The Language of Art History

Author(s): Michael Baxandall

Reviewed work(s):
Source: New Literary History, Vol. 10, No. 3, Anniversary Issue: I (Spring, 1979), pp. 453-465
Published by: The Johns Hopkins University Press
Stable URL: http://www.jstor.org/stable/468922 .
Accessed: 19/11/2012 09:45
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].

The Johns Hopkins University Press is collaborating with JSTOR to digitize, preserve and extend access to
New Literary History.

http://www.jstor.org

This content downloaded by the authorized user from 192.168.52.63 on Mon, 19 Nov 2012 09:45:56 AM
All use subject to JSTOR Terms and Conditions

The Language of Art History

Michael Baxandall

I. Dialogue Declined
WEIRD THING about the last ten yearshas been quite how many

art historians have been beating their breasts about the

"theoretical inadequacies" of the activity,and New Literary
Historyhas admirablyregisteredthat thudding, with a more representativespread of opinion than any of the art historians'own journals I see. To take threetypes:Kurt W. Forster,1who representsa line
found in a rather fuller and sharper form in the journal Kritische
Berichte,
deplores our formalism,our assimilationof art historyto the
of
history ideas, our breathless affirmativenessabout the works we
study,our concentrationon high art at the expense of genres like the
filmand the poster, our lack of self-awarenessabout our own preconceptions and their social roots, our failure to develop a genuine
social-historicalapproach: "The only means of gaining an adequate
grasp of old artifactslies in the dual critique of the ideology which
sustained theirproductionand use, and of the currentculturalinterests that have turned works of art into a highlyprivileged class of
consumer and didactic goods." James S. Ackerman,2by contrast,sees
the root of our trouble in a hybrid philosophical base: "Without
knowingit, my colleagues have grounded theirmethod in the tradition of nineteenth-century
positivismconceived to justify scientific
empiricism."But then we have absurdly taken into this an unconscious value systeminherited from the Neoplatonic idealism of the
Renaissance. No wonder,then,if we are tornbetween formand content,the social and the aesthetic,historyand criticism.What we need
to do is to "replace the present irrationalcollage of traditionsthat
constituteour basic value premiseswithconsciouslyarticulatedprinciples that correspond to what we actually believe." We should
evaluate art, and in the light of something called "the concept of
humane values," preliminarily described. David Rosand3 offers
moderate recommendationsin a line running immediatelyfrom an
influential article by Leo Steinberg called "Objectivity and the
Shrinking Self,"4 which worries about us compromising our individual selves in the attemptto see other men's or periods' worksfrom
Copyright? 1979 by New Literary
History,The Universityof Virginia

This content downloaded by the authorized user from 192.168.52.63 on Mon, 19 Nov 2012 09:45:56 AM
All use subject to JSTOR Terms and Conditions

454

NEW LITERARY

HISTORY

theirpoint of view. Quite apart fromthe danger of doing our Selves

harm, a pretensionto historicalobjectivityis liable to shrivelthe faculty through which worthwhileperception of art happens. "Until
[Rosand says] art historiansrecognize their responsibilitiesas critics
our inquiryintothe art of the past willremain incompleteand thatart
will be only partiallyaccessible to us."
I had betteradmit at once that I cannot get along withthissort of
thingand have no intentionofjoining the discussion.For one thing,I
have not much confidence in conclusions drawn from serial
generalizationat the level I and mostart historiansseem equipped to
practiceit: one may as well be blunt about that.Then, I do not at all
like the tone of the debate, which seems oddly hortatory and
peremptory:I dislike being admonished. On the other hand, what I
do like is there being a manifold pluralityof differingart histories,
and when some art historiansstarttellingother art historianswhat to
do, and particularlywhattheyare to be interestedin, myinstinctis to
scuttle away and existentially measure a plinth or reattribute a
statuette. It seems to me there must be some misunderstanding
among art historiansabout what "theory"is.
Then again, the discussion makes me wonder whetherwe are not
being too grand, disablinglyso. By origin we are rather lightweight
people. The literarycritichas ancientrootsin the lectureroom and in
the commentaryand disputationthatwere the lectureroom's genres:
thatsets him his own problems,I would guess, but it is imposingin its
way. We do not have thisbackground nor thissortof long-established
cultural function,but we do have a good natural vulgar streak. In
everygroup of travelers,everybunch of touristsin a bus, there is at
least one man who insistson pointingout to the othersthe beauty or
interestof the thingstheyencounter,eventhoughtheotherscan see the
too:we are thatman, I am afraid,aufond. Of course,otherroles
things,
have attachedthemselvesto thisbasic one, augmentingthe man in the
bus-the rhetoricaldescriber,the paid cicerone, the friendsdiscussing objects in a portfolioor cabinet,a littleof the antiquarycollector
and archaeologist,even a touch of the historian,and some others,
too-but the sum is modestand stillsociallyambiguous as to role. The
academicizing-upof the activityis a quite recentthing,and it is a pity
if it goes to our heads. In particularit is a great pityif it leads us to
confuse subjectswithsyllabi: I suspect thatbecause we are nowadays
offeringourselves as a liberal education in the lecturerooms, we are
temptedto strikeuntenablygrandiose attitudesand then bleat if the
subject does not comfortablyprop us up. For myself,I would prefer
to remain the augmented man in the bus who-if he can stop talking
long enough to have a reflectivemoment-must wonder firstat his

This content downloaded by the authorized user from 192.168.52.63 on Mon, 19 Nov 2012 09:45:56 AM
All use subject to JSTOR Terms and Conditions

THE LANGUAGE

OF ART HISTORY

455

own nerve in verbalizingat other people about objects they can already see. I do worryabout that.
And thisis reallythe main reason formynot being able to enterthe
current dialogue: the problems that are apparently tormentingmy
colleagues do not seem to be the problems I meet. The issues I most
worryabout in art history-a term I use interchangeablywith art
criticism-fallintotwo main groups. One group is connected withthe
prettygratuitousact of matchinglanguage withthe visual interestof
works of art; that is our staple. The other group is connected with
how one can and cannot state relationshipsbetween the characterof
worksof art and theirhistoricalcircumstances,but I shall hardlyget
around to these here. In fact,the remarksin the issue of New Literary
Historydevoted to "Literaryand Art History"thatchimed mostclosely
with my interestscame not from the art historiansbut fromJohn
Passmore's Commentary.5One of the thingshe said was: "it is very
difficultto say a great deal about a painting,except by talkingabout
its relationshipsto somethingelse, whetherto other paintings,other
arts,contemporarysocial movements,contemporarybeliefs,or contemporaryideas," and what followsstartsfroman attemptto gloss the
firstpart of this fromthe practicallevel.

II. Limitationsof the Lexicon

The specificinterestof the visual arts is visual, I take it,and one of
the art historian'sspecific facultiesis to find words to indicate the
character of shapes, colors, and organizations of them. But these
words are not so much descriptiveas demonstrative-I am not sure
how firmlywe have grasped the implicationsof this. Unlike a travel
writeror the man who writesabout exhibitionsin a newspaper,we are
not primarilyconcerned to evoke the visual character of something
never seen by our audience. 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 marginallyas a visualization derived from
knowledgeof otherobjects of the same class, and though the formof
our language maybe informative-"thereis a flowof movementfrom
the lefttowardsthe center"-its action is likelyto be a sort of verbal
pointing.What distinguishesit frommanual pointingis mainlythat
along withdirection("leftto center")goes a categoryof visual interest
("flowof movement").We are proposing thatour audience compare
the one withthe other.
It is this that goes some way towards extenuating the terrible
crudeness of our language. If I apply half-a-dozensimple termsof

This content downloaded by the authorized user from 192.168.52.63 on Mon, 19 Nov 2012 09:45:56 AM
All use subject to JSTOR Terms and Conditions

456

NEW LITERARY

HISTORY

visual interest(a phrase I am not going to define) to the pencil I am

writingwith-"long," "thin,""shiny,""green,""of hexagonal section,"
"with one conical end"-that is a quite inadequate description: to
someone who did not have experience of pencilsit would not carryan
accurate image, and equally to someone who did have such experience some of the termswould be otiose. But if my purpose is not to
describe but ratherto indicate(a) to someone who has seen it (b) such
kinds of visual interestas I am findingin itjust now, then the halfdozen termsdo cover about a thirdof what I have to offer.My blunt
words (e.g. "green") are sharpened forme because what I have done
is to instigate,or offerto instigate,a guided act of inspectionof the
particularobject by the hearer, and he knows reallythatthatwas my
intention.Neitherof us expects him to think,if he does elect to follow
my prompting,"Oh, not red then": rather, he will elaborate and
refinemycategory"green" for himself.Of course the matteris more
complicated than this,but the immediate point is that the art historian's use of language invitesthe receiverto supply a degree of precision to broad categories by a reciprocal referencebetween the word
and the available object. It is ostensive.
But my pencil is an untypicallysimple object, which is why I could
cover so much of itsvisual interestwithso fewwords. If I tryto do the
same even to my typewriter
("square," "mat,""gray,"and so on), I get
less far: the words cover less of what I findinterestingin it. If I tryto
do it witha paintingor a sculpture,I will hardlyget anywhereat all:
directdescriptivetermscan cover verylittleof the interestone wishes
to indicate. I can use them-it is not vacuous to point to
Michelangelo'sMoses as "square"-but the fitbetween sense and reference is now becoming very loose, and I can only use them by assuming that my hearer will interpretthem in a sophisticated and
specialized way: he must supply a great deal in the way of mental
comparison withother worksof art,of experience of the previous use
of such words in art criticism,6
tact.The
and of general interpretative
words have become thingsof a ratherdifferentkind.
Indeed, if one is not careful, the lack of the right,or adequately
determinate,word reduces one to someone just making a schwarmerischnoise; itbecomes quite unclear whyone should be takingiton
oneself to address other people about the picture at all. A thing the
practiceof art criticismquicklyteaches one is that the European languages discriminateveryfinelyin some areas (e.g., underlyingEuclidean form) and very coarsely in others (e.g., seen surface texture):
this has its own fascinationas an object of study,but it also sets a
practical problem because there is a limitto how much one can enlarge the lexicon by coiningand borrowing.It is not so much thatone

This content downloaded by the authorized user from 192.168.52.63 on Mon, 19 Nov 2012 09:45:56 AM
All use subject to JSTOR Terms and Conditions

THE LANGUAGE

OF ART HISTORY

457

wants to avoid generating academic gobbledygook as that novel

coinages and loanwordsare culturalorphans, not properlypartof the
collectiveframeworkof our thinking.Thus, I would verymuch like to
have genuine access to the Nigerian Yoruba criticaltermdiddn,7which
indicatesa degree of smoothbut not glossyluminosityin the surfaces
of sculpture, closely related to the contrast of these with sharp
shadows and edges: it would cover much of an interestI find importantin some German wood carvingsI study.But didonis a fragmentof
a complex of Yoruba criticalconceptsand takes itsrichmeaning from
just this set of relations.Even for my privateexploratorypurposes I
cannot possess it except in a crude and shallow,a dissociated way: to
go in forsuch inkhornismsheavilyin public would be intolerableand
sapping. Besides, such energy as I and my hearers can spare for
enteringinto alien criticalconcepts a littleI like to save forthe critical
conceptsof the culture I am studying:I feel entitledto a fewof these,
for reasons I shall not be discussinghere.

III. Three Kinds of Indirectness

But in factmostart-critical
language is not of such directdescriptive
background as "green" or "square"; rather,it is variouslyoblique or
tropical. And while there seems nothing to be said for workingout
any very crisp or general classificationof the types of indirectartcriticalwords, it will suit my purpose here to group them in three
rough divisionsor moods.
(I) Some words seem to point to a kind of visual interestby making
a comparison of some sort,often by metaphor: "rhythmic,""fugal,"
"dovetailing,""a forestof verticals,""striplike"-these words used of a
picture work comparatively.Among them I will also include words
like "square" in the extended use involved in calling Michelangelo's
Moses "square": thus, "Apollo and two of the Muses... forminga
broad triangle."And a special class of comparativewords (I. bis,let us
say) referto representationalworksof art as if the thingsor persons
represented were actual: "agitated" figures or "calm" or "spirited"
figures.(II) Some words characterizethe work of art in termsof the
action or agent that would have produced them: "tentative,""calculated," "sensitive," "elaborate," "difficult,""skilled," this or that
"treatment" or "development" or "virtuosity."(III) Some words
characterizea work of art by describingits action on the beholder or
his reaction to it: "imposing,""unexpected," "striking,""disturbing,"
"unpleasant,"thisor that"effect,""a feelingof crowding."One could
referloosely to these moods as (I) comparativeor metaphorical,(II)

This content downloaded by the authorized user from 192.168.52.63 on Mon, 19 Nov 2012 09:45:56 AM
All use subject to JSTOR Terms and Conditions

458

NEW LITERARY

HISTORY

causal or inferential,and (III) subject or ego words, and mightvisualize them in a field like this:
Similia
(I)

The maker
(II)

= [The object]

The beholder
(III)

Matterof representation
(I. bis)
But of course they are all projections of the subject, the speaking
beholder,as we all know perfectlywell. Equally theyare nearlyall in a
weak sense metaphorical,though some of the metaphors are more
educated than others.
There is much thatcould be said in a softeningway about this,if I
did not want to keep the types broad. Clearly a historyof use will
loosen the relationof a word to itsoriginal basis: "monumental,"say,
is a moribund metaphor that has leftmonumentssome way behind,
and it would be foolish to make a thing of "interesting"being an
ego-word. Clearly,too, many words partake of more than one type:
"dry," for instance,can be used in comparative, I. bis, causal (secco
rather thanfrescohandling), and subjectiveways, sometimesequivocally,and is a trickyword all round. It is also clear that roughlythe
same general area can often be pointed to with differenttypes of
words: say, (I) stormy,(I. bis) excited,(II) excited,(III) exciting.(The
between I. bisand
example, by the way,alertsus to the verbal affinity
II, which has much to do withour vulnerabilityto the "physiognomic
fallacy"8or Winckelmannsyndrome.)Above all, thereis the pointthat
in any piece of actual art criticismall thisis going on on several tiers.
My examples were mainly single words, but sentences are framed
withinone type or another, and paragraphs and books are weighted
overall towardsone or another: I am happy to classifywithmyclasses
at any of these levels. All the examples in the last paragraph were
taken from Heinrich Wolfflin'saccount in Classic Art of Raphael's
Camera della Segnatura. If anyone looks at those pages he willfind,I
think,that theircharacter is determinedby an overall dominance of

This content downloaded by the authorized user from 192.168.52.63 on Mon, 19 Nov 2012 09:45:56 AM
All use subject to JSTOR Terms and Conditions

THE LANGUAGE

OF ART HISTORY

459

types I. bis and II. Withinthis general characterall the kinds of language I have mentioned, including what I have rather simplycalled
"direct" language ("round," "large in proportion," "surrounded,"
"profile"),are in play. It is the patternof thishierarchythatgives the
individualcritica physiognomy.It is a traitof Wolfflin's,forinstance,
that withina sentence of Type III, reportingan impression,there is
often a Type II word as core: he tends to have an impression of a
cause, honest man. I am not sensitive, I should say here, to the
suggestionthat the differencesin words are purely formaland that
somewhere between sense and referencetheir origins are sloughed
off,words becoming denatured from their class once they are presented withincontinuousdiscourse. When reading art criticism,I do
not findthisto be so. On the contrary,I am pleasurablyconscious of
the constantlyveeringorientationsin the good critic'sdance towardsa
determinatedemonstrativeact. But what does strikeme is
sufficiently
that his need to stringhis words into discourse raises a problem of
another kind, which I have not space to discuss but wish at least to
state.

IV. The Problem of Linearity:

Words about Words and Words about Shapes
The art-criticallexicon is normallyassembled into consecutivelanguage of some sort. (Notionally,I suppose, one could assemble single
categories of visual interest,presyntacticalejaculations, in a nonsequential, galactic pattern on the page, but this would be affected.)
This raises problems that I can best accent here by pointing to the
contrastwithliterarycriticism.Literarycriticismis words about words
where art criticism,as has often been pointed out, is words about
shapes. Many differences-the dissimilaritiesbetween art criticism
and literary criticism seem much more interesting than the
similarities-follow from this, but the one I want to point to now
comes out of the shape of language, its dependence on syntagmatic
muscle, the factthat words have to be assembled in a linear progression.
A piece of literature,being language, is itselfa linear affairled from
here to there,or fromnow to later. A poem or storyhas a beginning
and an end and an authenticsequence in between. We may perceive
many nonlinear patternsunderlyingeither a sentence or the whole,
antitheticalsyntaxor narrativesymmetries;there are also likelyto be
many retracingmoments of rereading and referringback. But the
linear progress of the text is comprehended in these excursions and

This content downloaded by the authorized user from 192.168.52.63 on Mon, 19 Nov 2012 09:45:56 AM
All use subject to JSTOR Terms and Conditions

460

NEW LITERARY HISTORY

withstandsthem. If a critic'saccount of Wuthering

Heightsor Sarrasine
involves him in pointing to bits of it out of order, this is all right
because the directionalmovementof the book is strongenough for
his activitynot to be misunderstood.He is emerging here and there
fromthe stream,walkingback along the bank, and gettingin again to
floatalertlydown a particularstretchonce more. When the literary
criticdoes engage witha particularstretchof a text,his language can
pace its language, each linearly progressive. It is irritatingthat my
point is weakened here by the failureof manyliterarycriticsto make
athleticuse of theiradvantage, no faultof mine, but the possibilityis
there and is used in the literarycriticismI most envy fromover the
fence-to offer a hostage, Empson on Donne's "A Valediction of
Weeping." And in any case I thinkthe point is not so much that the
literarycriticcan workin parallel withhis textas thatthe textand our
reception of it have a robust syntagmaticprogression of their own
whichthe linear sequence of an expositioncannot greatlyharm. The
language of the descriptivecriticcan run with,run away and back,
run round the firmlyprogressinglanguage of the text,like an active
dog on a walk witha man.
A pictureon the other hand, or our perceptionof it, has no such
inherentprogressionto withstandthe sequence of language applied
to it. An extended description of a painting is committed by the
structureof language to be a progressiveviolationof the patternof
perceivinga painting.We do not see linearly.We perceivea pictureby
a sequence of scanning,but withinthe firstsecond or so of thisscanning we have an impression of the whole-that it is a Mother and
Child sittingin a hall, say,or a sortof geometricizedguitaron a table.
What followsis the sharpeningof detail, notingof relationships,perception of orders, and so on. And though the sequence of our scanning is influencedas to patternby both general scanning habitsand
particularcues in the picture,it is not comparable in regularityand
controlwithprogressthrougha piece of language. One consequence
of this is that no consecutive piece of verbal ostension, linear language, can matchthe pace and gait of seeing a pictureas itcan match
the pace of a text: the read text is majesticallyprogressive,the perception of a picturea rapid irregulardartingabout and around on a
field.There are various waysof meetingthe problem. One can work
the ostensivenessof one's language hard, so as to draw the hearer
into his own active act of perception for his attentionto
sufficiently
shiftaway fromone's own. One can also shun expositorysequences
thatlook like representationsof perceiving,e.g., descriptions,in favor
of ones thatassimilatethemselvesto thinking.The historyof art historyoffersmany other techniques,too.

This content downloaded by the authorized user from 192.168.52.63 on Mon, 19 Nov 2012 09:45:56 AM
All use subject to JSTOR Terms and Conditions

THE LANGUAGE

OF ART HISTORY

461

V. InferentialCriticism
I thinkI have been makingthreekindsof suggestion:first,thatthe
art-criticallexicon is stronglyostensive; second, that art-criticallanguage is largely and variouslyoblique, and at more than one level;
third,that the linear formof our discourse is curiouslyat odds with
the formof itsobject,whetherthisis considered to be the workof art
itselfor our experience of it. These seem to me basic facts of artcriticallife,and one would like to come to some sort of constructive
termswith them. Four hundred years of terriblygood and verydiverse European art criticismcertainlysuggest that there are ways of
doing so. It seems characteristicof the best art criticsthat they have
developed theirown waysof meetingthe basic absurdityof verbalizing about pictures: they have embraced its ostensive and oblique
characterpositively,as it were, as well as bouncing theirdiscourse out
of the pseudodescriptiveregisterthatcarriesthe worstlinear threat.I
repeat that they have done this in many differentways; about all
Vasari and Baudelaire have in common is conspicuous success. This
really seems somethingto insiston in the present climate of discussion: the linguisticfacts of our life may be general and pressing,
preliminaryconditionsone may well wantto take account of in working out a way of doing whateverit is one wantsto do, but theydo not
direct us to one kind of art history.
For instance, I am anxious not to suggest that there is a simple
affinitybetween the orientationof a critic'sovert interestand the
orientationof a mood of language-between, say,those of us who like
occupyingourselves withthe circumstancesin whichworksof art are
made, on the one hand, and inferentiallanguage on the other. What
worriesme about much criticismthat offersitselfas social-historical
analysisof art, including several of the people praised by Kurt Forster,is preciselyan un-self-awareType III qualityat the lowestverbal
level marshaled at a higherlevel in large a priori Type II patternssoftimpressionssloshingabout in hard causal schedules. For contrast
one can read the early books of Adrian Stokes9for local inferential
muscle, however subject-assertivethe total manner and effect.But,
for reasons that are only partlyverbal, it is particularlythe role of
inferentiallanguage I am curious about, and if I had not already used
up most of the five thousand words I was asked to write,it is this I
would now be going on to discuss,the strengthsI thinkitconfersand
the problems I am sure it sets. I have enough words leftto assertone
of the strengths,as a sort of summarythrowingdown of a gage.
Words inferentialas to cause are the main vehicleof demonstrative
precisionin artcriticism.They are activein two distinctsenses. Where

This content downloaded by the authorized user from 192.168.52.63 on Mon, 19 Nov 2012 09:45:56 AM
All use subject to JSTOR Terms and Conditions

462

NEW LITERARY

HISTORY

ego-words are formallyand often substantiallypassive, reporting

somethingdone by the work of art to the speaker as patient,causal
words deal in inferred actions and agents. At the same time they
involve the speaker in the activityof inferringand the hearer in the
activityof reconstructingand assessingthe patternof implication.For
my taste, I will say, all this activityis cheerful and absolutely more
wholesome than a lot of comparing of impressions,howeverhumane
or unshrunken,but the real point is that it seems to yield adequately
determinate and properly stimulatingostensive words. One of the
details mydescriptionof the pencil on p. 456 omittedwas the sort of
scalloped edge of the green paint at the point where it meets the
conical end. If I wanted to, I could registerthis quite sharply and
economicallyby inferringcause-the blade of a sharpener revolving
circularlyat an angle of 15? to a hexagonal cylinder.I do not thinkI
could registerit withego language at all: my Self is too uncertaina
qualityto myhearers foritsreactionto a scalloped edge to registerthe
scalloped edge or itsvisual interest-unless itsshare is indeed to infer
the revolvingblade. In a more complex way the same is true of art
criticism,where a mature inferentialvocabularyin fullplay can have
formidable demonstrative precision and punch. The eighteenthcentury critic Shen Tsung-hsien?1-to dramatize the matter with
somethingexotic-gives a glimpse of the resources classical Chinese
criticismhad for inferentialcharacterizationof the painter's brush
marks: among much else he distinguishesbetween wrist-dominant
and finger-dominantstrokes; between dead and live strokes,in the
sense that there is variation of power withinthe single live stroke;
between dragged marks and slipperymarks, splashed-inkones and
broken-inkones, betweenthe marksof a straightbrush and those of a
slantingone, betweencuttingstrokesand led strokes;he can speak of
an individual brushstrokehaving a center or core and opening and
closing phases, and he could wonder how far the closing phase of a
strokecarries the suggestionof furtherdevelopment; he could even
characterizea brush markbythe noise the strokewould have made, as
a "sousing" noise. Of course, there are reasons for the activenessof
this language: both Shen Tsung-hsien and his readers were themselves active users of the calligraphicbrush so that there was a firm
background of referencein everyone'sexperience. But stillit is enviable language: to findanythingcomparable in Europe, one mustgo to
thingslike Delacroix's occasional remarksin hisjournals on the technique of Rubens-remarks addressed by a painter to a painter. We
cannot compete with it in this area, but there are other areas of inference we can work towards, including-to twistJohn Passmore's
remarka little-"relationships.. . to other paintings,other arts,con-

This content downloaded by the authorized user from 192.168.52.63 on Mon, 19 Nov 2012 09:45:56 AM
All use subject to JSTOR Terms and Conditions

THE LANGUAGE

OF ART HISTORY

463

temporarysocial movements,contemporarybeliefs,or contemporary

ideas."

VI. Issues
I wanted to get thisfarbecause it may suggestmore clearlywhythe
dialogue declined at the startis one I cannot engage with,a matterof
embarrassment obviously. I have been suggesting that making
inferences-as well as making comparisons and talking about
oneself-is an inherentpart of art-criticaldemonstration,and in the
last paragraph I pointed to one of the reasons why I consider that
language inferentialabout cause is very importantto art criticism.
Now this means, for one thing,that I cannot naturallyaddress the
dialogue's typicalissue of Historyand Criticism.If one values what I
have been calling inferentialcriticism,critical "tact" and historical
"grasp" appear as verymuch the same thing. Inferringcauses I take
to entailbeing historical:equally one cannot conceive of eitherhistory
or inferencebeing accurate withoutcriticalacuteness. Clearly history
and criticismare differentinflectionsof attention-inquiryas against
judgment, then as against now, how as againstwhat,and so on-but I
have no purpose in drawing a line between them, and without a
purpose it is hard to know where the line is to be drawn. I accept that
others may have such a purpose.
For another thing,it means that I am insensitiveto the admonishmentsof the humane-value and unshrunken-Selfpeople. Inferential
criticismentails the imaginativereconstructionof causes, particularly
voluntary causes or intentions within situations. It is repetitively
pointed out and is clearly true that we cannot fullyreconstructand
interiorizethe habitsof thoughtand language of a past culture: there
is no possibilityof recreatingthe culturalcomponent in the medium
artist'sintentionor beholder's perception
of, say, a sixteenth-century
because we cannot make ourselvesintosixteenth-century
men, even if
we wanted to, not least because we cannot shed our own cultural
habitsand values. This seems so obvious, it is hard to understandwhy
it is stillstated so oftenand withsuch an air of discovery.But to see it
as an argument against exerting
oneselftowardsreconstructingan old
artist's intention and its medium1"-"the foredoomed effort of
positiviststo interpret past art 'in its own terms"' (Ackerman), I
suppose-seems odd. One mightas well dissuade a man fromtraining
to run by pointingout thathe willnever run his distancein no timeat
all. Just as we all ambulate, we all infercauses and intentions:it is a
disposition much too deep and diffused in us to be excised, as our

This content downloaded by the authorized user from 192.168.52.63 on Mon, 19 Nov 2012 09:45:56 AM
All use subject to JSTOR Terms and Conditions

464

NEW LITERARY HISTORY

language (one has been insisting)declares. What we are going to do

anyway,one could say, we reallyare entitledto enjoy tryingto do as
well as we can, while well aware we cannot do it completely.Then, to
make the cognate objection that by seeking "objectivity"-and sensitive critical inference does demand that we seek something like
this-we are starvingthe Self, somehow denying it full humane expansion or perhaps making it cognitivelyderacine, seems to me to
involve a quite dispiritingnotion of Self, something too fragile or
weak to indulge freelyand deliberatelyits curiosityabout the How
and Why of what and whom it meets: I cannot persuade myselfmy
Self is that frail. Inferentialcriticism,one could say, is active selfassertion. By this sort of remark I do not think to dispose of the
positivepracticeof self-elaboratingcritics,whichis one natural move
from the linguisticbase I have been tryingto sketch,but simplyto
declare that if theywant to warn me off my cause-elaboratinginterests, theyneed a ratherdifferentsort of argument.As it stands,the
issue seems to me a nonissue.
Though I have been makingratheran elaborate pointof havingno
vocationor statusforurging people to courses, some thingsdo interest me more than other things. For reasons I have been impolitely
open about, I do not thinkart historianshave been at theirinteresting
best recentlywhen talkingmethod. If historiansor literarycriticsor
anthropologistsasked me where the best methodologicalaction is to
be found in English-language art historynowadays-they never do
ask me that: theyask where the good art historyis-I would have to
point to two areas: on the one hand such implicitlyreflectivepractitionersof actual art historyas David Summers,12to name one among
several,and on the other hand such writersof authenticaestheticsas
Richard Wollheim.In the lastten yearsI have not enjoyed the ground
in between. Yet clearlyart historiansmustthinkand talk about what
theydo, and here mytasteis verymuch forpeople disposed to discuss
quite modestlythe specificproblemsof art criticism,in detail and on
the technicallevel. I have referredto two kinds of such problemswhat happens when one matches words or historicalcircumstances
withthe visual interestof worksof art-but there are others ripe for
airing: the notoriouslyheterogeneousrange of relationshipswe lump
together under the heading of artistic influence, or the general
tyrannyof art history'sdiachronic thrust(there is a limitedbut real
case for sometimeswritingart historybackwards),or our muddled
notion of the medium, and more. I do not mean we should look at
nothing but ourselves: I am almost as aware as the methodological
men that we can learn fromother historicaldisciplines,literarycriticism,or anthropology,or indeed fromthe philosophyof history,but

This content downloaded by the authorized user from 192.168.52.63 on Mon, 19 Nov 2012 09:45:56 AM
All use subject to JSTOR Terms and Conditions

465

THE LANGUAGE OF ART HISTORY

it would be good to get art history'speculiarityjust clear enough to

know roughly what sort of activityone is projecting the lessons
learned fromthem in or on to.
UNIVERSITY OF LONDON

NOTES
3 (Spring
of Values?" NewLiterary
1 "CriticalHistoryof Art,or Transfiguration
History,
1972), 459-70.
2 "Toward a New Social Theory of Art," New LiteraryHistory,4 (Winter 1973),
315-30.
3 "Art Historyand Criticism:The Past as Present,"New LiteraryHistory,5 (Spring
1974), 435-45.
4 Daedalus, 98 (Summer 1969), 824-36.
5 "Historyof Art and Historyof Literature:A Commentary,"New Literary
History,3
(Spring 1972), 575-87.
6 "Square" has a rather splendid history,in fact; its use in Greek and Latin art
criticismhas been investigatedin an almost overingenious but exhilaratingpaper by
Silvio Ferri,"Nuovi contributiesegeticial 'Canone' della sculturagreca," Rivistadel R.
e Storiadell'Arte,7 (1940), 117-39.
Istituto
d'Archaeologia
7 For didonand itscontext,Robert FarrisThompson, "Yoruba ArtisticCriticism,"in
The TraditionalArtistin AfricanSocieties,ed. Warren L. d'Azevedo (Bloomington, Ind.,
1973), esp. pp. 37-42.
on a HobbyHorse
8 For which,E. H. Gombrich,"Art and Scholarship,"in Meditations
(London, 1963), p. 108, coining the term; and also "On PhysiognomicPerception,"
ibid.,p. 51.
9 ParticularlyThe QuattroCento(London, 1932) and StonesofRimini(London, 1934).
The remarkablecomparisonbetweencarvingand modeling"conception"in the latteris
ofAdrianStokes,ed.
included in the Pelican edition of TheImagein Form:SelectedWritings
Richard Wollheim(London, 1972), pp. 147-83. The kind of quality I have in mind is,
froman account of Donatello's Dead ChristwithAngelsin the Victoria& AlbertMuseum
(Wollheim,p. 168): "To Donatello, changes of surfacemeantlittlemore than lightand
of plasticorganization.The bottom of the angels'
shade, chiaroscuro, the instruments
robes is gouged and undercutso as to providea contrastto the open planes of Christ's
nude torso.The layersof the stone are treatedwholesale. Though some of the cuttingis
manner of approach.In brief
beautifulin itself,the reliefbetraysa wilful,preconceived,
of adjoining surfaces,
the compositionis not so muchfoundeduponthe interrelationship
as upon the broader principlesof chiaroscuro" (my italics).
10 There are translatedexcerpts-all I know of the author-in Osvald Siren, The
Chineseon theArtofPainting(Peking, 1936), pp. 224-33, and Lin Yutang, The Chinese
TheoryofArt(London, 1967), pp. 169-219.
11 An analogous stance is better and more fullydescribed, as "actor-oriented,"by
CliffordGeertz,The Interpretation
of Cultures(New York, 1973), pp. 13-16.
12 Two of whose articlesare brilliantlysustained examples of inferentialcriticism:
"Maniera and Movement: The Figura Serpentinata,"The Art Quarterly,35 (1972),
269-301; and "Figurecomefratelli: A Transformationof Symmetryin Renaissance
n.s., 1, No. 1 (1977), 59-88.
Painting,"The ArtQuarterly,

This content downloaded by the authorized user from 192.168.52.63 on Mon, 19 Nov 2012 09:45:56 AM
All use subject to JSTOR Terms and Conditions