Mind Association

Art and the "Object of Art"

Author(s): Paul Ziff
Source: Mind, New Series, Vol. 60, No. 240 (Oct., 1951), pp. 466-480
Published by: Oxford University Press on behalf of the Mind Association
Stable URL: http://www.jstor.org/stable/2251144 .
Accessed: 20/06/2014 18:06

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].

Oxford University Press and Mind Association are collaborating with JSTOR to digitize, preserve and extend
access to Mind.

http://www.jstor.org

This content downloaded from 195.34.79.223 on Fri, 20 Jun 2014 18:06:46 PM

All use subject to JSTOR Terms and Conditions
 II.-ART AND THE "OBJECT OF ART"

BY PAUL ZIFF
PERHAPS the most persistentmy-th in presentday estheticsis
the notionthat whenwe discussa workofart we are nottalking
about a paintingb1utabout some " illusory" or " imaginary"
thing sometimescalled the " object of art" or the " esthetic
object ". I can best explain what this mythis by quotinga
statementof Samuel Alexander.
Morethan once I have pointedout how in the beautifulobject the
significanceis suppliedin partfromthe artist'smind; how it is he
whomakestheflatMadonnaseem,as Mr.Berensonputsit,a tangible
three-dimensional being,orwhogivesdivineplayfulness totheHermes,
or motionand dance to the motionless maidensin the pictureof the
Spring,or who findsthe perfect,the only fittingword,to express
a meaningthat springsfromhim, .. And I have contrasted
theobjectofartwiththemereperceptwherealso halfcomesfromthe
perceiver'smindand halffromwhat he directlysees: the coloured
movingshape is perceivedto be a man,thoughsightalone without
memorydoes not say so. The contrast. . . is this: the characters
we imputeto the object perceived,if we perceivecorrectly, really
do belongto the objectand maybe sensedthereon properoccasion;
the colouredshape is the visiblesurfaceof a man; but in the work
of art thereis alwaysillusion: the Hermesis not divineonly but
seemsso,andthegirlsintheSpringarenotinmotion.Atthesametime,
I haveadded,theartisticillusionis unlikeordinary perceptualillusion,
forthat-illusiondisappearsto betteracquaintance,is recognized to
be an illusion. Whereasthe illusionis ofthe essenceof the workof
art-ceases, therefore,to le illusionand makesthe objectsignificant.
Pieces,p. 259.
-Pilosoplical and Literary
Let me state the kind of thingAlexanderis talkingabout; for
even thoughhe expresseshimselfwith admirableclarity,it is
always usefulto statea pointin as simplea manneras is possible.
And let us take some particularexample,say, a Cezanne still-
life of some apples. When we describe Cezanne's painting
we may say 'the apples are solid, round full-volumes-like
tangible three-dimensionalthings. The painting has great
depth.' But a paintingis a thin stripof canvas coveredwith
tiny pellets of pigment. The canvas is flat,but the work of
art has great depth? Thus Alexanderarrivesat the viewthat
the workof art is distinctfromthe painting; forthe characters
we imputeto the paintingdo not reallybelongto it. We speak
of solid voluminousapples, but there are only thin pellets of
pigmenton the surfaceof the canvas. Thus thereis an illusion,
466

This content downloaded from 195.34.79.223 on Fri, 20 Jun 2014 18:06:46 PM

All use subject to JSTOR Terms and Conditions
 rAUL ZIFF: ART AND THE OBJECT OF ART 467
and this illusionis the objectofart,that whichwe call beautiful,
that whichwe judge, criticise,evaluate,and in general,discuss.
It is as simpleas this.
Beforeexamining thisargumentweshouldnoticethatAlexander
is not alone amongestheticiansin adoptingsuch a view. For
analogous reasons Collingwoodhas claimed that the work of
art is an "imaginary object ". And, in a similarvein, S. C.
Pepper contendsthat the workof art is our " perceptions" of
the painting. De Witt Parker,followingBosanquet,held quite
similarviews. Of course,not all these estheticiansagree that
thereis an illusioninvolvedin our perceptionof a painting,but
they all maintain,along with Alexander,that there is a sig-
nificantdiscrepancybetweenthe paintingand the workof art.
That is to say, they agree with Alexanderin claimingthat the
characterswe imputeto the paintingdo not reallybelongto the
painting,but rather,are charactersof the work of art. Just
what the workof art is held to be, variesfromone philosopher
to another; for Collingwoodit is an "imaginary object ";
Peppercontendsthatit is a " seriesofintermittentperceptions";
Alexauderinsistsit is an " illusion"; and so forth. But all
agreethat the paintingis not the workof art. Indeed,it would
not be too faramissto say thatthe prevalentopinionin esthetics
is in generalaccordancewith Alexander'saccount. And this
is why it is importantto show and to show clearly,how very
mistakenthis viewis.

I
Firstofall we mustconsiderthe notionthatthereis an illusion
involved in observinga painting. This particular point is,
as we have alreadymentioned,somewhatpeculiarto Alexander.
For although other estheticiansclaim there is a discrepancy
between the characterswe impute to the painting and the
charactersit actually possesses,few are preparedto maintain
that thereis any illusionat workhere. Thus, even thoughwe
maysucceedin dispelling theillusionthatthereis illusioninvolved
in observinga painting,we shall not have dispelledthe more
significantand more momentousillusion that there is some
discrepancybetweenthe characterswe imputeto the painting
and the charactersthe paintingactuallypossesses. Nonetheless,
it is fruitfulto begin our discusssion
by pointingout that all
talk of illusionis whollymistaken. For althoughit does not
followfromthe fact that thereis no illusionat workhere,that
the characterswe impute to the paintingreally do belong to

This content downloaded from 195.34.79.223 on Fri, 20 Jun 2014 18:06:46 PM

All use subject to JSTOR Terms and Conditions
468 PAUL ZIFF:

the painting,it would followfromthe fact-if it werea fact-

that there is an illusionat work here that the characterswe
impute to the paintingdo not really belong to the painting.
Thusit is quiteimportant to showthatthereis no illusionat work
here,even thoughshowingthis does not proveall that we want
to prove. To showthat all talk of illusionis quite mistakenwe
must take pains to make clear just what is meant in speaking
of an illusion.
A travellercrossinga desertmay suddenlysee a clear lake
spread out beforehim. He staresat it in amazement,rubs his
eyes and looks again. The waters beckon him; he rushes
forwardand plungeshis hand intohot sand. He was the victim
of a mirage-an illusion. He thoughttherewas waterthere,it
looked like water,but in fact it was only sand. This is one
example of what an illusionis like. Suppose someonesuflers
fromhallucinations, and further supposehe sees an illusoryapple
on a table beforehim. He thinksan apple is there,he sees an
apple, but whei, he reaches out to grasp it, there is nothing
to be grasped. He is deceived,forhe too is the victimof an
illusion. Miragesand hallucinationsare but two typesof visual
illusions. There are othersstill more common. But thereare
otherthingswhichare sometimes looselyspokenofas " illusions"
whichare usuallynot illusionsat all. For example,if we asked
someone to describewhat he sees when he looks at railroad
trackshe mightsay' Theylookas iftheyconvergeinthedistance'.
We are all familiarwith this sort of descriptionof railroad
tracks. Andifanyonesaid thiswe shouldn'tfeelanythingwrong
withwhat he was saying. For,in a way,theydo look as if they
converge,even thoughwe all knowthat theydo not. You can
looklongand hardat railroadtracks,sqUintyoureyesand peeroff
into the distance,but theystill look as if they converge,even
though they do not in fact converge. But, ordinarily,there
is nothingillasoryabout lookingat railroadtracks,forno one is
deceived. If a. child,say, wereto look at the railroadtracksand
ask 'how do the trains run on the tracks when they come
together? ' in such a case we wouldfeelinclinedto speak of an
illusion. This pointca,n,perhaps,be made clear by considering
the matterin a slightlydifferent fashion. Suppose a prankster
laid out somerailroadtracks which,instead of runningparallel
to one another,divergedin sucha manneras to makethemlook
parallel. Would anyone who stood at the rightpoint on the
tracksand looked offinto 'The trackslook as
- the distancesay
if they run parallel' ? Someone could describewhat he saw
in that way, but, most likely,he would be too amazed to say

This content downloaded from 195.34.79.223 on Fri, 20 Jun 2014 18:06:46 PM

All use subject to JSTOR Terms and Conditions
 ART AND THE "OBJECT OF ART
) 469
anythingat all. Indeed, he would thinkhe was suflering from
an illusion. But supposehe said to someonewhowas notlooking
at these queer tracks' The trackslook as iftheyrun parallelto
one another'. The personwhoheardthiswould,mostlikely,not
be the least astonished. E.g. a railroadforemancould ask one
of his crewto take a look at the trackstheyhad just laid to see
if they wereparallel. He mightbe told that ' The trackslook
as if they run parallel', and such a responsewould not be apt
to makehimstarein amazement. The pointis thatwe can and do
describewhat we see when we look at ordinaryrailroadtracks
by saying either' The tracks look as if they run parallel' or
' The trackslookas iftheyconverge'. We shouldnotordinarily
say that either descriptionis the reportof an illusion. For
whetheror no we say a particulardescriptionis the report
of an illusion depends on whetherthe person who gives the
descriptionof what he sees is likelyto be deceivedor not.
Let us nowconsiderwhathappenswhenwe observea painting,
and whetherthere is anythingin such a situationwhich can
properlybe spokenofas an illusion. Suppose,whilelookingat a
Cezanne still-life,we say ' It has great depth. The apples
picturedare full;solid volumes.' This is the kind of statement
we makewhen,say,we are in a paintinggalleryand are discussing
the painting. In sayingthisI do notwishto suggestthat we are
apt to make this statementonly in such a situation,forthis is
not true. There are many situations,many difTerent ones, in
which we make exactly the same statement. All I wish to
suggestis that at least one ofthe situationswherewe make such
a statementis theoneI havejust mentioned. Whenwe areinthe
paintinggallerydiscussingthe painting,and whenwe say 'The
paintinghas great depth', are we deceivedinto believingthat
we.could walk throughthe canvas, put our hands in it, move
aroundinside? Does the canvas look as thoughwe could walk
into and throughit ? The canvas looks as thougb.it has great
depth? Anotherway of puttingthis questionis to ask: does
the framearound the paintinglook, say, like the frameof an
open window? Or like a doorwaythroughwhichwe could pass
freely? We oftendo say that a paintingis verymuch like a
window; a windowthroughwhichwe can see all kdndsofstrange
and fascinatingthingswhichcannotbe seenthroughany ordinary
window. But whenwe see an ordinaryopen windowit really
does look as thoughwe can pass throughthe window. But when
we see a paintingwhichwe sometimesspeak of as a window
throughwhichto viewtheworld,it does notreallylook as though
we could pass our hands and feetthroughit. Not at all. The

This content downloaded from 195.34.79.223 on Fri, 20 Jun 2014 18:06:46 PM

All use subject to JSTOR Terms and Conditions
 470 PAUL ZIFF:

canvas looks flat just as it is flat. Aie we to. suppose when

Alexanderfirstsaw the " flatMadonna" he speaks of,that he
couldn'tsee thatthepictureoftheMadonnawas paintedon a flat
canvas ?-that the painted canvas didn't look flat to him ?
Did he have to go up and look closelyat the canvas to see that it
actually was flat,and that it would be quite futileforhim to
attemptto clamberthrough? This seems most unlikely. For
the onlytype of paintingwherethe canvas doesn't look flatis
whatwe call a workof " trompel'oeil "-and the nameis tailor-
made; forit is a type of paintingwherethe artisthas amused
himself,and possiblyhis audience,by creatingopticalillusions.
When we firstlook at this sortof workwe are deceivedand fail
to realisethatwe are lookingat a painting. In observinga work
of trompel'oeil we are apt to suppose that the pictureof an
apple is not a picturebut a real apple. The apple looks as
thoughwe couldtake it in ourhands. And peoplesometimesare
deceived. Justas childrenmightbe deceivedby the illusionof
railroadtracksconverging and mighthave to be takenfora walk
along the tracksto be convincedthattheydo not in fact con-
verge,so peoplesometimesmusttouchthe canvas, mustpeerat
it, to convincethemselvesthat the pictureof an apple is indeed
just that, not a real apple after all. But neitherCezanne's
paintingsnortheRenaissance, Madonnasareworksoftrompel'oeil.
If Alexanderwas correctin claimingthat " the illusion is of
the essence of the workof art " and if this kind of illusionis
the kind he is talking about then there would be Do way of
distinguishing a workoftrompel'oeil fromany otherwork-but
in factwe do makesuch a distinction. (It is truethat it is hard
to draw a clear cut line betweenthoseworkswhichare properly
spoken of as worksof trompel'oeil and those which are not.
For here, as everywhere,there are borderlinecases. Thus
red mergesimperceptiblyinto blue-but who would on that
accountsay thatredis blue ?) It is truethatthereare paintings
which produce optical illusions,and the fact that they do is
adequatelyindicatedby the wa.ywe label them. For in calling
them worksof " trompel'oeil " we wishto signalisethat these
paintingsdo deceivethe eye,and in thisrespectare quite unlike
otherpaintingswhichare not worksoftrompel'oeil. Thus it is
quite true to say that thereare some paintingswhichproduce
opticalillusions,and ifthiswas whatAlexanderwas sayingthen
he would be perfectlycorrect. However, it is completely
obviousthat thisis not whatAlexanderwas saying. For I take
it that whenhe says " theillusionis ofthe essenceofthe workof
art ", he is sayingthateveryworkofarthas something to do with

This content downloaded from 195.34.79.223 on Fri, 20 Jun 2014 18:06:46 PM

All use subject to JSTOR Terms and Conditions
 ART AND THE "OBJECT OF ART" 471
an illusionand notmerelythatthereare someworksofartwhich
have somethingto do with an illusion. Thus unless we are
preparedto maintainthe false viewthat all paintingsare works
of trompel'oeil, and not merelythat some are, we have, as yet,
seenno evidencewhichwouldwarrantAlexander'sstatementthat
" theillusionis ofthe essenceoftheworkofart ". And we must
notethat eventhoughthereare somepaintingswhichhave some-
thingto do withan illusion,these paintingsare relativelyfew
in number,indeed,quite scarce,and they are of slightinterest
to the esthetician. By farthe greatestnumberofpaintingsare,
in this respect,like Cezanne's canvases. The canvases are flat
and theyappear flat,they are seen as flat,they look flat. No
one is everdeceived.
We have seen that one importantcharacteristic of illusionsis
not applicableto paintings. Thisis thefactthatwe are deceived
by illusions,but we are not deceivedby any ordinarypainting.
This is, in itself,sufficient
to indicatethat all talk of illusionis
quite misleading. But we need not stop here,for,as we press
the matter,it becomesmoreand moreapparentthat therecan
be no questionof an illusion. A personon the desertmightbe
familiarwithmiragesand thusmightnot be deceivedby whathe
sees. Thus, in this case, looking at the paintingmight be
analogous to lookingat themirage,forin neithercase is there
any deception. But thesimilarity betweenthetwo cases begins
and endsat thispoint. No one has to studythemirage,examine
and growfamiliarwithit beforehe says 'It looksas ifthereis a
lake in the distance'. For a singleglance will sufficeto reveal
the lake in the distance. But this is totallyunlikewhat occurs
whenpeoplelook at a Cezannepainting. It may take sometime
to see the depthin one ofhis paintings; the observermustfirst
grow familiarwith the painting,get to know it by carefully
examiningthe structure,composition,design,and so forth. Of
course,thereare manypaintingswhichdo notrequiresuchexten-
sive study. For example, even a quick look at Rembrandt's
NightWatch will suffice to show that it is a pictureof a group
offigures, somein frontofothers. But thereare manypaintings
forwhicha briefglance will not sufficeto disclosethe volumes
and spaces and, withsucha painting,we mustfirstgetto knowit
quite well beforewe can see its depth. An analogywithmusic
may be usefulhere. This processof gettingto knowa -painting
has its counterpartin the processof gettingto knowa piece of
music. When we firsthear a somewhatcomplexwork,say a
Bach violinpartita,we may have difficulty in recognising the
themesand in followingthe variations. When we do knowthe

This content downloaded from 195.34.79.223 on Fri, 20 Jun 2014 18:06:46 PM

All use subject to JSTOR Terms and Conditions
472 PAUL ZIFF:

workwe can thenhearthe themesclearly. Can we say thatthis

processofgettingto knowa workofartis theprogressive growth,
of an illusion? If to see the volumesand space in a painting
is to have an illusion then it seems we should describethis
processof gettingto knowa paintingas the progressivegrowth
of an illusion,the deliberatecultivationof an hallucination.
But ifthisis an illusionit certainlyis unlikeanythingelse which
we ordinarilycall an illusion. Indeed, it is so unlikeanything
whichis ordinarily calledan illusionthatit seemsquitegrotesque
to speak ofit in this way. Finally,afterhavingmanagedto see
the depth in a Cezanne painting,we can oftenrevertto our
originalway of seeingit. That is, sometimeswe can chooseto
see it as havingdepthor to see it as flat. But who can do this
withan illusion? For the illusionof the lake in the distanceis
not something that can be dispelledat will. And the same thing
is true of all illusions. But this is totallyunlikethe so-called
" illusion" ofspace in the Cezannestill-life. For in lookingat a
Cezanne still-lifewe can oftenchoose eitherto see the painting
as flator to see it as havingdepth.
So farwe have failedto produceany evidenceforAlexander's
statement" the illusionis of the essence of the workof art ".
But we cannotleave the matterlike this. For, fromwhat has
so far been said, it would seem that Alexanderwas an utterly
confusedthinkerand had not the slightestidea of what an
illusionwas like. And this,of course,is absolutelyfalse. I do
think Alexanderis mistakenin what he says, but I do not
think it is a foolishor a naive mistake. And if we cannot
presentwhat he is sayingin a more plausible light,this only
points to a failurein our analysis. However,I thinkwe can
make his positionseem quite plausible if we approach it in a
certainmanner. It was mentionedabove that we oftendo say
a paintingis very much like a window. And we argued that
lookingat a paintingis, except perhapswith worksof trompe
l'oeil, not at all like lookingat an open window. And I still
believethat what we said is quite true. But we failedto point
out that lookingat a paintingmay seem verymuchlikelooking
at a closedwindow,and it may also seemverymuchlike looking
through a closed window. For example, someone standing
beforea closed windowmay look throughthe glass at, say, the
landscape outside. But he may also look at the glass of the
window. Similarly, someonestandingbeforea paintingmaylook
as it were,throughthe canvas and see a new and fascinating
landscape. But he may also look at the canvas. Now whenwe
look at a closedwindow,especiallyiftheglassis somewhatdusty,

This content downloaded from 195.34.79.223 on Fri, 20 Jun 2014 18:06:46 PM

All use subject to JSTOR Terms and Conditions
 ART AND THE "OBJECT OF ART 473
we do not thinkwe can pass throughthe window,forwe know
quite well that the glass will preventus fromdoing this. It
does notlook as thoughwe couldactuallypass throughthe glass,
and the only timeit would look that way is whenthe window
looks just like an open window. Nonetheless,we can look
throughthe glass,and we knowthat whatwe see is on the other
side of the glass. Thus lookingat a paintingseemsverymuch
like lookingat a somewhatdustywindow. But when we look
throughthe canvas we knowthat whatwe see is noton the other
side of the canvas. For what we see whenwe look throughthe
canvas is no-whereand no-when. The canvas looks as though
thereis somethingbeyondit, but in factthereis nothing. Thus
-thereis an illusion. Whenwe lookthrougha windowwe look at,
say, the landscape. But whenwe look throughthe canvas we
look at the estheticobject,the workof art. Thus the workof
artis akinto theRed Queenin thatit too existsonlyon the other
side of the glass. And this is a verypersuasivepictureof what
does transpirewhenwe look at paintings.
But even thoughthis account is persuasivewhen looked at
in one way,ifwe approachit witha criticaleyeits persuasiveness
can be dispelled. For, in fact, looking througha canvas is
quite unlikelookingthrougha glass, and this despitethe fact
that an analogy can be made out between them. When we
look througha glass-even a dustyone-it does look as though
we could walk aroundin the space beyondthe glass, as though
thereis a place forhead and hands and feeton the otherside of
the glass. But whenwe look througha canvas it does not look
at all like this. The onlytimeit would look like this is when
we look at worksoftrompel'oeil. And we have alreadypointed
outthattheseare veryspecialand rarecases. If whenwe looked
throughwhat we took to be a window,evena dustywindow,we
foundthat therewas nothingoutside,no otherside to the glass,
we shouldbe astounded. For, in such a case, we assumethat it
did look as thoughtherewas somethingon the otherside of the
glass. It would be like looking across the room througha
dustywindowat what we took to be the house acrossthe street
only to findthat some astoundingartisthad paintedthe whole
scene on our windowpane. In such a case it reallywouldlook
as thoughtherewas somethingon the otherside ofthe glass. I
submitthat this is not at all like what does happen when we
look at an ordinarypainting,or even whenwe look throughan
ordinarypainting. And it is no good sayingthat if we put the
paintingin the windowframein place of the glass and if we
arrangedthelightingin a certainway and ifwethenhad a person
31

This content downloaded from 195.34.79.223 on Fri, 20 Jun 2014 18:06:46 PM

All use subject to JSTOR Terms and Conditions
474 PAUL ZIFF:

look at it who didn't know what had been done, thenit would
look to him as thoughtherewas somethingon the otherside of
the painting. FFor all this is to say no morethan that if the
paintingdidn'tlook the way it does in factordinarily look,then
it mightlook as thoughtherewas somethingon the otherside of
the painting. And forall I know perhapswe could play such
trickswith paintings,but whetherit could be done or 'not is
totallyirrelevantto whatI am saying. For in sayingthat when
we look througha paintingit does not look as thoughthereis
somethingon the otherside of the painting,I am not talking
about what mightbe the case. I am talkingonly about what
is the case, and the factthat somethingelse mightbe the case
doesn'tin the least matter.
Thus, we may repeat,we have not as yet foundany evidence
whichwould warrantAlexander'sstatementthat " the illusion
is ofthe essenceofthe workofart ". Perhapsthisis due to the
fact that, as Alexanderstates, " the artisticillusionis unlike
ordinaryperceptualillusion". And we must admit that it is
unlikeordinaryperceptualillusion,verymuchunlikeit. Indeed,
so verymuchunlikeanythingwe ordinarily speak ofas an illusion
that, as yet, we haven't seen the slightestreason forspeaking
of it as an illusion. However,we need not labour this point.
It is clear that Alexanderdid thinksomethingdreadfullyqueer
was going on and he used the word 'illusion ' to describeit.
And one ofthethingsthathe,alongwithsomeotherestheticians,
finds dreadfullyqueer is the way we ordinarilytalk about
paintings. Here we have the real question.

II
The real question beforeus is, to use Alexander's terms,
whetherthe characterswe imputeto a paintingreallydo belong
to the painting or not. Alexander,along with some other
estheticians,claimsthat theydo not. Thus theyclaimthat we
oftensay thingsabout paintingswhichare not in facttrueofthe
paintings. Of course,theydo not claimthat none ofthe things
we say about paintingsare in facttrueofthe paintings,but only
that some of the.thingswe say about paintingsare not in fact
true. And what is more,it is the importarnt and interesting
thingswe say about paintingsthat are not in fact true. For
example,such statements- as 'The painting has great depth',
'The apples picturedare fullof solid volumes', etc.,are held to
to see one reason
be not in facttrue. Now it is not too difficult
why they are inclinedto say this. There are times when we

This content downloaded from 195.34.79.223 on Fri, 20 Jun 2014 18:06:46 PM

All use subject to JSTOR Terms and Conditions
 ART AND THE OBJECT OF ART 475
point to a paintingby Cezanne and say ' It has great depth'.
And thereare timeswhenwe pointto the same paintingand say
' It is flat'. Alexander,havingca.ughthold of thesetwo, quite
different, descriptionsis troubled how to p-utthem together.
The paintingcan'tbothhave greatdepthand be flat-this sounds
like a contradiction.Hence, he reasons, when we say the
paintinghas depth we are suffering froman illusion.
Now I want to say that a Cezanne paintingreallydoes have
great depth. We are not victimisedby an illusion,nor are we
suffering froman hallucination,whenwe say of the apples in a
Cezanne still-life,that one apple is in frontof the other. Nor
are we speakingof what seems to be the case instead of what
actually is the case. One apple is in frontof the other,the
paintingdoes have great depth. Thus there is nothingto be
explainedaway. But thereis a confusionwhichmustbe elimin-
ated. Let therebe no mistakeabout what is being said here.
Alexandersaysthatthe Madonnais flat,theHermesis notdivine
but onlyseemsso, the girlsin the Springare notin motion-they
too only seem so. Now I cannot say whetherthe Hermes
Alexanderspeaks of is divine or not,forI do not know which
statue he is referring to. But if, when he speaks of the girls
in the Spring,he is referring to Botticelli'sPrimavera,then,of
course,the girlsare in motion. Thereis no questionofit merely
seemingto be the case, it is the case. Furthermore, let it be
clearthatit is thepaintingI am talkingabout,notsome" illusory
object" or"imaginary object" or a "series of inteimitt ent
perceptions".
There are many ways of describinga painting. We can say
'The paintingis flat' and we can also say 'The paintinghas
great depth'. We can say 'It is a paintingof apples, all of
whichare about the same size, withsomeclose up and othersoff
in the distance'. We can also say 'It is a paintingofapples,all
of whichare different sizes, some large and placed towardsthe
bottomofthecanvas,and otherssmallat thetop ofthe canvas '.
(We can also describewhatwe see and speak ofhowthe painting
looks, but the use of this sort of descriptionis not germaneto
our problem.) Difficulties arise only if we suppose that all of
these descriptions have the same use, onlyif we confusethe use
of these variousdescriptions.
We can imaginea curioussortofgamethatmightbe playedhere
to clarifyour problem. Suppose we set ourselvesthe task of
describinga Cezannepainting,and we wishedto describeit in all
sortsofways. First,we mightsay the paintingis a flatpigment
coated strip of canvas mountedon wooden stretchers. If we

This content downloaded from 195.34.79.223 on Fri, 20 Jun 2014 18:06:46 PM

All use subject to JSTOR Terms and Conditions
476 PAUL ZIFF:

wereto go to a carpenterand ask himto builda crateforshipping

thepaintingwe mightsay to him' The crateneedn'tbe verywide;
it will be used to ship an ordinarypainting'. It would be im-
portantin thiscase forthe carpenterto realisethatthe painting
we wish to ship is made of a flat stripof canvas mountedon
wooden stretchers,for some paintingsare executed on gesso
panels, some on masonite,and some on plaster slabs, and so
forth. The widthof the crate requiredmightvary in different
cases. Secondly,we mightsay that paintingis a stripof linen
canvas coated with pigmentscontainingmanganese oxides,
iron hydroxides,and so forth. This would be importantand
relevantto a chemistengagedin cleaningthepainting. It might
also be of interestto someoneplanningto purchasethe painting
for he would probablywish to know whetherthe colours are
especiallyapt to fade. Thirdly,we mightsay it is a painting
of apples, all of whichare differentsizes,some large and placed
towardsthe bottomofthe canvas, and otherssmalltowardsthe
top. This wouldbe like describinga pictureof two people,one
closeup and the otheroffin the distance,as a pictureofone very
big personand anotherverytinyperson. Even this queer sort
of descriptionoccasionallyhas a use. For example,it would be
relevantto describ-e the pictureas consistingof large and small
apples to an art studentintenton copyingthe painting. E.g.
we mightsay to him ' You have drawnthe apple towardsthe
top of the canvas muchtoo large. The one in the paintingyou
are supposed to be copyingis a very small apple.' Fourtbly,
we mightsay it is a paintingof apples, all of whichare about
the same size, withsome close up and othersoffin the distance.
Thisis,perhaps,themostfamiliar kindofdescription ofa painting.
When we ask 'What is the paintingof ? ' this is one sort of
answerthat mightbe relevant. In short,it is what we usually
call a descriptionofthe subject-matter ofthe painting. Fifthly,
we might say it is a painting with strong two-dimensional
movementscontrastedwitha diagonalthree-dimensional move-
ment. Or we mightsay simply,the paintinghas great depth.
This is the sort of descriptiona person who criticises,judges,
evaluates,or simplyappreciates,the paintingwouldbe primarily
interestedin. There are of course still many other kinds of
descriptionsthat mightbe given, but these are sufficient for
our purposes.
Lest there be any misunderstanding of what has been said,
let me say explicitlythat in providingexamples of how these
variousdescriptions are sometimesused I do not,by any means,
wishto suggestthat theyare used onlyin the ways mentioned.

This content downloaded from 195.34.79.223 on Fri, 20 Jun 2014 18:06:46 PM

All use subject to JSTOR Terms and Conditions
 ART AND THE "OBJECT OF ART 477
For this is plainlyuntrue. Nor do I wish to suggestthat the
examplesgivenare to be taken as typicalinstancesof how these
various descriptionsare used. Perhaps they are typical in-
stancesand perhapstheyare not. I cannotsay, forI am not at
all surehowitcouldbe showneitherthattheyaretypicalinstances
or that they are not typical instances. And whetherthey
are or are not typicalinstancesis of no importancehere. For
all I wishto suggestis that at least one kindof situationwhere
we use such descriptions is of the kind we have just mentioned.
And ifthisis grantedthenthatis quitesufficient forthe purposes
of this discussion.
One way ofrelievingthe puzzleabout the apparent" conflict"
betweenthe firstand fifthdescriptionsis to realise that we
have a similar " conflict" betweenthe third and fourthdes-
c-riptions.In onebreathwe can speakoftwoapplesinthepicture
as being different sizes, one large and one small. In another
breathwe can also speak of the apples as beingthe same size,
one in frontof the other. We shall be in Alexander'sdilemma
ifwefeelthatonlyone ofthesedescriptions is thetruedescription.
If we feel inclinedto say ' How can both of these accountsbe
true ? Eitherthe two apples picturedare the same size or they
are not', the same puzzle arises. For what shall we say ?
Shall we say the two apples are the same size but they look
different? Or shall we say they are different sizes but they
look alike ? If we say theyare the same size, in what way do
theylook diflerent ? Only in that one is closeup and the other
is offin the distance. When we say they are the same in size
they also look the same size. If we say they are different in
size,in whatway do theylook alike ? For whenwe say theyare
diflerentsizes we can also say they look different sizes. All
that we shall have done,in introducing descriptionsof how the
apples look, is to have introducedanotherpair of apparently
"conflicting" descriptions. For now insteadofthe descriptions
'The apples are the same size ' and ' The apples are different
sizes' we have the descriptions'The apples look the same size'
and ' The apples look different sizes'. Thus we are not one jot
betteroffthan we were at the start. (And this is one way of
seeingthatthe use ofdescriptions ofwhatwe see and howthings
look is not germaneto our problem.)
What we must realise here is that these apparently" con-
flicting" descriptions, the third'withthe fourth,the firstwith
the fifth,do not in fact conflictwithone another. They could
conflictonly if they were used in the same ways, in the same
situationand forthe same purposes. This is not the case, and

This content downloaded from 195.34.79.223 on Fri, 20 Jun 2014 18:06:46 PM

All use subject to JSTOR Terms and Conditions
478 PAUL ZIFF:

thatit is not must-bequite clear. If we contrastthe description

'The paintingis flat'-when we use this descriptionto inform
a carpenterabout the widthof a crate requiredforshippinga
painting,withthe description' The paintinghas greatdepth'-
when we use this descriptionto informan art-criticabout the
structuireof a painting,we can see that these two descriptions
have nothingto do with one another. It would be absurd to
tell the carpenter'The paintinghas greatdepth' forhe would
supposea verywidecratewas required. Andit wouldbe equally
absurdto tell the art-critic' The paintingis flat' forhe wouldbe
astoundedt,ohear that a Cezanne still-lifehad no depth. He
mightwell reply'Are you sure? Cezannewas a masterin his
treatmentof space. Perhapsthe paintingis not an original? '
'If an art-critictellsus 'Many ofGauguin'spaintings,unlikethose
of Cezanne, are flat' are we to say the critic is confused?
Doesn't he knowthatall paintingsare flat? And ifwe told this
to the critiche would probablyreplyin a justifiablyirritable
tone that he had no time to waste on this nonsense. If we
didn't learn enoughabout art whenwe werechildrento under-
stand what he is talkingabout we had bettergo back to school
and learn some more beforewe attemptto philosophiseabout
art.
The pointis that the description'The paintingis flat' which
we use in speakingabout a paintingto an art-criticis totally
unlike the description'The paintingis flat' which we use in
speakingabout a paintingto a carpenter. If we do not take
the notiontoo seriouslyit is usefulhere to speak of a family
of descriptions; in this way we can speak of membersof the
carpenter'sfamily of descriptions,and membersof the art-
critic'sfamily,membersof the chemist'sfamily,and so forth.
Now it is veryimportantto realisethat quite oftenmembersof
the carpenter'sfamilylook verymuchlike membersof the art-
critic's family. We have already pointed out that the des-
criptions'The paintingis flat' and 'The paintinghas depth'
which belong to the critic'sfamilyhave their doubles which
belongto thecarpenter'sfamily. And notethatthe descriptions
'The paintinglooks flat' and 'The paintinglooks as if it has
depth' whichbelongto the carpenter'sfamilyalso have their
doubleswhichbelongto thecritic'sfamily. And thereare many
othersets of doubles; e.g. 'The paintingis top-heavy', 'The
paintingis fragile', and so fortli,all have theirdoubles. But
we ordinarily have no troublein tellingthemapart. The reason
forthisis thattheyrarelyassociate. Membersofthe carpenter's
familyhardly ever stray into an art galleryfor,if they do

This content downloaded from 195.34.79.223 on Fri, 20 Jun 2014 18:06:46 PM

All use subject to JSTOR Terms and Conditions
 ART AND THE OBJECT OF ART 479
they are usually thrownout by the criticwho findsthem dull
company. The only time a memberof the carpenter'sfamily
everlingersin an artgalleryis whenhe has beendraggedin bythe
heels by some philosopherto be mated with a memberof the
critic'sfamily. Is it any wonderthat such a union engenders
a contradiction ? But this is no excuse Lorthe philosopherto
claim that the membersof the critic'sfamilyare all bastards.
When we say 'The paintingis flat' and thenin anotherbreath
add 'but the paintinghas greatdepth' we are mixingup these
two, quite different, descriptionsin a horriblyconfusedmanner
so as to yield a dilemma. Thereis in fact no conflictbetween
these two descriptions ; thus thereis no need to tryto explain
away one in favourof the other.
Of course,in sayingthereis in fact no conflictbetweenthese
two desciiptionsI do not wishto suggestthat everytimesome-
one says 'The paintingis flatbut the paintinghas greatdepth'
the case thatin factthereis no conflict. For this
it is necessarily
wouldbe verymuchlikesayingthat it is necessarily thecase that
no one in factcouldcontradicthimself. Andthiscertainlyis not
true. People are not in the habit of deliberatelypronouncing
explicitcontradictions, but thisis notto say that theycould not
do so iftheychose. But thepresumption is thatifanyonedoes,
in the ordinaryrunof things,say 'The paintingis flat but it
has greatdepth' he is not deliberatelypronouncingan explicit
contradiction,but rather,is mixingup two, quite different,
descriptions,perhaps deliberately,so as to yield an apparent
paradox. In such a case there is in fact no conflictbetween
thesetwo descriptionsand, thus,thereis no needto attemptto
explainaway oneinfavouroftheother. Even moresignificantly,
thereis no need,and indeedit is a seriousmistake,to supposethat
there is some unique object correspondingto each different
description. It is an error,and an errorwhichhas vitiateda
good deal of recent esthetics,to postulate some illusory or
imaginaryobject to be the workof art. The ordinarypainting
hangingin the muiseum is the workof art,and not someillusion
or hallucination. Therearen'ttwothingsbeingreferred to when
we say, in the carpenter'sshop,'The paintingis flat', and when
we say, in the gallery,' The paintinghas greatdepth'. Thereis
just one, and it is the painting. Thereare twodescriptions, not
twoobjects.
In concludingthis paper I should like brieflyto indicateone
import-ant point that we have not mentioned. And although
wehave notmentionedit,thispointis quiteimportant forthereis
verygood reasonto believethat it is the sourcefromwhichall

This content downloaded from 195.34.79.223 on Fri, 20 Jun 2014 18:06:46 PM

All use subject to JSTOR Terms and Conditions
 480 PAUL ZIFF: ART AND THE " OBJECT OF ART"

theseotherpuzzlesstem. To see just whatthisis, it is essential

to noticethat Alexander,along withotherestheticians, does not
feel inclined to explain away the description'The painting
is flat' but he does wishto explainaway the description'The
paintinghas great depth'. This is quite illuminating, forthe
factthat he does wishto explainaway 'TThepaintinghas great
depth' but does not wish to explainaway ' The paintingis flat'
showsthat he is troubledby the one description in a way that he
is not by the other. Thus ifwe wishto get at the sourceofthis
difficulty,it is not sufficient to indicate,as we have done, that
these two apparentlyconflictingdescriptionsdo not in fact
conflictand thattheyhave totallydifferent uses. What we must
do is eliminatethe difficulty estheticiansfindin connectionwith
our use of the description' The paintinghas great depth'.
And this difficulty is, of course, how to settle disputes over
whetheror no a paintingdoes have greatdepth,howto verifythe
statement'The paintinghas greatdepth', and so forth. Thus
I am sayingthat estheticianswho tryto explain away the des-
cription'The paintinghas greatdepth' do so because theyare
inclinedto believe that a disputeover the depth of a painting
cannot be resolved. That is, they are inclinedto say that in
such a situationthereis nothingthat can be done to settlethe
issue,that we cannotestablisheitherthatthepaintingdoes have
depth or that it does not have depth. Thus, instead of saying
'The paintinghas great depth' they feel we should say, if we
wishto speakproperly, ' The paintingseemsto have greatdepth'.
Thusthefundamental questionis: can weverify thatthepainting
has great depth? And the answeris that, of course,we can.
However, I do not propose to argue this point here. The
questionhas beenmentionedonlyto indicatethatit is a problem
which must be dealt with beforethat ghost of esthetics,the
mysteriousestheticobject,can finallybe laid to rest.
CornellUniversity,
Ithaca,N. Y.

This content downloaded from 195.34.79.223 on Fri, 20 Jun 2014 18:06:46 PM

All use subject to JSTOR Terms and Conditions