style would have made it an obvious style for them to think
of for their novels, and it would have been a derivative style for them to
adopt if they did adopt it. Derivativeness can be evidence of obviousness,
and therefore can be evidence that something is not imaginative.
Sixth, we can explain the difference between what is imaginative, what
is unimaginative, and what is merely not imaginative. Something can fail
to be imaginative without being unimaginative. The imaginative and the
unimaginative are opposed as contraries, not contradictories. The choc-
olate-chip cookie recipe with balsamic vinegar is not imaginative. But it is
not unimaginative either. With my account, we can explain what the
difference between these three concepts consists in. If something is far
from an obvious f to think of, it may fail to be imaginative, but it will not
be unimaginative. Being unimaginative requires being obvious.
Seventh, we can explain why imaginativeness is associated with the
imagination. Thinking of possibilities is central to my account of imagina-
tiveness. In this respect, my account resembles those accounts of imagination
according to which imagination is an ability to think of possibilities. For

example, Alan White argues that to imagine something is to think of it as
possibly being so.61 Whatever the right account of imagination may be,
imagination is at least associated with the ability to think of possibilities. The
fact that thinking of possibilities is involved in imaginativeness therefore
explains why imaginativeness is associated with imagination.
My account also suggests a reason why many imaginative people have
impressive powers of imagination. The power to imagine many possibil-
ities accurately and in detail can enable a person to have plausible beliefs
about more possibilities than someone who could think of fewer possibil-
ities, or someone who could imagine the same possibilities less accurately
or in less detail. A cook who can imagine, accurately and in detail, how
different combinations of ingredients would taste will be more likely to
have plausible beliefs about whether those combinations would taste
good. A writer who can imagine, accurately and in detail, the psychology
of a certain type of person will be in a better position to form plausible
beliefs about whether various ways of continuing a story about such a
person would be good. Indeed, imagining a characters psychology in
detail can make one more likely to think of many good ways of continuing
the story in the rst place.
We should not conclude from this that one imagines whenever one
does or produces something imaginative. I have said that imagining is not a
sufcient condition of doing or producing something imaginative. It is not
a necessary condition either. Take the case of the imaginative reply
delivered by the comedian to the heckler. I said above that the comedian
thought of a reply. But it does not seem that, if the comedians reply was
imaginative, she must have imagined. She neednt have imagined the reply
itself, or have imagined replying, and the reply need not express something
she imagined. So exercising imaginativeness does not require imagining.
But for reasons my account makes clear, a powerful imagination can help
to make one more imaginative.
Finally, we can explain why imaginativeness is associated with inspiration.
An important part of our conception of inspiration is its non-voluntary
character. Traditional images of the inspired person being acted upon or
controlled by another force, such as the Muses, capture this vividly. One
61
Alan R. White, The Language of Imagination (Oxford: Blackwell, 1990), p. 184.
reason why imaginativeness is associated with inspiration is that it involves

thinking-of, and thinking of something is often not voluntary. You can
suddenly think of a solution to a problem without having tried to do so, or
while giving your attention to something else. There are many expressions
for reporting the experience of non-voluntarily thinking of something, and
this reects how common this experience is. In many of these expressions,
what is thought of is characterized in active terms, while the person who
thinks of it, or her mind, is characterized as inert or passive. A solution
can suddenly come to mind, suggest itself to you, present itself to you,
occur to you, strike you, or hit you. It is therefore natural to associate
imaginativeness with inspiration.
6. The imaginativeness of persons

We can use this account to say what it is for a person to be imaginative.
Many philosophers have held that the imaginativeness of persons is an
ability.62 This is mistaken. If imaginativeness were an ability, then presum-
ably it would be the ability to act or omit or think (etc.) imaginatively. But
a person could have this ability without exercising it. And one cannot be
an imaginative person without ever acting, omitting, or thinking (etc.)
imaginatively. Therefore, imaginativeness is not an ability. The imagin-
ation is presumably an ability, but imaginativeness is not.
Imaginativeness is, rather, a propensity. In at least in one sense of
propensity, if a person has a propensity to f, then she has f-ed. If
someone has a propensity to commit crime, for instance, then she has
committed crime. My account of the imaginativeness of acts, omissions,
and products shows what kind of propensity the imaginativeness of per-
sons is. If that account is correct, a person who does something imagina-
tive, produces something imaginative, omits to do something it is
imaginative to omit to do, etc., thinks of something it is imaginative to
think of. This strongly suggests that imaginativeness is a propensity to think
of what it is imaginative to think of.
62
See ibid., p. 185; Margaret A. Boden, Creativity in a Nutshell, in Creativity and Art:
Three Roads to Surprise (Oxford: Oxford University Press, 2011), p. 29; Berys Gaut, Creativity
and Skill, in Michael Krausz, Denis Dutton, and Karen Bardsley, eds, The Idea of Creativity
(Leiden: Brill, 2009), p. 95. Gaut has since changed his view on this.
6. THE IMAGINATIVENESS OF PERSONS 81
There is another reason to believe this. In many instances, both thinking

of an act and performing the act one has thought of can be imaginative.
A way of rearranging the furniture can be both an imaginative rearrange-
ment to think of and an imaginative rearrangement to undertake. A chess
move can be both an imaginative move to think of and an imaginative
move to make. But a person who has a propensity both to think of acts it is
imaginative to think of, and to perform those acts, is no more imaginative
than she would be if she only had a propensity to think of those acts.
A person who is constantly thinking of imaginative rearrangements of the
furniture, but has no interest in actually rearranging the furniture in those
ways and does not do so, would not be more imaginative if she also had a
propensity to rearrange the furniture in the ways she thinks of. A chess
player who always thinks of imaginative moves but never makes them may
lack condence or a competitive streak, but she is no less imaginative than
she would be if she also had a propensity to make those moves.
We would expect this to be otherwise if the imaginativeness of persons
were simply a propensity to do or produce or omit to do (etc.) what it is
imaginative to do, produce, omit to do (etc.), and not, specically, a
propensity to think of what it is imaginative to think of. If the chess player
in my example had a propensity not only to think of those moves, but also
to make them, she would have a greater propensity to do what it is
imaginative to do. But she would not be more imaginative, as we would
expect if the less specic characterization of imaginativeness were correct.
This suggests that the imaginativeness of persons is a propensity to think of
what it is imaginative to think of.
We already have an account of being an imaginative f to think of. I said
that something is an imaginative f to think of only if:
(1) It is an unobvious f to think of,
and
(2) It is plausible to believe that it is reasonably likely to be an
achievement.
This allows us to characterize the imaginativeness of persons in more
detail. The imaginativeness of persons is a propensity to think of unob-
vious acts, omissions, products (etc.) which it is plausible to believe are
reasonably likely to be achievements.
7. Imaginativeness and the endowments

of good critics
According to my account, imaginativeness is a certain kind of aptitude for
thinking-of. This marks a basic difference between imaginativeness and
many other endowments, including the endowments I have so far identi-
ed as endowments of good critics.
Some of these endowments are not aptitudes at all. Knowledgeability,
I argued, is an endowment of a good critic. But whatever propositional
knowledge is, it is presumably not an aptitude for anything. So, too, being
disposed to respond in appropriate affective or conative ways can indicate
sensitivity, intelligence, and understanding, but it seems mistaken to describe
it as an aptitude.
Among the endowments of a critic that are aptitudes, imaginativeness
contrasts with good judgement. I identied several matters concerning
which the critic must have good judgement: she must be a good judge of
what needs to be communicated to the readership to enable them to
appreciate the work better, of how to communicate it to them so as to
enable this, and of the appreciative relevance of features, parts, represented
elements, responses, and facts. To make a judgement about such matters is
not to think of something.
Perceptiveness is another example of an aptitude for something other
than thinking-of. Indeed, my account of imaginativeness enables us to see
more clearly the differences between perceptiveness and imaginativeness
differences that might otherwise be hard to articulate.63 For there are
important similarities between perceptiveness and imaginativeness. Like
imaginativeness, perceptiveness is related to obviousness. Perceptive people
grasp what is not obvious. And plainly, like imaginativeness, perceptiveness
enables people to become aware of what is, at least potentially, valuable.
They grasp what is potentially useful, illuminating, relevant, and so forth.
Perceptiveness differs from imaginativeness in that it is an aptitude for
acquiring knowledge, not an aptitude for thinking-of. It can be perceptive
to see, to observe, to note, to grasp, to recognize, to register, or to hear that
63
Philosophers sometimes mistake instances of perceptiveness for instances of imagina-
tiveness. Some of Peter Strawsons examples of imaginativeness, such as the sensitive observer
of a personal situation seeing that situation as one of humiliation for one party and triumph for
another, are actually cases of perceptiveness. See P. F. Strawson, Imagination and Percep-
tion, in Freedom and Resentment and Other Essays (London: Methuen, 1974), p. 61.
7. IMAGINATIVENESS AND THE ENDOWMENTS OF GOOD CRITICS 83
something is (or is not) so. It may be perceptive of a museum curator to

recognize that a certain item is a fake. As I noted in Chapter 2, it was
perceptive of Frank Kermode to see that the Fools line in King Lear is an
allusion to a common saying. To take another example, R. R. R. Smith,
in a discussion of the Belvedere Torso, perceptively observes that the
famous Torso is seated, not on a lion-skin of Herakles, but on a panther-
skin, the regular animal wear of satyrs.64 This supports his hypothesis that
the Torso is a satyr, and not, as has been believed, Herakles.
Perceptiveness is not only an aptitude for acquiring knowledge, but an
aptitude for acquiring knowledge in certain ways rather than others. It
cannot, for example, be perceptive to acquire knowledge through testi-
mony. Certainly, it may be perceptive to gure out that someone is telling
you that p, or that she is trying to tell you that p. But it cannot be perceptive
of you to learn that p from someone else.
Perceptiveness is not only an aptitude for acquiring knowledge in
certain ways. It is also an aptitude for acquiring knowledge of facts of
certain kindsnamely, unobvious facts. Recall my distinction between
the kinds of obviousness in question when we use the constructions:
(I) It is obvious that p.
(II) Such-and-such is an obvious f to f.
Imaginativeness is opposed to the second kind of obviousness. But percep-
tiveness is opposed to the rst kind. If it was perceptive of a person to acquire
the knowledge that p, then it was not obvious that p. That is, it was not
evident that p. It was perceptive of the museum curator to see that the
artwork is a fake, because it was not obvious that it is a fake. It was perceptive
of Kermode to recognize that the Fools line is an allusion to the common
saying about fools and kings, because it was not obvious that it is.65
It is therefore clear how imaginativeness and perceptiveness differ.
A person is perceptive in virtue of being good at acquiring, in certain
ways, unobvious items of knowledge. A person is imaginative in virtue of
being good at thinking of (apparent) achievements that are not obvious
64
R. R. R. Smith, Hellenistic Sculpture (London: Thames and Hudson, 1991), p. 133.
65
My claim does not imply that it cannot be obvious to the person herself that p, in the
context in which she acquired the knowledge that p. It may have been immediately obvious
to Smith, when he saw it, that the Belvedere Torso is seated on a panther-skin. But that is
consistent with its not having been obvious, in that context, that the Belvedere Torso is seated
on a panther-skin.
achievements, or apparent achievements, to think of. Thinking of an

achievement is a different kind of attainment than acquiring knowledge.
An aptitude for the one is therefore different from an aptitude for the
other. This is not to deny that knowledge acquisition can be one aim of
thinking-of (e.g., the attempt to think of explanations). Nor is it to deny
that something (e.g., a critics remark) can be both imaginative and
perceptive. But what makes it perceptive will be different from what
makes it imaginative.
8. The role of imaginativeness in criticism

We are now in a position to say what the role of imaginativeness in
criticism is. In my account of the endowments of good critics, I began
by saying what good critics are good at, given what the aims of criticism
are. Notably, a good critic is good at making various kinds of judgement
(e.g., of appreciative relevance), and at acquiring knowledge of various
kinds of fact (e.g., about properties of the work), on the basis of her
awareness of the work.
Is she, in addition, good at thinking of things? Imaginativeness is an
aptitude for thinking-of. To determine its role in criticism, we must rst
answer the question: what are good critics good at thinking of ?
My account of criticism suggests a two-part answer. First, good critics
are good at thinking of ways of communicating effectively. As I argued in
Chapter 2, the critic must be articulate. She must communicate what
appreciation of the work can involve in such a way as to be fairly readily
understood by her readership. She must also communicate in such a way as
to enable the readership to appreciate the work better. She must therefore
think of ways of communicating that will achieve these aims.
Second, the critic needs to be good at thinking of ways of appreciating
the work better than the readership would be likely to appreciate it if they
were aware of the work without her criticism. One appreciates a work by
responding appropriately to its features, parts, or represented elements for
appropriate reasons. You can come to appreciate a work better than you
did before in two ways. First, you can respond to an object of an
appropriate response in an appropriate way in which you did not respond
to that object before. Second, you can respond appropriately to an object
of an appropriate response for an appropriate reason for which you did not
8. THE ROLE OF IMAGINATIVENESS IN CRITICISM 85
respond in that way to that object before. Therefore, for a critic to think of
ways of appreciating the work better is for her to think of (i) appropriate
responses to some object O, which the reader is unlikely to have had to O,
or of (ii) appropriate reasons for an appropriate response R to O, but for
which the readership is unlikely to have had R to O. In short, the critic
thinks of appropriate responses and appropriate reasons.
So good critics are indeed good at thinking of things. They are good at
thinking of ways of communicating effectively and ways of better appre-
ciating a work. To say this, however, is not yet to establish the role of
imaginativeness in criticism. Imaginativeness enables a critic to think of
unobvious things.
It is evident that many effective ways of communicating are not obvious
ways of communicating to think of. Mirs description of his response to
Courbets Stormy Sea, quoted in Chapter 2, is an example: One feels
physically drawn to it, as by an undertow. It is fatal. Even if this painting
had been behind our backs, we would have felt it. This is far from an
obvious way of describing this response to think of, but it is a very effective
description.
It is also clear that many ways of better appreciating a work are
unobvious ways of better appreciating it to think of. That the 33 cantos
of the Paradiso and Purgatorio have a salient relation to the number of the
persons of the Trinity is not an obvious reason to admire the design of the
Divine Comedy to think of. But admiring its design for this reason is a way
of better appreciating the work. Another example is a comment Keats
made about a line by Shakespeare, according to Leigh Hunt:
Mr Wordsworth found fault with the repetition of the concluding sound of
the participles in Shakspeares line about bees:
The singing masons building roofs of gold.
This, he said, was a line which Milton would never have written. Mr Keats
thought, on the other hand, that the repetition was in harmony with the
continued note of the singers, and that Shakspeares negligence (if negligence
it was) had instinctively felt the thing in the best manner.66
That the repetition of the concluding sound of the participles is like the
continued buzzing sound bees make is certainly not an obvious reason to
66
Leigh Hunt, Retrospective Views of Keats, in G. M. Matthews, ed., John Keats: The
Critical Heritage (London: Routledge, 1971), p. 244.
admire the line to think of. But it is plausible that appreciation can involve
admiring it for this reason. It was highly imaginative of Keats to think of
this way of better appreciating Shakespeares line.
I conclude, therefore, that the role of imaginativeness in criticism is
twofold. First, it enables the critic to think of ways of better appreciating a
work that are unobvious ways of better appreciating it to think of. Second,
it enables her to think of effective ways of communicating that are
unobvious ways of communicating to think of.
Note that the role of imaginativeness is not to enable the critic to think
of what it is plausible to believe would be ways of better appreciating the
work or effective ways of communicating. My account allows that it can
be imaginative to think of what are actually ineffective ways of communi-
cating to ones audienceprovided it was plausible to believe they had a
reasonable chance of being effective; and it can also be imaginative to
think of what it was plausible to believe were ways of better appreciating a
work even if they are not actually ways of better appreciating the work.
Imaginative failure in criticism is certainly possible. But to characterize its
role in criticism, we must characterize the kind of success it enables the
critic to bring off, as I have done here.
This characterization of the role of imaginativeness in criticism shows
that imaginativeness enhances aptitudes I have already identied. I said that
good critics are articulate: that is, good at expressing themselves so as to be
fairly readily understood. Imaginativeness enhances articulacy, by enabling
the critic to think of unobvious effective ways of communicating. So too,
I said that good critics are good at thinking of responses appreciation can
involve: clearly, imaginativeness enhances this aptitude as well, by enab-
ling them to think of appropriate responses that are not obvious responses
to think of.
As I observed above, several aspects of criticism have seemed to phil-
osophers to indicate a signicant role for imaginativeness in criticism. In
the remainder of this book, I will focus on one of these aspects: namely,
the prevalence of metaphor in critics descriptions of artworks. I will
consider the role of metaphor in criticism in the light of my account
of the aims of criticism and the role of imaginativeness in it. My aim in
the remainder of this book is to explain why metaphor is so prevalent
in art-critical descriptions.
4
Metaphor and Likeness
The conclusions I have reached in the rst three chapters will help us to
explain the role of metaphor in criticism. But we must also have an
account of metaphor. We must establish what people who use metaphor
are communicating. This is what I will attempt in this chapter.
Outside philosophy, metaphors are commonly believed to be similes
with the word like or as removed. This is what many people are taught
at school. According to this view, Romeo communicates the same thing
by saying, Juliet is the sun, as he would have communicated by saying,
Juliet is like the sun, though the effects on the audience may be different.
Inside philosophy, this view is almost universally rejected by those
working on metaphor. The so-called comparison theory or simile
theory of metaphor is generally thought to have been defeated by numer-
ous objections. According to many philosophers working on metaphor
today, likeness does not play a role in determining what is communicated
with a metaphor or in enabling readers to grasp what is communicatedor
at least, it does not do so in every case.
I hold that this widespread view is wrong. Likeness always plays a role in
determining what is communicated with metaphor and in enabling
readers to grasp it. The many objections to comparison theories of meta-
phor fail to undermine this claim. But comparison theories of metaphor
are not correct either. They are much closer to the truth than is commonly
allowed. But a trope can be based on likeness without being a comparison.
I do not claim that all metaphors are comparisons, but that likeness always
plays a certain role in determining what is communicated and in enabling
us to grasp it. In this chapter, I will explain what that role is.
There are several well-known questions about metaphor I will not
attempt to answer. The claim I will defend is about what is grasped by
someone on the receiving end of a successful act of communication with
metaphor. It is also a claim about what enables such a person to grasp what
88 METAPHOR AND LIKENESS
she does. But I will not claim (or deny) that what is grasped by such a
person is the speakers meaning, semantic content, what is said, what is
conversationally implicated, what she is caused to notice, etc. Similarly,
my claim is also not a claim about whether what is grasped is the meta-
phors meaning or content. I am not concerned with classifying, under any
of these heads, what is grasped by the metaphor-users audience.
It is unnecessary, for the purpose of explaining why metaphor is so often
used in criticism, to take a stand on these matters. Most theorists of
metaphor can agree with me without compromising whatever view they
have on the matters mentioned in the last paragraph. My account is, in this
respect, a modest account of metaphor. Accordingly, I call the claim I will
defend the Minimal Thesis about metaphor. Though this claim is
modest, its truth has gone unrecognized in contemporary debates about
metaphor. This has had damaging consequences in many philosophical
discussions of metaphor, as we shall see.
Hereafter, I will use the phrase what is grasped by a reader who
understands the metaphor, rather than the more cumbersome what is
grasped by a reader on the receiving end of a successful act of communi-
cation with metaphor. The more cumbersome phrase is less liable to
mislead, as some might hold that a reader who understands a metaphor is
just a reader who sees what it means or grasps what is said. Since I do not
wish to commit to a claim about what a metaphor-user says or what
metaphors mean, I do not wish to be taken to be making a claim about
what a reader understands when understanding is taken to be specically a
matter of grasping what is said or seeing what the metaphor means. From
now on, a reader who understands the metaphor is to be taken as a
reader on the receiving end of a successful act of communication with
metaphor.
The structure of this chapter is as follows. In the rst section, I introduce
some concepts that I will employ in stating the Minimal Thesis. I then
state the thesis, and clarify and elaborate it. In the second section, I discuss
prominent objections to other theories of metaphor. I explain why my
account avoids these objections. I address objections that do apply in the
third section. I conclude, in the nal section, by providing further support
for the Minimal Thesis by pointing out a range of facts that it allows us to
explain.
1. THE MINIMAL THESIS 89
1. The Minimal Thesis

We often speak of whole sentences or other semantically complex expres-
sions, used in a certain way, as metaphors. We might describe the whole
sentence, Juliet is the sun, used in the way Romeo uses it, as a metaphor.
Obviously, however, not every expression in this sentence is being used
metaphorically, for Juliet is not being used metaphorically. I will call
the expression(s) in the metaphor that are used metaphorically the meta-
phorical element(s) of the metaphor.1 For example, the word sun is a
metaphorical element of Romeos metaphor. A metaphor can contain
only metaphorical elements.2 Prince Hal calls Falstaff woolsack while
asking him a question; but he would still have used a metaphor if he had
said only Woolsack! upon seeing him.3
Many kinds of expression, in many grammatical forms and positions, can
be the metaphorical element of a metaphor. For example, a noun or noun
phrase can be a metaphorical element, and there are no apparent restrictions
on its case. Nouns used metaphorically can appear in subject position (The
undiscovered country, from whose bourn/No traveller returns, puzzles the
will4), as part of a predicate (Juliet is the sun), and in the vocative case
(Woolsack!). If a verb or verb phrase is used metaphorically, there is no
apparent restriction on the verbs mood, as the imperative mood of cast
used metaphorically in this proverb from the Book of Ecclesiastes suggests:
Cast thy bread upon the waters: for thou shalt nd it after many days.5
Adjectives used metaphorically can occur in the grammatically predicative
position (Silence is golden) or in the grammatically attributive position
(a big fat lie). Adverbs, too, can be used metaphorically. Emily Dickinson
wrote a poem beginning: To ght aloud, is very brave.6 The variety of
forms the metaphorical element can take partly explains the commonly
1
Others have drawn distinctions similar to this (e.g., Monroe C. Beardsley, Metaphorical
Senses, Nos 12/1 (1978), p. 3).
2
Contra Max Black, Metaphor, Proceedings of the Aristotelian Society 54 (19541955),
p. 275; and Beardsley, Metaphorical Senses, p. 3.
3
William Shakespeare, Henry IV: Part One, ed. David Bevington (Oxford: Oxford
University Press, 1987), 2.4.129.
4
William Shakespeare, Hamlet, ed. G. R. Hibbard (Oxford: Oxford University Press,
1987), 3.1.8081.
5
Ecclesiastes 11:1 (King James version).
6
Emily Dickinson, To ght aloud, is very brave, in The Poems of Emily Dickinson, ed.
R. W. Franklin (Cambridge, MA: Harvard University Press, 1999), no. 138, p. 70.
noted fact that metaphors come in far more forms than the simple subject-
predicate type exemplied by Juliet is the sun.
Still, there appear to be restrictions on what expressions can be used
metaphorically. The interjection Ah! can be used sarcastically, but it is
hard to imagine metaphorical uses of it. Perhaps the only ground for saying
that such determiners as the, a, any, and both can be used metaphor-
ically is that they can determine a noun used metaphorically. And it is
doubtful that such adverbs and adverbials as very, also, primarily,
when, where, why, however, and therefore can be used metaphor-
ically, at least if they are also used adverbially.7
In many metaphors, the metaphorical element is used to characterize
something. In calling Falstaff woolsack, Hal characterizes him as fat. One
can characterize a child as very well-behaved by calling her an angel, or a
beer as tasting very good by describing it as nectar. I will call what is
characterized with the metaphorical element the subject of the metaphor.
Metaphors in which a metaphorical element is used to characterize some-
thing dominate philosophical discussions of metaphor. But not all metaphors
are of this kind. In metaphorical questions, the metaphorical element may
not be used to characterize anything. In the question, Is Juliet the sun? Juliet
is not characterized with sun, but sun is used metaphorically. Metaphors in
the antecedent or the consequent of a conditional do not necessarily charac-
terize anything with the metaphorical element either. The metaphor used by
someone who says, If Juliet is the sun, then you should marry her, does not.
There are presumably many other kinds of case in which this is so.
The notion of characterizing something is broader than that of predicating
a property of something. Many expressions can be used to characterize
something as having a certain property without being used to predicate a
property. For example, it is possible to characterize someone with an expres-
sion by using that expression to address her. One can characterize a deity as
mighty when one addresses him by saying O mighty one, even though one
does not predicate mightiness of him. Adjectives used in the attributive
position can characterize something as having certain properties, even
though, so used, they do not predicate properties. Accordingly, when I say
that the metaphorical element of a metaphor can be used to characterize, this
is not equivalent to the claim that it can be used to predicate properties.
7
For a way in which they could be used metaphorically but not adverbially, see the
discussion of anthimeria, below (p. 120).
Among metaphors in which a metaphorical element is used to character-

ize a subject, there are various cases to be aware of. First, some only
characterize their subjects as having certain properties, and some characterize
their subjects as lacking certain properties.8 If Mercutio had told Romeo,
Juliet is not the sun, he might characterize her as lacking certain properties
(e.g., those which Romeo had characterized her as having). If there are
negative properties, such as not being beautiful or primeness (the property
numbers can have), then perhaps such metaphors also characterize their
subjects as having negative properties. But the contrast between metaphors
that only characterize something as having properties, and metaphors that
characterize something as lacking properties, would remain. Second, if we
negate a metaphor in which the metaphorical element is used to characterize
something as lacking some property, the resultant metaphor can characterize
the subject as having that property. No man is an island is an example. As
used by Donne, this characterizes men as having signicant relationships
with others. It does this by negating a metaphor (of men as islands) that
would have characterized them as lacking signicant relationships with
others. Third, the content of someones speech, thinking, or propositional
attitudes can be given with a metaphor, as in Romeo believes that Juliet is
the sun. This metaphor characterizes Romeo as believing that (for example)
Juliet is far superior to those around her.
The claim I will defend is not a claim about all metaphors. It is about
metaphors in which a metaphorical element is used to characterize a
subject as having certain non-negative properties, and in which this is
not done by forming a metaphor from a metaphor that would have
characterized it as lacking certain properties, or by giving the content of
the subjects speech, thinking, or propositional attitudes with a metaphor.
However, if I am right about how these metaphors are based on likeness, it
will be evident that other metaphors are also based on likeness.
In the case of metaphors that characterize something as having a prop-
erty, there is an important connection between understanding the meta-
phor, on the one hand, and the properties the subject is characterized as
8
In earlier work, I claimed: If metaphors express or communicate anything, then
presumably they characterize something. . . . This being so, the metaphor must characterize
it as having some property or properties. You cannot characterize anything without charac-
terizing it as having some property ( James Grant, The Dispensability of Metaphor, British
Journal of Aesthetics 50/3 (2010), p. 271). I now retract these claims. In Chapter 5 of this book,
I have amended the argument in which these claims originally gured.
having, on the other. If a property is one that the subject is characterized

as having, then coming to understand the metaphor, or coming to under-
stand it better, can at least partly consist in grasping that the subject has that
property. Grasping this is at least part of what being on the receiving end of
a successful act of communication with this metaphor amounts to. This is
true of non-metaphorical characterizations as well. Coming to understand
the utterance of O mighty one consists, at least partly, in grasping that the
god is mighty.9
Before stating the Minimal Thesis, we also need to make some distinc-
tions between types of property. Suppose that this tomato is red, and this
stop sign is too. The stop sign therefore has the property, being like this
tomato. I will call this type of property a likeness. One property that makes
the stop sign like the tomato is redness. I will call this type of property a
likeness-maker.10 I will also describe properties as likeness-makers for a
certain likeness. Thus, redness is a likeness-maker for the likeness, being like
this tomato. Sometimes, when we speak of the likenesses or the similar-
ities between objects, we are talking about the properties that make them
alikethat is, the likeness-makers. But I will not use the word likeness in
this way.
We should also distinguish between likenesses, such as being like this
tomato, and determinates of likenesses, such as looking like this tomato, tasting
like this tomato, being shaped like this tomato, and so forth. This distinction
between being like N and V-ing like N, where V-ing is something other
than being, tends to be overlooked, as we shall see.11
9
Note that this is not a claim about what one is required to grasp in order to
understand the metaphor. It is a claim about what coming to understand the metaphor,
or coming to understand it better, can partly consist in. One can be required to grasp a
certain thought in order to understand an utterance without its being true that grasping
that thought is part of what coming to understand the utterance can consist in. We may
need to grasp the thought in order to grasp what is communicated, but the thought itself
may not be part of what is communicated. What is distinctive about a property that
something is characterized as having is that coming to understand the characterization, or
coming to understand it better, can consist at least partly in grasping the thought that the
thing has that property.
10
In choosing this term, I do not mean to imply support for truthmaker theory.
11
It is worth distinguishing both likeness-makers and determinates of a likeness. This is
worth doing even though it may be that determinates of a likeness can be likeness-makers
for that likeness. Whether this is possible depends on whether, for example, Audrey
Hepburn looks like Audrey Hepburn, France is shaped like France, and so forth. If
Audrey Hepburn has the property, looking like Audrey Hepburn, then looking like Audrey
Hepburn is both a determinate of the likeness, being like Audrey Hepburn, and a likeness-
Lastly, I want to distinguish between likeness-makers and properties

the possession of which is a way of possessing a likeness-maker. The
determinate-determinable relation can be used to illustrate this distinction.
If redness is a likeness-maker for the likeness being like this tomato, then
scarlet is a property the possession of which is a way of possessing this
likeness-maker. Being scarlet is a way of being red. But if the tomato is not
scarlet, then scarlet is not itself a likeness-maker for this likeness. Rather,
being scarlet is a way of possessing a likeness-maker for this likeness.
In what follows, instead of the ungainly phrase, properties the posses-
sion of which is a way of possessing a likeness-maker, I will simply
use ways of possessing a likeness-maker. But strictly, it is possessing the
property, not the property itself, that is a way of possessing the likeness-
maker. I explain below why I shall be appealing to this relation, rather than
simply to the determinatedeterminable relation, which illustrates it.
I can now state the thesis I wish to defend in this chapter. The rst part
of the Minimal Thesis about metaphors of the kind I distinguished above is
the following claim:
Each property a metaphors subject is characterized with the metaphor-
ical element as having is either:
(1) A likeness indicated by the metaphorical element, or
(2) A determinate of such a likeness, or
(3) A likeness-maker for such a likeness, or
(4) A way of possessing a likeness-maker for such a likeness.
For short, in what follows I will say, the properties the metaphor attributes
to its subject instead of the properties the metaphors subject is character-
ized with the metaphorical element as having. This claim about the
properties attributed by a metaphor suggests a claim about part of what
enables us to understand a metaphor. To understand metaphors, we must
employ our knowledge of what likeness is indicated by the metaphorical
element, and our knowledge of what properties are likeness-makers
maker for this likeness. Whatever the correct view on this matter is, it does not affect the
truth of the claim I will defend. And it is worth speaking of both kinds of property,
whatever the correct view is. There is still an important distinction between determinates
of a likeness, on the one hand, and properties that are likeness-makers for, but not
determinates of, that likeness, on the other. And even if some determinates of a likeness
are likeness-makers for that likeness, not all are. If this tomato does not talk, then talking
like this tomato is not a likeness-maker for the likeness, being like this tomato.
for, determinates of, or ways of possessing likeness-makers for, that like-

ness. This claim about understanding is the second part of the Minimal
Thesis.
If this is right, then likeness plays an essential role in determining what is
communicated with metaphors, and in enabling us to understand them.
Contrary to what is often assumed, the traditional idea that metaphor is
essentially related to similarity is right. The Minimal Thesis therefore needs
defending. Before I defend it, however, I must make several clarications
about it.
First, the Minimal Thesis, being a disjunctive claim, leaves it open
whether every metaphor that attributes properties attributes a likeness. It
may be, for example, that the properties attributed by some metaphors are
all likeness-makers for a likeness indicated by the metaphorical element,
and the likeness indicated is not attributed. But in such a case, we would
still employ our knowledge of what properties are likeness-makers for an
indicated likeness in order to understand the metaphor.
Second, the Minimal Thesis states a constraint on what properties can
be attributed by a metaphor. But it is not a formula for determining what
properties are attributed by a metaphor.
For instance, I am not making any claim about what determines which
likeness is indicated by the metaphorical element. In particular, I am not
assuming that, if E is the metaphorical element, then the likeness indicated
by it is always being like (an) E. For example, it is possible to characterize
someone as aggressive by describing him metaphorically as a gorilla, even
though (as both speaker and audience may know) gorillas are not aggres-
sive. It is plausible, and consistent with the Minimal Thesis, to claim that
the likeness indicated by the metaphorical use of gorilla is being like a
stereotypical gorilla rather than being like a gorilla. So I do not assume that
there is a formula applicable in every case for moving from the metaphor-
ical element to the likeness indicated by it; nor is the Minimal Thesis an
attempt to state one. But it is normally fairly easy to work out what the
indicated likeness is.
Similarly, I am not making a claim about what determines which
likeness-makers, determinates of likenesses, or ways of possessing like-
ness-makers are attributed. No metaphor that attributes likeness-makers
attributes every likeness-maker for the likeness it indicates. Romeos
metaphor does not characterize Juliet as a star, or as withering the crops
when there is no rain. And the same expression, used as a metaphorical
element, can attribute different likeness-makers to different subjects, even

when it indicates the same likeness. We attribute different likeness-makers
for the likeness being like the sun when we describe Juliet as the sun than we
do when we describe Achilles as the sun. The question therefore arises of
what determines which likeness-makers are attributed. This is an import-
ant question, but the Minimal Thesis does not commit one to any
particular answer to it.
The third point concerns likenesses. There are two varieties of likeness,
and it is important to be clear on the difference between them. We
sometimes attribute to an object a likeness to a particular item. When
you say, She is like her father, you imply that there is a particular person
she is like. Other likenesses are different. You can say, Richard is like a
lion, without implying that there is some particular lion he is like. The
question, Which lion is he like? is obviously out of place here.12
So too, a metaphor such as Richard is a lion can indicate a likeness to
a lion, without its being the case that it indicates a likeness to a particular
lion. If the Minimal Thesis is correct, it is never out of place to ask about
a metaphor, What is the likeness indicated by the metaphorical element
a likeness to? In this case, the answer is: A lion. But this should not
be mistaken for the claim that it is never out of place to ask, What
particular item is the likeness indicated by the metaphorical element a
likeness to?
Fourth, I hold that metaphors can attribute not only likeness-makers,
but determinates of likenesses and ways of possessing likeness-makers,
because of cases like the following. Suppose that someone says, The
forms in all of Kandinskys paintings are alive with movement. It would
be implausible to think that understanding a metaphorical use of this
sentence, in most normal contexts, involves grasping that the forms in all
of Kandinskys paintings seem to leap out at the viewer, or that they seem
to dance, or that they seem to move quickly across the canvas, etc. Rather,
understanding this metaphor involves grasping that the forms look like
something alive with movement. Likeness-makers that are not determi-
nates of being like something alive with movement are not attributed.
I also hold that ways of possessing a likeness-maker are sometimes
attributed by a metaphor. For example, the metaphor Sally is a block of
12
See John Hyman, The Objective Eye: Color, Form, and Reality in the Theory of Art
(Chicago: University of Chicago Press, 2006), pp. 6466.
ice attributes emotional unresponsiveness to Sally. Unresponsiveness is a

likeness-maker for the likeness, being like a block of ice (physical rigidity
being a kind of unresponsiveness). And being emotionally unresponsive is
a way of being unresponsive. But emotional unresponsiveness is not itself a
likeness-maker for the likeness, being like a block of ice.
In some metaphors, other words in the sentence help us to work out
which ways of possessing a likeness-maker are attributed. The word
bloom does this in Wallace Stevenss line:
Moisture and heat have swollen the garden into a slum of bloom.13
The garden is not (or at least, not merely) characterized as being in a state
of disorder. Being in a state of disorder is a property that makes it like a
slum. Rather, it is characterized as being overgrown, which is a way of
being in a state of disorder, but not a property that makes it like a slum. We
work this out at least partly because of the presence of the word bloom.
I leave it open whether all such cases are cases of determinates of deter-
minable likeness-makers. How the determinatedeterminable relation
should be characterized, and whether and how it should be distinguished
from relations like the realizerrealized relation, are vexed questions that
have received a steadily increasing amount of attention in recent years.
If, as some claim, a property is a determinate of another property only
if, necessarily, whatever instantiates the rst instantiates the second (e.g.,
necessarily, whatever is scarlet is red), then it is doubtful that the determin-
ate-determinable relation covers enough cases.14 For example, it is not
obvious that being overgrown would count as a determinate of being in a state
of disorder. For it is not obvious that whatever is overgrown is in a state of
disorder. Therefore, I employ instead the notion of a property that is a way
of possessing another property.
Fifth, it may be that some metaphors attribute properties that the
metaphor-user did not intend to attribute. For example, Stephen Yablo,
among others, argues that sometimes a speaker uses metaphor to prompt
her reader to discover things she did not herself have in mind. Her sense of
13
Wallace Stevens, Banal Sojourn, in The Collected Poems of Wallace Stevens (New York:
Alfred A. Knopf, 1954), p. 62, l. 4.
14
For a good summary of some of the main characteristics attributed to the determinate-
determinable relation, see Jessica Wilson, Determination, Realization and Mental Caus-
ation, Philosophical Studies 145 (2009), p. 151.
the potential metaphorical truthfulness of a form of words outruns her sense

of the particular truth(s) being expressed.15 It may be correct to regard the
metaphor itself as communicating more than the metaphor-user would
have realized, as with certain poetic metaphors.
Sixth, it may be that, for many metaphors and for many properties, it is
unclear whether the metaphor attributes them. For instance, Stanley
Cavell is frequently quoted explaining Juliet is the sun in this way:
Romeo means that Juliet is the warmth of his world; that his day begins
with her; that only in her nourishment can he grow.16 Now, it is fairly
clear that Romeo does not mean that his day begins with Juliet. However, if
we allow that a metaphor may attribute properties that its user did not
intend to attribute, then it is not clear (to me, at least) either that Romeos
metaphor does, or that it does not, communicate that Romeos day begins
with Juliet. The sort of elaboration in which Cavell engages hereto
which some poetic metaphors, at least, seem to lend themselvesis likely
to turn up numerous properties that are neither clearly attributed nor
clearly not attributed.
A similar but distinct seventh point is this. It may be that at least some
metaphors attribute indenitely many properties to their subjects. It may
not be clear when we would be done listing all of the properties that are
attributed by the metaphorwhether or not each of the properties already
on our list clearly is attributed. Simon Blackburn, among others, has
claimed that there is no single list of literal thoughts which cashes in
certain metaphors.17 Considering metaphors provided with Cavell-style
explanations, Blackburn writes that the metaphor is in effect an invitation
to explore comparisons. But it is not associated with any belief or inten-
tion, let alone any set of rules, determining when the exploration is
nished.18 Again, endorsement of the Minimal Thesis does not commit
one to a view on this. It commits one to the claim that each of the
15
Stephen Yablo, Does Ontology Rest on a Mistake?, in Things (Oxford: Oxford
16
Stanley Cavell, Aesthetic Problems of Modern Philosophy, in Must We Mean What We
Say?: A Book of Essays (Cambridge: Cambridge University Press, 1976), pp. 7879.
17
Simon Blackburn, Spreading the Word: Groundings in the Philosophy of Language (Oxford:
18
Ibid. See also Severin Schroeder, Why Juliet is the Sun, in Semantik und Ontologie, ed.
Mark Siebel and Mark Textor (Frankfurt am Main: Ontos Verlag, 2004), pp. 95100, for
recent (and sceptical) discussion of claims for metaphors open-endedness.
properties attributed will be of one of the four kinds mentioned in the

statement of the Minimal Thesis.
Finally, I have claimed that understanding a metaphor involves know-
ing what at least some of the properties it attributes are. Perhaps, for some
metaphors, there are certain properties you must know to be attributed in
order to count as understanding the metaphor at all. Perhaps, for other
metaphors, you need merely know sufciently many of the properties the
metaphor attributes in order to understand the metaphor, even though
there is no property such that you must know that it is attributed to
understand the metaphor. And nally, if some metaphors attribute indef-
initely many properties, understanding them is surely a matter of degree. If
Cavells explanation of Juliet is the sun is right, you have not failed to
understand the metaphor if you do not know that the metaphor attributes
the properties he claims are attributed; but presumably you would under-
stand the metaphor better or more fully, other things being equal, if you
did know this.
The fact that likenesses ought to have such an important role in an
account of metaphor is easily obscured by the contemporary philosophical
literature. Much of this literature is hostile, in particular, to the view that
metaphors are comparisons. The Minimal Thesis may remind many of this
view. It is not, however, a version of the comparison theory. The com-
parison theory is a claim about what metaphors mean and/or what
metaphor-users say. The Minimal Thesis, as I have stressed, is a claim
about neither.
Moreover, I do not support the comparison theory. As I shall explain
below, many of the objections usually thought to defeat the comparison
theory fail. It is much closer to the truth than many philosophers today
believe. But the comparison theory is nevertheless false.
There is a simple refutation of the comparison theory. A comparison is a
statement of similarity. But not all metaphors are statements. The meta-
phor used by someone who asks, Is Juliet the sun? is not a statement.
Again, not all metaphors are statements of similarity. The metaphors used
by someone who says, If Juliet is the sun, then you should marry her, or
Juliet is not the sun, are not statements of similarity. Therefore, not all
metaphors are comparisons. Therefore, the comparison theory is false.
Nevertheless, because so many other objections have been levelled
against the comparison theory, one might think the Minimal Thesis is
2. INAPPLICABLE OBJECTIONS 99
vulnerable to them. It is therefore necessary to point out which of these

objections are, and which are not, applicable to the Minimal Thesis. In the
next section, I will explain how the Minimal Thesis avoids many of these
objections. In section 3, I will identify, and argue against, the principal
objections that do apply to the Minimal Thesis.
2. Inapplicable objections
The objections raised against the comparison theory of metaphor attack
different views that are sometimes not distinguished. First, some apply to
the view that a metaphor is just an attribution of a likeness. On this view,
Romeos metaphor means the same as a literal utterance of Juliet is like
the sun, where such a literal utterance is understood as only attributing
the likeness being like the sun, and not likeness-makers. Second, some
objections apply to the view that a metaphor is an attribution of both a
likeness and likeness-makers for that likeness. On that view, what is
communicated by Romeos metaphor is what is communicated by:
Juliet is like the sun in that she is a sustainer of all that is good in life,
far superior to those around her, [etc.]. So not only do some of the
objections to the comparison theory not apply to the Minimal Thesis:
some of them do not apply to all versions of the comparison theory
either.
Let us rst consider those objections that apply only to the claim that
metaphors are merely attributions of likenesses. Davidson says that all
views that identify either the literal meaning or the gurative meaning
of a metaphor with the literal meaning of a related simile
share a fatal defect. They make the hidden meaning of the metaphor all
too obvious and accessible. In each case the hidden meaning is to be found
simply by looking to the literal meaning of what is usually a painfully
trivial simile. This is like thatTolstoy like an infant, the earth like a
oor. It is trivial because everything is like everything, and in endless ways.
Metaphors are often very difcult to interpret and, so it is said, impossible
to paraphrase. But with this theory, interpretation and paraphrase typically
are ready to the hand of the most callow.19
19
Donald Davidson, What Metaphors Mean, in Inquiries into Truth and Interpretation
(Oxford: Oxford University Press, 1984), p. 254. See also Samuel Guttenplan, Objects of
Metaphor (Oxford: Oxford University Press, 2005), p. 209.
There are, in fact, two objections in this passage. The rst might be
called the Argument from Difculty. If the literal or gurative meaning of
a metaphor were the literal meaning of some related simile, then it would
be easy to interpret and to paraphrase metaphors. To do so, one would
only have to identify the related simile. This is typically easy. But it is often
very difcult to interpret and paraphrase a metaphor. Therefore, neither
the literal nor the gurative meaning of a metaphor is the literal meaning
of some related simile.
The second is what might be called the Argument from Triviality. If the
literal or gurative meaning of a metaphor were the literal meaning of
some related simile, then metaphors would usually be trivial. For the
related simile is usually trivialindeed, it is usually painfully trivial
because everything is like everything. But metaphors are not usually trivial.
The Argument from Difculty does not apply to the Minimal Thesis.
First, the Minimal Thesis does not commit one to the claim which Davidson
thinks has the absurd consequence that metaphors are typically easy to
interpret and paraphrasenamely, the claim that the literal or gurative
meaning of a metaphor is the literal meaning of a related simile. As I have
said, the Minimal Thesis does not commit one to any claim about what a
metaphor means. It also does not commit one to the similar claim that a
metaphor communicates only what some related likeness-ascription does.
One might think that this claim, too, would have the same absurd conse-
quence, at least if one thinks that it is typically easy to identify the related
likeness-ascription. But the Minimal Thesis does not commit one to
the claim that a metaphor communicates only what some likeness-ascription
does. It is consistent with the Minimal Thesis to hold, and I do hold,
that metaphors attribute likeness-makers, determinates of likenesses, and
ways of possessing likeness-makers. So even if it is typically easy to identify
the related likeness-ascription, but not typically easy to paraphrase and
interpret metaphors, this poses no problem for a supporter of the Minimal
Thesis.
The Argument from Triviality is also inapplicable to the Minimal
Thesis, and for similar reasons. Even if every ascription of a likeness is
trivial, not every ascription of a likeness-maker, determinate of a likeness,
or way of possessing a likeness-maker, is trivial. So the Minimal Thesis
does not commit one to the claim that metaphors are usually trivial.
It is also worth noting that Davidson is wrong about similes here. To be
trivial is to be unimportant or uninteresting, which is reected in
Davidsons remark about a metaphors corresponding simile usually being

painfully trivial. But even if it is true that everything is like everything,
not all similes are unimportant or uninteresting. This simile from the Song
of Songs, in which a man addresses his lover, is not:
Thy teeth are like a ock of sheep that are even shorn, which came up from
the washing; whereof every one bear twins, and none is barren among them.20
This may be a weird and perhaps slightly hilarious simile to our ears. But it
is not trivial. Much great poetry contains a wealth of interesting and
powerful similes.
Moreover, Davidson gives us no reason to believe that the similes that
correspond to metaphors, in particular, are usually uninteresting. The point
that everything is like everything does not establish this. Neither do Da-
vidsons examples. Tolstoy is like a great moralizing infant is hardly an
uninteresting simile; and it is not difcult to imagine a context (e.g., on the
space shuttle) in which The earth is like a oor would be interesting too.
Presumably, the similes that correspond to metaphors are usually somewhat
less concise than the metaphors, as they are when they are formed by adding
like or as. But this, too, does not show that they are trivial.
Davidson claims not only that similes are trivial. He also claims that they
are trivially true.21 When we say that something is trivially true, we say
something about how evidently true it is. When Davidson says that similes
are trivial because everything is like everything, he is making a point
about how evidently true they allegedly are. But this claim is wrong too.
We shall see later that not all similes are trivially true.
William Lycan objects that some metaphorical statements are too
convoluted to be parsed as similes.22 He gives this example from Hamlet
(1.3.116117):
When the blood burns, how prodigal the soul
Lends the tongue vows.
He comments:
20
Song of Songs 4:2.
21
John Hyman pointed out to me that Davidson is apparently running together the
notion of being trivial and that of being trivially true.
22
William G. Lycan, Philosophy of Language: A Contemporary Introduction, 2nd edn (New
York: Routledge, 2008), p. 183.
It is not literally about anyones blood, and blood cannot literally burn . . .
the soul is probably itself being used metaphorically, and, even if not, souls
cannot literally lend anything to tongues; but tongues is not being used to
mean tongues, either, and vows are not the sorts of things that can be lent. So
any simile theorist faces the daunting task of translating all of those things at
once into resemblance talk. . . . A rst pass might be: When x, which is like a
persons blood, does something resembling burning, how prodigally y, which
is like a persons soul, does something similar to lending some things that are
vowlike to z, which resembles a persons tongue. We are not much the
wiser. And renement is needed, because for the blood metaphorically to
burn is probably something distinctive to a bloodlike substance, not for it to
do something that resembles the literal burning of, say, a piece of wood.23
Lycans objection here has inuenced others. Marga Reimer and Elisabeth
Camp also endorse it, and claim that it poses a problem for all versions of
the comparison theory.24 We might describe it as the Convolutedness
Objection.
There are two reasons why, as Lycan puts it, we are not much the wiser
as a result of explanations like the one he offers. First, in this explanation, we
are not told what the subjects of the metaphor are. Assume for the sake of
argument that all of the expressions Lycan claims are being used metaphor-
ically in this passage are indeed being used metaphorically. (This assump-
tion is false, but I will discuss that point in the next section.) Lycans
explanation does not tell us what is being characterized with blood,
burns, soul, lends, vows, or tongue. As I mentioned above, there
may be no expression in a sentence used metaphorically that literally
designates the subject(s) of the metaphor. Normally, an illuminating ex-
planation of a metaphor of that kind will have to tell us what the metaphors
subject is. Lycans explanation does not do this.
Second, his explanation does not tell us which likeness-makers, deter-
minates of likenesses, or ways of possessing likeness-makers these subjects
are characterized as having. This, too, is often what we want from an
explanation of a metaphor. In fact, Lycans explanation only tells us what
likenesses are indicated by the metaphorical elements. But we typically
know this already when we read a metaphor, if we know which
23
Ibid.
24
Marga Reimer and Elisabeth Camp, Metaphor, in Ernest Lepore and Barry C. Smith,
eds, The Oxford Handbook of Philosophy of Language (Oxford: Oxford University Press, 2006),
pp. 851853.
expressions are the metaphorical elements. So it is no wonder that we are

not much the wiser.
In view of this, it is clear that the Convolutedness Objection does not
apply to the Minimal Thesis. Whatever may be true of comparison
theories, the Minimal Thesis certainly does not commit one to the claim
that an explanation which only tells us what the likenesses indicated by the
metaphorical elements are, but which fails to tell us what the subjects are
or what other properties are attributed to them, will be illuminating.
I now turn to objections that apply to the claim that a metaphor means
what some attribution of both a likeness and likeness-makers means.
John Searle provides several objections applicable to any comparison
theory that claims that likenesses are attributed. He argues that in the
production and understanding of metaphorical utterances, there need not
be any two objects for comparison.25 He develops this point in two
directions.
First, he claims: If I say
13. Sally is not a block of ice,

that, I take it, does not invite the absurd question, Which block of ice is it
that you are comparing Sally with, in order to say that she is not like it? 26
Here, Searle says, there is no object for comparison in that the question,
Which particular item are you comparing her with?27 is out of place.
Therefore the metaphor cannot be a comparison. We can call this the
Particularity Objection.
Second, Searle claims that there are true metaphorical assertions [S is P]
for which there are no objects to be designated by the P term, hence the true
metaphorical statement cannot falsely presuppose the existence of an object
of comparison.28 He gives the example of the metaphorical assertion of
Sally is a dragon, which can be true even though no dragons exist. It could
25
John R. Searle, Metaphor, in Expression and Meaning: Studies in the Theory of Speech Acts
(Cambridge: Cambridge University Press, 1979), p. 87.
26
Ibid., pp. 8788.
27
As Schroeder points out, Searle misrepresents the comparison theory here as claiming
that metaphors are comparisons of A with B, rather than as claiming that metaphors are
comparisons of A to B. See Schroeder, Why Juliet is the Sun, pp. 8283.
28
Searle, Metaphor, p. 88.
not be true if it were a comparison, because it would falsely presuppose that

dragons exist. We can call this the Non-Existence Objection.
The Particularity Objection has been forestalled by the distinction
I mentioned above between two varieties of likeness. You can attribute
a likeness to a thing even though the question, To which particular item
are you likening this thing? is out of place. Many examples of literal
likeness-ascriptions show this. The fact that this question is out of place in
the case of many metaphors therefore does nothing to show that they do
not attribute likenesses.
The Non-Existence Objection is inapplicable because, as I have
stressed, the supporter of the Minimal Thesis does not have to claim that
the likeness indicated by the metaphorical element is attributed to the
subject. So we do not even need to challenge the implausible claim to
which Searle is committed by the Non-Existence Objection: namely, that
it cannot be literally true that Sally is like a dragon, since dragons do not
exist.29 Even if it cannot be literally true that Sally is like a dragon, a
supporter of the Minimal Thesis can allow that the metaphor, Sally is a
dragon can communicate a truth. For it could still be true that Sally
possesses likeness-making properties for the likeness being like a dragon.
The metaphor can attribute these properties (and not the likeness itself ) to
Sally. It is just that, if Searle is right, Sally would not be like a dragon if she
does possess those properties. A likeness-maker for the likeness being like an
X would have to be understood as a property that would make Y like an X
only if Y possesses it and Xs exist. Something can possess properties of
which this is true even if no Xs do exist.
Robert Fogelin raises a different objection applicable to any theory that
treats metaphors as attributing likenesses, but it has the same upshot as the
Non-Existence Objection. Fogelin argues that, even when the likeness
ascribed is between existing objects, many likeness-ascriptions are literally
false.30 I will explain his grounds for this claim below. The important
point, for the moment, is that these false likeness-ascriptions include
ascriptions of the likenesses indicated by the metaphorical element of
many metaphors. Therefore, if we treat such metaphors as attributing
29
For criticism of the conception of likeness-ascriptions that Searle adopts here, see
Hyman, The Objective Eye, pp. 6466.
30
Though, importantly, he holds that some of these are true if taken guratively (Robert
J. Fogelin, Figuratively Speaking, 2nd edn (Oxford: Oxford University Press, 2011), p. 84).
these likenesses, we shall have to conclude that these metaphors commu-

nicate a falsehood. But many such metaphors do not communicate a
falsehood. Therefore, they do not attribute likenesses. We may call this
the Figurative-Likeness Objection.
Again, the disjunctive character of the Minimal Thesis renders this
objection inapplicable. The supporter of the Minimal Thesis is free to
maintain that such metaphors attribute only properties related to the
likeness indicated, and not the likeness itself. Indeed, nothing would be
lost if the supporter of the Minimal Thesis stipulated that likeness, as she
uses it, is simply to be an expression for the sharing of properties, whether
or not likeness and its cognates are already expressions of this kind in
English. Fogelins argument is that is like and is similar to do not in fact
have this meaning. As he says: I do not deny . . . that, given any two
entities (existent, nonexistent, or mixed), it is always possible to nd
something that is true of both; I deny . . . that this shows they are similar.31
Nevertheless, it is not necessary to make this stipulation, because Foge-
lins argument does not establish this conclusion. Fogelin takes certain
observations made by the cognitive psychologist Amos Tversky to show
that not every claim that A is like B is literally true. Tversky observes that
we say an ellipse is like a circle, not a circle is like an ellipse, and we
say North Korea is like Red China rather than Red China is like North
Korea.32 Fogelin claims that this point refutes Davidsons claim that
everything is, after all, similar to everything else.33
In Fogelins view, A is similar to B just in case A has a sufciently large
number of Bs salient features.34 Thus, it is literally true that a road grader is
like a bulldozer, for like bulldozers, road graders are also used to push about
large quantities of dirt, the chief difference being that road graders have their
blades beneath their chassis rather than in front of them.35 But it is literally
false that Margaret Thatcher is like a bulldozer, for she does not have a
sufciently large number of a bulldozers salient features. Nevertheless,
there are people who would consider [this simile] true if taken gura-
tively.36 Fogelin regards metaphors as elliptical similes, but he argues that
similes are literally false, though they may be true if taken guratively.
31
Ibid., p. 60.
32
Amos Tversky, Features of Similarity, Psychological Review 84/4 (1977), p. 328.
33
Fogelin, Figuratively Speaking, p. 63.
34 35 36
Ibid., p. 76. Ibid., p. 84. Ibid.
One problem with Fogelins argument is that the linguistic data he cites
do not establish that the likeness-ascriptions in question are false. The fact,
if it is a fact, that we do not say that a circle is like an ellipse establishes only
that this claim is not (normally) assertible. Fogelin needs more argument to
show that A circle is like an ellipse and An ellipse is like a circle differ not
only with respect to their assertibility conditions, but with respect to their
truth conditions.
A second problem is that other linguistic data support the conclusion
that likeness-ascriptions of the kind Fogelin mentions do not differ in truth
conditions. Suppose someone says (what also seems to support Fogelins
view), He is nothing like his father. We could reply, Well, strictly and
literally speaking, he is like his father: he is male and he has blond hair.
Normally, it would be pedantic to say this, and what we say would be
irrelevant. We would no doubt be told this if we offered this reply. But we
would not be told that we were mistaken. This kind of reply seems to be
available for all of the cases Fogelin cites. Strictly and literally speaking, a
circle is like an ellipse, Red China is like North Korea, and Margaret
Thatcher was like a bulldozer: circles are closed plane gures, Red China is
an Asian country, and Margaret Thatcher destroyed much. In many
contexts, it may be misleading or otherwise conversationally inappropriate
simply to assert that a circle is like an ellipse. Someone who insists in such a
context that, strictly and literally speaking, a circle is like ellipse misses this.
But the fact that she is only open to this criticism, and not to the criticism
that she is wrong, indicates that Fogelins account of the truth conditions
of likeness-ascriptions is mistaken.
Interestingly, there is a large class of similes that are literally false, and
uncontroversially so. But the examples Fogelin cites are not of this kind,
and the reason why they are false is not the reason Fogelin gives. Not all
similes take the form A is like B. Many take the form A Vs like B or
A Vs as B Vs, with a verb other than is. That is, many similes attribute
determinates of likenesses. And many similes of this form are literally false,
for in many such cases, B does not V, and therefore A does not V like
B. Describing his time in government, Alastair Campbell wrote: most of
the time I felt like a round peg in a round hole.37 Here, if you said, Well,
37
Alastair Campbell, Questions over David Camerons Judgment are Justied, The
Guardian, 21 January 2011, <http://www.guardian.co.uk/media/2011/jan/21/alastair-
campbell-david-cameron> accessed 14 June 2012.
strictly and literally speaking, he did feel like a round peg in a round hole:
he felt such-and-such, you would be wrong, and not merely a pedant. So
Davidson is indeed mistaken to claim that all similes are true.38 But he is
not mistaken for the reasons Fogelin gives.
Searle raises an objection applicable only to comparison theories that
treat metaphors as attributing both likenesses and likeness-makers. Searles
objection is, in effect, that such theories wrongly construe all metaphors as
communicating that the subject possesses the likeness in virtue of possessing
the other properties attributed. According to such theories, Romeos
metaphor does not just communicate that Juliet is like the sun and that
she is beautiful, stands out from her surroundings, etc. It communicates that
Juliet is like the sun in that she is beautiful, stands out from her surroundings,
and so forth. But not all metaphors do communicate a claim of this form.
For example, a metaphorical assertion of Richard is a gorilla can be true
because it is true that Richard is erce, nasty, prone to violence, and so
forth, even though it is false that Richard is like a gorilla in virtue of being
erce, nasty, prone to violence, and so forth.39 Therefore, the metaphor
cannot be communicating the latter, false claim. A hearer might rely on his
familiarity with the falsehood that gorillas are erce and nasty to work out
that the metaphor communicates that Richard is erce and nasty. But it
does not follow from this fact about his procedures of comprehension that this is
part of the speakers utterance meaning of the metaphor.40 We might de-
scribe this as the Argument from Dissimilarity.
The Minimal Thesis avoids the Argument from Dissimilarity for two
reasons. First, it does not commit one to the view that metaphors attribute
likenesses. Second, as I said above, a supporter of the Minimal Thesis is free
to maintain that the likeness indicated by the metaphorical element of the
gorilla metaphor is not being like a gorilla but being like a stereotypical gorilla.
So if Richard is erce, then it is true that Richard is like a stereotypical
gorilla in virtue of being erce. A supporter of the Minimal Thesis can
therefore claim both that the metaphor is communicating this and that the
metaphor does not communicate a falsehood.41
38
Davidson, What Metaphors Mean, p. 257.
39
Beardsley makes a similar point in Monroe C. Beardsley, The Metaphorical Twist,
Philosophy and Phenomenological Research 22/3 (1962), p. 294.
40
41
Schroeder defends the comparison theory from this objection along similar lines,
and I have modelled my response here on his. See Schroeder, Why Juliet is the Sun,
pp. 8385.
Finally, Searle objects that comparison theories do not explain why

metaphors attribute certain likeness-makers but not others. He writes that,
if a defender of the simile theory
wants to insist [regarding Sally is a block of ice] that blocks of ice are literally
unresponsive, then we need only point out that that feature is still insufcient
to explain the metaphorical utterance meaning of 4 [Sally is a block of ice],
because in that sense bonres are unresponsive as well, but
22. Sally is a bonre
has a quite different metaphorical utterance meaning from 4.42
Let us call this the Failure-to-Explain Objection.

As I noted above, the question of why a metaphor attributes certain
properties, but not others, that are related to the indicated likeness in the
ways I have described, is an important one. But as I also stress, the Minimal
Thesis is not an attempt to answer it. It is a claim about what kinds of
properties are attributed by metaphors.
An ad hominem point here is that Searles own account of metaphor is
vulnerable to the Failure-to-Explain Objection. Searle suggests nine prin-
ciples for computing a metaphorical meaning S is R for a metaphor S is P.
These principles, however, are themselves insufcient to explain the meta-
phorical utterance meaning of most of the metaphors they apply to. For
instance, Searles second principle is:
Things which are P are contingently R. . . . If the metaphor works, the
property R should be a salient or well known property of P things.
39. (MET) Sam is a pig
will be taken to mean
39. (PAR) Sam is lthy, gluttonous, and sloppy, etc.43
However, living in mud and being raised for their meat are also salient and
well-known properties of pigs, but 39 does not attribute these properties
to Sam. Searles principles fail to explain why it does not.
Thus, none of these eight objections is applicable to the Minimal Thesis.
At least some are unsuccessful objections to comparison theories. But they
do not even apply to the view I defend. Nevertheless, there are several
42
43
Ibid., p. 107.
3. APPLICABLE OBJECTIONS 109
objections to other theories of metaphor that do apply, or can easily be

made to apply, to the Minimal Thesis.
3. Applicable objections
Richard Moran writes: resemblance and similarity are both symmetrical
relations: if A resembles B, then B resembles A. Hence, if metaphor were
some kind of assertion of resemblance, we should be able to reverse any of
the parts without loss or change of meaning.44 Clearly, however, we cannot
do so, for the metaphor of the lovers lips as cherries has a very different
meaning than the metaphor of cherries as lips. Similarly, Kendall Walton
writes: Many metaphors are not reversible. Life is hell is different from,
Hell is life. But similarity is presumably symmetrical. Life resembles hell in
exactly the respects that hell resembles life. This should make us wary of
construing metaphor in terms of similarity.45 Let us call this the Revers-
ibility Objection.
I will mention two reasons why the Minimal Thesis does not imply that
metaphors are reversible without a change in what is communicated.
First, reversing a metaphor of the form A is B would change the subject
of the metaphor. The reversed metaphor would attribute properties to a
subject to which the original metaphor did not attribute them. According
to the Minimal Thesis, Her lips are cherries attributes certain properties to
her lips. These properties are likeness-makers for the likeness, being like
cherries. But this metaphor does not attribute these properties to cherries.
Her lips are the subject, and cherries are not. The reversal of that meta-
phorCherries are her lipswould have cherries for its subject. It
would attribute to cherries likeness-makers for the likeness, being like her
lips. Obviously, this would be to communicate something different. If we
communicate that her lips are a deep red, for example, we do not
communicate that cherries are a deep red.
44
Richard Moran, Seeing and Believing: Metaphor, Image, and Force, Critical Inquiry
16/1 (1989), p. 93.
45
Kendall L. Walton, Metaphor and Prop Oriented Make-Believe, European Journal of
Philosophy 1/1 (1993), p. 48. Beardsley and Guttenplan have also endorsed versions of this
objection. See Beardsley, The Metaphorical Twist, p. 297; Guttenplan, Objects of Metaphor,
p. 211.
This is so even though, according to the Minimal Thesis, the original,

unreversed metaphor relies on the readers knowledge that cherries are a
deep red. To gure out what the original metaphor communicates about
her lips, you must draw on your knowledge that cherries are a deep red.
But the original metaphor does not attribute deep redness to cherries, even
though it exploits the knowledge that they have this property.
Second, not only would the reversed metaphor attribute properties to a
different subject. In the vast majority of cases, it would attribute different
properties to that subject. Calling Elvis Mother Teresa would communi-
cate something very different about Elvis than calling Mother Teresa
Elvis would communicate about her. You might attribute compassion
in the rst case but not the second, and celebrity status in the second case
but not the rst. Compassion is a likeness-maker for the likeness, being like
Mother Teresa, and celebrity status is a likeness-maker for the likeness, being
like Elvis. Of course, if Elvis was compassionate, then compassion is also a
likeness-maker for the likeness being like Elvis. But it does not follow, and it
would rarely be true, that you would attribute compassion to Mother
Teresa by calling her Elvis.
An objection often raised against certain forms of the comparison theory
is that, as Reimer and Camp argue,
the similarities that we most naturally cite in explaining what a metaphors
corresponding simile means are often themselves gurative. Consider the
opening lines of Sylvia Plaths (1961) poem Mirror: I am silver and exact /
I have no preconceptions. Presumably the protagonist is here describing
herself metaphorically as a mirror; on the simile theory she thus means that
she is like a mirror. One natural elaboration of what this simile means is that
she reects the world around her, but the key word reects here is itself
obviously metaphorical. We seem to have fallen into a vicious explanatory
circle.46
Searle, Lycan, and Samuel Guttenplan also raise this objection against
forms of the comparison theory.47 It would apply to the Minimal Thesis
as well. The claim would be that the original metaphor does not attribute
the property of reecting, which is a likeness-maker for the likeness being like
a mirror, to the protagonist. Yet the most natural way of explaining the
46
Reimer and Camp, Metaphor, p. 852.
47
Searle, Metaphor, pp. 9597; Lycan, Philosophy of Language, p. 180; Guttenplan,
Objects of Metaphor, pp. 3233, 209, 270.
mirror-metaphor is with the reection-metaphor. Let us call this the

Figurative-Likeness-Maker Objection.
This is not a compelling objection. It is certainly true that it is often most
natural to explain a metaphor or its corresponding simile with another
metaphor. But this does not support the conclusion that the original
metaphor attributes no likeness-makers that the subject literally has. If
the second metaphor attributes some of the likeness-makers attributed by
the rst metaphor and literally possessed by the subject, we would expect
to be able to explain the rst metaphor with the second. It might even be
natural to explain it with the second metaphor. Metaphors often come to
mind more readily than non-metaphorical ways of making the same point,
even when there exist non-metaphorical ways of making the same
point. Indeed, suppose it were not just very natural to explain the original
metaphor with the second, but impossible to explain it any other way.
Even this would not show that the original metaphor does not attribute
likeness-makers that the subject literally has. It might only reect the fact
that these likeness-making properties, which the subject literally has,
cannot be attributed without using one of those two metaphors.
All the objection shows is that the mirror-metaphor does not attribute
to the poems speaker the likeness-making property of reecting (the
property mirrors literally have). It is true that this particular likeness-
maker is not attributed by the mirror-metaphor. But this is not relevant.
What the objection needs to show is that the mirror-metaphor does not
attribute any likeness-makers for the likeness indicated. It does not show
this. Moreover, it is not true of this metaphor. The speaker characterizes
herself as being apt to represent the world accurately and neutrally without
distortions. This is a likeness-maker for the likeness being like a mirror (as
commonly conceived ). Whether or not mirrors do represent their surround-
ings, they are certainly thought of as being apt to do so accurately and
neutrally without distortions.
Davidsons paper contains, or at least provides the materials for, what we
might call the Non-Propositionality Objection. Davidsons denial that
metaphors have a special meaning beyond the literal is his best-known
claim about metaphor. But he also writes:
The central error about metaphor is most easily attacked when it takes the
form of a theory of metaphorical meaning, but behind that theory, and
statable independently, is the thesis that associated with a metaphor is a
denite cognitive content that its author wishes to convey and that the
interpreter must grasp if he is to get the message. This theory is false as a
full account of metaphor, whether or not we call the purported cognitive
content a meaning.48
Davidson then goes on to present arguments against this theory, even

when it does not take its most easily attacked form of a theory of
metaphorical meaning. I will here consider whether the arguments Da-
vidson presents undermine the Minimal Thesis.
What Davidson says above is just that the following is false as a full
account of metaphor:
(C) There is a denite cognitive content associated with the metaphor

that its author wishes to convey and that the interpreter must
grasp if he is to get the message.
But in the discussion that follows, it emerges that Davidson does not just
hold that (C) is false as a full account of metaphor, but that it is false
simpliciter.
The rst thing to note is that there are several implications of (C) to
which the Minimal Thesis does not commit one. It does not commit one
to the claim that
(C1) What the interpreter must grasp, in order to understand a meta-

phor, is something the author wishes to convey.
This, I take it, would follow if metaphors carry a message, a view which
Davidson denies in several places.49 As we have seen, it also does not
commit one to the claim that
(C2) There is a cognitive content, apart from the cognitive content

that would be expressed by using the sentence literally, such that
the interpreter must grasp it in order to understand the metaphor.
This is implied by (C)s talk of a denite cognitive content that the
interpreter must grasp.
What the Minimal Thesis does commit one to is (C)s implication that
48
49
See ibid., pp. 260263.
(C3) The reader must grasp a proposition, apart from the proposition
that would be expressed by using the sentence literally, in order
to understand a metaphor.
(C) has this implication because Davidson regards cognitive content as
propositional content. The Minimal Thesis commits one to the claim that
a reader must grasp propositions in order to understand a metaphor, and
that these are propositions apart from the proposition that would be
expressed by using the same sentence literally. The Minimal Thesis does
not commit one to the claim that these propositions are the content of the
metaphor. But according to the Minimal Thesis, a reader who understands a
metaphor that characterizes the subject as having certain properties grasps
that the subject has such-and-such properties. Coming to understand the
metaphor consists at least partly in grasping such a proposition. The
question is, then, whether Davidson has an argument against (C3). If he
does, then he has an argument against the Minimal Thesis.
Davidson says of (C):
It should make us suspect the theory [namely, (C)] that it is so hard to decide,
even in the case of the simplest metaphors, exactly what the content is
supposed to be. The reason it is often so hard to decide is, I think, that we
imagine there is a content to be captured when all the while we are in fact
focusing on what the metaphor makes us notice. If what the metaphor makes
us notice were nite in scope and propositional in nature, this would not in
itself make trouble; we would simply project the content the metaphor
brought to mind on to the metaphor. But in fact there is no limit to what
a metaphor calls to our attention, and much of what we are caused to notice
is not propositional in character. When we try to say what a metaphor
means, we soon realize there is no end to what we want to mention.50
Further on, Davidson elaborates the point about what we notice not being
propositional:
Its not only that we cant provide an exhaustive catalogue of what has been
attended to when we are led to see something in a new light; the difculty is
more fundamental. What we notice or see is not, in general, propositional in
character. Of course it may be, and when it is, it usually may be stated in fairly
plain words. . . . Seeing as is not seeing that. Metaphor makes us see one thing
as another by making some literal statement that inspires or prompts the
insight. Since in most cases what the metaphor prompts or inspires is not
50
Ibid., pp. 262263.
entirely, or even at all, recognition of some truth or fact, the attempt to give
literal expression to the content of the metaphor is simply misguided.51
It is difcult to say exactly which of the several, subtly different points

made in these well-known passages are meant as attacks on (C3), and
which are not. Rather than attempt to answer this question, I will simply
discuss whether they are effective as attacks on (C3).
First, it poses no challenge to (C3) merely to say that some of what
metaphors make us notice is non-propositional. To hold that coming to
understand a metaphor consists at least partly in grasping propositions does
not commit one to holding that the metaphor makes one notice nothing
but propositions.52 Nor does it pose a challenge to (C3) to say that what
metaphors make us notice is not nite in scope. Finally, it does not
undermine (C3) to say that metaphors make us see one thing as another.
Coming to understand a metaphor can still partly consist in grasping
propositions even if metaphors make us see one thing as another.
Seeing-as is not seeing-that, but it is not incompatible with seeing-that.
Or, for that matter, with noticing-that and grasping-that.
To undermine (C3), then, Davidson must rest his case on his claim
that, sometimes, what a metaphor makes us notice is not even at all
propositional. The cases in which what we notice is merely not entirely
propositional will not do. If there are cases in which the metaphor does not
make us notice anything propositional, then coming to understand
such metaphors could not consist in grasping propositions (assuming that
we grasp a proposition only if we notice something propositional). The
question is, then: what evidence does he provide for the existence of such
cases?
The only evidence Davidson provides is his claim that it is so hard to
decide, even in the case of the simplest metaphors, exactly what the content is
supposed to be. If what the metaphor makes us notice were nite in scope
and propositional in nature, he says, this would not be hard to decide. In that
case, we would simply project the content the metaphor brought to
mind on to the metaphor. But this often is hard to decide. Therefore, what
such metaphors make us notice is not propositional in nature and nite in
scope.
51
Ibid., p. 263.
52
I am assuming, for the sake of argument, that there is such a thing as noticing
propositions.
There are two problems with this argument as an attack on (C3). First, it
does not rule out the possibility that what such metaphors make us notice
is propositional in nature, yet not nite in scope. Its conclusion is only
that what we are made to notice is not both propositional and nite in
scope.
Second, it is false that, if what the metaphor made us notice were
propositional in nature, it would not be hard to decide exactly what the
content is supposed to be. A possibility I mentioned earlier serves as a
counterexample to this claim.
For many metaphors, there are many propositions which are such
that (i) the metaphor makes us notice them, and (ii) it is hard to decide
whether they are part of the content of the metaphor. For example, the
proposition that Romeos day begins with Juliet seems to be of this kind.
Romeos metaphor made Cavell notice this proposition. But it is hard to
decide whether it is part of the content of Romeos metaphor. Accord-
ingly, it is hard to decide exactly what the content of the metaphor is
supposed to be.
Being made to notice propositions and nding it hard to decide
whether they are part of the content of the metaphor are perfectly
compatible. Davidson gives us no reason to believe that, as he says, we
would simply project whatever content the metaphor brings to mind on
to the metaphor, if it made us notice something propositional. We
certainly do not project whatever content a literal utterance brings to
mind on to that utterance. So even if it is hard to decide exactly what
the content of a metaphor is supposed to be, this does not establish that
what some metaphors get us to notice is entirely non-propositional. It
therefore does not establish that coming to understand metaphors does not
involve grasping propositions apart from what would be expressed by
using the same sentence literally.
I conclude, therefore, that Davidson provides no successful argument
against the claim that coming to understand metaphors partly consists in
grasping propositions.
The nal objection I will discuss here is what I regard as the most challen-
ging of the objections that apply to the Minimal Thesis. The objection
is simply that there are many counterexamples to the claim that metaphor
is based on likeness. Several philosophers, it is commonly thought, have
identied numerous metaphors that are not understood by reference to
likenesses.
I will attempt both to show that this objection is mistaken, and to

explain why it appears plausible. All the apparent counterexamples of
which I am aware are of four kinds. Showing that they belong to these
kinds will both disarm objections that appeal to them, and make it clear
why these cases have seemed to be counterexamples to the view that
metaphors are based on likenesses.
Searle gives many candidate counterexamples. As we have seen, he holds
that, though similarity often plays a role in the comprehension of metaphor,
the metaphorical assertion is not necessarily an assertion of similarity.53
However, he also provides examples of metaphors in which, he claims,
similarity does not play a role even in their comprehension: For example,
the numerous spatial metaphors for temporal duration are not based on
literal similarities. In time ies, or the hours crawled by, what is it that
time does and the hours did which is literally like ying or crawling?54
These seem to be clear counterexamples both to the position Searle expli-
citly targets and to the Minimal Thesis.
They are not. We say, Time ies, when we are talking about cases in
which time seems to elapse quickly. We say that time ies when youre
having fun, or that the years y by after you reach a certain age. You do
not understand Time ies unless you grasp that seeming to elapse quickly is
attributed to time.
Now, seeming to elapse quickly is not a likeness-maker for the likeness
being like something that ies. Flying things do not seem to elapse at all. This
is why this metaphor appears to be a counterexample. However, seeming to
elapse quickly is a way of possessing the property, seeming to be quick. And
this property is a likeness-maker for the likeness, being like something that
ies. Many things that y seem to be quick. This is often because they are
quick: they move quickly. So Time ies is not a counterexample to the
Minimal Thesis. It is merely a metaphor that attributes a way of possessing
a likeness-maker for the likeness indicated by the metaphorical element.
Several of Searles other examples are also cases of this kind. I have
already discussed Sally is a block of ice, and a similar reply deals with his
example of sexual frigidity.55 Sexually frigid people are sexually unre-
sponsive, and being sexually unresponsive is a way of being unresponsive,
which is a likeness-maker for the likeness, being like something frigid. Searle
53
Searle, Metaphor, p. 88. 54
Ibid., p. 99.
55
Ibid., p. 98.
also gives the example of sweet disposition, writing: Of course, sweet

dispositions and sweet things are both pleasant, but much more is conveyed
by the metaphor than mere pleasantness.56 This is true. But, in the rst
place, some of the other properties attributed by the metaphor are also
shared by sweet-tasting things. A sweet disposition, like something sweet-
tasting, is such as to please immediately and easily. And, in the second place,
the metaphor may convey that the disposition is pleasant in certain ways,
some of which are distinctive of human dispositions and are not ways in
which sweet-tasting things can be pleasant. For example, sweet dispositions
are endearing. Being endearing is a way of being pleasant; but pleasant
sweet-tasting things do not endear themselves to us. Searles examples of
warm welcome and lukewarm friendship conform to the same model.
Warm welcomes, like many warm things (replaces, beds, etc.), are com-
forting; describing a welcome as warm may also be a way of characterizing
it as friendly, which is a way in which welcomes (though not replaces) can
be comforting. So Searle concludes much too quicklyand, indeed,
makes concessions inconsistent with the claimthat similarity does not
even play a role in the comprehension of these metaphors.
Some apparent counterexamples, however, are of a different kind. The
explanation of these cases is that they are combinations of metaphor and
another way of extending the use of expressions.
This is true of Searles example, I am in a black mood. Severin
Schroeder criticizes Searles use of this as a counterexample to comparison
theories by arguing that
to call dangerous or depressing things black is based on the fact that
darkness (which looks black) is typically felt to be dangerous or depressing.
This is a fairly straightforward similarity. However, in many gurative uses
of the word black . . . the exploitation of such fundamental similarity is
combined with metonymical shifts. Thus my black mood is not, like a black
night, something that makes me gloomy, for it is my being gloomya
metonymical shift from cause to effect.57
Schroeders reply is on the right lines. Black mood is indeed an example

of both metaphor and another way of extending the use of an expression.
As I shall explain below, the relevant likeness is not that between literally
56
Ibid., p. 99.
57
Schroeder, Why Juliet is the Sun, p. 91.
black causes of gloom and metaphorically black causes of gloom. But

Schroeders basic point here is right.
Black mood is a combination of metaphor and causal analogy. In a
causal-analogical extension of a use of an expression, the expression is
applied to something that causes, or is caused or affected by, what the
expression attributed or applied to in its previous use. Aristotle pointed out
this way of extending the use of healthy: Everything which is healthy is
related to health, one thing in the sense that it preserves health, another in
the sense that it produces it, another in the sense that it is a symptom of
health.58 Aquinas gives the examples of healthy medicine and healthy
urine.59 We attribute the property of being such as to cause, contribute to,
or maintain health in the former case, and the property of being caused by
health in the latter case. Similarly, Grays Elegy describes the sound of
sheep bells at evening as drowsy tinklings that lull the distant folds.60
The tinkling sounds are such as to cause drowsiness.
Black in black mood is used both metaphorically and by causal
analogy. Black moods are like moods or experiences caused by black
things of various kinds, such as black clouds and darkness. Similarly,
sharp pains are like pains caused by sharp things, and prickly pains are
like pains caused by prickly things. In each case, there is a likeness to
something caused by things the expression literally applies to.
A third class of apparent counterexamples consists of cases in which
the use of the expression in question is indeed an extension of a pre-
existent use of that expression, but is not a metaphorical extension. In
these cases, it is not even a combination of metaphor and some other way
of extending the use of an expression.
One case of this kind is Lycans use of the lines from Hamlet as a
counterexample to comparison theories:
Lends the tongue vows.
58
Aristotle, Metaphysics, trans. W. D. Ross, in The Complete Works of Aristotle, ed. Jonathan
Barnes (Princeton: Princeton University Press, 1984), vol. 2, 1003a. See also Aristotle, Topics,
trans. W. A. Pickard-Cambridge, in The Complete Works of Aristotle, vol. 1, 106b.
59
Thomas Aquinas, Summa Theologiae: Questions on God, ed. Brian Davies and Brian
Leftow, trans. Brian Davies (Cambridge: Cambridge University Press, 2006), Part 1, Question
13, Article 10, p. 161.
60
Thomas Gray, Elegy Written in a Country Church Yard, in Christopher Ricks, ed.,
The Oxford Book of English Verse (Oxford: Oxford University Press, 1999), p. 278, l. 8.
As we saw, Lycan takes blood, burns, soul, lends, tongue, and vows
to be used metaphorically here. This is mistaken.
This line occurs early in the play when Ophelia is trying to persuade
Polonius that Hamlet loves her. She says that Hamlet conrmed his
expressions of love With almost all the holy vows of heaven.61 Polonius
replies dismissively:
I do know
Lends the tongue vows.62
Polonius is here alluding to the theory of humours, according to which the

blood of people in a passionate state is hot. Hence blood is not being used
metaphorically. According to the theory alluded to, it is literally true that
the blood of such people is hot. Allusion is not a species of metaphor,
though some allusions are metaphorical. We allude to beliefs in divine
providence when we say, Someone up there must like me, to communi-
cate that we have been very lucky. But we do not speak metaphorically
when we say this, even if we are atheists and so do not mean it literally.
Burns is perhaps being used hyperbolically, although believers in humoral
theories also believed that psychological imbalances could be caused by
humours literally burning within the blood and releasing vapours.63 The
soul is either not being used guratively at all, or is used as a metonym for
the person.64 In either case, it is not a metaphor. The context makes clear
that vows is being used literally, for Polonius and Ophelia are talking
about Hamlets vows. The tongue is used as a synecdoche, the gure of
speech in which an expression for a part of some whole is used to stand for
the whole itself, as in All hands on deck. In this case, the tongue is used
to stand for the person.
Lastly, lends is indeed being used metaphorically, and so is prodigal.
The use of prodigal here combines metaphor and a gure of speech
61
Shakespeare, Hamlet, 1.3.114.
62
Ibid., 1.3.1151.3.117.
63
See Cynthia Marshall, Cosmology and the Body, in Donna B. Hamilton, ed.,
A Concise Companion to English Renaissance Literature (Oxford: Blackwell, 2006), p. 221.
64
In metonymy, we apply to a thing a word or phrase for something associated with that
thing, as when we ask someone to address the chair, wonder how the markets will react, or
accuse the White House of a cover-up.
known as anthimeria. In anthimeria, an expression standardly used as one

part of speech is used as a different part of speech, as when we say we had a
good cry.65 In Poloniuss line, the adjective prodigal is used as an adverb.
The metaphorical use of these expressions here is consistent with the
Minimal Thesis. People who lend prodigally readily lend to people who
are unlikely to honour their debts. Therefore, such lenders readily put
people in a position to benet from something with regard to which those
people have obligations they are unlikely to honour. This property is a
likeness-maker for the likeness being like someone who lends prodigally.
Poloniuss metaphor attributes a way of possessing this likeness-maker.
Polonius communicates that he knows how readily, when a person is in
a passionate state of mind, his soul causes him to swear vows he is unlikely
to honour. Readily causing someone to swear vows he is unlikely to
honour is a way of readily putting someone in a position to benet from
something with regard to which he has obligations he is unlikely to
honour.
A common complaint made against many philosophers writing on
metaphor, and especially against defenders of comparison theories, is that
they conne themselves to simple, hackneyed subject-predicate meta-
phors and ignore the rich, complex metaphors of so much actual poetry.
Lycan concludes his discussion of the above example by commenting: It is
no wonder that simile theorists have in the main stuck to simple subject-
predicate examples.66 But the Minimal Thesis can accommodate poetic
metaphors. The above discussion suggests, not that poetry undermines the
claim that metaphor is based on likeness, but that when we look at poetry,
we should expect to nd many rhetorical devices other than metaphor.
Many major poets of the past received an intensive education in rhetoric,
which drilled them in the use of many gures of speech that most people
now are not taught. The injunction to take the language of actual poetry
seriously when thinking about metaphor is good advice. But following it
65
See T. V. F. Brogan, Anthimeria, in Alex Preminger and T. V. F. Brogan, eds, The
New Princeton Encyclopedia of Poetry and Poetics (Princeton: Princeton University Press, 1993),
p. 74.
66
Lycan, Philosophy of Language, p. 183. For similar complaints about philosophers
examples, see Roger White, The Structure of Metaphor: The Way the Language of Metaphor
Works (Oxford: Blackwell, 1996); Guttenplan, Objects of Metaphor, p. 93.
requires familiarizing ourselves with these non-metaphorical ways of

extending the use of expressions.
Lastly, it is worth distinguishing a special case of this kind. These are
cases in which a metaphorical use of an expression has itself been extended
in a non-metaphorical way. It is easy to assume that all extensions of uses of
expressions are extensions of literal uses of expressions. But this is not so.
You can also extend non-literal uses of expressions (which are themselves
extensions of some other use). In the cases I have in mind, an established
metaphorical use of an expression, or an imaginable metaphorical use of an
expression, has been extended in a non-metaphorical way. The resulting
use of the expression is not metaphorical. But because it is an extension of a
metaphorical use, it is easy to mistake it for one.
Searles example of bitter person is a case of this kind. My discussion
of causal analogy gives us the resources with which to explain bitter
person. For a bitter person is someone marked in certain ways by the
effects of bitter experiences. Bitter experiences are like experiences caused
by bitter-tasting things. Therefore, the use of bitter to characterize
experiences, like the use of black to characterize moods and the use
of sharp and prickly to characterize pains, is a combination of metaphor
and causal analogy. It is a metaphorical and causal-analogical extension
of the use of bitter to characterize bitter-tasting things. The use of bitter
to characterize persons, in turn, is a causal-analogical extension of
the use of bitter to characterize experiences. The expression applied
non-literally to the experiences is applied to persons affected by such
experiences.
So these four objections do not defeat the Minimal Thesis. And the
apparent counterexamples I have discussed can be accounted for in these
four ways. Moreover, the account in each case makes it evident why they
have been mistaken for counterexamples to the claim that all metaphors
are based on likeness. This discussion should reduce our condence that
metaphors are too various for them all to be based on likeness. It highlights
the fact that there are far more non-metaphorical ways of extending uses of
expressions than most of us are disposed to recognize immediately. It also
shows that we tend to use the term metaphor to describe examples of
non-metaphorical ways of extending uses of expressions when we do not
recognize them.
4. What the Minimal Thesis explains

I will now conclude my account of the Minimal Thesis by discussing four
of the explanatory benets it brings.
First, the Minimal Thesis explains how metaphor differs from other
ways of extending the use of expressions. As I have stressed, to speak
metaphorically is to extend the use of an expression in a certain way.
Moreover, this way of extending the use of an expression is not the way in
which we extend the use of expressions when we employ sarcasm, causal
analogy, metonymy, synecdoche, anthimeria, and so forth. The Minimal
Thesis explains how metaphor differs from these.67
Depending on the account we should give of simile, the Minimal Thesis
may not explain how metaphor differs from simile. If similes attribute
likenesses, likeness-makers, determinates of likenesses, and ways of pos-
sessing likeness-makers, then the Minimal Thesis does not explain how
metaphor differs from simile. But in that case, we could appeal to the claim
that similes attribute these properties by using like, as, or other words
that can be used literally to attribute likenesses and their determinates.
Second, the Minimal Thesis explains why it is often unclear whether a
given expression is being used literally or metaphorically. One can imagine
people reasonably disagreeing over whether, for example, a horseshoe is
a kind of shoe or only metaphorically a shoe, whether a tree surgeon is a
kind of surgeon or only metaphorically a surgeon, whether genetic engin-
eering is a kind of engineering, and so on. This kind of uncertainty is
possible because metaphors attribute likeness-makers for a likeness indi-
cated by the metaphorical element. We may be unsure whether the
properties these expressions attribute not only make something like a
surgeon, a shoe, engineering, etc., but actually make it an instance of
that kind. If something has the right likeness-makers for the likeness being
like a duck, then it is not just like a duck: it is one.
Third, the Minimal Thesis enables us to explain why some metaphors
are open-ended, or at least why they appear to be. The Minimal Thesis
identies a constraint on the properties a metaphor can attribute. The
properties must be either likenesses indicated by the metaphorical element,
67
Comparison theorists have claimed similar advantages for their theories. See Robert
J. Fogelin, Metaphors, Similes, and Similarity, in Jaakko Hintikka, ed., Aspects of Metaphor
(Dordrecht: Kluwer, 1994), pp. 3033; Schroeder, Why Juliet is the Sun, pp. 9294.
4. WHAT THE MINIMAL THESIS EXPLAINS 123
or related to such likenesses in at least one of the three ways I identied. This
excludes a great deal. But there are also indenitely many properties this
does not exclude. Suppose that, rather than being asked to explain Romeos
metaphor, we were simply asked to think of which properties the sun has or
is thought of as having, or which ways of possessing such properties, it would
be relevant or appropriate for Romeo to attribute to Juliet in the context. If
we were given that task, it would be no surprise if we found that, to borrow
Davidsons phrase, there is no end to what we want to mention,68 or that
there is no point at which we feel no more properties could be thought of.
So if this is, roughly, the task we set ourselves when we try to interpret
certain metaphors, that would explain why they seem open-ended.
Finally, the Minimal Thesis enables us to explain why some metaphors
seem indispensable. As I shall discuss in the following chapter, many hold
that we sometimes need metaphor to express or communicate certain
thoughts. There are, I believe, several reasons why metaphors can seem
indispensable. I will mention some in Chapter 5. Here I will note that if
the Minimal Thesis is right, it should be unsurprising that many metaphors
seem indispensable to many people. When we think of an apt comparison,
we often cannot think of a better description that does not involve
comparison. You might be able to think of no better way of describing
how a persons hands feel than by saying they feel like dead leaves; how
a beer tastes than by saying it has hints of bread and caramel; or how a
person wears his hair than by saying he wears it like Einstein wore his.
It can easily seem that any description you can think of that does not
involve comparison would leave out something important that these
descriptions capture. Likening one thing to something else can seem to
be an indispensable way of expressing what we want to express.69 If the
68
69
Austin made a similar point. As he put it, in many situations certain words function as
adjuster-wordswords, that is, by the use of which other words are adjusted to meet
the innumerable and unforeseeable demands of the world upon language. . . .
Vocabularies are finite; and the variety of possible situations that may confront us is
neither finite nor precisely foreseeable. . . . One day we come across a new kind of
animal, which looks and behaves very much as pigs do, but not quite as pigs do; it is
somehow different. . . . What we could do, and probably would do first of all, is to say,
Its like a pig. (Like is the great adjuster-word, or, alternatively put, the main
flexibility-device by whose aid, in spite of the limited scope of our vocabulary, we can
always avoid being left completely speechless.) ( J. L. Austin, Sense and Sensibilia, ed.
G. J. Warnock (Oxford: Oxford University Press, 1962), pp. 7374)
Minimal Thesis is right, it is unsurprising that similar claims are often made
about metaphors.
I will use the Minimal Thesis to explain why metaphor is used so
frequently in criticism. Before providing my explanation, however,
I must address an important question that frequently arises in discussions
of critics use of metaphor. This is the question of whether some metaphors
are, in fact, indispensable.
5
The Dispensability
of Metaphor
In many discussions of the use of metaphor in criticism, the issue of

metaphors dispensability or indispensability for thinking, expressing, com-
municating, or discovering various truths often arises. Some philosophers
would take the explanation of the prevalence of metaphor in criticism to be
that the thoughts critics express or communicate are of a kind that cannot
be communicated or expressed without metaphor. Roger Scruton, for
example, holds that in our most basic apprehension of music there lies a
complex system of metaphor, which is the true description of no material
fact. And the metaphor cannot be eliminated from the description of music,
because it is integral to the intentional object of musical experience.1 The
metaphors Scruton has in mind include, among others, metaphors of action
affecting our experience of rhythm, metaphors of movement affecting our
apprehension of melody, and metaphors of open and closed in our percep-
tion of chords. Nick Zangwill also holds that metaphorical aesthetic de-
scriptions are ineliminable. He claims that this is because such judgments
are based on an inner response or feeling, and thoughts about the qualita-
tive character of mental states cannot be expressed without metaphor unless
they are thoughts about very general characteristics of those mental states,
such as their being painful.2
1
Roger Scruton, Understanding Music, in The Aesthetic Understanding: Essays in the
Philosophy of Art, 2nd edn (South Bend, IN: St Augustines Press, 1998), p. 97. See Malcolm
Budd, Musical Movement and Aesthetic Metaphors, in Aesthetic Essays (Oxford: Oxford
University Press, 2008), pp. 154170, for what I regard as decisive objections to Scrutons
views.
2
Nick Zangwill, The Metaphysics of Beauty (Ithaca: Cornell University Press, 2001),
pp. 172174.
126 THE DISPENSABILITY OF METAPHOR
Many other philosophers have held that metaphor is indispensable for

various other purposes as well. Let us call the Indispensability Thesis the
claim that:
We use at least some metaphors to think, to express, to communicate,
or to discover what cannot be thought, expressed, communicated, or
discovered without metaphor.
Versions of this thesis are advocated in a remarkable variety of areas: one nds
it supported not only by aestheticians, but also by metaphysicians, linguists,
philosophers of language, philosophers of mathematics, philosophers of sci-
ence, and philosophers of religion. But one nds it more often asserted than
argued for. I wish to examine arguments for it. I will argue that support for the
Indispensability Thesis is based on several confusions. Although I do not show
that the thesis is false, I provide eight grounds for suspicion of our sense (if we
have it) that some metaphors are indispensable for the purposes claimed by
advocates of the Indispensability Thesis. Accordingly, we should be sceptical
of explanations of the prevalence of metaphor in criticism that claim critics
express thoughts that cannot be expressed without metaphor.
1. Metaphor and make-believe

Stephen Yablo defends a version of the Indispensability Thesis by linking
metaphor to make-believe.3 Inuenced by Kendall Walton,4 Yablo points
out that the real properties of props used in a game of make-believe can
help determine what we are to imagine in that game. For example,
suppose we are playing a game in which a certain mud-cake is used to
represent a pie and a certain hollow tree-stump is used to represent an
oven. If the mud-cake is, in reality, too big to t in the tree-stump, we are
to imagine in the game that the pie is too big to t in the oven.
Because of this, we could make an assertion within a game as a way of
giving voice to a fact holding outside the game: the fact that the props are in
such and such a condition, viz., the condition that makes [what we
3
Stephen Yablo, How in the World?, in Thoughts (Oxford: Oxford University Press,
2008), pp. 191220; Stephen Yablo, Does Ontology Rest on a Mistake?, in Things (Oxford:
4
See Kendall L. Walton, Metaphor and Prop Oriented Make-Believe, European Journal
of Philosophy 1/1 (1993): 3956.
1. METAPHOR AND MAKE - BELIEVE 127
pretend to assert] a proper thing to pretend to assert.5 You could say, The
pie is too big to t in the oven, to give voice to the real-world fact that the
mud-cake is too big to t in the tree-stump.
According to Yablo, metaphors are like pretend-assertions used for this
purpose. A metaphor suggests a game of make-believe in which what is
described metaphorically is used as a prop. If we describe Italy as a boot,
that suggests a game in which Italy is used as a boot-prop. By using a
metaphor, we represent what we describe as having properties that would
make our utterance appropriate in a game suggested by the metaphor.6
A metaphor is pretence-worthy when the object described does have
properties that would make our utterance appropriate in such a game. For
example, Crotone is in the arch of the Italian boot is pretence-worthy
because Italy and Crotone have properties that make it appropriate to
imagine, in a game in which Italy is used as a boot-prop, that Crotone is in
its arch. What Yablo calls the metaphorical content of a metaphor is
given by the worlds in which the same sentence, meaning the very same
thing, is pretence-worthy.7
Yablo argues that some metaphors are representationally essential. That
is, there is no way to access the ensembles of worlds picked out by their
shared property of legitimating a certain pretence except via metaphor:
The language might have no more to offer in the way of a unifying principle
for the worlds in a given content than that they are the ones making the
relevant sentence ctional. It seems at least an open question, for example,
whether the clouds we call angry are the ones that are literally F, for any F
other than such that it would be natural and proper to regard them as angry if
one were going to attribute emotions to clouds. Nor does a literal criterion
immediately suggest itself for the pieces of computer code called viruses, the
markings on a page called tangled or loopy, the glances called piercing, or the
topographical features called basins, funnels, and brows.8
One problem with Yablos argument, as presented, stems from the fact
that the same expression can be used to attribute different properties on
different occasions when it is used metaphorically. For example, Muhammad
Ali called Joe Frazier a gorilla as an insult. But a primate scientist might use
the same expression as a compliment, to characterize someone as gentle and
5
Yablo, Does Ontology Rest on a Mistake?, p. 131.
6
Yablo, How in the World?, p. 219.
7
Yablo, Does Ontology Rest on a Mistake?, pp. 133134.
8
Ibid., pp. 134135.
peace-loving. So there is no F such that (i) all the people we call gorillas are
literally F, and (ii) we represent them as being F whenever we call them
gorillas metaphorically. But obviously, this does not show that we ever use
gorilla to represent people as having a property that cannot be attributed to
them by speaking literally. Making the same point about angry as applied to
clouds does not show that angry clouds is an indispensable metaphor, either.
Yablos claim must therefore be understood as the claim that there is a
metaphorical use of angry in which clouds are represented as having a
certain property,9 but it is unclear how we would attribute this property by
speaking literally. In this use of angry, (i) we represent clouds as having a
certain property; (ii) their having this property would make it appropriate
to describe them as angry in a game of make-believe; and (iii) it is utterly
unclear how we would attribute this property by speaking literally. The
point is not that no literal criterion suggests itself for all clouds that can be
called angry in all metaphorical uses of the term. The point is that no
literal criterion suggests itself for all clouds that can be called angry in this
metaphorical use of the term.
To assess Yablos argument, then, we need to identify the use of angry
he has in mind. And the problem is that no use that suggests itself is one in
which the property we represent clouds as having can be attributed only
with metaphor. For example, we sometimes call clouds angry to indicate
that they look like storm-clouds. Looking like storm-clouds is plainly a
property that can be attributed by speaking literally. Again, we sometimes
call clouds angry to indicate that they look like something expressive of
anger, as when a cloud looks like an angry face. But this property, too, can
be attributed without metaphor.
I take it that the latter use of angry is, in fact, the one Yablo has in
mind. This use can easily seem indispensable if one disregards the possibil-
ity that angry, in this use, represents clouds as looking like something
expressive of anger. Philosophers are liable to disregard this possibility,
given the widespread opposition to the idea that metaphors are based
on likenesses. If one does disregard it, and one considers all the clouds
that can be called angry in this use of the term, one is apt to be struck by
the enormous variety among themby the apparent lack of a unifying
principle for them, as Yablo puts it. Looking like something expressive of anger
is, after all, a multiply realizable property. A cloud can look like something
9
Or properties. I omit this qualication in what follows.
1. METAPHOR AND MAKE - BELIEVE 129
expressive of anger in virtue of many different shapes, and in virtue of

many different patterns of illumination and shadow on its surface. We
would be hard put to identify a literally ascribable property that the
metaphor could be attributing to all these clouds, if we disregard the
property of looking like something expressive of anger.
Yablos other examples also attribute multiply realizable properties that
can be attributed without metaphor. And like the use of angry cloud I just
considered, they attribute multiply realizable determinates of likenesses.10
Some of these examples may now be dead metaphors; but it is clear what
they conveyed when they were live. To call a piece of computer code a
virus is (or was) to represent it as propagating itself like a virus; to call
markings on a page tangled or loopy is to represent them as looking like
(a picture of ) something tangled or loopy; to call a glance piercing is to
represent it as affecting us as something piercing does; and describing
topographical features as basins, funnels, or brows involves characterizing
these features as being shaped like basins, funnels, or brows. This suggests that
in these cases, too, these metaphors appear indispensable because we disre-
gard the possibility that these metaphors attribute determinates of likenesses,
and because the determinates they attribute are multiply realizable.
Yablo also argues that a nominalist about numbers would regard the
following as a metaphor:
(T) The number of planets divided by the number of stars is 2.4.

And he claims that the nominalist would say that T
provides us with access to a content more literally expressed by
(U ) There are 12 planets and 5 stars or 24 planets and 10 stars or . . .
And now here is the rub. The rules of English do not allow innitely long
sentences; so the most literal route of access in English to the desired content is
T, and T according to the nominalist is a metaphor.11
However, it is false that T is an indispensable metaphor if nominalism

about numbers is true. T is not a metaphor at all. To be a metaphorical use
of number, the use of number in T would have to be an extension of
some other use of number. As I stressed in Chapter 4, to use a term
metaphorically is to extend the use of a term in a certain way. But the use
10
See Chapter 4, section 1.
11
Yablo, Does Ontology Rest on a Mistake?, p. 135.
of number in T is not an extension of a use of number at all. The use of

number with which nominalists are concerned is not one we understand
by drawing on our knowledge of some other use of number. It is an
unextended use of this word. If nominalism about numbers is true and T is
false, T might, of course, give us access to a truth. But whether nominalism
about numbers is true or false, T is not a metaphor.
Yablo also holds that some metaphors are procedurally essential. He
argues that sometimes, the metaphor-user is not (or not merely) using
metaphor to communicate a certain message she has in mind, but to prompt
her audience to discover things of which she herself may be unaware:
Someone who utters S in a metaphorical vein is recommending the project of
(i) looking for games in which S is a promising move, and (ii) accepting the
propositions that are Ss inverse images in those games under the modes of
presentation that they provide. The overriding principle here is make the most of it;
construe a metaphorical utterance in terms of the game or games that retromap it
onto the most plausible and instructive contents in the most satisfying ways.12
By using S metaphorically, then, we in effect recommend that our audi-

ence follow a certain procedure for discovering certain instructive prop-
ositions. In calling such metaphors procedurally essential, Yablo implies
that metaphor is needed to carry out this procedure.
Metaphor is not, however, needed for this. We do need the concept of
a game of make-believe to carry out the procedure Yablo describes. But
we do not need to use S metaphorically in order to think of games within
which S is a promising move, nor to think of what plausible and instructive
propositions would have to be true for it to be a promising move. For S to
be a promising move within a game is for it to be appropriate to use S
literally within a game of make-believe involving the object as a prop. At
least, judging by what Walton and Yablo say, for Crotone is in the arch of
the boot to be pretence-worthy is for it to be appropriate to pretend to
assert, literally, that Crotone is in the arch of the boot. To use S literally
within a game is, obviously, not to use it metaphorically within the game.
Neither is it to use S metaphorically outside the game. An actor playing a
character who is speaking literally is not himself speaking metaphorically.
Therefore, if I invite you to think of games in which S is a promising
move, I am inviting you to think of scenarios in which it would be
appropriate to use S in a certain non-metaphorical way. If there is a
12
Ibid., pp. 137138.
2. METAPHOR IN CRITICISM 131
sound argument to show that, in order to think of scenarios in which S is

used in this non-metaphorical way, or to invite someone to do so, we must
use S metaphorically, Yablo does not provide it. Even if Yablo and Walton
are right that metaphors suggest games of make-believe, nothing they say
shows that we need metaphor to think of the games metaphors suggest.
Therefore, even if these games are indispensable for some purpose meta-
phors achieve by suggesting them, it does not follow that metaphor is too.
2. Metaphor in criticism
Berys Gaut holds that metaphors used in art criticism are often indispens-
able.13 Thinking of the metaphor, according to Gaut, is often the only way
to have the experience the metaphor provides. This is because of the role
of metaphor in classication,14 and because the way we classify affects
how we experience things.
He asks us to imagine that a critic gets us to attend to various properties
of a Kandinsky by describing it as alive with movement:
The metaphor classies together a motley bunch of properties: properties of
vibrancy, subdued violence, extreme contrasts of saturation and hue, having
jagged edges, acentric composition, a sense of uctuation in pictorial depth,
and so on. . . . How does one decide how to extend this list? There is such
diversity here that we have no sense of how to carry onexcept by use of the
master-metaphor of being alive with movement. And certainly there is no
reason to classify together these diverse properties other than because of their
connection to the metaphor. So the metaphor cannot be discarded: it guides
our ability to group these properties with each other, grounds our sense that
they belong together.
Further, we are aware not just that these properties belong together, but
also that what makes this the case is that they are all connected to the
metaphor.15
So Gauts claim is that we need the metaphor of the Kandinsky as alive

with movement in order to:
(a) tell us how to carry on extending the list of properties he mentions;
(b) ground our sense that they belong together; and
(c) tell us why they belong together.
13
Berys Gaut, Metaphor and the Understanding of Art, Proceedings of the Aristotelian
Society 97 (1997): 223241.
14 15
Ibid., p. 230. Ibid.
The basis for this is the claim that there is no reason to classify together these
diverse properties other than because of their connection to the metaphor.
But this claim is false. There is another reason to classify together the
properties the metaphor classies together: they are all connected to the
property, being alive with movement. They are connected to it in various ways.
Some of the properties of the forms are also properties of creatures alive with
movement (e.g., causing a sense of uctuation, violence). Others are prop-
erties in virtue of which the forms share properties with creatures alive with
movement: extreme contrasts of saturation and hue, for example, can make
forms seem to leap out at us, and vibrant colours can arrest our attention.
Still others are properties the picture shares with pictures of things alive with
movement: acentric composition can be used (along with other features) to
show that what is depicted is moving fast, and jagged edges can be used to
suggest erratic motion, or the path of something moving erratically.
If the properties the metaphor draws to our attention are all connected
to the property of being alive with movement, then this undermines
Gauts argument for the Indispensability Thesis. For in that case, knowing
what properties are connected to the property of being alive with move-
ment, and being able to identify them in the Kandinsky, would:
(a) tell us how to carry on extending the list of properties he
mentions;
(b) ground our sense that they belong together; and
(c) tell us why they belong together.
And one can have such knowledge and exercise such an ability without
the metaphor of the Kandinsky as alive with movement.
Gaut comes close to acknowledging that the properties a metaphor
classies together are connected to something other than the metaphor.
He writes:
A person who classied together all and only artworks we call sad, but
denied any connection between them and sadness, would have failed to grasp
the aesthetic property we were indicatingwould have failed to grasp the
sadness of these things, and so would have missed what was of primary
interest to us. Hence there could not be a person whose experience and
understanding of a work was as ours is, but who did not have a grasp of the
metaphor in terms of which we classify features of the work.16
16
Ibid., pp. 230231.
3. METAPHOR IN SCIENCE 133
The conclusion does not follow. The fact (if it is a fact) that you must accept
that there is a connection between certain artworks and sadness in order to
grasp the sadness of those works does not show that you need a metaphor to
grasp their sadness. It only shows that you need the concept of sadness to do so.
3. Metaphor in science
Elisabeth Camp argues that not everything that can be meant can neces-
sarily be given literal expression, even in a private language.17 Sometimes,
according to Camp, we need metaphor to express, and not merely to
communicate to others, what we use it to express.
Developing an argument advanced by Richard Boyd,18 Camp asks us to
suppose that we are scientists investigating sub-personal cognitive pro-
cesses. We want to identify a certain kind of causally efcacious property.
However, we dont know much about properties of this kind. We know
something about the propertys causal relations, but not enough to dene
it in functional terms; nor can we identify the property ostensively.
This is where metaphor comes in. Camp writes:
We can still make theoretical and experimental progress, though, by thinking
metaphoricallyfor example, by exploiting the metaphor of memory stor-
age and retrieval as the opening of a computer le. . . . Research progresses, in
part, by investigating specic candidate similarities that might underwrite the
analogical equations that are implicit in such metaphors. As we establish some
similarities and rule out others, our cognitive access to the properties under
investigation becomes more fully and literally conceptualized. At some point,
if investigation progresses well, we may well be able to dispense with the
metaphor in favor of a new, literally applicable concept. But at this early stage
of our inquiry, the metaphor plays an essential role in xing what we are
thinking about.19
This is a case, Camp claims, in which metaphor is needed to express what

we use it to express.
This argument is unconvincing. We do not need metaphor to do what
Boyd and Camp describe at any stage of the process of inquiry they
17
Elisabeth Camp, Metaphor and That Certain Je Ne Sais Quoi , Philosophical Studies
129/1 (2006): 17.
18
Richard Boyd, Metaphor and Theory Change: What is Metaphor a Metaphor For?,
in Andrew Ortony, ed., Metaphor and Thought, 2nd edn (Cambridge: Cambridge University
Press, 1993), pp. 481532.
19
Camp, Metaphor and That Certain Je Ne Sais Quoi , pp. 1718.
envisage. Consider what the scientist is doing. She is investigating whether

memory has certain properties, which it would share with the process of
opening a computer le if it has them. Her research is guided by the
hypothesis that memory does share properties, or relevant properties, with
the process of opening a computer le. She is trying to establish, for
various properties that memory would share with this process if it has
them, whether memory does have them.
To conduct her research in this way, the scientist does need the concept,
opening a computer le. She also needs to know what properties the process
of opening a computer le has. But she does not need metaphor in order
to think of properties memory would share with the process of opening a
computer le if it has them. Nor does she need metaphor in order to
investigate whether memory does have them.
Camp responds to an objection like this one. She considers the objec-
tion that we might make explicit the implicit analogical equation through
which the metaphor xes the property we want to investigate by using a
literal description, such as:
(8) The property of cognition that causes memory retrieval in a

manner that is analogous in some theoretically relevant respect to
opening a folder in a computer program.20
Her response is that identifying the denotation of a literal description like (8)
requires the same cognitive capacity as the original metaphor does. We still
need to identify which particular similarities are relevant, and then construct
a positive concept of the appropriate property on that basis.21 But this
response does not vindicate the Indispensability Thesis. If correct, it shows
at most that a capacity we need in order to think of or understand metaphor
is also needed, in this situation, in order to x what we are thinking about. It
does not show what her argument purports to show: that metaphor itself, in
this situation, is needed in order to x what we are thinking about.
4. Why metaphors seem indispensable

A pattern is emerging here. In several of the examples these philosophers
have chosen, a certain concept is needed in order to do what these
philosophers claim metaphor is needed for. In Boyd and Camps example,
20 21
Ibid., p. 18. Ibid.
4. WHY METAPHORS SEEM INDISPENSABLE 135
the scientist needs the concept of opening a computer le in order to

conduct her research in the way described. In the second example from
Gaut, we need the concept of sadness in order to grasp the sadness of music.
In Yablos rst example, we need the concept of anger in order to represent
clouds as looking like something expressive of anger. In each case, meta-
phor is not needed for the purpose in question, but a certain concept is.
These concepts all gure in the metaphor that is said to be indispensable
for the purpose in question. The concept of opening a computer le
gures in the metaphor of memory storage and retrieval as the opening
of a computer le; the concept of anger gures in the metaphor of the
clouds as angry; and the concept of sadness gures in the metaphor (if it is a
metaphor) of the music as sad. We might say that, when we use these
metaphors, we apply these concepts, but we apply them metaphorically.
This suggests an explanation of why metaphors have seemed indispens-
able to at least some supporters of the Indispensability Thesis. I suggest
they are mistaking the indispensability of a concept for the indispensability
of a metaphor in which that concept gures. In Boyd and Camps
example, for instance, the following things are true:
(a) We need to use the concept of opening a computer le in order to

conduct research in the way they describe.
(b) We can conduct research in this way by applying this concept meta-
phorically.
As my discussion in the last section shows, however, the following is not true:
(c) We need to apply this concept metaphorically in order to conduct

research in the way they describe.
The same is true, mutatis mutandis, of the other examples. In these cases,
too, we need to use a certain concept to achieve a certain end; and it is
possible to achieve this end by applying this concept metaphorically. This
misleads some into thinking that we need to apply the concept metaphor-
ically in order to achieve the end in question. This is why I say that they
have mistaken the indispensability of a concept for the indispensability of a
metaphor in which that concept gures.
This diagnosis is supported by a second consideration. Advocates of the
Indispensability Thesis not only say that metaphor is needed for a certain
purpose when, in fact, it is the concept that gures in the metaphor that is
needed. The cases in which some of them allow that metaphor is
dispensable for some purpose are cases in which this concept is dispensable
for that purpose.
Scruton, for example, writes: we must distinguish among metaphors
between luxuries and necessities. . . . I can spell out homo homini lupus [man
is a wolf to man], for instance, by describing the known facts of mans
aggression towards his fellows . . . For all intents and purposes, it is . . . dis-
pensable.22 Frank Sibley discusses a metaphorical description of a wine: it
will never win a race but its a wonderful little jogger. He comments: We
know exactly the prosaic meaning of this last one: not top class but a
satisfying day-to-day tipple you wont get tired of. Here the metaphor
performs no irreplaceable function; it is dispensable.23 These metaphors
are indeed dispensable. And there is indeed a difference between these
metaphors and others. But the difference is not that these metaphors are
dispensable and others are not.
When we use these metaphors, if Scruton and Sibley are right about
them, we are not primarily interested in the fact that the item described
metaphorically shares certain properties with what the concept that gures
in the metaphor literally applies to. Rather, we are primarily interested in
the fact that what is metaphorically described has these properties. A typical
user of homo homini lupus, for example, is not primarily interested in the fact
that people share the property of aggressiveness with wolves. Rather, she is
mainly interested in the fact that people are aggressive towards each other.
Similarly, the wine critic is not primarily interested in communicating that
the wines failure to be top-class, and its being consistently good neverthe-
less, are properties it shares with wonderful little joggers who will never
win a race. Her point is that it has these features.
Sometimes, however, we are primarily interested in the items sharing
of features when we use a metaphor. Berninis colonnade around St Peters
Square has been described as the arms of the Church, embracing her
ock. It has a shape of a kind that embracing arms also have. The point of
describing it as the arms of the Church, however, is not merely to draw
attention to that shape. It is to point out that the colonnade shares that
shape with pairs of embracing arms. Appreciating the colonnade involves
22
Roger Scruton, The Aesthetics of Music (Oxford: Oxford University Press, 1997),
pp. 52, 91.
23
Frank Sibley, Tastes, Smells, and Aesthetics, in Approach to Aesthetics: Collected Papers on
Philosophical Aesthetics, ed. John Benson, Betty Redfern, and Jeremy Roxbee Cox (Oxford:
5. CLARIFICATIONS 137
noticing this, for the building expresses welcome by sharing a shape with
arms in a gesture of welcome.
Now, what is true of the metaphors Scruton and Sibley discuss is that, to do
what we are mainly interested in doing with them, we do not need to use the
concepts that gure in the metaphors. To point out that people are aggressive
towards each other, we neednt use the concept of a wolf. This leads some
philosophers to say that, in such cases, we neednt use the metaphor to commu-
nicate what we want. As it happens, this is true; but it is the same confusion of
metaphors with the concepts that gure in them that leads them to say it.
By contrast, we need to use the concept of embracing arms to point out
that the colonnade shares a certain shape with pairs of embracing arms. But
we do not need to apply this concept metaphorically to point this out.
So there is a genuine distinction to be made among metaphors with
respect to the dispensability of the concept that gures in the metaphor.
But if the concept that gures in the metaphor is indispensable, it does not
follow that the metaphor is too.
5. Clarications
It is important to be clear about what is at issue when metaphors are said to
be indispensable.
First, there are trivially true versions of the Indispensability Thesis, and
these are plainly not at issue. For some things we use metaphor to think, to
express, to communicate, or to discover, it is trivially true that we need
metaphor in order to think, express, communicate, or discover them. For
example, we obviously cannot discover, without at least thinking of
metaphor, that a given metaphor draws our attention to a certain feature
or fact. The modality of the version of the Indispensability Thesis being
advocated also affects how interesting that version is. Friends of indispens-
ability sometimes distinguish their position from the view that metaphor is
sometimes needed to communicate something we merely happen to lack
non-metaphorical means of communicating.24 If we lack such non-
metaphorical means, but could easily develop them (and do so without
metaphor), that would not establish the truth of an especially interesting
24
See Max Black, Metaphor, Proceedings of the Aristotelian Society 55 (19541955): 280
282, and his More about Metaphor, Dialectica 31/3 (1977): 439. Compare John R. Searle,
Metaphor, in Expression and Meaning: Studies in the Theory of Speech Acts (Cambridge:
Cambridge University Press, 1979), p. 114.
version of the thesis. Beyond saying this, however, advocates of the thesis
do not tend to make clear the modality of the version they advocate.
Second, the Indispensability Thesis is not a claim about the manner in
which we think, express, communicate, or discover things when we use
metaphor to do so. Where f-ing ranges over thinking, expressing, com-
municating, and discovering, it is not the view that we could not f in such-
and-such a way without metaphor. It is the view that what we f with some
metaphors cannot be fed without metaphor.
One might hold, to take one example, that a single metaphor can
communicate many propositions, and that these propositions are empha-
sized to different degrees when we use the metaphor to communicate
them. One might also hold that there is no other way to communicate
those propositions with just that distribution of emphasis. An oft-quoted
complaint Max Black makes about the attempt to state the content of
certain metaphors in plain language is that when we attempt to do so, the
metaphors implications, previously left for a suitable reader to educe for
himself, with a nice feeling for their relative priorities and degrees of
importance, are now presented explicitly as though having equal
weight.25 The Indispensability Thesis is not the view that metaphor is
indispensable for communicating in this, or any other, manner.
Third, I have so far said nothing about the paraphrasability of metaphors.
Several philosophers hold that at least some metaphors cannot be para-
phrased. One might think that this claim implies the truth of the Indis-
pensability Thesis. This would be a mistake. In fact, the claim that some
metaphors cannot be paraphrased does not imply the truth of the Indis-
pensability Thesis, although the claim that we use metaphor to communi-
cate or to express what cannot be communicated or expressed without
metaphor (which is a version of the Indispensability Thesis) implies that
metaphors cannot be non-metaphorically paraphrased. There are several
philosophers whose views commit them to the claim that metaphors
cannot be paraphrased, but not to the Indispensability Thesis.
Davidson, for example, appears to hold that a paraphrase would give the
non-literal meaning or special cognitive content of the metaphor. But
according to him, metaphors have no non-literal meaning or special
cognitive content. Therefore, they cannot be paraphrased.
He writes:
25
Black, Metaphor, p. 293.
I agree with the view that metaphors cannot be paraphrased, but I think this is
not because metaphors say something too novel for literal expression but
because there is nothing there to paraphrase. . . . Metaphor can, like a picture
or a bump on the head, make us appreciate some factbut not by standing
for, or expressing, the fact.
If this is right, what we attempt in paraphrasing a metaphor cannot be to
give its meaning, for that lies on the surface; rather we attempt to evoke what
the metaphor brings to our attention.26
Davidson is committed to the view that metaphors cannot be paraphrased

(although he acknowledges that there is a point to the activity we call
paraphrasing). But he is not committed to the Indispensability Thesis. He
does not claim that there is a content expressed or communicated by
metaphor that no non-metaphorical paraphrase can express or communi-
cate. He claims that there is no special cognitive content expressed or
communicated by metaphor. From this, it follows that metaphors cannot
be paraphrased, in Davidsons sense. But it also follows that metaphors are
not indispensable for the expression or communication of such a content.
Of course, if Davidson had said that what metaphor brings to our attention
cannot be discovered without metaphor, then he would be committed to
the Indispensability Thesis; but he makes no such claim.
To take another example: Samuel Guttenplan regards it as wrong to
take paraphrasing X to be more or less equivalent to saying what
X tells us or means .27 You can say what a photograph tells us, but you
cannot paraphrase one. Rather, following the OED, Guttenplan holds
that a paraphrase is a re-statement of the sense of a passage in other
words.28 Since it is a re-statement of a thought in other words than those
in which the original expressed the thought, one can only paraphrase
something that expressed a thought in words in the rst place. And
since a photograph, whatever it tells us, is not itself in words, it is
inappropriate to paraphrase it.29
On Guttenplans view, metaphors also do not express thoughts in
words. Rather, it is words in the metaphor that call on [an] object, and
it is what Guttenplan calls the proto-predicate, object included, which
26
Donald Davidson, What Metaphors Mean, in Inquiries into Truth and Interpretation
(Oxford: Oxford University Press, 1984), pp. 246, 262.
27
Samuel Guttenplan, Objects of Metaphor (Oxford: Oxford University Press, 2005), p. 18.
28
Ibid.
29
Ibid.
conveys a message, not the words themselves.30 The burden of Gutten-

plans theory of metaphor is to explain these ideas and to show how
speakers can use objects in this way. It is unnecessary to enter into the
details of it here. The point is that he draws from his theory the moral that,
as in the case of photographs, it would be bizarre to ask someone to
express this same information in other words. . . . Since the speaker is using
an object, not words, to convey a message, it makes no sense even to try to
paraphrase a metaphor in the strict sense of the term.31
It would follow from the truth of Guttenplans theory that metaphors
cannot be paraphrased in the sense he identies. But here too, it would not
follow that metaphor is indispensable. One would need to argue that the
thoughts expressed by metaphor could not be expressed without meta-
phor. Establishing that the thoughts expressed by metaphor are not ex-
pressed by the metaphor in words, and therefore that nothing can count as
expressing these thoughts in other words, does not show this.
It sometimes goes unrecognized that the impossibility of paraphrase
does not entail the indispensability of metaphor for communication or
expression.
Camp, for instance, holds that a paraphrase of a metaphor should state only
the content of the speakers intended illocutionary act, and therefore should
not . . . include contents the speaker merely insinuated, or merely caused her
hearer to entertain.32 Moreover, in a paraphrase the content is stated in a
literal and explicit fashion: that is, the paraphrase should enable an otherwise
linguistically competent speaker to understand the original utterances
content simply in virtue of understanding the meanings of the paraphrasing
sentences constituent terms and their mode of combination.33
She defends the idea that certain metaphors cannot be paraphrased (at
least in certain circumstances) by arguing that various plausible candidates
fail to meet the criteria she sets for paraphrase. But if these arguments
succeed in showing that such metaphors cannot be paraphrased in her
sense, they still do not establish the Indispensability Thesis. They show, at
most, that in many circumstances we communicate with metaphor some-
thing we cannot communicate by making a statement satisfying the criteria
she sets for being a paraphrase. Camp, however, takes her argument to
establish the stronger conclusion that we sometimes cannot communicate
certain contents without metaphor.
30 31
32
Camp, Metaphor and That Certain Je Ne Sais Quoi, p. 2. 33
Ibid.
She considers this example:
(7) When he nally walked out the door, I was left standing on the
top of an icy mountain crag, with nothing around me but thin cold
air, bare white cliffs, and a blindingly clear blue sky.
Here, she says, the speaker is claiming to have experienced a specic
property, one for which the language has no existing expression, and one
which the hearer has not (let us suppose) experienced himself.34 Under
these circumstances, Camp grants, the speaker herself is still not forced to
speak metaphorically.35 The speaker could have said:
(72) I felt an emotion which was like the way it would feel physically
to stand on top of an icy mountain crag . . .
Statements like these, however, still rely at least implicitly on the original
metaphor, and so they fail to provide explicit formulations of the speakers
meaning:
if like expresses a substantive relation which holds just in case a particular,
contextually salient similarity holds between the two objects . . . then (72)
implicitly builds those similarities into its content. It may then succeed
in capturing the speakers intended content, but it arguably also fails to be
fully explicit, in much the way that Hes ready fails to specify its implicit
argument.36
Thus, (72) violates the requirement that a paraphrase must be an explicit

statement of the metaphors content.
This argument does not show that (72) relies at least implicitly on the
original metaphor. It shows, at most, that (72) implicitly builds [certain]
similarities into its content. Implicitly building similarities into its content
is different from implicitly relying on a metaphoreven a metaphor
which has those similarities as part of its content. It is true that, if (72) has
this content built into it only implicitly, then (72) is not an explicit
statement of the metaphors content. It would therefore fail to satisfy
one of Camps requirements for a paraphrase. But that does not show
that (72) relies on the metaphor, or that it fails to communicate the content
of the metaphor. Indeed, if it has the metaphors content implicitly built
into it, then it succeeds.
34 35 36
Ibid., p. 11. Ibid. Ibid., p. 12.
She also considers a paraphrase in which we are

citing the relevant similarities explicitly, as in
(73) I experienced an emotion which is like the physical feeling of standing
on an icy mountain crag . . . in respects i, j, k. . . .37
The problem with this is that,

construed as a paraphrase, (73) attributes unintended content to the speaker.
In uttering (7), the speaker isnt making any claims about what icy mountain
crags are like or about their relation to her emotional stateshes just
characterizing her emotion, using shared attitudes about icy mountain crags
to do so. Her intended claim has the form: When he left, I felt that way.38
(73), in short, violates the requirement that a paraphrase must state only the
content of the speakers intended illocutionary act.
Again, this may show that (73) fails to be a paraphrase. It does not show that
by using (73) we would fail to communicate what the metaphor does. It shows
at most that, if (73) does communicate what the metaphor does, then it also
communicates more besidesnamely, the claims about icy mountain crags.
Moreover, if the problem with (73) is that it does not just communicate
that the emotion has features i, j, and k, but also communicates that it shares
these features with the physical feeling of standing on an icy mountain crag,
then it is unclear why this cannot be easily corrected. If we managed to cite
these properties explicitly in (73) by talking of respects i, j, k . . . , it is
unclear why we cannot attribute these features to the emotion without also
claiming that they are shared by the feeling of standing on a crag.
Finally, if the metaphor-user manages to communicate that the emo-
tion has these features by using shared attitudes about icy mountain crags
to do so, it is also unclear why a speaker could not use these shared
attitudes to communicate the same thing without metaphor. It is unclear,
for instance, why she could not communicate the same thing with a simile.
Suppose the speaker had said: When he left, my emotions were like the
feelings of someone left standing on the top of an icy mountain crag, with
nothing around her but thin cold air, bare white cliffs, and a blindingly
clear blue sky. It is far from clear that she would not have communicated
to a hearer what she would have communicated with the metaphor. The
fact that this simile would not be an explicit statement of what the shared
37 38
Ibid. Ibid.
features are does not prevent it from communicating what they are to her
audience. The metaphor, after all, is not an explicit statement of what
those features are, either.39
Not all of these points would constitute problems if Camp wanted to
establish only that there are unparaphrasable metaphors. But she also wants
to establish that there are indispensable metaphors. Camp holds that
metaphor is needed for successful communication in situations like that
in which she imagines (7) being used: ones in which the language lacks an
expression for the property the metaphor-user has in mind, and in which
the hearer has not experienced the property (which, she says, prevents the
speaker using any demonstrative that would enable the hearer to identify
the property). She takes the impossibility of paraphrase in these situations
to show that, even if she avails herself of all possible literal means of
coining a word for the property, the speaker
could not introduce that word into the language, because her hearer would
be in no position to comprehend itnot as a result of linguistic incompe-
tence, or irrationality, but just from a lack of worldly experience. . . . It is of
course true that after the speaker has gotten her hearer to identify the relevant
property by metaphorical means, she can then introduce a new term which
denotes it. . . . But because the metaphor here plays an essential role in
dening the new term, this possibility cannot be used to show that metaphor
in general is theoretically eliminable. Although each particular metaphor can
eventually be eliminated, the situation exemplied by (7) can always arise
anew for a different property.40
But as I have argued, nothing Camp says about paraphrase shows that all
possible literal means will fail to enable the hearer to identify the relevant
property, even in the kind of situation exemplied by (7). She shows at
most that her candidate paraphrases fail to be paraphrases, not that they fail
to enable the hearer to identify the relevant property without relying
39
Camp rejects other candidate paraphrases on the grounds that, contrary to her second
requirement for paraphrase, an otherwise linguistically competent, rational hearer could no
longer understand the paraphrasing sentence simply in virtue of his basic linguistic compe-
tence and rationality. . . . He would need to engage in just the sort of interpretation called for
by the original metaphor (ibid., p. 14). The same kind of objection as I raised above applies
here. The fact that basic linguistic competence and rationality would not be enough to enable
a hearer to understand a candidate paraphrase may show that it is not a paraphrase. It does not
show that the hearer could not understand the candidate paraphrase at all. And the fact that a
hearer of the candidate paraphrase would need to engage in the same sort of interpretation
demanded by the metaphor does not show that the speaker would need to rely on the
metaphor to be understood.
40
Ibid., pp. 15, 16.
implicitly on the metaphor. And if she does not establish that metaphor is
essential for communication in this situation, she also does not establish that
metaphor plays an essential role in dening a new term for the property.
What implies the truth of the Indispensability Thesis, then, is not the thesis
that metaphors cannot be paraphrased. It is the claim (i) that metaphors do
communicate or express something, and the claim (ii) that there is no non-
metaphorical way of communicating or expressing what the metaphor does.
Of course, if paraphrase just is the expression by non-metaphorical means
of what the metaphor expresses, or the communication by non-metaphorical
means of what the metaphor communicates, then the impossibility of
paraphrase plus claim (i) together imply the truth of the Indispensability
Thesis. But this is what paraphrase must be in order for the denial of the
possibility of paraphrase and claim (i) to imply the truth of the Indispens-
ability Thesis. And it is, as we have seen, not universally acknowledged that
this is what a paraphrase is.
The nal clarication I wish to make concerns the value of metaphor.
Nothing I have said casts any doubt on the idea that some metaphors
express what they do more beautifully, more powerfully, or more suc-
cinctly than any other form of words could. That view, being one about
the manner in which metaphors express what they do, is distinct from the
Indispensability Thesis. Many writers seem to hold that they must defend
the Indispensability Thesis in order to defend the claim that metaphor is of
great value. If metaphor is one among several possible ways of communi-
cating or expressing what it does, then (the assumption seems to be) it is of
minor importance. Scruton insists that metaphors are indispensable not
merely because they are part of some unique literary experience,41 and
Black stresses that metaphor provides more than the incidental pleasures
of stating guratively what might just as well have been said literally.42
Such impatience with the idea that great metaphors are valuable
merely because of their power, beauty, vividness, and so forth is curious.
Pointing out that a piece of writing is an imaginative, beautiful, vigorous,
clear, and concise way of communicating something, as many metaphors
are, is normally sufcient to show that it is a very valuable way of
communicating. If someone were to show that there is a dull, laboured,
and rambling way of communicating the same thing, we would not
41
Scruton, The Aesthetics of Music, p. 91.
42
Black, More About Metaphor, p. 441. See also Black, Metaphor, pp. 282, 293.
6. PROSPECTS FOR THE INDISPENSABILITY THESIS 145
conclude that the original way of putting things is of little value. We

would certainly not conclude that what we communicated might just as
well have been said in the dull way. Perhaps it is because philosophers are
so often concerned with questions of truth and knowledge that they are
inclined to defend metaphor by arguing that it is essential for the expres-
sion or discovery of certain truths.
6. Prospects for the Indispensability Thesis

I have not, of course, demonstrated that the Indispensability Thesis is false.
Rather, I have shown that various arguments for it do not succeed, and
I have distinguished it from a variety of related claims. Some of these other
claims are plausible, and the Indispensability Thesis acquires an air of
plausibility when it is not clearly distinguished from them. As I said at
the beginning, however, many philosophers do not even attempt to argue
for the Indispensability Thesis. They simply assert that metaphors are
indispensable. This being so, it may be that some will respond to my
discussion so far in the following way: perhaps these arguments do fail, but
is it not obvious that some metaphors are indispensable? Arguments for the
Indispensability Thesis, it may be felt, are unnecessary. Consider complex
poetic metaphors such as:
Cast thy bread upon the waters: for thou shalt nd it after many days.43
selfwrung, selfstrung, sheathe- and shelterless,

thoughts against thoughts in groans grind.44
There is shadow under this red rock,

(Come in under the shadow of this red rock),
And I will show you something different from either
Your shadow at morning striding behind you
Or your shadow at evening rising to meet you;
I will show you fear in a handful of dust.45
43
Ecclesiastes 11:1 (King James Version).
44
Gerard Manley Hopkins, Spelt from Sibyls Leaves, in Gerard Manley Hopkins: The
Major Works, ed. Catherine Phillips (Oxford: Oxford University Press, 2002), p. 175, l. 14.
45
T. S. Eliot, The Waste Land, in The Complete Poems and Plays (London: Faber and Faber,
1969), p. 61, ll. 2530.
Is it not obvious that we could not express what these express in any
other way?
I have not shown that these philosophers are wrong. What I hope to
have shown is that asking rhetorical questions like this is not good enough.
Advocates of the Indispensability Thesis do need to provide arguments for
it. If the arguments that are given for it so often turn out to be based on
confusion, we have reason to be suspicious of our sense (if we have it) that
very apt, striking, or complex metaphors are indispensable for the purposes
claimed by advocates of the Indispensability Thesis. I have so far provided
six grounds for suspicion.
First, I have shown that the indispensability of a concept of something in
terms of which metaphor is explained (e.g., the concept of a game of
make-believe) is sometimes confused with the indispensability of meta-
phor itself. Second, the indispensability of the concept that gures in a
metaphor is often confused with the indispensability of metaphor itself.
Third, the fact that we happen to lack expressions with which to attribute
a certain property non-metaphorically, and the fact that we can attribute it
with metaphor, would not by themselves establish an interesting version of
the Indispensability Thesis. Fourth, the fact that we use some metaphors to
f in a manner in which we could not f without metaphor would not
show that they are indispensable for f-ing what we f with them. Fifth,
the unparaphrasability of metaphor is sometimes confused with the indis-
pensability of metaphor. Sixth, the fact that some metaphors are tremen-
dously valuable ways of communicating and expressing things is consistent
with the claim that they are dispensable for these purposes.
In short, supporters of the thesis need to argue for it because they need
to show that they are avoiding these common confusions. It is not simply
obvious that certain metaphors are indispensable, because it is not simply
obvious that our sense of the indispensability of certain metaphors is not
due to one of these confusions. Argument is required to show this.
There are also more general grounds for suspicion. For ease of expos-
ition, I will focus on the case in which metaphor is claimed to be indis-
pensable for expressing something; but the points I will make can be made,
mutatis mutandis, about communication.
I have said that metaphors are indispensable for expressing what they do
only if they do express something. In addition, to know that a given
metaphor is indispensable for expressing what it does, we need to know
what it expresses. If we do not know what a given metaphor expresses, we
6. PROSPECTS FOR THE INDISPENSABILITY THESIS 147
cannot claim that metaphor is needed in order to express what the given
metaphor does.
This consideration shows that not just any complex or poetic metaphor
can be used as evidence for the Indispensability Thesis. It must be one that
we understand: we must know what it expresses. But many difcult
metaphors are difcult precisely because it is unclear what they express.
Philosophers often write as though we nd difcult metaphors puzzling
only because we nd it hard to put them into other words. In fact, we often
puzzle over difcult poetic metaphors because we are unsure what they are
expressing. Take Eliots metaphor: I will show you fear in a handful of
dust. Perhaps some people know what this expresses. But I expect that for
many of us it is not clear. And if we do not know what it expresses, then we
do not know that it expresses something inexpressible without metaphor.
Since many metaphors are like this, this is a seventh reason to be suspicious
of the impression that some metaphors, surely, are indispensable.
Suppose, then, that we have examples of metaphors that express some-
thing, and we do know what they express, and they seem indispensable.
We would still need an argument for their indispensability, for the
following reason.
If we know what a metaphor expresses, such that we understand the
metaphor, and the metaphor characterizes, then we know what property
(or properties) it characterizes its subject as having or lacking. We can
identify the property. Knowing what is expressed by metaphors that do
not characterize still involves being able to identify certain properties. For
example, understanding the metaphorical question, Is Juliet the sun?,
involves knowing what properties Juliet would have if the answer to the
question is Yes. Again, if we understand the metaphor, we can identify
these properties. An ability to identify relevant properties is also involved
in understanding other metaphors that do not characterize (e.g., some
metaphors in the antecedent or consequent of a conditional).
Given these facts, a tempting but over-hasty argument against the
Indispensability Thesis would be to say this: if you can identify a property,
then you can coin a non-metaphorical expression for ita name for it or a
predicate or an adjective with which we can characterize something as
having it. To think otherwise is comparable to thinking that there are
particulars we can identify but cannot name. Consequently, any metaphor
that expresses something and is understood is dispensable. In the case
of any such metaphor, we can identify the relevant properties, and
therefore we can coin a non-metaphorical expression for the property.

And if we can coin a non-metaphorical expression for the property, then
we can use it to express whatever we expressed with the metaphor. We
can use the term to characterize the metaphors subject as having or lacking
the property, to ask the same question as was asked with the metaphor, and
so forth. So the Indispensability Thesis is falseor, at least, it is not true of
any metaphor we understand.
This is certainly a challenge that needs to be met by a supporter of the
Indispensability Thesis. But I call this argument over-hasty because it
overlooks the possibility that we sometimes need metaphor in order to
identify certain properties in the rst place, or to enable others to identify
them. To my knowledge, no advocate of the Indispensability Thesis does
explicitly and directly claim that there are properties such that, even when
we have identied them, we cannot then name them or coin a predicate
or adjective with which to characterize something as having them.46
Camp shows greater awareness of this problem than many. She is careful
to claim only that we sometimes need metaphor in order to identify a
property or to enable others to do so. As we have seen, however, her
arguments do not establish this.
So this is an eighth reason why advocates of the thesis must provide
arguments for it. It is very implausible that we cannot coin a non-meta-
phorical expression for a property once we have identied it. If any
metaphors are nevertheless indispensable for communication or expres-
sion, they must be needed (either by the metaphor-user or her reader) in
order to identify certain properties. It is not simply obvious that they are
needed for this, and so it is not simply obvious that metaphors are indis-
pensable.
None of this demonstrates that the Indispensability Thesis is false. Its
prospects, however, do not look good.
46
Zangwill, however, comes close to saying this, and appears to be committed to it. He
writes: My view is that there are some aspects of the world which cannot be described without
metaphor, for I think that the world has properties that are literally indescribable. And I think
that thoughts about those properties cannot be linguistically expressed without metaphor.
(Zangwill, The Metaphysics of Beauty, p. 174).
6
Metaphor and Criticism
My aim in this chapter is to answer the question: why is metaphor so often

used in criticism? As I said in Chapter 3, the prevalence of metaphor in
criticism has been explained in many ways. It has usually been thought
to reveal something important about aesthetic experience and artistic
expression. Kant wrote that we call buildings or trees majestic and stately,
or plains laughing and joyful; even colours are called innocent, modest,
soft, because they excite sensations containing something analogous to the
consciousness of the state of mind produced by moral judgements.1 Kant
takes the kinds of descriptions he mentions to support his view that we
experience the beautiful as a symbol of the morally good. E. H. Gombrich
is struck by the fact that as long as criticism has existed, critics have used
metaphors to express their approval or disapproval. They have branded
colour combinations as vulgar or exalted forms as dignied, have
praised the honesty of one artists palette and rejected the meretricious
effects of others.2 Gombrich explains this as a reection of how, in art, a
visual quality may be experienced as the equivalent of a moral value.3
Stuart Hampshire regards critics use of metaphor as evincing the kind of
attention characteristic of aesthetic contemplation, and Nelson Goodman
argues that reection on metaphorical descriptions of artworks reveals
how artistic expression differs from forms of artistic representation such
as depiction.4
1
Immanuel Kant, Critique of Judgement, ed. Nicholas Walker, trans. James Creed Meredith
and Nicholas Walker (Oxford: Oxford University Press, 2007), Ak. 5: 354.
2
E. H. Gombrich, Visual Metaphors of Value in Art, in Meditations on a Hobby Horse and
Other Essays on the Theory of Art (London: Phaidon, 1963), pp. 1415.
3
Ibid., p. 14. See also E. H. Gombrich, On Physiognomic Perception, in Meditations on a
Hobby Horse and Other Essays on the Theory of Art, pp. 4748.
4
See Stuart Hampshire, Logic and Appreciation, in William Elton, ed., Aesthetics and
Language (Oxford: Blackwell, 1959), pp. 166167; Nelson Goodman, Languages of Art: An
Approach to a Theory of Symbols, 2nd edn (Indianapolis: Hackett, 1976), ch. 2.
150 METAPHOR AND CRITICISM
My approach to this question is to consider what critics achieve by using

metaphor. Whatever the use of metaphor in criticism reects about
aesthetic experience or other matters, it certainly reects something
about criticism. It enables critics to achieve what they are trying to achieve.
What, then, are they trying to achieve? And what makes metaphor such an
effective way of achieving it?
Other philosophers have appealed to the Indispensability Thesis to
explain the prevalence of metaphor in criticism. On such views, critics
are trying to express or communicate facts that cannot be expressed or
communicated without metaphor (for instance, certain facts about how
artworks look and sound, and about our responses to them). I will not
appeal to this claim. Rather, I will appeal to my account of the aims of
criticism (defended in Chapter 2) and to the Minimal Thesis (defended in
Chapter 4). I will begin by defending the Minimal Thesis against those
who would object that it does not apply to art-critical metaphors. I will
then draw a distinction between two kinds of art-critical metaphor. This
distinction has not, to my knowledge, been previously recognized; but
drawing it is essential to understanding the function of metaphor in
criticism. I will then provide my own explanation of metaphors preva-
lence in criticism.
1. What the critic communicates

The rst part of the Minimal Thesis is this claim:
Each property a metaphors subject is characterized with the metaphor-
ical element as having is either:
(1) A likeness indicated by the metaphorical element, or
(2) A determinate of such a likeness, or
(3) A likeness-maker for such a likeness, or
(4) A way of possessing a likeness-maker for such a likeness.
The second part of the Minimal Thesis is the claim that, to understand
a metaphor, we employ our knowledge of what properties are of
kinds (1)(4).
If the Minimal Thesis is correct, the same is true of metaphorical
descriptions in criticism. A critic who describes music metaphorically as
chattering attributes to the music the likeness being like something chattering,
1. WHAT THE CRITIC COMMUNICATES 151
and/or properties related to this likeness in one of the above ways. She
might, for instance, communicate that the music sounds like something
chattering. To understand the metaphor, the critics readers must draw on
their knowledge of what likeness is indicated by chattering and what
properties are related to this likeness as determinates of it, likeness-makers
for it, or ways of possessing likeness-makers for it.
Views inconsistent with the Minimal Thesis are common in aesthetics.
Some believe that, in criticism, metaphors are not used to attribute any
property to the objecteven if they are so used elsewhere. I will call this
view anti-realism about metaphor.
Some aestheticians go further. They claim that the fact that aesthetic
descriptions are often metaphorical lends support to an anti-realist under-
standing of aesthetic descriptions in general. John Bender says that one thing
that makes it difcult to argue that aesthetic properties are real is that many
of them are metaphorical,5 by which he presumably means that (apparent)
ascriptions of these properties are often metaphorical. Frank Sibley writes:
I include [among aesthetic descriptions], moreover, those remarks, metaphor-
ical in character, which we might describe as apt rather than true, for these
often say, only more strikingly, what could be said in less colourful language.
The transition from true to apt description is a gradual one.6
In the same paper, Sibley also says that he poses the question of objectivity
in aesthetics as a question about the truth and aptness of remarks, rather
than as a question about the possession of properties by objects. One thing
that leads him to do this is the existence of metaphorical aesthetic descrip-
tions. He explains: while we might replace the question Is she graceful?
by talk of properties, we might feel less happy, with metaphorical remarks,
saying that a work has the property of gemlike re or marmoreal hardness
(though we might say it has properties that make these descriptions apt).7
In a later paper, Sibley says straight out that gurative descriptions are apt
rather than true.8
5
John W. Bender, Aesthetic Realism 2, in Jerrold Levinson, ed., The Oxford Handbook of
Aesthetics (Oxford: Oxford University Press, 2003), p. 80.
6
Frank Sibley, Objectivity and Aesthetics, in Approach to Aesthetics: Collected Papers on
Philosophical Aesthetics, ed. John Benson, Betty Redfern, and Jeremy Roxbee Cox (Oxford:
7
Ibid., p. 72.
8
Frank Sibley, Making Music Our Own, in Approach to Aesthetics, p. 152.
These remarks suggest that Sibley endorses anti-realism about meta-

phor. It is a familiar claim that what we say with metaphor is, normally, not
true. But if Sibleys view was only that a metaphor-user does not normally
say, but may communicate, something true, he would be unlikely to
describe this as the view that metaphors are apt rather than true. Similarly,
if Sibley believed only that metaphors are not used to attribute the
property the object is said to have, though they are used to attribute
properties, he would be unlikely to cite the existence of metaphorical
aesthetic remarks as a reason for eschewing all talk of properties in favour
of talk of the aptness of remarks. So I take it that, although Sibley wrote
little on metaphor, he was an anti-realist about metaphor.
However, the reasons given in Sibleys (admittedly cursory) discussion
are bad reasons to embrace anti-realism. He says that we might describe
[metaphorical aesthetic descriptions] as apt rather than true, for these often
say, only more strikingly, what could be said in less colourful language.
But the fact that we could say something in language less colourful than
the language we actually use is, of course, no reason at all to believe that
our actual remark is apt rather than true. A remarks colourfulness has
nothing to do with its truth.
Similarly, although it is perhaps true that we might feel less happy, with
metaphorical remarks, saying that a work has the property of gemlike re or
marmoreal hardness, that is no reason to doubt that some property is being
attributed. It is just that we cannot be attributing the property we would
have been attributing had we been speaking literallywhich should come
as no surprise, since we are not speaking literally. We would be equally
unhappy saying that a brave person has the property of being a lion. That,
however, is no reason to doubt that we attribute bravery to the person
when we describe her metaphorically as a lion.
So the considerations Sibley advances do not provide good reasons to
endorse anti-realism about metaphor. The most developed and inuential
argument for anti-realism about art-critical metaphors is provided by
Roger Scruton. According to Scruton, a metaphor attributes no property
at all to the work.9 Rather, we use art-critical metaphors to describe
something other than the material world, namely, how the world seems,
from the point of view of the active imagination.10
9
Roger Scruton, The Aesthetics of Music (Oxford: Oxford University Press, 1997), p. 154.
10
Ibid., p. 91.
Like Davidson, Scruton holds that an expression used metaphorically

means exactly what it would mean if used literally. When we say meta-
phorically, The music is sad, sad means exactly what it would mean if
we were speaking literally. Clearly, however, we are not attributing literal
sadness to the music.
It follows, Scruton thinks, that we are not attributing any other property
to it, either. To say that the word ascribes, in this use, another property, is
to say that it has another sensein other words that it is not used
metaphorically but ambiguously.11 So we claim that we are attributing
another property to music on pain of denying that expressions used
metaphorically have the same sense as they would have if used literally.
But that is unacceptable. It follows that the word sad attributes to the
music neither the property that is possessed by sad people, nor any other
property. It therefore attributes no property at all.12
There are several problems with Scrutons position. First, it is false that
to say that the word ascribes, in this use, another property, is to say that
it has another sense. The same predicate can be used in the same sense to
attribute different properties on different occasions. You do not have to
use a word in a different sense to attribute a different property.
A parallel will illustrate this. By parity of reasoning, we could use
Scrutons assumptions to show that a speaker who describes someone
sarcastically as friendly is not attributing unfriendliness to heror, indeed,
any property at all. For clearly, friendly used sarcastically means what it
means when used literally. Equally clearly, the speaker does not believe that
the person described is friendly. So she is not attributing friendliness to
her. Nor is she attributing any other property. To hold that she is
would commit one to the claim that friendly is ambiguous. Indeed, one
would have to say that one meaning of friendly is unfriendly, which is
absurd. Therefore, you attribute no property at all when you describe
someone sarcastically as friendly.
Obviously, this argument is faulty. The sarcastic person is attributing
unfriendliness to the person she describes. But she is not using any word in
a different sense than she would be if she were speaking non-sarcastically.
She is attributing unfriendliness by other means.
11 12
Scruton might make the following reply: Granted, it is possible to use a

word in its usual sense to attribute a property that is not attributed when
we use it in this sense and speak literally. But in order to show that this is
actually the case with sad as applied to music, we must identify some
plausible candidate for a property that is being attributed here. And this, he
might claim, we cannot do.
However, even if this is true of the notoriously puzzling example of sad
music, it is not true of a vast range of art-critical metaphors. In these cases,
there are very plausible candidates for properties that are being attributed.
I pointed out in Chapter 3 that connoisseurs of Chinese jades distinguish
between the colours of spinach, lychee-esh, and mutton-fat jade.
These metaphors characterize these different kinds of jade as being
coloured like these substances. There are numerous examples of art-critical
metaphors for which there are plausible candidates for the properties being
attributed. So if Scrutons case rests on the example of sad music, he does
not have adequate support for his conclusion.
Finally, suppose one did show that neither a likeness to sad people, nor
any properties related to this likeness in ways (2)(4), are attributed to
music with expressive terms such as sad. This would pose a problem for
the Minimal Thesis only if sad music is a metaphor in which being like a
sad person is the likeness indicated by the metaphorical element. If another
likeness is indicated by the metaphorical element, then it does not matter if
neither being like a sad person, nor any related properties, are attributed to
the music. If Scruton is right that words do not acquire new senses when
used metaphorically, then it seems that this likeness is indicated only if sad
is used metaphorically with the sense it has when literally applied to sad
people, and not used metaphorically with, say, the sense it has when
literally applied to sad gestures or sad feelings.
But, in the rst place, it is far from obvious that sad is used metaphoric-
ally here with the sense it has when applied literally to sad people. If this
were so, then applying it to music would be an example of personication.
We personify the weeping willow when we describe it as sad. But we do
not appear to be personifying music when we speak of sad music.13
13
This is so even if Jerrold Levinson is right that we hear sad music as an expression of
sadness (as a sad gesture of some sort) and imagine a sad persona in the music (see Jerrold
Levinson, Musical Expressiveness as Hearability-as-Expression, in Contemplating Art: Essays
in Aesthetics (Oxford: Oxford University Press, 2006), pp. 91108). If these claims are right,
Second, it is not obvious that sad is here used metaphorically at all.14 It

is certainly not as obviously metaphorical as The weeping willow is sad.
Moreover, to assume that expressive terms are metaphorical is to assume
that several theories of artistic expression are false. Philosophers often
introduce the claim that expressive terms are metaphorical as though
this assumption begs no relevant questions. But this is not so. If a
certain version of the arousal theory of expression (to take one example)
is correct, then to call music sad is to say that it is such as to cause
sadness. In that case, sad is being used only as a causal analogy. It is
like healthy as applied to food: as I said in Chapter 4, we attribute,
not health, but the property of being such as to cause, contribute to,
or maintain health when we call food healthy. And if we are using
sad only by causal analogy, then we are not using it metaphorically.
Healthy food is no metaphor (not even a dead one). Expressive terms
are clearly metaphors only if a range of theories of artistic expression are
clearly false.
So the possibility that expressive terms do not attribute likenesses
or related properties should not concern a supporter of the Minimal
Thesis. There are independent grounds for hesitating to regard them
as metaphors anyway. We saw in Chapter 4 that there are many non-
metaphorical ways of extending the uses of expressions. If applying sad
to music is an extension of some pre-existent use of sad, that by itself
is poor evidence that it is a metaphorical extension. Numerous clear
cases of metaphor in criticism are accurately described by the Minimal
Thesis.
I conclude that these arguments do not show that the Minimal Thesis
fails to apply to art-critical metaphors. Anti-realism about art-critical
metaphors, widespread as it is in aesthetics, is an obstacle to understanding
why critics frequently use metaphor. However, before explaining why
they do, I should make two important clarications.
then the most plausible view about the sense of sad as applied to the music (rather than sad
as applied to the persona imagined in the music) is that it has the sense it has when applied to
an expression of sadness, such as a sad gesture.
14
Several philosophers, at least, have doubted it. See R. A. Sharpe, Philosophy of Music: An
Introduction (Chesham: Acumen, 2004), pp. 102108; Paul Boghossian, Explaining Musical
Experience, in Kathleen Stock, ed., Philosophers on Music: Experience, Meaning, and Work
(Oxford: Oxford University Press, 2007), p. 123.
First, the Minimal Thesis does not necessarily establish that critics
metaphors attribute aesthetic properties. That depends on what aesthetic
properties are. Realism about aesthetic metaphors may not be sufcient to
establish realism about aesthetic properties.
This is important because Scruton, for one, wants to establish anti-
realism about metaphor partly because he wants to establish anti-realism
about aesthetic properties. The assumption seems to be that, if these
metaphors are used to attribute any properties, they are used to attribute
aesthetic properties. That assumption is not obviously correct. It depends,
again, on what aesthetic properties are. If the likenesses, likeness-makers,
or other properties attributed with aesthetic metaphors are not themselves
aesthetic properties, then an aesthetic anti-realist can happily accept the
Minimal Thesis. Anti-realism about aesthetic metaphor may not be neces-
sary to establish anti-realism about aesthetic properties.
Second, the aesthetic realist could also accept the Minimal Thesis, even
if the likenesses or related properties attributed by art-critical metaphors
are not themselves aesthetic properties. It is consistent with my position to
say that speakers attribute properties in addition to the properties identied
by the Minimal Thesis when they use some particular metaphor. Critics
may often imply, for example, that the work is aesthetically interesting in
virtue of having the likenesses or likeness-makers attributed. However, if
they do attribute such properties, they do not do so in virtue of using a
metaphor, but in virtue of something else (e.g., contextual factors).
Acceptance of the Minimal Thesis, then, does not by itself commit one
either to aesthetic realism or to aesthetic anti-realism.
2. What interests the critic

So much, then, for what the critic communicates. A further important
point is this. To say that critics attribute likeness-makers for a certain
likeness is not to imply that the critic is always interested in the fact that
those properties give the object that likeness. Sometimes, the critic is
indeed interested in those properties because they give the object the
likeness. However, sometimes she is interested in these properties, but
not for this reason. Some examples will make this clear.
The critic is often interested in the fact that certain properties give the
work a certain likeness. As I said in Chapter 5, Berninis colonnade around
2. WHAT INTERESTS THE CRITIC 157
St Peters Square has been compared to a pair of arms embracing the

pilgrims. A certain shapecall it Smakes the colonnade like a pair of
embracing arms. A critic who describes the colonnade as a pair of arms
embracing the pilgrims would probably not be interested only in the fact
that the colonnade has S. She is also interested in the fact that S makes the
colonnade resemble a pair of embracing arms. The colonnade expresses
welcome by resembling arms that do.
Sometimes, by contrast, the likeness-makers are of interest, but not
because they give the work the likeness. Victor Hugo says of Hamlet: In
this tragedy . . . everything oats, hesitates, delays, staggers, becomes dis-
composed, scatters, and is dispersed. Thought is a cloud, will is a vapour,
resolution a crepuscule; the action blows each moment in an inverse
direction, man is governed by the winds.15 Take the metaphor, will is a
vapour. Hugo is interested in the fact that Hamlets will has certain
properties. For instance, his will is continually changing. These properties
make his will like a vapour. However, Hugo is not interested in these
properties because they make his will like a vapour. The likeness to a vapour
does not have the sort of importance here that the likeness of the colon-
nade to embracing arms has.
Another example is art historians description of the drapery in late fth-
century Greek vase-painting as being drawn in the spaghetti style.16 As
Figure 6.1 shows, this description is used because there are many lines of
drapery-folds drawn close together, making the drapery resemble spa-
ghetti. The critic is not interested in the fact that the drapery resembles
spaghetti. It is hard to imagine what relevance that could have to the
appreciation of ancient Greek vase-paintings. Rather, her interest is in a
certain way of looking, which makes the drapery look like spaghetti.
I should mention a signicant sub-class of this second kind of metaphor.
Frequently, when the critic is interested in the likeness-maker, but not
because it gives the object the likeness, the likeness-maker is the property
of being such as to provide an experience, or elicit a response, of a certain
kind. Keith Miller describes Alan Hollinghursts novel The Line of Beauty
in this way: Faintly perfumed and of fractal complexity, Hollinghursts
15
Victor-Marie Hugo, William Shakespeare (extract), trans. A. Baillot, in Jonathan Bate,
ed., The Romantics on Shakespeare (London: Penguin, 1992), p. 350.
16
Whether or not the phrase the spaghetti style is itself a metaphor, one could certainly
communicate the same thing by describing the drapery as spaghetti.
Figure 6.1. Attic red-gure pyxis decorated with women and erotes or cupids.
Close to the Meidias Painter. End of 5th century bc. # Ashmolean Museum,
University of Oxford. Detail. Adapted with permission.
prose endows Nick with a rounded, ironical inner life.17 The prose is like
something faintly perfumed in a key respect: it is such as to provide an
experience of a certain kind, which something faintly perfumed also
provides. The property of being such as to provide that kind of experience
is the likeness-maker attributed to the prose. That it is such as to provide
this kind of experience is of greater interest here than the fact that it is
like something faintly perfumed in virtue of doing so. Metaphors that tell
us the kind of response a work elicits loom large in certain kinds of
criticism.
There is, then, a distinction to be made between metaphors used by
critics. Sometimes, the critic is interested in the fact that the subject of her
metaphor has the likeness-makers, or ways of possessing them, that she
attributes to it. But she is not interested in the fact that the subject has the
likeness indicated, or a determinate of it, in virtue of having these likeness-
makers. Sometimes, by contrast, she is interested in this fact. This distinc-
tion will enable us to explain why metaphor is so prevalent in criticism.
17
Keith Miller, People Who Cant Love People, The Times Literary Supplement, 9 June
2006, p. 22.
3. WHY CRITICS USE METAPHOR 159
3. Why critics use metaphor

In Chapter 2, I argued that one criticizes an artwork only if one aims to
communicate:
(a) what parts, features, or represented elements appreciation of it
can involve responding to; or
(b) what responses appreciation of it can involve; or
(c) what appropriate reasons for these responses there are.
Communicating facts of this kind is, I argued, a constitutive aim of criticism.
I also argued that aiding appreciationthat is, enabling ones readers to
appreciate the work better than they would be likely to if they were aware
of the works features by appropriate means without having read the
criticismis a non-constitutive aim of criticism. Being such as to achieve
this helps make a piece of criticism good as criticism.
I will argue that using metaphor is a particularly effective way of
achieving both of these aims. Establishing this will explain why critics
often use metaphor.
First consider those cases in which critics are interested in the fact that
the likeness-makers give the metaphors subject the likeness. Why are they
interested in this fact? I suggest that it is because appreciation of the work
can involve acquiring or conrming, by appropriate means, the know-
ledge that these properties give the subject of the metaphor the likeness.
Appreciating Berninis colonnade can involve seeing that the shape makes
it like a pair of embracing arms. That is plainly why the critic is communi-
cating that it is the case.
This suggests a partial explanation of metaphors prevalence in criticism.
As I observed in Chapter 2, critics often try to convey to their readers what
cognitive responses appreciation of a work can involve. Appreciation often
can involve acquiring or conrming by appropriate means (e.g., by per-
ceptual means) the knowledge that certain properties give something a
certain likeness (hereafter, for the sake of brevity, I will mostly speak only
of perceiving this). Metaphors enable critics to indicate a likeness and
attribute likeness-makers for that likeness. They thereby enable critics to
convey that appreciation can involve perceiving that the likeness-makers
give the subject the likeness. That is one reason why metaphor is so
common in criticism.
If anything needs defending in this explanation, it is presumably the

claim that appreciation often involves acquiring or conrming, by appro-
priate means, the knowledge that certain properties give something a
certain likeness. There is, however, a great deal of evidence that this is so.
The appreciation of literature often can involve this. Allusions are a very
large class of examples. Appreciating Eliots lines that begin a description of
a woman in an unhappy marriage,
The Chair she sat in, like a burnished throne,
Glowed on the marble,18
involves recognizing that their wording makes them like the beginning of
Enobarbuss description of Cleopatra when she rst meets Antony:
The barge she sat in, like a burnished throne
Burned on the water.19
Nor is the evidence from literature limited to allusions. Henry Jamess short
novel The Aspern Papers is about the narrators efforts to get the unpublished
letters of a famous poet from an old woman and her niece. At one point, the
narrator meets the niece in her garden and tries to get her to relinquish the
papers. It has been pointed out that this scene is like the temptation of Eve
by the serpent in the garden of Eden. Appreciating the story can involve
recognizing how this incident is like the incident in the Bible.
There are similar cases in the visual arts. In a discussion of the arrange-
ment of the Apostles in Raphaels cartoon for a tapestry, The Miraculous
Draught of Fishes (Figure 6.2), Kenneth Clark comments that the Apostle
Zebedee . . . seated in the stern [on the extreme right], is intended to recall
an antique river god.20 In some Greek temples, a sculpture of a reclining
gure, often identied as a river god, was placed at either end of the line of
gures in the temples pediments.21 Appreciating Raphaels painting can
involve seeing this resemblance.
18
T. S. Eliot, The Waste Land, in The Complete Poems and Plays (London: Faber and Faber,
1969), p. 64, ll. 7778.
19
William Shakespeare, The Tragedy of Anthony and Cleopatra, ed. Michael Neill (Oxford:
Oxford University Press, 1994), 2.2.198199.
20
Kenneth Clark, Raphael: The Miraculous Draught of Fishes, in Looking at Pictures
(London: John Murray, 1960), p. 64.
21
For example, Figure A from the west pediment of the Parthenon. For an image,
see <http://commons.wikimedia.org/wiki/File:West_pediment_A_Parthenon_BM.jpg>
accessed 14 June 2012.
Many consumer products exploit likenesses. Toothbrushes, for example,

often have an aerodynamic shape to make them look like pieces of advanced
technology. Jonathan Woodham, in a work on twentieth-century design,
discusses the design of the tailns of the Cadillac Fleetwood in the 1950s.
He remarks that the detailing . . . relates to the contemporary fascination
with the Jet Age, the rear lights simulating rocket-burn.22
Appreciation does not only involve perceiving likenesses that the artist
intends the audience to notice. Frederick Hartt describes Christs head in a
crucix by Coppo di Marcovaldo in this way: The closed eyes are treated
as two erce, dark, hooked slashes, the pale mouth quivers against the
sweat-soaked locks of the beard, the hair writhes like snakes against the
tormented body.23 Hartt is describing how the details of the picture
combine to produce a total effect of the greatest expressive power.24
Whether or not Coppo intended it, appreciating the expressive power of
this work can involve seeing that the parts of the painting depicting the
eyes are like slash-marks, and that the locks of Christs hair are shaped like
writhing snakes. Similarly, in an account of the development of Panathe-
naic prize amphorae, vases given as prizes to the victors in athletic contests,
John Beazley remarks: The Burgon vase is stout and squat; let us compare
it with some later Panathenaics. . . . In London B 134, by the Euphiletos
Painter, about 530, the neck is shorter, the body longer, and the whole
vase gives a deeper impression of collected power. In London B 133, by
the Eucharides Painter, about 480, the shape is even stronger and more
compact.25 London B 133 gives a deeper impression of collected power,
and the shapes of later vases are described as stronger, because the later
vases are more like a taut, strong human body than the stout and squat
Burgon vase is.26 Seeing that the shape makes it like something strong can
22
Jonathan M. Woodham, Twentieth-Century Design (Oxford: Oxford University
Press, 1997), p. 114. For an image, see <http://commons.wikimedia.org/wiki/File:Cadillac_
Fleetwood_1959_4.jpg> accessed 14 June 2012.
23
Frederick Hartt, History of Italian Renaissance Art, 3rd edn (London: Thames
and Hudson, 1987), p. 43. For an image, see <http://commons.wikimedia.org/wiki/File:
Coppo_di_Marcovaldo_001.jpg> accessed 14 June 2012.
24
Ibid.
25
J. D. Beazley, The Development of Attic Black-Figure (Berkeley: University of California
Press, 1986), p. 82.
26
For the Burgon vase, see <http://commons.wikimedia.org/wiki/File:Panathenaic_
amphora_BM_B130.jpg> accessed 14 June 2012. For London B 133, see <http://www.
britishmuseum.org/research/search_the_collection_database/search_object_details.aspx?
objectId=399634&partId=1> accessed 14 June 2012.
be involved in appreciating it. These likenesses to strong bodies are

especially relevant, given the function of these vases as prizes for athletes.
I conclude that we can be condent in the partial explanation I have
provided of the prevalence of metaphor in criticism. Appreciation often
can involve perceiving that certain properties give something a certain
likeness. One reason why critics frequently use metaphor is to convey that
appreciation of a certain work can involve such a perception.
I turn now to cases in which critics are interested in the fact that the
subject of her metaphor has the likeness-makers, or ways of possessing
them, that she attributes to it, but not in the fact that it has the likeness
indicated, or a determinate of it, in virtue of having these likeness-makers.
I said above that metaphors in this class include both cases in which the
property attributed is the property of providing an experience of a certain
kind, and cases in which it is not. Let us rst take cases in which it is not.
In these cases, it seems clear, appreciation can involve perceiving that
the subject of the metaphor has the property attributed, but cannot involve
perceiving that it gives the subject the likeness indicated. Appreciating
Greek vase-paintings in the spaghetti style can partly consist in perceiving
that the drapery-folds are painted in a certain pattern. But it would be
ridiculous to suppose that appreciation of them can partly consist in
perceiving that this pattern makes the drapery like spaghetti. So an explan-
ation of why critics use such metaphors suggests itself. Critics commonly
use metaphors of this kind to convey that appreciation can involve
perceiving that the subject of the metaphor has the likeness-makers, or
ways of possessing likeness-makers, attributed. But they do not use them
to convey that appreciation can involve perceiving that the likeness-
makers give the subject the likeness indicated or a determinate of it.
Consider now cases in which the likeness-maker attributed is the
property of being such as to provide an experience of a certain kind, as
when Miller calls Hollinghursts prose faintly perfumed. Here the ex-
planation must be different. It is implausible to say that appreciating the
prose can involve perceiving that it provides this kind of experience.
Rather, appreciating the prose involves having this kind of experience.
Miller is using this metaphor to communicate that this is a response that
appreciating the novel can involve. Metaphors that attribute the likeness-
making property of being such as to provide an experience of a certain
kind are prevalent because critics use them to convey that appreciation can
involve having such experiences.
Both of these cases, however, raise a question not answered by the

explanations I have given so far. We have established what critics are doing
in each case and why they are doing it: communicating that the subject has
certain likeness-makers or ways of possessing them (to convey that appre-
ciation can involve perceiving that the subject of the metaphor has those
properties), and communicating that the work provides a certain kind
of experience (to convey that appreciation can involve having that kind of
experience). Why, however, is the critic communicating these things by
using metaphor? As I have stressed, appreciation, in these cases, does not
involve becoming aware of the likeness indicated by the metaphor. For
example, it does not involve perceiving that the drapery is like spaghetti or
that the prose is like something faintly perfumed. So why bring in
spaghetti or perfume at all? Why indicate a likeness that is irrelevant to
appreciation in order to communicate something that is relevant to appre-
ciation? To answer this question, I must rst point out two things about
these metaphors.
Consider the following passage by John Ruskin, who describes arriving
in Venice by boat and seeing the long ranges of columned palaces,each
with its black boat moored at the portal,each with its image cast down,
beneath its feet, upon that green pavement which every breeze broke into
new fantasies of rich tessellation, and observing how the front of the
Ducal palace, ushed with its sanguine veins, looks to the snowy dome of
Our Lady of Salvation.27
The rst notable feature of the metaphors in this passage is that they are
very specic descriptions. Take the metaphor of breezes breaking the
waters surface into fantasies of rich tessellation. There are many ways
water looks when breezes blow across its surface. There are fewer ways it
looks when breezes blow across its surface and make it look like something
broken into many pieces. And there are still fewer ways it looks when
breezes blow across its surface and make it look like something broken into
pieces forming a rich mosaic with the colours of Venetian palaces and a
greenish tint.
Ruskins metaphor communicates that the waters of Venice have
properties making them look like that. And to characterize them this
way is to characterize them very specicallyespecially in comparison
27
John Ruskin, The Stones of Venice (London: George Allen, 1900), vol. 2, p. 3.
with many other, more obvious alternative descriptions. The more specic
description is the more informative; and Ruskins metaphor is an unusually
informative description of the way the waters of Venice look.
Metaphors used to communicate that the object elicits a certain
response can also be very specic. The response itself can be characterized
very specically.
An example is Clarks description of The Miraculous Draught of Fishes
(Figure 6.2). After a night spent without catching anything, the Apostles,
on Christs command, are hauling up their nets, which are suddenly full of
sh. Clark writes:
A rhythmic cadence runs through the whole composition, rising and falling,
held back and released, like a perfectly constructed Handelian melody. If we
follow it from right to left . . . we see how the river god, like a stoker, drives
us into the group of heroic shermen and how the rich, involved movement
of this group winds up a coil of energy; then comes an artful link with the
standing Apostle, whose left hand is backed by the shermans billowing
drapery, and then St Andrew himself forming a caesura, a climax in the line,
which holds us back without lessening our momentum. Then, at last, the
marvellous acceleration, the praying St Peter to whose passionate movement
all these devices have been a preparation, and nally the comforting gure of
Christ, whose hand both checks and accepts St Peters emotion.28
Consider the metaphors of the rhythmic cadence and of the standing St

Andrew forming a caesura. A caesura is a pause near the middle of a line of
poetry. Elaborating the rhythmic-cadence metaphor, Clark communicates
with the caesura metaphor that an appropriate response to this part of the
painting is like a response to a caesura in a line of poetry. He does not,
however, communicate that it is like this response merely in that it is one
of pausing. He characterizes it much more specically than that. We pause
here after having followed the line of Apostles from the right, our gaze
moving along naturally as we attend to salient parts like the heads,
shoulders, arms, and hands, in turn. Our attention to these salient parts
as we move along is like our attention to the stressed syllables in a line of
poetry, which are spaced at regular intervals in an intelligible pattern,
propelling our attention along as we read. And we pause at St Andrew
without nding the pause jarring, despite the fact that it interrupts the
prior movement of our attention.
28
Clark, Raphael: The Miraculous Draught of Fishes, pp. 6465.
Figure 6.2. Raphael. The Miraculous Draught of Fishes. 15151516. Bodycolour on

paper laid into canvas. Victoria and Albert Museum, London. The Royal Collec-
tion # 2012, Her Majesty Queen Elizabeth II courtesy Victoria and Albert
Museum. With permission.
The second thing to note about metaphors of this kind, in addition to

their specicity, is that they tend to cause a reader to have, or to imagine or
recall having, certain experiences. What makes Ruskins metaphor vivid
and evocative is that it tends to cause a reader to imagine seeing the waters
of Venice. Clarks description accompanies a reproduction of Raphaels
painting. It causes us to look at the picture and try to have, or imagine
having, the kind of response he is communicating that the painting elicits.
Other metaphors I have mentioned have similar effects. When you rst
hear of the spaghetti style, and a (reproduction of a) painting in that style is
visible, you are likely to look at it to see what is being attributed.
Many have held that there is a causal connection between metaphor and
perception. As we have seen, Davidson and Scruton, among others, think
there is such a connection between metaphor and perceiving-as. Many
have also held that metaphor causes us to imagine having certain experi-
ences. Aristotle says that liveliness is got by using the proportional type of
metaphor and by making our hearers see things.29 George Orwell writes
that a newly invented metaphor assists thought by evoking a visual
image,30 while Richard Moran cites numerous philosophers and writers
who have held such a view.31
However, few have seen why there is this connection. Moran discusses
the temptation among those he cites to suppose, not only that metaphors
cause us to imagine, but that having certain mental images is what
constitutes the full understanding of a metaphor.32 That temptation is
certainly to be resisted. But what is true is that, often, one cannot gure
out what properties the metaphor attributes unless one perceives, recalls
perceiving, or imagines perceiving the subject of the metaphor. Many
metaphors are more or less impenetrable until you take a look at
(or imagine or recall seeing) the subject, and see what properties make it
like what the metaphor communicates that it is like. In many contexts,
you need to see (or imagine or recall seeing) that the subject has certain
likeness-making properties in order to tell that the metaphor attributes
them.33 This is not what understanding the metaphor consists in: rather, it
is often what enables us to understand metaphors.
It seems clear that perceiving and imagining perceiving play this role in
our coming to understand the metaphors considered above. We look at
the vase-painting and try to see what the speaker means by describing it as
being drawn in the spaghetti style. To gure out what kind of visual
appearance Ruskin is claiming the waters present, we try to imagine seeing
water that looks like what Ruskin communicates that Venices waters look
like. We also use imagination, while perceiving the reproduction of the
painting, to gure out what kind of response Clark claims the painting
elicits.
This is not, I stress, a claim about every property attributed by every
metaphor. For example, it would obviously not apply to properties one
cannot perceive that something has. Moreover, it would not normally be
29
Aristotle, Rhetoric, trans. W. Rhys Roberts, in The Complete Works of Aristotle, ed.
Jonathan Barnes (Princeton: Princeton University Press, 1984), vol. 2, 1411b.
30
George Orwell, Politics and the English Language, in Essays, ed. Bernard Crick
(London: Penguin, 2000), p. 350.
31
See Richard Moran, Seeing and Believing: Metaphor, Image, and Force, Critical
Inquiry 16/1 (1989), pp. 8994.
32
Ibid., p. 92.
33
The same is often true of guring out what ways of possessing likeness-makers the
metaphor attributes.
true of the likeness indicated by a metaphor that (imagining or recalling)

perceiving is the easiest way of guring out that it is indicated or attributed.
Rather, we often need perception, perceptual memory, or perceptual
imagination to gure out what other properties are attributed.
Why is this so? The answer is not hard to nd. For many expressions
used metaphorically, but especially those of which this claim is true, we
cannot rely wholly or even partly on our familiarity with past metaphorical
uses of them to gure out what properties they attributeas we often can
rely on familiarity with past literal uses of an expression to gure this
out when it is used literally. Certainly, with some metaphorical uses of
expressions, we can do this. Expressions like pig, lion, and block of ice
normally attribute certain likeness-makers rather than others when used
metaphorically, at least when the subject of the metaphor is human. In
these cases, we can rely on familiarity with past metaphorical uses of the
expression.
Many metaphorical uses of expressions, however, are not like this. This
may be because we never have encountered a metaphorical use of that
expression before. Vivid or interesting metaphors are often novel. Alter-
natively, it may be because we have never encountered that expression
used metaphorically to attribute the properties it attributes on this occa-
sion. The same expression can be used metaphorically in different contexts
to attribute different properties. We attribute different properties when we
call John Major grey than we do when we speak of a grey area in
morality or law.
So a metaphor may have for its metaphorical element an expression we
have never seen used metaphorically before, or an expression used meta-
phorically to attribute properties we have never known it to attribute
before. If so, then we cannot rely entirely, or at all, on familiarity with past
metaphorical uses of the expression to gure out what likeness-makers it
attributes. We need some other way of guring this out. In such cases,
perceiving the metaphors subject, or imagining or recalling perceiving it,
is sometimes the only way, or the easiest way, of guring out what
likeness-makers are attributed.
This explains why it is novel metaphors that have been singled out for
their connection with perceiving and imagining perceiving. In the quota-
tion given above, for example, Orwell characterizes newly invented
metaphors as evoking visual images; similarly, Scruton writes that dead
metaphors achieve nothing, but living metaphors change the way things
are perceived.34 Examples of metaphors that prompt us to perceive, or to

imagine perceiving, support belief in such a connection. They tend to be
novel, like Ruskins and Clarks, rather than clichd.
To recapitulate: I said above that there are two reasons why critics often
use metaphor in cases where appreciation does not involve perceiving that
the (ways of possessing) likeness-makers attributed give the subject the
likeness indicated. First, in such cases, appreciation often does involve
perceiving that the subject has the likeness-makers or ways of possessing
them. Second, in such cases, appreciation often involves having the
response which the metaphor communicates that the subject elicits. This
raised the question of why critics use metaphor at all, given that they
indicate or attribute likenesses that are irrelevant to appreciation. We are
now in a position to answer this question.
The rst reason why metaphor is used is to cause readers to have
responses appreciation can involve. Critics use metaphor to get their
readers to perceive, or to imagine or recall perceiving, that the subject
has the properties attributed. They also use it to cause readers to have,
imagine having, or recall having other responses, which the metaphor
conveys that the subject can elicit. Getting ones readership to perceive,
and to have other responses appreciation can involve, is a way of enabling
them to appreciate the work better. Aiding appreciation, in turn, is an aim
of much criticism.
I have not much discussed the claim that critics often want to get their
readers to imagine or to recall perceiving. It is a commonplace in aesthetics
that critics try to get their readers to perceive. But philosophers less often
note that critics try to get their readers to imagine or recall perceiving. But
clearly they often do attempt this. As I have argued, in many contexts
critics do not presuppose that their readers are in a position to perceive
what they describe. Many passages in Ruskin and Pater illustrate this, as do
many reviews (e.g., of a theatrical production that has nished its run).
And critics may also describe works they do assume their readers have
perceived, in order to get them to recall perceiving what the critic wants to
discuss.
One can aid appreciation by getting ones readers to recall having a
perception or other response involved in appreciation. As I argued in
34
Roger Scruton, Beauty (Oxford: Oxford University Press, 2009), p. 124.
Chapter 2, one neednt be experiencing a work when one comes to

appreciate it better. One can come to appreciate a lm better in discussing
it afterward with friends. Similarly, if the critic tells you of a reason to
admire a certain detail of the work, recalling seeing that detail might
enable you to nd it intelligible that the reason stated is indeed a reason
to admire that detail. You might then admire the detail for that reason, and
thereby appreciate the work better. Alternatively, you might appreciate
the work better simply by recognizing that the reason stated is a reason to
admire it. As I observed in Chapter 2, recognizing or conrming by
appropriate means that a certain response or reason is appropriate can itself
be a response appreciation can involve. Appreciation often can involve
such meta-responses.
Causing someone to imagine perceiving or having another kind of
response can also aid appreciation. Sometimes, this may be because
imagining having a certain response or perception is itself a response
appreciation can involve. To return to examples used earlier, it is plaus-
ible that Beethoven was able better to appreciate music he composed
while deaf by accurately imagining hearing it performed; and one might
be able to acquire some appreciation of Malevichs Black Square, or certain
works of conceptual art, by accurately imagining perceiving them. In
other cases, imagining having a certain response, though not itself a
response appreciation can involve, can enable you to have such responses.
For instance, imagining, like recalling, can enable you to have appropri-
ate meta-responses to a work you have perceived. Being able to imagine
having a certain response to a work can enable you to nd it intelligible
that it is an appropriate response to the work.
Even when the critics aim is not to aid appreciation, it makes sense that
a critic would often want to get readers to imagine having the responses
appreciation can involve. It is understandable that a critic reviewing a work
that her readers can no longer experience would try to get them to
imagine perceiving or otherwise responding to the work.
In short, by speaking in a way her reader cannot fully understand
without perceiving, imagining perceiving, or recalling perceiving, the
critic impels the reader to perceive, imagine, or recall what she wants
them to. Similar points can be made about responses other than percep-
tions. This, then, is a reason why metaphor is often used in criticism even
when the likeness indicated by it is irrelevant to appreciation.
This brings us to the second reason why critics often use metaphors of
this kind. I said that metaphors can characterize their subject, or the
response elicited by it, very specically, especially compared with more
obvious alternative descriptions. This by itself will often be a reason to
prefer the metaphor to the more obvious alternatives. Moreover, on
account of their capacity to be specic, metaphors can enable us very
accurately to imagine or recall perceiving their subject.
Obviously, a critic who wants to cause the reader to imagine or recall
experiencing the subject of the metaphor wants to cause her to imagine or
recall this experience as accurately as possible. It is clearly possible to
imagine experiences of objects more or less accurately.35 If the object is
a red square, for example, then you more accurately imagine seeing it if
you imagine seeing a red square than you do if you imagine seeing an
otherwise identical black square. This is so even if the square you have
imagined seeing does not possess the shade of red possessed by the actual
square. You have still imagined the experience of seeing the actual square
more accurately than when you imagine seeing a black square, even
though you have not imagined this experience with perfect accuracy.
The more specic a description is, the more informative it is. Therefore,
assuming the reader can imagine perceiving that the object has the prop-
erties attributed by a more specic description, the critic can be sure of the
reader getting more right when she uses the more specic description than
she can be when she uses a less specic description. A reader might, of
course, imagine the experience of seeing the red square with perfect
accuracy if she is only told that what she is to imagine seeing is a coloured
shape. But the critic obviously does not ensure this by using this descrip-
tion. Describing the object as a red square, by contrast, ensures at least
that the reader imagines something square and red. Therefore, given that
the critic wants to cause the reader to imagine, as accurately as possible, the
experience of perceiving the object, she does well to get the reader to
imagine perceiving that the subject has the properties attributed with a
more specic description.
So it is not just that metaphors often prompt a reader to imagine or
recall perceiving their subjects. Metaphors often prompt a reader to
35
For the sake of brevity, I will hereafter discuss only the case of imagining perceiving.
What I will say also applies mutatis mutandis to recalling perceiving, and to imagining or
recalling having other responses to the subject of the metaphor.
4. CONCLUSION 171
imagine or recall this experience very accurately. A reader who imagines

perceiving that the waters of Venice have the properties attributed by
Ruskins metaphor imagines with great accuracy the experience of per-
ceiving the waters of Venice. Ruskin ensures that she imagines this with
greater accuracy than he would if he prompted her to imagine perceiving
that the waters have the properties attributed by more obvious, less
imaginative alternative descriptions (e.g., saying that the waters sparkle in
the sunlight).
This account not only provides reasons why metaphors are common in
criticism when the likeness indicated by them is irrelevant to appreciation.
Plausibly, when the likeness is relevant, metaphors are often used because
they have these effects. For in these cases, too, critics often try to have these
effects on their readers. This is an additional reason why metaphors of the
rst kind are prevalent in criticism.
4. Conclusion
We now have an explanation of metaphors prevalence in criticism.
(1) Critics often attribute properties, in describing artworks, to
convey that appreciation can involve perceiving that the object
has those properties. Appreciation, in turn, often involves per-
ceiving that certain properties give the object a certain likeness.
By using metaphor, critics can give us to understand that certain
properties give something a certain likeness, and thereby convey
to us that appreciation can involve perceiving that this is so. That
is one reason why critics frequently use metaphor.
(2) Critics often want to cause readers to perceive that the object has
certain properties, when appreciation involves perceiving that
it has those properties; or to imagine or recall this experience
accurately. So too, they often want to cause readers to have, or to
accurately imagine or recall having, other kinds of response that
appreciation involves having. Using metaphors, especially novel
ones, is an effective way of achieving both of these goals. Such
metaphors are often hard to understand without perceiving,
imagining perceiving, or recalling perceiving the object. Using
one therefore prompts a reader to perceive, imagine, or recall
what the critic wants her to. The same is true, mutatis mutandis, of
responses that are not perceptions. Metaphors can also be very

specic, and this can ensure that the reader recalls or imagines the
relevant experience very accurately.
I argued in Chapter 3 that the role of imaginativeness in criticism is to
enable the critic to think of unobvious ways of communicating effectively
and unobvious ways of better appreciating a work. This distinction is
reected in the distinction I have drawn between the two kinds of art-
critical metaphor.
Imaginative art-critical metaphors in which the likenesses are irrelevant
to appreciation are imaginative ways of communicating effectively.
Describing the vase-painters style of painting drapery as the spaghetti
style was an imaginative way of communicating how the drapery in their
paintings look. It was an unobvious way of communicating to think of,
and it is effective.
In the case of imaginative art-critical metaphors in which the likeness is
relevant to appreciation, it is imaginative of the critic to think of the way of
better appreciating the work she thinks of. That the colonnades shape
makes it like a pair of embracing arms is not an obvious reason to take an
interest in it to think of. Hence taking an interest in it for this reason is an
imaginative way of better appreciating it to think of.
Conclusion
Criticism of the arts is a major part of our cultural life. Critics help
determine which lms and plays get seen and which books get read, and
criticism commonly affects our experience and evaluation of paintings,
poems, music, the urban environment, fashion, and much else. But as
I said in the introduction, many people nd it rather mystifying how critics
do what they do. Critics cannot explain what they do by identifying
denite procedures, rules, or algorithms which they consult and which
ensure they get things right if they follow them. As in other areas in which
this is the casesuch as, notably, the creation of artit can be mysterious
how those who are good at it succeed. One aim of this book has been to
make criticism less mystifying. If my arguments have succeeded, they have
given us a better understanding of what critics do and how they do it.
The rst step was to consider what appreciation involves. Appreciating an
artwork involves having appropriate perceptual, cognitive, cogitative, affect-
ive, or conative responses to the right aspects of a work for the right reasons.
Much more, of course, could be said about appreciation; but clarifying even
this much about its nature allows us to identify an aim shared by all criticism.
A critic aims to communicate what appreciation can involve responding to,
what responses appreciation can involve, or what appropriate reasons there
are for such responses. This, I argued, is a constitutive aim of criticism.
Knowing this much enables us to understand many other features of
criticism. It explains why a critic is required to be acquainted with a work
in the same way (e.g., perceptually) as one must be to appreciate it. It helps
us see how criticism differs from similar forms of discourse about art, such
as art history, which have different aims and are subject to different
requirements. Finally, identifying this constitutive aim prompts us to
look at the question of the aims of criticism itself in a new light. It prompts
us to distinguish criticisms constitutive aims from its non-constitutive
aims. Aiding appreciation, I argued, is a non-constitutive aim of criticism.
174 CONCLUSION
Knowing both of these aims helps us to make sense of further aspects of

criticism. First, it helps us to see what unites a great variety of things critics
do, such as describe responses to works, and evaluate, interpret, and
explain works. Critics do much of what they do in order to achieve the
aims of criticism I identied. They describe responses and interpret in
order to communicate what appreciation involves; they guide perception
to aid appreciation. In other cases, critics achieve these aims of criticism in
order to achieve something further. Critics sometimes communicate facts
about appreciation to help their readers decide what works to experience;
often, too, they communicate such facts to aid appreciation. Seeing this
allows us to understand both what is wrong and what is right about the ve
rival views on the aims of criticism in the philosophical literature.
Second, knowing these aims makes it clear how critics produce good
criticism. If I am right about appreciation, there are three basic kinds of
change critics can cause to help us better appreciate a work. They can affect
what we respond to, how we respond, and why we respond. Accordingly,
the endowments that make someone good at criticism are those that make a
person good at effecting these changes by achieving the constitutive aim of
criticism. I singled out nine endowments that make critics good at this.
Metaphor is no less mystifying than criticism. A second primary concern
of this book has been to shed light on it. Why some metaphors are so
effective, how they work, and what their distinctive effects are, are ques-
tions of perennial interest. They have received especially close attention in
recent decades. My arguments give us new answers to these questions.
Three widely held views have made a correct understanding of meta-
phor difcult to attain: anti-realism about metaphor, the Indispensability
Thesis, and the denial that all metaphors are based on likenesses. I have
argued that these views are false or poorly supported. Instead, I have based
my account of metaphor on the Minimal Thesis about what metaphors
communicate. The Minimal Thesis can withstand an array of objections
and putative counterexamples, as we have seen. And just as my view of the
aims of criticism puts us in a position to explain many other aspects of
criticism, so the Minimal Thesis enables us to explain many other aspects
of metaphor.
The Minimal Thesis helps us understand how metaphors guide our
thinking. Much has been made of the impact of metaphor on thought.
The history of science and the history of philosophy both bear witness to
the hold a metaphor can exert on our minds. The metaphor of the mind as
CONCLUSION 175
a computer, of the body as a machine, of knowledge as a structure

supported on foundations, and many others have exercised a powerful
inuence on our thinking.
We fail to see how metaphors inuence thought if we ignore the Minimal
Thesis. Metaphors prompt us to seek likeness-makers for an indicated
likeness. As we saw in Chapter 5, the cognitive scientist guided by the
computer-le metaphor is looking for properties that would make memory
like a computer le. This may lead her to discover properties of memory she
would not otherwise have looked for. It may lead her to postulate further
likeness-makers to explain properties she discovers. And it may, of course,
cause her to overlook important differences as well (recognizing the limita-
tions of a useful metaphor can also constitute progress). We can thus explain
metaphors inuence on our thinking without making the vague claim that
metaphors make us think of one thing in terms of another, the unwar-
ranted claim that our concepts are themselves metaphorical, or the false
claim that metaphor is indispensable for thinking in these ways.
We also have a better understanding of what makes some metaphors
effective. Some metaphors cause us not only to seek likeness-makers for an
indicated likeness, but to perceive or imagine perceiving that the subject
has them. Some characterize their subjects very specically. Both features
make metaphor especially well-suited to the critics task.
My conclusions also have a bearing on other questions about the effects
of metaphor. For example, it has been common since Davidson to appeal
to perceiving-as to explain various features of metaphor. My account,
however, has explained many of the same features (such as metaphors
open-endedness and the difculty of paraphrase) without reference to
perceiving-as. When I have discussed perception, I have appealed only
to the connection between metaphor and perceiving that something is the
case. We should reconsider what, if anything, is explained by the claim
that metaphors make us perceive one thing as another.
Finally, the third principal aim of this book has been to provide us with
a better understanding of the nature of imaginativeness and its role in
criticism. Imaginativeness is a propensity to think of unobvious achieve-
ments. It contrasts with perceptiveness, which is an aptitude for acquiring
the knowledge that p when it is not obvious that p. The role of imagina-
tiveness in criticism is to enable critics to think of unobvious ways of better
appreciating a work and unobvious ways of communicating effectively.
Imaginative art-critical metaphors exemplify both aspects of this role.
176 CONCLUSION
Many are unobvious ways of effectively communicating what appreci-

ation involves. And in many cases, what they communicate is itself an
unobvious way of better appreciating the work to think of.
This says what imaginativenesss role in criticism is. It does not say how
large its role is. I will conclude with some remarks about how my account
of the role of imaginativeness in criticism should guide reection about
how much scope there is for imaginativeness in criticism.
Claims about the role of imaginativeness in criticism are often coupled
with the claim that the critics task offers broad scope for imaginativeness,
or even that it requires her to be imaginative. For example, it follows,
from Scrutons claim that aesthetic descriptions express an experience of
perceiving the work imaginatively, that critics need to be imaginative to
provide aesthetic descriptions. Again, much of the appeal of Roland
Barthess claim that some texts are scriptible is due to the suggestion that
there is broad scope for imaginativeness on the part of the reader. Barthes
writes: Why is the scriptible our value? Because the goal of literary work (of
literature as work) is to make the reader no longer a consumer, but a
producer of the text.1 A purely lisible text, by contrast, denies the reader
access to the pleasure of writing and would leave him no more than the
poor freedom either to accept or reject the text.2
My explanation of the role of imaginativeness in criticism can contrib-
ute to an assessment of these inuential views. On the one hand, it clearly
is not true that a critic always needs to be imaginative in order to achieve
the aims of criticismfor example, in order to aid the readers appreci-
ation. Depending on the readership being addressed, thinking only of
obvious ways of better appreciating the work and of communicating
effectively may be adequate.
On the other hand, if there are many unobvious ways of better appre-
ciating a work (for instance, because there are many unobvious but
appropriate responses to it), or many unobvious yet effective ways of
communicating what the ways of better appreciating it are, then there is
much scope for imaginativeness in criticizing the work. But whether
philosophical argument can show that one of these conditions is satised
by many, most, or all works of art is another question. Perhaps the best
prospect for someone who wants to argue that the role imaginativeness
1
Roland Barthes, S/Z, trans. Richard Howard (Oxford: Blackwell, 1990), p. 4.
2
Ibid.
CONCLUSION 177
plays in criticism is (or can be) a large one is to consider a certain claim
often made about great artworks, but less often examined. This is the claim
that the works of art we value most lend themselves to criticism by
generation after generation. It is not just that we continue to value
themthat, in other words, they stand the test of time. It is that critics
keep nding new things to say about them. For many works, it seems
absurd to suppose critics will ever nish thinking of the responses to them
that appreciation can involve having, or nish thinking of appropriate
reasons for such responses. Of course, the claims made for the inexhaust-
ibility of great artworks may be exaggerated. But even if they are, the truth
they exaggerate may provide reason to believe that there is broad scope for
imaginativeness in criticism of such works. The critical attention lavished
on canonical works certainly suggests that, at any given time, there are
indeed many unobvious ways of better appreciating them that remain to
be thought of.
Coleridge said of Shakespeare: You feel him to be a poet inasmuch as,
for a time, he has made you onean active creative being.3 If my account
here has been successful, it helps us to make sense of the thought that
engaging with art can make poets out of critics.
3
Samuel Taylor Coleridge, Lectures on Shakespeare and Milton in Illustration of the
Principles of Poetry, in Lectures 18081819: On Literature, ed. R. A. Foakes (London:
Routledge and Kegan Paul, 1987), p. 251.
Bibliography
Acocella, Joan, Mozart Moves, The New Yorker, 20 August 2007, <http://www.
newyorker.com/arts/critics/dancing/2007/08/20/070820crda_dancing_acocella>
accessed 14 June 2012.
Addison, Joseph et al., The Spectator, ed. Donald F. Bond, 5 vols (Oxford: Oxford
University Press, 1965).
Alison, Archibald, Essays on the Nature and Principles of Taste, 2nd edn (Boston:
Cummings and Hilliard, 1812).
Aquinas, Thomas, Summa Theologiae: Questions on God, ed. Brian Davies and Brian
Leftow, trans. Brian Davies (Cambridge: Cambridge University Press, 2006).
Aristotle, The Complete Works of Aristotle, ed. Jonathan Barnes, 2 vols (Princeton:
Princeton University Press, 1984).
Metaphysics, trans. W. D. Ross, in The Complete Works of Aristotle, ed. Jonathan
Barnes (Princeton: Princeton University Press, 1984), vol. 2, 15521728.
Rhetoric, trans. W. Rhys Roberts, in The Complete Works of Aristotle, ed.
Jonathan Barnes (Princeton: Princeton University Press, 1984), vol. 2, 2152
2269.
Topics, trans. W. A. Pickard-Cambridge, in The Complete Works of Aristotle, ed.
Jonathan Barnes (Princeton: Princeton University Press, 1984), vol. 1, 167277.
Arnold, Matthew, The Function of Criticism at the Present Time, in Lectures and
Essays in Criticism, ed. R. H. Super (Ann Arbor: The University of Michigan
Press, 1962), pp. 258285.
Austin, J. L., Sense and Sensibilia, ed. G. J. Warnock (Oxford: Oxford University
Press, 1962).
Barthes, Roland, S/Z, trans. Richard Howard (Oxford: Blackwell, 1990).
Baudelaire, Charles, The Salon of 1846, in Selected Writings on Art and Literature,
ed. and trans. P. E. Charvet (London: Penguin, 1992), pp. 47107.
Baxandall, Michael, The Language of Art History, New Literary History 10/3
(1979): 453465.
Beardsley, Monroe C., The Metaphorical Twist, Philosophy and Phenomenological
Research 22/3 (1962): 293307.
Metaphorical Senses, Nos 12/1 (1978): 316.
Aesthetics: Problems in the Philosophy of Criticism, 2nd edn (Indianapolis: Hack-
ett, 1981).
The Aesthetic Point of View: Selected Essays, ed. Michael J. Wreen and Donald
M. Callen (Ithaca: Cornell University Press, 1982).
180 BIBLIOGRAPHY
Beardsley, Monroe C., Critical Evaluation, in The Aesthetic Point of View: Selected
Essays, ed. Michael J. Wreen and Donald M. Callen (Ithaca: Cornell University
Press, 1982), pp. 316331.
What Are Critics For?, in The Aesthetic Point of View: Selected Essays, ed.
Michael J. Wreen and Donald M. Callen (Ithaca: Cornell University Press,
1982), pp. 147164.
What Is an Aesthetic Quality?, in The Aesthetic Point of View: Selected Essays,
ed. Michael J. Wreen and Donald M. Callen (Ithaca: Cornell University Press,
1982), pp. 93110.
Beattie, James, Of Taste, and its Improvement, in Selected Philosophical Writings,
ed. James A. Harris (Exeter: Imprint Academic, 2004), pp. 161182.
Beazley, J. D., The Development of Attic Black-Figure (Berkeley: University of
California Press, 1986).
Bender, John W., Aesthetic Realism 2, in Jerrold Levinson, ed., The Oxford
Handbook of Aesthetics (Oxford: Oxford University Press, 2003), pp. 8098.
Bennett, M. R. and P. M. S. Hacker, Philosophical Foundations of Neuroscience
(Oxford: Blackwell, 2003).
Black, Max, Metaphor, Proceedings of the Aristotelian Society 54 (19541955):
273294.
More About Metaphor, Dialectica 31/3 (1977): 431457.
Blackburn, Simon, Spreading the Word: Groundings in the Philosophy of Language
(Oxford: Oxford University Press, 1984).
Boden, Margaret A., Creativity in a Nutshell, in Creativity and Art: Three Roads to
Surprise (Oxford: Oxford University Press, 2011), pp. 2940.
Boghossian, Paul, On Hearing the Music in the Sound: Scruton on Musical
Expression, The Journal of Aesthetics and Art Criticism 60/1 (2002): 4955.
Explaining Musical Experience, in Kathleen Stock, ed., Philosophers on
Music: Experience, Meaning, and Work (Oxford: Oxford University Press,
2007), pp. 117129.
Borges, Jorge Luis, The False Problem of Ugolino, trans. Esther Allen, in The
Total Library: Non-Fiction 19221986, ed. Eliot Weinburger (London: Penguin,
2001), pp. 277279.
Boyd, Richard, Metaphor and Theory Change: What is Metaphor a Metaphor
For?, in Andrew Ortony, ed., Metaphor and Thought, 2nd edn (Cambridge:
Cambridge University Press, 1993), pp. 481532.
Bromberger, Sylvain, Why-Questions, in R. Colodny, ed., Mind and Cosmos:
Essays in Contemporary Science and Philosophy (Pittsburgh: University of Pitts-
burgh Press, 1966), pp. 86111.
Budd, Malcolm, Music and the Emotions (London: Routledge and Kegan Paul,
1985).
BIBLIOGRAPHY 181
Budd, Malcolm, The Aesthetic Appreciation of Nature: Essays on the Aesthetics of Nature
Aesthetic Essays (Oxford: Oxford University Press, 2008).
Aesthetic Realism and Emotional Qualities of Music, in Aesthetic Essays
(Oxford: Oxford University Press, 2008), pp. 171184.
The Characterization of Aesthetic Qualities by Essential Metaphors and
Quasi-Metaphors, in Aesthetic Essays (Oxford: Oxford University Press,
2008), pp. 142153.
Musical Movement and Aesthetic Metaphors, in Aesthetic Essays (Oxford:
Understanding Music, in Aesthetic Essays (Oxford: Oxford University Press,
2008), pp. 122141.
Burke, Edmund, A Philosophical Enquiry into the Origin of Our Ideas of the Sublime and
Beautiful, ed. Adam Phillips (Oxford: Oxford University Press, 1990).
Camp, Elisabeth, Metaphor and That Certain Je Ne Sais Quoi , Philosophical
Studies 129/1 (2006): 125.
Campbell, Alastair, Questions over David Camerons Judgment are Justied, The
Guardian, 21 January 2011, <http://www.guardian.co.uk/media/2011/jan/
21/alastair-campbell-david-cameron> accessed 14 June 2012.
Carroll, Nol, On Criticism (New York: Routledge, 2009).
Casey, John, The Language of Criticism (London: Methuen, 1966).
Cavell, Stanley, Aesthetic Problems of Modern Philosophy, in Must We Mean
What We Say?: A Book of Essays (Cambridge: Cambridge University Press,
1976), pp. 7396.
Clark, Kenneth, Raphael: The Miraculous Draught of Fishes, in Looking at Pictures
(London: John Murray, 1960), pp. 6373.
Clunas, Craig, Jade Carvers and Their Customers in Ming China, The Bulletin of
the Friends of Jade 6 (1989): 3346.
Cohen, Ted, On Consistency in Ones Personal Aesthetics, in Jerrold Levinson,
ed., Aesthetics and Ethics: Essays at the Intersection (Cambridge: Cambridge Uni-
versity Press, 1998), pp. 106125.
Coleridge, Samuel Taylor, Lectures on Shakespeare and Milton in Illustration of
the Principles of Poetry, in Lectures 18081819: On Literature, ed. R. A. Foakes
(London: Routledge and Kegan Paul, 1987), pp. 235263.
Cooke, Brandon, Imagining Art, British Journal of Aesthetics 47/1 (2007): 2945.
Danto, Arthur C., The Fly in the Fly Bottle: The Explanation and Critical
Judgment of Works of Art, in Unnatural Wonders: Essays from the Gap Between
Art and Life (New York: Farrar, Straus and Giroux, 2005), pp. 354368.
Davidson, Donald, What Metaphors Mean, in Inquiries into Truth and Interpretation
182 BIBLIOGRAPHY
Davies, Stephen, Musical Meaning and Expression (Ithaca: Cornell University Press,
1994).
Music and Metaphor, in Musical Understandings and Other Essays on the
Philosophy of Music (Oxford: Oxford University Press, 2011), pp. 2133.
De Clercq, Rafael, Melody and Metaphorical Movement, British Journal of Aes-
thetics 47/2 (2007): 156168.
De Quincey, Thomas, On the Knocking at the Gate in Macbeth, in Confessions of
an English Opium-Eater, and Other Writings, ed. Grevel Lindop (Oxford: Oxford
University Press, 1985), pp. 8185.
Desgodets, Antoine Babuty, Les Edices Antiques de Rome: Dessins et Mesurs trs
Exactement (Paris: Jean Baptiste Coignard, 1682).
Dickinson, Emily, To Fight Aloud, Is Very Brave, in The Poems of Emily Dick-
inson, ed. R. W. Franklin (Cambridge, MA: Harvard University Press, 1999),
p. 70.
Du Bos, Jean-Baptiste, Rexions critiques sur la posie et sur la peinture, 3 vols (Paris:
Pierre-Jean Mariette, 1733).
Eliot, T. S., The Waste Land, in The Complete Poems and Plays (London: Faber and
Faber, 1969), pp. 6180.
Elkins, James, What Happened to Art Criticism? (Chicago: Prickly Paradigm Press,
2003).
Fogelin, Robert J., Metaphors, Similes, and Similarity, in Jaakko Hintikka, ed.,
Aspects of Metaphor (Dordrecht: Kluwer, 1994), pp. 2339.
Figuratively Speaking, 2nd edn (Oxford: Oxford University Press, 2011).
Forsyth, Angus and Brian McElney, Jades from China (Bath: The Museum of East
Asian Art, 1994).
Fried, Michael, Courbets Realism (Chicago: University of Chicago Press, 1990).
Gaut, Berys, Metaphor and the Understanding of Art, Proceedings of the Aristotelian
Society 97 (1997): 223241.
Creativity and Imagination, in Berys Gaut and Paisley Livingston, eds, The
Creation of Art: New Essays in Philosophical Aesthetics (Cambridge: Cambridge
Creativity and Skill, in Michael Krausz, Denis Dutton, and Karen Bardsley,
eds, The Idea of Creativity (Leiden: Brill, 2009), pp. 83103.
Gerard, Alexander, An Essay on Taste, 3rd edn (Delmar, NY: Scholars Facsimiles
and Reprints, 1978).
Gombrich, E. H., Meditations on a Hobby Horse and Other Essays on the Theory of Art
(London: Phaidon, 1963).
On Physiognomic Perception, in Meditations on a Hobby Horse and Other
Essays on the Theory of Art (London: Phaidon, 1963), pp. 4555.
Visual Metaphors of Value in Art, in Meditations on a Hobby Horse and Other
Essays on the Theory of Art (London: Phaidon, 1963), pp. 1229.
BIBLIOGRAPHY 183
Goodman, Nelson, Languages of Art: An Approach to a Theory of Symbols, 2nd edn

(Indianapolis: Hackett, 1976).
Gould, Stephen Jay, A Biological Homage to Mickey Mouse, in The Pandas
Thumb: More Reections in Natural History (New York: Norton, 1980),
pp. 93107.
Grant, James, The Dispensability of Metaphor, British Journal of Aesthetics 50/3
(2010): 255272.
The Value of Imaginativeness, Australasian Journal of Philosophy 90/2 (2012):
275289.
Gray, Thomas, Elegy Written in a Country Church Yard, in Christopher Ricks,
ed., The Oxford Book of English Verse (Oxford: Oxford University Press, 1999),
pp. 278281.
Guttenplan, Samuel, Objects of Metaphor (Oxford: Oxford University Press, 2005).
Guyer, Paul, Kant and the Claims of Taste, 2nd edn (Cambridge: Cambridge
Values of Beauty: Historical Essays in Aesthetics (Cambridge: Cambridge Uni-
versity Press, 2005).
The Origins of Modern Aesthetics: 17111735, in Values of Beauty: Historical
Essays in Aesthetics (Cambridge: Cambridge University Press, 2005), pp. 336.
The Standard of Taste and the Most Ardent Desire of Society , in Values of
Beauty: Historical Essays in Aesthetics (Cambridge: Cambridge University Press,
2005), pp. 3774.
Hampshire, Stuart, Logic and Appreciation, in William Elton, ed., Aesthetics and
Language (Oxford: Blackwell, 1959), pp. 161169.
Hartman, Geoffrey H., How Creative Should Literary Criticism Be?, The New York
Times Book Review, 5 April 1981 <http://www.nytimes.com/1981/04/05/
books/how-creative-should-literary-criticism-be.html> accessed 14 June 2012.
Hartt, Frederick, History of Italian Renaissance Art, 3rd edn (London: Thames and
Hudson, 1987).
Hazlitt, William, Macbeth, in Selected Writings, ed. Jon Cook (Oxford: Oxford
Hopkins, Gerard Manley, Spelt from Sibyls Leaves, in Gerard Manley Hopkins:
The Major Works, ed. Catherine Phillips (Oxford: Oxford University Press,
2002), p. 175.
Hopkins, Robert, Sculpture, in Jerrold Levinson, ed., The Oxford Handbook of
Aesthetics (Oxford: Oxford University Press, 2003), pp. 572582.
Critical Reasoning and Critical Perception, in Matthew Kieran and Dom-
inic McIver Lopes, eds, Knowing Art: Essays in Aesthetics and Epistemology (Dor-
drecht: Springer, 2007) pp. 137153.
Horace, Ars Poetica, in The Satires and Epistles of Horace and Persius, ed. and trans.
Niall Rudd (Harmondsworth: Penguin, 1979), pp. 190203.
184 BIBLIOGRAPHY
Hugo, Victor-Marie, William Shakespeare (extract), trans. A. Baillot, in Jonathan

Bate, ed., The Romantics on Shakespeare (London: Penguin, 1992), pp. 349352.
Hume, David, Of the Standard of Taste, in Essays: Moral, Political, and Literary, ed.
Eugene F. Miller (Indianapolis: Liberty Fund, 1985), pp. 226249.
A Treatise of Human Nature, ed. David Fate Norton and Mary J. Norton
Hunt, Leigh, Retrospective Views of Keats, in G. M. Matthews, ed., John Keats:
The Critical Heritage (London: Routledge, 1971), pp. 240245.
Hyman, John, The Objective Eye: Color, Form, and Reality in the Theory of Art
(Chicago: University of Chicago Press, 2006).
Isenberg, Arnold, Critical Communication, Philosophical Review 58/4 (1949):
330344.
Johnson, Samuel, Selections from the Notes to the Edition of Shakespeares Plays,
in Samuel Johnson on Shakespeare, ed. H. R. Woudhuysen (London: Penguin,
1989), pp. 166248.
Milton, in The Lives of the Poets: A Selection, ed. Roger Lonsdale (Oxford:
Kant, Immanuel, Critique of Judgement, ed. Nicholas Walker, trans. James Creed
Meredith and Nicholas Walker (Oxford: Oxford University Press, 2007).
Kaufman, Daniel A., Normative Criticism and the Objective Value of Artworks,
The Journal of Aesthetics and Art Criticism 60/2 (2002): 151166.
Critical Justication and Critical Laws, British Journal of Aesthetics 43/4
(2003): 393400.
Kermode, Frank, Shakespeares Language (London: Penguin, 2000).
Kivy, Peter, Sound Sentiment: An Essay on the Musical Emotions (Philadelphia:
Temple University Press, 1989).
Kupperman, Joel J., Value . . . and What Follows (Oxford: Oxford University Press,
1999).
Lamarque, Peter, The Philosophy of Literature (Oxford: Blackwell, 2009).
Lehrer, Keith, Art, Self and Knowledge (Oxford: Oxford University Press, 2012).
Lessing, Gotthald Ephraim, Laocon: An Essay on the Limits of Painting and Poetry, ed.
and trans. Edward Allen McCormick (Indianapolis: Bobbs-Merrill, 1962).
Levinson, Jerrold, Musical Expressiveness as Hearability-as-Expression, in Con-
templating Art: Essays in Aesthetics (Oxford: Oxford University Press, 2006),
pp. 91108.
Lopes, Dominic McIver, Sight and Sensibility: Evaluating Pictures (Oxford: Oxford
Lycan, William G., Philosophy of Language: A Contemporary Introduction, 2nd edn
(New York: Routledge, 2008).
March, Ivan et al., The Penguin Guide to Recorded Classical Music 2010 (London:
Penguin, 2009).
BIBLIOGRAPHY 185
Marshall, Cynthia, Cosmology and the Body, in Donna B. Hamilton, ed., A

Concise Companion to English Renaissance Literature (Oxford: Blackwell, 2006),
pp. 217237.
Matravers, Derek, Art and Emotion (Oxford: Oxford University Press, 1998).
Mendelssohn, Moses, On the Sublime and Naive in the Fine Sciences, in
Philosophical Writings, ed. and trans. Daniel O. Dahlstrom (Cambridge: Cam-
bridge University Press, 1997), pp. 192232.
Miller, Keith, People Who Cant Love People, The Times Literary Supplement, 9
June 2006, p. 22.
Moran, Richard, Seeing and Believing: Metaphor, Image, and Force, Critical
Inquiry 16/1 (1989): 87112.
Mothersill, Mary, Critical Reasons, The Philosophical Quarterly 11/42 (1961):
7478.
Beauty Restored (Oxford: Oxford University Press, 1984).
Olsen, Stein Haugom, Criticism of Literature and Criticism of Culture, Ratio
22/4 (2009): 439463.
Orwell, George, Politics and the English Language, in Essays, ed. Bernard Crick
(London: Penguin, 2000), pp. 348360.
Passmore, John, Serious Art: A Study of the Concept in all the Major Arts (London:
Duckworth, 1991).
Pater, Walter, Studies in the History of the Renaissance, ed. Matthew Beaumont
Peacocke, Christopher, Experiencing Metaphorically-As in Music Perception:
Clarications and Commitments, British Journal of Aesthetics 49/3 (2009):
299306.
The Perception of Music: Sources of Signicance, British Journal of Aesthetics
49/3 (2009): 257275.
Music and Experiencing Metaphorically-As: Further Delineation, British
Journal of Aesthetics 50/2 (2010): 189191.
Pettit, Philip, The Possibility of Aesthetic Realism, in Eva Schaper, ed., Pleasure,
Preference and Value: Studies in Philosophical Aesthetics (Cambridge: Cambridge
Reimer, Marga, and Elisabeth Camp, Metaphor, in Ernest Lepore and Barry
C. Smith, eds, The Oxford Handbook of Philosophy of Language (Oxford: Oxford
Ruskin, John, The Stones of Venice, 3 vols (London: George Allen, 1900).
Schroeder, Severin, Why Juliet is the Sun, in Semantik und Ontologie, ed. Mark
Siebel and Mark Textor (Frankfurt am Main: Ontos Verlag, 2004), pp. 63100.
Scruton, Roger, Art and Imagination (London: Methuen, 1974).
The Aesthetics of Music (Oxford: Oxford University Press, 1997).
186 BIBLIOGRAPHY
Scruton, Roger, Understanding Music, in The Aesthetic Understanding: Essays in

the Philosophy of Art, 2nd edn (South Bend, IN: St Augustines Press, 1998),
pp. 89115.
Musical Movement: A Reply to Budd, British Journal of Aesthetics 44/2
(2004): 184187.
Beauty (Oxford: Oxford University Press, 2009).
Searle, John R., Metaphor, in Expression and Meaning: Studies in the Theory of
Speech Acts (Cambridge: Cambridge University Press, 1979), pp. 76116.
Shakespeare, William, Hamlet, ed. G. R. Hibbard (Oxford: Oxford University
Press, 1987).
Henry IV: Part One, ed. David Bevington (Oxford: Oxford University Press,
1987).
The Tragedy of Anthony and Cleopatra, ed. Michael Neill (Oxford: Oxford
Sharpe, R. A., Philosophy of Music: An Introduction (Chesham: Acumen, 2004).
Shelley, James, The Character and Role of Principles in the Evaluation of Art,
British Journal of Aesthetics 42/1 (2002): 3751.
Critical Compatibilism, in Matthew Kieran and Dominic McIver Lopes,
eds, Knowing Art: Essays in Aesthetics and Epistemology (Dordrecht: Springer,
2007), pp. 125136.
Shusterman, Richard, Wittgenstein and Critical Reasoning, Philosophy and Phe-
nomenological Research 47/1 (1986): 91110.
Sibley, Frank, Approach to Aesthetics: Collected Essays on Philosophical Aesthetics, ed.
John Benson, Betty Redfern, and Jeremy Roxbee Cox (Oxford: Oxford
Aesthetic Concepts, in Approach to Aesthetics: Collected Essays on Philosophical
Aesthetics, ed. John Benson, Betty Redfern, and Jeremy Roxbee Cox (Oxford:
Aesthetic and Non-Aesthetic, in Approach to Aesthetics: Collected Essays on
Philosophical Aesthetics, ed. John Benson, Betty Redfern, and Jeremy Roxbee
Cox (Oxford: Oxford University Press, 2001), pp. 3351.
Aesthetics and the Looks of Things, in Approach to Aesthetics: Collected Essays
on Philosophical Aesthetics, ed. John Benson, Betty Redfern, and Jeremy Roxbee
Making Music Our Own, in Approach to Aesthetics: Collected Essays on
Objectivity and Aesthetics, in Approach to Aesthetics: Collected Essays on
BIBLIOGRAPHY 187
Sibley, Frank, Originality and Value, in Approach to Aesthetics: Collected Essays on

Tastes, Smells, and Aesthetics, in Approach to Aesthetics: Collected Essays on
Silverstein, Shel, Slithergadee, in Uncle Shelbys Zoo: Dont Bump the Glump
(New York: Simon and Schuster, 1964).
Smith, R. R. R., Hellenistic Sculpture (London: Thames and Hudson, 1991).
Snowdon, Paul F., Peacocke on Musical Experience and Hearing Metaphorically-
As, British Journal of Aesthetics 49/3 (2009): 277281.
Stevens, Wallace, Banal Sojourn, in The Collected Poems of Wallace Stevens
(New York: Alfred A. Knopf, 1954), pp. 6263.
Strawson, P. F., Imagination and Perception, in Freedom and Resentment and Other
Essays (London: Methuen, 1974), pp. 4565.
Stroud, Barry, Hume (London: Routledge and Kegan Paul, 1977).
Summerson, John, The Classical Language of Architecture (London: Thames and
Hudson, 1980).
Tappolet, Christine, Through Thick and Thin: Good and Its Determinates,
Dialectica 58/2 (2004): 207221.
Tversky, Amos, Features of Similarity, Psychological Review 84/4 (1977): 327352.
van Fraassen, Bas C., The Scientic Image (Oxford: Oxford University Press, 1980).
Vendler, Helen, The Art of Shakespeares Sonnets (Cambridge, MA: Harvard
Walton, Kendall L., Mimesis as Make-Believe: On the Foundations of the Representa-
tional Arts (Cambridge, MA: Harvard University Press, 1990).
Metaphor and Prop Oriented Make-Believe, European Journal of Philosophy
1/1 (1993): 3956.
Weiss, Antonio, Harold Bloom, The Art of Criticism No. 1, The Paris Review
118, Spring 1991 <http://www.theparisreview.org/interviews/2225/the-art-
of-criticism-no-1-harold-bloom> accessed 14 June 2012.
White, Alan R., The Language of Imagination (Oxford: Blackwell, 1990).
White, Roger, The Structure of Metaphor: The Way the Language of Metaphor Works
(Oxford: Blackwell, 1996).
Wilde, Oscar, The Critic as Artist, in The Major Works, ed. Isobel Murray
Wilson, Jessica, Determination, Realization and Mental Causation, Philosophical
Studies 145 (2009): 149169.
Wollheim, Richard, Correspondence, Projective Properties, and Expression in
the Arts, in The Mind and Its Depths (Cambridge, MA: Harvard University Press,
1993), pp. 144158.
188 BIBLIOGRAPHY
Woodham, Jonathan M., Twentieth-Century Design (Oxford: Oxford University

Press, 1997).
Woolf, Virginia, The Faery Queen, in Hugh Maclean and Anne Lake Prescott,
eds, Edmund Spensers Poetry, 3rd edn (New York: Norton, 1993), pp. 672675.
Yablo, Stephen, How in the World?, in Thoughts (Oxford: Oxford University
Press, 2008), pp. 191220.
Does Ontology Rest on a Mistake?, in Things (Oxford: Oxford University
Press, 2010), pp. 117144.
Zangwill, Nick, The Metaphysics of Beauty (Ithaca: Cornell University Press, 2001).
Index
Acocella, Joan 18 Bender, John 151

Addison, Joseph 17n31, 556, 59 Bennett, M. R. 32
aesthetic experience 539, 645, 67, 149 Bernini, Gian Lorenzo 1367, 1567,
aesthetic properties 23, 31, 37, 54, 645, 159, 172
67, 156 Bible 89, 101, 145, 160
Alhambra 33 Black, Max 89n2, 137n24, 138, 144
Ali, Muhammad 127 Blackburn, Simon 97
Alison, Archibald 589 Bloom, Harold 62n35
allusion 31, 119, 160 Boden, Margaret 80n62
anthimeria 90n7, 11920 Borges, Jorge Luis 32
appreciation 2952, 845, 173 Boyd, Richard 1335
awareness of a work 2930, 347, Brogan, T. V. F. 120n65
449, 51, 169 Bromberger, Sylvain 25n54
cogitative responses 324, 169 Budd, Malcolm 33n6, 125n1
conative responses 33 Burgon vase 161
criticism and, see criticism, Burke, Edmund 55, 58
acquaintance requirement;
criticism, aims of Camp, Elisabeth 102, 110, 1335,
of culturally and historically distant 1404, 148
works 334, 52 Campbell, Alastair 106
emotional and affective responses 323, Carroll, Nol 1314, 1623
49, 512, 162 causal analogy 118, 121, 155
imaginativeness and 5365, 67 Cavell, Stanley 978, 115
knowledge 29, 314, 15962 Cervantes, Miguel de 51
meta-responses 334, 169 Clark, Kenneth 160, 1648
perception 301, 15962 Coleridge, Samuel Taylor 17n31, 177
pleasure 33n6, 549 Collingwood, R. G. 17n31
reasons for responses 357, 469, Colosseum 25, 41
845, 169 Coppo di Marcovaldo 161
Aquinas, Thomas 118 Courbet, Gustave 49, 85
Aristotle 17n31, 63, 118, 1656 criticism
Arnold, Matthew 5961 academic criticism 6
art history 17, 38 acquaintance requirement 3840,
Austin, J. L. 123n69 456, 173
aims of 528, 3952, 1734
Barthes, Roland 176 aiding appreciation 278, 4150,
Baudelaire, Charles 59 159, 168, 1714
Baxandall, Michael 7, 10n17 communicating facts about
Beardsley, Monroe 67, 15, 18, 23n44, appreciation 3952, 15963,
89nn12, 107n39, 109n45 1701, 1734
Beattie, James 55, 58 constitutive vs non-constitutive
Beazley, John 1612 aims 5, 3942, 173
Beethoven, Ludwig van 30, 169 evaluation, see evaluation
Belvedere Torso 83 explanation 237, 41
190 INDEX
criticism (cont.) Gerard, Alexander 55, 58

guiding perception 716, 412, Gettier, Edmund 756
46, 168 Goldscheider, Ludwig 913
helping readers choose 67, Gombrich, E. H. 149
1819, 41 Goodman, Nelson 149
description 456, 52, 54, 59, 615, 85 Gould, Stephen Jay 356
see also criticism, metaphor and Gray, Thomas 118
elucidation 1516, 19, 234, 401 Guggenheim Museum Bilbao 76n59
endowments of good critics 4552, Guttenplan, Samuel 99n19, 109n45,
826, 174 110, 120n66, 13940
imaginativeness and 5367, 826, Guyer, Paul 55, 578
172, 1757
interpretation 401, 524 Hacker, P. M. S. 32
metaphor and 54, 625, 67, 86, 125, Hampshire, Stuart 7, 149
1313, 135, 14972 Hndel, Georg Friedrich 59
reviews 67, 15, 278, 38, 168 Hardy, Thomas 24, 41
value of 5962 Hartt, Frederick 161
versus Humean judgement 502 Hazlitt, William 223
versus other discourses about Hollinghurst, Alan 1578, 162
art 3840, 42 Hopkins, Gerard Manley 145
Hopkins, Robert 30n1
Dal, Salvador 69n50 Horace 63
Dante 15, 32, 356, 85 Hugo, Victor 157
Danto, Arthur 23 Hume, David 502, 557, 59
Davidson, Donald 99101, 105, 107, Hunt, Leigh 85
11115, 123, 1389, 153, 165, 175 Hyman, John 95n12, 101n21, 104n29
De Quincey, Thomas 24 hyperbole 119
determinates and determinables 93, 96
Dickinson, Emily 89 imagination 32, 5562, 65
Donne, John 91 versus imaginativeness, see
Du Bos, Jean-Baptiste 55 imaginativeness, versus
imagination
El Greco 913 see also metaphor, imagination and
Eliot, T. S. 31, 145, 147, 160 imaginativeness 66, 6886, 175
Erasmus 48 of acts, omissions, and products 6880
Eucharides Painter 161 criticism and, see criticism,
Euphiletos Painter 161 imaginativeness and
evaluation 810, 1314, 1623, 27, 40, of emotions and attitudes 68n49
46, 60 inspiration and 7980
category-relative 1718, 20 obviousness 702, 778, 825
principles 810, 1718 of persons 68, 804
pro tanto 203 relativity of 689, 778
reasons 810, 1314, 1618, 203, 46 thinking 716, 7886
expression 534, 64, 1367, 149, 1535 value and 747, 82, 86
versus imagination 66, 789
Fogelin, Robert 1047, 122n67 versus originality 6970, 78
Frazier, Joe 127 versus perceptiveness 824, 175
versus the unimaginative and the
Gaut, Berys 66n47, 74, 80n62, 1313, 135 non-imaginative 78
Gehry, Frank 76n59 Isenberg, Arnold 717
INDEX 191
James, Henry 1516, 160 perception and 11314, 125, 1313,

Johnson, Samuel 17n31, 223, 1659, 1712, 175
44, 50n25 poetic and novel metaphors 11821,
1457, 1678
Kandinsky, Wassily 95, 131 science and 1335
Kant, Immanuel 33n6, 55, 579, specicity of 1634, 1702, 175
66, 149 subject of 90, 102
Kaufman, Daniel 1314 understanding metaphors 878,
Keats, John 856 914, 98, 107, 11116, 1468,
Kermode, Frank 45, 48, 83 1667
value of 1445, 16372, 175
Lamarque, Peter 24 versus other ways of extending the use
Leonardo da Vinci 46, 612, 67, 746 of expressions 11722, 155
Lessing, Gotthald Ephraim 58 what metaphors communicate 878,
Levinson, Jerrold 154n13 938, 11115, 1234
likeness, see metaphor, likeness and see also metaphor, Minimal Thesis
Lloyd, Harold 14, 17, 201 metonymy 119
Lorrain, Claude 59 Miller, Keith 1578, 162
Lycan, William 1013, 110, 11821 Milton, John 445, 59, 85
Mir, Joan 49, 85
make-believe 37, 12631, 146 Mitchell, Joan 23
Malevich, Kazimir 30, 169 Moran, Richard 109, 166
Marden, Brice 23 Morris, Mark 18
Marshall, Cynthia 119n63 Mothersill, Mary 11
Mendelssohn, Moses 58
metaphor nominalism 12930
anti-realism about 1516, 174
characterizing versus non- obviousness, see imaginativeness,
characterizing 902, 1478 obviousness
comparison theory 87, 98111, Olsen, Stein Haugom 6n3
11521, 122n67 Orwell, George 1667
criticism and, see criticism, metaphor
and Palazzo Ducale (Venice) 163
expressions used metaphorically 8990, Palazzo del T 478
1023, 11920 Parrhasios 38
imagination and 16572, 175 Parthenon 160n21
indispensability of 12348, Passmore, John 74
150, 1745 Pater, Walter 44, 46, 54, 612, 67, 168
why metaphors seem Penguin Guide to Recorded Classical
indispensable 1234, 1347 Music 1819
likeness and 923, 956, 1001, perceptiveness, see imaginativeness, versus
10311, 11618, 1289, 1413, perceptiveness
1568, 1667 Plath, Sylvia 110
mathematics and 12930 pleasure, see appreciation, pleasure
meaning of 878, 99101, 11112,
1389, 153 Raphael 160, 1645
Minimal Thesis 87124, 1506, 1745 Reimer, Marga 102, 110
open-endedness of 978, 1223 Rembrandt van Rijn 17
paraphrase and explanation of 1023, Romano, Giulio 478
111, 13844, 146 Ruskin, John 54, 61, 1638, 171
192 INDEX
Santa Maria della Salute 163 synecdoche 119

sarcasm 90, 153
Schroeder, Severin 97n18, 103n27, Tappolet, Christine 21
107n41, 11718, 122n67 testimony 83
Scruton, Roger 65, 67, 125, 1367, 144, see also criticism, acquaintance
1526, 165, 1678, 176 requirement
Searle, John 1034, 1078, 110, thinking, see imaginativeness, thinking
11618, 121 Turner, J. M. W. 61
Shakespeare, William 15, 19, 202, 24, Tversky, Amos 105
312, 3941, 45, 48, 83, 856,
8990, 101, 11821, 157, 160, 177 van der Weyden, Roger 14
Shelley, James 8n11 van Fraassen, Bas 25
Shriver, Lionel 68 Vendler, Helen 48
Shusterman, Richard 17n31
Sibley, Frank 7, 23, 31, 37n10, 57n12, 63, Walton, Kendall 37, 109,
70n53, 1367, 1512 126, 1301
Silverstein, Shel 36 Weitz, Morris 17n31
similarity, see metaphor, likeness and White, Alan 79, 80n62
simile 87, 99101, 1057, 11011, White, Roger 120n66
122, 1423 Wilde, Oscar 602, 66
Smith, R. R. R. 83 Wilson, Jessica 96n14
Sophocles 31, 35 Woodham, Jonathan 161
spaghetti style 157, 1623, Woolf, Virginia 6
1656, 172 Wordsworth, William 5960, 85
Spenser, Edmund 6
Stevens, Wallace 96 Yablo, Stephen 967, 12631, 135
Strawson, P. F. 82n63
Stroud, Barry 52n32 Zangwill, Nick 125, 148n46
Summerson, John 25, 41, 478, 51 Zeuxis 38

0199661790

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

0199661790

Uploaded by

Copyright:

Available Formats

The Critical Imagination

oxford philosophical monographs

Nietzsche and Metaphysics

Dispensability of Metaphor, British Journal of Aesthetics 50/3 (2010):

1.1 C. Elizabeth Grant. Drawing after Antoine Babuty Desgodets, Second

This view of criticism is of interest for several reasons. I will mention

communication in general. The language of criticism is not alone in at-

Wilde, and Roger Scruton, among others. I then develop my own

To characterize the role imaginativeness plays in criticism of the arts, we

1. Helping readers choose

Isenberg begins his argument by examining

(A) The truth of R offers no support for the value judgement.

(C) The best explanation of Rs function is that using it is a way of

and depend on some general claim, so it seems that R is perfectly

the qualities to which it somehow directs our attention.18 Presumably, the

Goldscheider does not use the expression wavelike contour in the

information. . . . We shall be concerned exclusively with the descriptive

such as description, classication, contextualization, elucidation, inter-

Thus, evaluation is an essential feature of criticism such that if a piece of

To illustrate such category-relative evaluation, Carroll uses a review by

As I have said, Carroll is right to hold both that it is possible to provide

This reply is weak because at most it establishes that, when there is no

appears to be criticism is a datum that any account of criticism must respect

For example, Samuel Johnson, discussing Othello, writes: The scenes

reasoned pro tanto evaluations. Often, it will indeed be necessary to provide

Sibley describes explanation as one of the central activities of critics,46

example of attempting to explain why Hamlet procrastinated, which

So too, some explanations provide answers to what-questions. You can

It would appear, then, that explanation plays a signicant role in

someone who appreciates it, without providing an explanation of why it

describe the work in such a way as to provide information that is useful to

objects of appreciation, some form of perceptual acquaintance with the

involve. You appreciate the work in virtue of responding in some of these

Acquiring knowledge by appropriate means is not the only kind of

a polished gem, being awestruck by the Alhambra, and cringing at a poorly

you from having an appropriate affective response yourself, but do not

the features in virtue of which he looks more childlike, explain our

2. Appreciation and the aims of criticism

explain the meaning of unfamiliar words are not necessarily critical

the content of interpretations and elucidations as their content. They tell

criticismthough not a constitutive aim. I said in Chapter 1 that an aim is

appropriate response (a part, feature, or represented element), R an appropriate

4. Endowments of good critics

In short, the critic must be articulate.

appropriate response. I suggest we can identify three endowments that

It might be obvious to someone with enough knowledge of the rules and

It is partly Kermodes knowledge of the adage that enables him to recognize

reason or an object of an appropriate response. As I will put it, she must be

(vii) A disposition to have appropriate responses the readership is

In my view, a disposition to have appropriate sentiments can certainly

We are almost in a position to say what role imaginativeness plays in

music as expressive of emotion we perceive imaginatively. Likewise, many

1. The pleasures of the imagination

and judge art. Humes account of the characteristics of true judges in Of

this is a beauty merely of imagination, and has no foundation in what appears

Kant, as is well known, also thinks pleasure in beauty derives from an

sublimity and majesty of creation,17 and stimulates the imagination into

Kants account of our responses to such artworks is similar to his account of

nd, is insufcient to excite these emotions.23 The emotions produced in

This emphasis in both Kant and Alison on the multitude of thoughts,

2. The critic as artist

This leads to an attack on Arnolds dictum that the proper aim of

A great critic, in Wildes view, will be attracted to such works as make

3. Criticism and metaphor