Sunteți pe pagina 1din 17

Oceania Publications, University of Sydney

A Critique of Alfred Gell on "Art and Agency" [Corrected title: A Critique of Alfred Gell on
Art and Agency]
Art and Agency: An Anthropological Theory by Alfred Gell
Review by: Ross Bowden
Oceania, Vol. 74, No. 4 (Jun., 2004), pp. 309-324
Published by: Wiley on behalf of Oceania Publications, University of Sydney
Stable URL: http://www.jstor.org/stable/40332070 .
Accessed: 01/09/2014 21:29

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 support@jstor.org.

Wiley and Oceania Publications, University of Sydney are collaborating with JSTOR to digitize, preserve and
extend access to Oceania.

http://www.jstor.org

This content downloaded from 146.96.128.36 on Mon, 1 Sep 2014 21:29:12 PM


All use subject to JSTOR Terms and Conditions
A CritiqueofAlfredGell on Artand Agency

Ross Bowden
La TrobeUniversity

ABSTRACT
Thisreviewarticleexaminesin detailtheargument in AlfredGell's posthumously
published
bookArtandAgency:an anthropological The reviewis dividedintotwoparts.In the
theory.
firstI summarisethemainargument of each chapterin turnand commenton someof the
author'smorequestionable
assumptions andconclusions. In thesecondI stepbackfromthe
individual andcomment
chapters on threeofthemoregeneralissuesthebookraises.

AlfredGell's posthumously publishedbook Artand Agency:an Anthropological Theory


(1998) has achieved something ofa cultstatusin modernanthropology. It is theworkwhich
everyanthropologist interested in artis assumedto haveread,and notto citeit,and espe-
ciallynotto citeitapprovingly, is regardedin somequarters as automatic groundsforcriti-
cism. Some indicationof thehighregardin whichthebook is held is indicatedby the
blurbson thebackcover.Publisher'sblurbscan be expectedto be complimentary butall of
thewritersin thiscase are respectedanthropologists. For instance,CarolineHumphrey,
Readerin AsianAnthropology at Cambridge, is quotedas sayingthatthebook 'completely
reshapestheanthropology of art',andMauriceBloch,a professor at theLondonSchoolof
Economics, that it the
'changes very basis of the way arthas been viewed in thehumansci-
ences'. The same highlyflattering tone is continuedin the ForewordwhereNicholas
Thomascontendsthatthebook 'mayamountto themostradicalrethinking of theanthro-
pologyofartsincethatfieldofinquiry emerged'(Thomas1998:ix).
UnlikeHumphrey et al. I am notpersuadedthatthisbookjustifiesthepraisetheyheap
on it (see also Layton2003). Like his otherworkthebook is unquestionably writtenwith
literarypanacheandGell presents hisargument in a commendably forthright manner. Some
mightevenbe attracted to thefashionably difficult language.Insteadof sayingthatviewers
oftendrawinferences fromartworks aboutthestatus,powerand intentions of thepeople
who createand displaythemGell assertsthat'indexes' (i.e. artworks)'motivate'(i.e.
prompt)'patients'(viewers)to make 'abductions'(inferences) about'social agency'.The
problemis thatwhenthereaderbeginsto digbeneaththesurfaceto tryto workoutprecise-
lywhatGell is saying,itsoonbecomesapparent is notgoldanalytically,
thatall thatglitters
andthatwhathe is arguingis eithermuchless originalthanitfirstseems,ethnographically
uninformed or simplyincorrect. (I illustrateall ofthesepointsbelow.).1
This reviewis dividedintotwo parts.In thefirstI summarise themainargument of
each chapterin turnand comment on someof theauthor'smorequestionableassumptions
and conclusions.In thesecondI stepback fromtheindividualchaptersand commenton
threeofthemoregeneralissuesthebookraises.

'ANTHROPOLOGICAL' APPROACHES TO ART


of enchantment
Gell beginsthebook,as he did his 1992 article'The technology and the
enchantment witha frontal
of technology', studiesof
attackon existing'anthropological'

Oceania 74, 2004 309

This content downloaded from 146.96.128.36 on Mon, 1 Sep 2014 21:29:12 PM


All use subject to JSTOR Terms and Conditions
GellonArtandAgency
ofAlfred
Critique

art.He arguesthatto theextentthatearlierstudiesfocussedon suchissuesas aesthetic val-


ues or theway artworks 'encode' culturally
significant 'meanings',theyare notso much
wrongas inherently unanthropological.For Gell, an anthropological studyby definition
focuseson 'social relationships'(1998:4; see also p.7); in thecase of anthropological
stud-
ies of artthismeansthesocial contextsin whichartworks are produced,circulated,and
received(p.3). His particularinterest
in thisbook,he tellsus,is inthewayartworks mediate
social agency(p.7), i.e. theway in whichviewersmake'abductions'on thebasis of them
abouttheintentions ofthosewhoproduceordisplaythem.The significance of suchobjects
in sociallifeis thattheygiveviewersaccessto other'minds'(pp.16etc.).

AGENT-PATIENTRELATIONSHIPS
Two keytermsin theanalysisare 'agent' and 'patient',or 'social agent'and 'recipient'
respectively. Gell presumablytakes the firstpair fromtraditionalgrammarwherethe
'agent' is theperformer of an actionand the 'patient'theperson(or otherentity)which
'suffers'or is thetargetof theaction.An agent,Gell writes,is any'thing'(e.g. an artwork
or a person)'who is seen as initiating
causal sequencesof a particular type,thatis, events
causedbyactsofmindorwillor intention.' (p.16; see also pp.17and 19). Again,'Whenev-
er an eventis believedto happenbecauseof an "intention" lodgedin thepersonor thing
whichinitiatesthecausal sequence,thatis an instanceof 'agency" (p.17). Personsare
alwaystheprimary agentsbutartworks andotherinanimate objectscan be agentsin a sec-
ondaryor indirectsense,foralthoughtheythemselves are notintentional beingstheyfre-
quentlyact as themediumsthrough whichpeople 'manifestand realize' theirintentions
(p.21). As suchtheyare 'extensions'of thepersonswhoseagencytheyexpress- partof
their'distributed' personhood(see especiallyChapter7). Agentand patientare relational
concepts:foreveryagenttheremustbe a patient, andviceversa(p.22).

THE ART NEXUS


Anysocial contextin whichan artwork mediatessocial agencyconstitutes whatGell calls
an 'artnexus'(pp.l2ff).In eachnexusfour'terms'(p.27) needtobe distinguished:
1. The 'index' - i.e. theartwork (or othermaterialentity)which'motivate[s] abductive
inferences, cognitiveinterpretations, etc'
2. The 'artist':theperson(or otherintentional being,such as a divinity)to whomis
'ascribed,by abduction, causal responsibility fortheexistenceand characteristics of
theindex'
3. The patientor 'recipient':'thosein relation to whom,byabduction, indexesareconsid-
eredtoexertagency,or who[reciprocally] exertagencyvia theindex'(see below);and
4. The 'prototypes': 'entitiesheld,by abduction, in theindex,oftenby
to be represented
virtueofvisualresemblance, butnotnecessarily.'
Not unreasonably, Gell contendsthatanyone of these'terms'can, accordingto context,
occupytheroleof 'agent'or 'patient'in an artnexus(p.13), and thegreater partof Chap-
ters2 to 4 is devotedto illustratingthis.HereI citeone ofhismorestraightforward illustra-
tions,thoughtheexampleis myown.In thecartoonillustrated in Fig. 1 a comfortably-off
manis shownstanding beforea portraitand holdinga glassof winealoft,toasting theper-
son depicted.Although thecartoonlacksa captionthefactthatit appearedin thebusiness
sectionofa newspaper impliesthatthebeardedlook-alikein theportrait was thefounder of
a familybusinessofwhichthespectator is a grateful In Gell's termstheviewer
beneficiary.
oftheportrait is thepatient,thepainting theindex,andthepersondepictedtheprototype. In
Chapter3 Gell arguesthatwhenviewersof artworks respondprimarily to whatis depicted,
as in thesituation rather
illustrated, thanhow it is depicted,theprototype exercisesgreater

310

This content downloaded from 146.96.128.36 on Mon, 1 Sep 2014 21:29:12 PM


All use subject to JSTOR Terms and Conditions
Bowden

Fig. 1 . Untitled cartoon by Reg Lynch, fromthe Melbourne newspaper The Age,
Money Managers section, 31 Aug. 2002. Reproduced with the artist's permission.

agency in relationto the patientthanthe producerof the prototype,i.e. the artist.But in sit-
uations where a viewer responds to an artworkprimarilybecause of how it depicts some-
thingthenthe artistexercises greateragency than the prototype.As an example, Gell cites
the well known case of Winston Churchill who so disliked the way the painterGraham
Sutherlanddepicted him in a commissioned portraithe refusedto have the paintingin his
house (p.34).
Gell uses complex, scientific-lookingformulaeto express the way in which the differ-
ent termsin an art nexus can exercise greateror lesser agency according to context(e.g.
pp.52-56). In the situationdepicted in the cartoon,for instance,the greateragency of the
prototypein contrastto the artistcould be expressed in the formulareproducedbelow. The
extensions'-A' and '-P' signifythe agent and patientroles respectivelyplayed by different
termsin thisartnexus.

[[[Prototype-A]- Artist-A]- Index-A] ^ Recipient-P


If the artist,in contrast,had exercised greater agency than the prototype,this could be
expressed by reversingthe positions of 'Artist' and 'Prototype' in the same formula.Gell
termsin an art nexus can exer-
also uses 'tree' diagrams to illustratethe way the different
cise greateror lesser agency accordingto context(pp.54-65). Whetherthese,or formulaeof
the kind illustratedabove, actually capturethe social complexitiesof art nexuses is some-
thingI leave to othersto decide. My own view is thattheydo not.

VISUAL COMPLEXITY AND ORIGIN STORIES


In Chapter5 the argumentshiftsfromartnexuses and agent-patientrelationshipsin general
to the firstof threeaccounts of what it is about artworksthatmakes themeffectivemedia-
torsof social agency,i.e. what it is about themthat'motivates' viewers to make abductions

311

This content downloaded from 146.96.128.36 on Mon, 1 Sep 2014 21:29:12 PM


All use subject to JSTOR Terms and Conditions
CritiqueofAlfredGell on ArtandAgency

of agency.Gell does notsay so explicitly buthe impliesbothin thischapterandthefollow-


ing twothatthefeaturesthatmakeartworks effective mediators of agencyare also those
thatgivethemtheirvalue aesthetically, i.e. theirvalueas art (as sculptures,as paintings,
and so on). One majorproblemwiththebookanalytically, however,is thattheaccountshe
gives in these three chapters differ in a number of ways and arenotnecessarily compatible.
Even moresurprisingly he implies,withoutadvancinganyevidenceto supporttheclaim,
thatwhatpeoplevalue aboutartworks as art is thesame cross-culturally. Nothingin fact
couldbe further fromthetruth, forevena cursorysurveyoftheethnographic evidencewill
revealthatradicallydifferent criteria areused in different societiestojudgequalityin art.I
return tothistopicbelow.
In Chapter5, entitled'The origination of theIndex',Gell advancestheview thatthe
aspectof artworks thatenablesthemto act as effective mediators of social agencycross-
culturally is their technical complexity (p.72). Technical complexity 'captivates'viewers,
'captivation' beingboththe'primary formof artistic agency'(p.69) andtheprimary source
of aestheticenjoyment (see also pp.78-80).The particular featurethatviewersfindcapti-
vatingis thefactthattheyareunableto explainhow an artwork cameintobeingtechnical-
ly.This includes an inability to reconstruct mentally sequenceofphysicalactsthatled to
the
its creation,and theviewer'sinability easilyto reproducethosesequencesof acts - and
hencecreatesimilarobjects.As Gell putsit: 'Anyobjectthatone encounters in theworld
invitesthequestion'how did thisthinggetto be here?" Encounters withthemosthighly
regardedartworks automatically involve'playingout theirorigin-stories mentally, recon-
structing their histories as a sequence of actions performed by another agent (the ora
artist),
multitude of agents,in theinstanceof collectiveworksof artsuch as cathedrals'(p.67).
Objectsthatare notmystifying technically, he claims,are never valued as art'because
nobody attends to their as a
making particularly salientfeature oftheiragency'(p.68).
Gell illustrates hisargument in severalways.One is byreference tothefactthathe,as a
Sundaypainter, repeatedly triedbutfailedtopaintafterthemanner ofVermeer, andinpartic-
ularfailedto replicatethemaster'sdazzlingvisual effects.Another(p.71) relatesto the
allegedlymystifying complexity ofthebright polychrome designson theprowsoftheocean-
going canoes used in southeast New Guinea in Kula exchanges,an examplehe also uses in
thearticle'The technology ofenchantment...' referred to above(pp.69-71).The visualcom-
plexityand brillianceof thesecanoe prows,Gell argues,is designed'to captivate'(p.68)
theirviewersbothmotivationally andaesthetically andinducethemagainsttheirwilltohand
overmuchmorevaluableexchangegoodsthantheyinitially intended tooffer.
In thesecondpartofthisreviewI comment atlengthon theequationGell makesinthis
chapter betweenaesthetic powerandtechnological complexity. The onlycomment I wishto
makehereconcernstheway he handlestheethnographic evidence.The brightly painted
Kula canoe prowthatplayssucha pivotalrolebothin theargument in thischapterand in
the earlierarticlereferred to - the same canoe prow in bothcases (see 1998:70 and
1992:ColourPlate 1) - has clearlybeen freshly paintedand decorated.It was probably
also photographed at itshomeportbeforeit had ventured anywhere- let aloneouton to
theopensea in thecourseof a Kula voyage.The prow,in otherwords,is in pristine condi-
tionand at itsmostimpressive 'aesthetically'. I have never sailed offthe southeast coast of
New GuineabutI can imaginethatKula canoeson majorinter-island trading voyagesfre-
quentlyencounter choppyseas. Ifthisis thecase,theforceofthewavesbreaking overtheir
prows would undoubtedly wash much of the earth pigment offtheir polychrome splash-
boards,and probablydamagemanyof theotherornaments. These canoes,in otherwords,
wouldarriveat theirvariousportsof call in farfrompristine condition. Gell failsto men-
tionthisfactor to considerthepossibility thatit is notthedecorations as suchthatinduce
trading partners to partwithbetterqualityvaluablesthantheyhad intended to giveup but
other, more obvious,factors, such as the prestige and wealth of the visitors or thesize and
politicalpower of the groups to which they belong. Given his argument this is a majorover-

312

This content downloaded from 146.96.128.36 on Mon, 1 Sep 2014 21:29:12 PM


All use subject to JSTOR Terms and Conditions
Bowden

sightforit means thatthe claims he makes forthe impactof these artworkson theirowners'
tradingpartners- and indeed forart in general in this chapter - is nothingmore than an
hypothesisforwhichno concreteevidence is provided.
In Chapter6, 'The Critiqueof the Index', the focus of the argumentshiftsto non-figu-
rativeor 'abstract' art,somethingwhich Gell equates, astonishinglynaively,with non-rep-
resentationalart.Much non-figurative artis non-representational but in indigenoussocieties
of the kind Gell is consideringthe greatbulk of non-figurative art is representational.This
considerationapart,Gell offersin this chaptera new argumentin relationto what makes
artworkseffectivemediatorsof social agency - and, by implication,valuable aesthetically.
Here he argues that non-figurativemotifsare effectivemediators of agency not because
viewers cannot reconstructhow they were made but because they form visual puzzles
which cannot be deciphered easily or quickly (pp.73ff). Gell does not make the point
explicitlybut he appears to be implyingthatthe more puzzling a design is visually the more
highlyvalued it will be as art.Much of thischapteris devoted to a discussion of mazes and
othercomplex optical forms,all of which Gell amusinglydubs 'mind-traps'(Gell 1998:80).
Visually complex non-figurativeforms,he argues, often also have apotropaic roles. For
instance,in partsof India women place such motifsat the doors of theirhouses to prevent
malign spirits from entering and harming the occupants. They serve as 'demon-traps'
(pp.84-85) because the spiritsstop at the thresholdto tryto decipherthe designs. As inter-
estingas the materialin this chapteris ethnographicallyGell fails totallyto show how it is
relevantto the studyof art more generally.It throwsno light on how people in different
societies conceptualise what Gell and other analysts call 'art' or the criteriathey use to
judge qualityin art.
In Chapter 7, 'The distributedperson', the focus shiftsto representationalart (some-
thingwhich Gell naively equates withfigurativeimagery).Here he advances a thirdtheory
forwhy artworksserve as effectivemediatorsof social agency - and, by implication,why
they are valued aesthetically(e.g. p.97). For Gell, representationalimagery serves as an
effectivemediatorof social agency because the act of representationitselfconferspower
over the entityrepresented(p. 102) - even if this power is only imaginary.It does this by
-
capturingpartof the entityrepresented,notablyits image or simulacrum partof the enti-
ty's 'distributed' self - and 'binds' thatpart of its to
identity the artwork (or otherindex of
agency). By conferring(imaginary)power over the entitydepicted the maker of the repre-
sentationbecomes entitledto benefitsof various kinds fromthe entitydepicted (see espe-
cially pp.lO2ff). Gell devotes the greaterpart of this chapternot to a discussion of what
would normallybe called 'art' but to two institutionswhich he believes operate according
to the same representationallogic. One is idol-worship,an institutionwhich he says, with
commendable candour, he would preferto explain ratherthan rechristenfor reasons of
politicalcorrectness(p.96). The otheris whathe calls 'volt' sorcery.Gell definesidolatryas
physical interactionwith inanimateentities(such as sculptures)which representdivinities
for the purpose of receiving the blessings those supernaturalbeings supposedly confer
(p.135). In partsof South Asia, he reports,the act of makinga representationof a divinity,
in returnforwhich the divinityis obliged to confera blessing on the devotee, mightconsist
of nothingmore thanforminga visual image of it, such as by layingeyes on an idol depict-
ing it. Gell uses the termvolt sorceryforthe practiceof harmingvictimsfroma distanceby
manipulatingrepresentationsof them. The act of representingan intendedvictim 'binds'
partof the target'sdistributedpersonto the index and therebyenables the sorcererto injure
or kill the victim(pp.96ff) - by inflictinginjuriesof thekind the sorcererwould like to see
inflictedon thetargetof the sorceryon thatpartof his or her identityattachedto the index.
AlthoughGell undertakesno analysis of the social role of artin modernWesternsoci-
eties he neverthelessfeels able to assertthatthe culturalvalue placed on representational art
in the West can be explained in the same way as idolatry(or volt sorcery).Indeed, forGell
the value thatmodernWesternersplace on the act of contemplatingartworksin galleries

313

This content downloaded from 146.96.128.36 on Mon, 1 Sep 2014 21:29:12 PM


All use subject to JSTOR Terms and Conditions
ofAlfred
Critique GellonArtandAgency

andmuseumsis nothing morethana formofidol-worshipstrictly comparable to thatfound


in, say,SouthAsia. Earlierin thebook,forinstance,he wrote,'[I]n theNationalGallery
[London],even if we do notcommitfull-blown idolatry,we do vergeon it all thetime'
(p.62). In thischapter he asserts,'We have neutralizedour idols by reclassifyingthemas
art;butwe perform obeisancesbeforethemeverybitas deep as thoseof themostcommit-
tedidolaterbeforehiswoodengod.' (p.97).
Gell's lengthydiscussionof idol-worship in SouthAsia and thePacificis certainly
engaging. One of its more intriguingaspects his accountof theway representations
is of
divinities are madeto symboliseintentional beingsthrough a contrastbetweentheirinner
andouterparts,a metaphor, in hisview,forthatbetweenmindandbody.Buthisclaimthat
theculturalvaluethatmodernWesterners place on art,andtheenjoyment theyderivefrom
it in and
contemplating galleries museums, can be explainedin the same wayas idolatry or
sorcery is notsupported byanyconcreteevidence.If Gell seriously believedthatan institu-
tionsuchas idol-worshipprovidedan appropriate analyticalframework forunderstanding
thepracticeofgallery-going, andthevaluethatmodernWesterners place on artaesthetical-
ly,he wouldneedto do a lot morethanmerelynotethatrepresentational imageryis com-
monto bothinstitutional settings.Among other things,he would need to show that'devo-
tees' inbothcases:
1. conceptualise theobjectstheycontemplate, and value themculturally, in thesame(or
verysimilar)ways- i.e. thatgallery-goers conceptualise 'art'in thesamewayas idol-
atersconceptualise'idol'
2. conceptualise theiremotionalandothercognitive responsestorepresentational imagery
in thesameway
3. conceptualise theprototypes representedbytheobjectsthey'worship'inthesameway
4. conceptualise thebenefitsconferred on theviewersof suchobjectsin thesame way,
and
5. use thesamecriteria tojudgetherelativemeritsoftheobjectstheycontemplate (what-
everthesemightbe inthecase ofidols)
Needlessto say,Gell does noneofthesethings.
A secondobviousproblemwithGell's argument in thischapteris thatit deals only
withrepresentational imagery. MuchWestern artis representational butmuchis not,espe-
cially themost recent.At best, his
therefore, argument to
appliesonly one formofWestern
art.Furthermore, hisattempt to likenthepracticeofviewingartinWestern galleriesto idol-
atry(and sorcery)also suffers fromthefactthatmanyof theobjectsdepictedin Western
representationalimagery arenotanimate, andhencediffer radicallyfromthesentient, inten-
tionalbeingsthatfigurein systemsofidolatry andvoltsorcery. Gell is facedwiththeprob-
lem,therefore, of explaininghow inanimate entitiesdepictedin Westernart,suchas land-
scapesand still-lifes,
can be conceptualised as intentionally conferring benefitson viewers
in ways thatparallelthebenefitsthatdivinitiesconferon devoteesas a consequenceof
beingrepresented visually.Theremightbe parallelsbetweentheact of contemplating an
artwork in a Westerngallery and idol-
worship in South Asia or the Pacific,but if there are
theyaremuchtoodistant tobe ofanyinterest analytically.

STYLE
At theconclusionof Chapter7 Gell makesan astonishing admissionandone thatis actually
verydamaging forhisargument as a whole.He statesthatthereasonhe hasfocusedup tothis
pointin his bookexclusively
on thetransactional contexts in whichartworks arefound,and
paid no attention
to suchtopicsas themeanings theymight encode,is notbecauseto do so
wouldbe unanthropological, whichis whathe initially argued,butbecausehe has hitherto
focussedon individualartworks,i.e. objectstakenin isolationfromeach otherethnographi-

314

This content downloaded from 146.96.128.36 on Mon, 1 Sep 2014 21:29:12 PM


All use subject to JSTOR Terms and Conditions
Bowden

cally. When objects are takenin isolation,he says, the analyst'is always going to emphasize
therelationalcontextat theexpense of artisticor aestheticform,the networkof agent/patient
relations'in the vicinity'of the workof art' (p.153). But if sets of artworksbecome the sub-
ject of analysis,especially sets whose membersare unifiedby a commonstyle,otherinterpre-
tive opportunities open up. One is thatit allows foran investigationof the problematiccon-
cept of style.In the followingchapter,accordingly,Gell turnsto the problemof style.In that
chapterhe also addressesthequestionof whetherthe structural principlesthatunderliea par-
ticularartstyleoperatein otherareas of the same society,and whetherartstylesin thissense
'thematizeand make cognitivelysalient' wider'culturalparameters'(p.157; see also p.216).
Entitled 'Style and culture', Chapter 8 is divided into three parts. In the firstGell
attemptsto define 'style'. In the second he attemptsto illustratehis understandingof this
concept throughan examinationof Marquesan art.In the thirdhe attemptsto show thatthe
principlesthatunderliethe stylisticunityof Marquesan artalso underliethe Marquesan sta-
tus system,and thatin theMarquesas, at least, the artdoes thematize'essential' (p.157) cul-
turalparameters.
Regrettably,none of the ambitious goals that Gell sets himself in this chapter is
achieved. In the firstsection he fails to provide a methodologicallyuseful definitionof
style.The best thathe can do is offerseveral high-levelglosses of the term,such as by say-
ing thatstyleis the 'harmonicprinciple' that 'unites works of art into groups' (p.157). His
failureto providea usefuldefinitionis a consequence of the factthathe ignoresvirtuallyall
of the genuinelyvaluable work on stylethathas been done in recentyears by anthropolo-
gists, art historiansand especially archaeologists (see, for example, the various essays in
Conkey and Hastoff1990). The only theoreticalworkon stylehe does discuss in detail is an
anthropologicallyuninformedessay by thephilosopherWollheim(1979).
Even thoughhe failsto providea methodologicallyusefuldefinitionof styleGell never-
theless attemptsto illustratewhat he understandsby the termin the second section through
an examinationof Marquesan art.The workhe examines is not theentirecorpus of Marque-
san art (which, of course, is still being added to) but the limitedset of motifsillustratedin
Karl von den Steinen'sthree-volumeThe Marquesaners and theirart,publishedin 1925. On
thebasis of theseobjects Gell contendsthatwhatgives Marquesan artits stylisticunityis the
fact thateach distinguishablemotif(a notion he does not define) can be transformed into
every other motif through one or more steps involving, at each stage, what he refers to as
minimalvisual changes. This process is based on what he calls the 'principle of least differ-
ence' (Gell 1998:218). More particularly, Gell arguesthateach motifcan be transformed into
every other through the application of one or more of four 'transformation' rules (p.170);
these,he acknowledges,parallelthosedevised a generationearlierby proponentsof thecom-
ponentialapproachto the studyof kinshipterminologies(e.g. Lounsbury1964).
In markedcontrastto thetransformation rulesdevised by proponentsof thecomponential
to
approach kinshipterminologies, which can at least be applied rigorously, Gell's rulesare so
vaguely formulated that there is no way of applying them consistently or in a testableway;
of how theyallegedlyoperateare therefore
Gell's own illustrations ad hoc and arbitrary.

A B
Fig. 2. 'Plane of
rotation' one Marquesan motif(A) intoanother(B), adapted from
Gell 1998, Fig. 8.7/2,p.172.

315

This content downloaded from 146.96.128.36 on Mon, 1 Sep 2014 21:29:12 PM


All use subject to JSTOR Terms and Conditions
GellonArtandAgency
ofAlfred
Critique

This can be illustrated withreference to whathe refersto as 'rigidmotionsin theplane'


(Gell 1998:170).Gell claimsthatthemotiflabelledA in Fig. 2 can be transformed intothe
one labelledB byrotating each ofA's twolowerlimbs90 degrees,clockwisein thecase of
theleftleg (thefigure'sright)and counter-clockwise in thecase of theother.These rota-
tionsillustrate whathe meansby minimalvisualchanges(p.172). But as anyonecan see,
motifA is nottransformed intoB simplyby rotating itslowerlimbs,for(1) thetwohave
quitedifferent hands,(2) thefeatures ofthefaceareindicateddifferently and(3) theheadin
motifB hasbeendetachedfromthefigure'strunk. In thecase ofthehands,A's haveno dig-
its,butB's lefthand(i.e. thefigure'sright)has threeandtherighthas twoseparatedby an
unattached ovoidform.The differences in thehandsalso entailthatthetwomotifs beginto
takeon quitedifferent structures. A is characterised by verticalreflection symmetry - a
commonstylistic feature in Oceanicart- butB is not.A motifdisplaysverticalreflection
symmetry ifa straight linecan be drawn'vertically'through it(i.e. parallelto itslongaxis)
suchthatone halfcan be 'foldedover' on to theotherso thatitcoincideswithit.Gell pre-
sumablyconsidered thesevisualdifferences, otherthanthoserelating tothelegs,tobe irrel-
evantforhispurpose,buthe givesno reasonforwhytheyshouldbe ignored.Suchrelative-
ly minordifferences, furthermore, can havea majorimpacton theoverallstyleofa bodyof
designs and can give riseto major differences betweentheartstylesofdifferent peoples,or
ofthesamepeopleat twodifferent periodsin theirhistory. Thisexampleillustrates theway
Gell tendsto ignoreeverything thatis inconvenient forhis analysisand also pointsto the
factthathiswholeapproachto stylein thissectionis unsystematic andbasedon littlemore
thanguesswork.
The argument in thethirdsectionofthesamechapteris no moreconvincing. Therehe
claimsthattheprincipleof 'least difference' whichprovidesthe 'axis of coherence'in
Marquesanartstylistically (pp.218-19) also underliestheMarquesanstatussystem.Each
person in thishighly stratifiedsociety,he asserts,is separatedin statusfromeveryotherby
one or moregenealogicalsteps;thesegenealogicalstepsaretheequivalent, in Gell's view,
of the one or moreminimalvisual differences thatdistinguishdifferent visual motifs.
Becausetheartandthestatussystemarestructured according to the same principle, thatof
least difference, Marquesan art thematizes and makes cogniti vely salient wider cultural
parameters (p.157).
The problemwithhis analysis,however, is thatGell offers no coherent accountofhow
theMarquesasstatussystemis actuallystructured. Whathe does say,furthermore, shows
thathe fundamentally misunderstands therelationship betweenstatusbasedon genealogical
positionand politicalpower.This becomesespeciallyapparentwhenhe triesto correlate
aspectsof theartwiththestatussystem.Like manyothersbeforehim(e.g. Sahlins1966;
Oliver1974; Weiner1992) Gell mistakenly believesthatpoliticalpowerin thisand other
Polynesian societies can be directlycorrelated withgenealogicalrank.In fact,social status
throughout Polynesia had two quitedistinctand separableaspectsto it.One consistedofthe
ascribedstatusthatgenealogicalpositionconferred (thoughthiswas relativeand depended
on groupmembership); theotherconsistedof thesocial prestigethatpeople,and groups,
acquiredthrough outstanding achievements, such as success in intertribal warfare.In the
Marquesas, as elsewhere in Polynesia, ascribed status based on rank and achieved prestige
based on achievement wereconceptualisedin quitedifferent ways: in termsof tapu,or
'contagioussacredness'as Gell himself glossesit,in thecase oftheformer, andmanainthe
case of the latter.It was theoretically possible,furthermore, to have tapu (status)butno
mana(prestige), andmanawithout tapu. Just as status and prestige were acquiredin differ-
ent ways theycould also be lost in different ways. Prestige(mana), based on personal
achievement, was lostthrough failure,orlackofcontinuing success.Status(tapu)couldnot
be lostthrough lack of successin someendeavour, butit couldthrough deliberate or acci-
dentalstatus-breaking of variouskinds.For instance,beingcapturedin war and made a
slave strippeda man,even thehighestrankingchief,of his tapuand status.A personof

316

This content downloaded from 146.96.128.36 on Mon, 1 Sep 2014 21:29:12 PM


All use subject to JSTOR Terms and Conditions
Bowden

highstatuscould also be 'polluted'through rituallyunregulated contactwithpersonsor


objectsof muchlower status,especiallycooked food whichthroughout Polynesiawas
regarded as theantithesis of tapu(see Bowden1984).To protect theirinherited, butvulner-
able,status,therefore, of and
people rank, especially men who were more tapu a sex than
as
women,observednumerous personal restrictions,including the manner in whichtheyate,
wheretheysat,whotheycameintophysicalcontactwith,andhowtheywerepositioned in
termsof heightrelativeto thoseof lowerrank(foran illustration see Smith1989:147).
Throughout Polynesiaeffective politicalleadershipwas acquiredand held exclusivelyon
thebasisofdemonstrated personalskillin intergroup contexts,especiallywarfare, an activi-
ty fromwhichwomenwere excluded - except in the role of cheerleaderor victim.
Becausepoliticalleadership was basedon personalskillrather thanrank,themostpowerful
politicalleaders- thementypicallyreferred to as 'chiefs' (Sahlins 1966) - werenot
uncommonly menoflow rank,evenformer slaves,whohadrisentopositionsofgreatpolit-
ical powerbecauseoftheirabilityas leadersin war.Leadershipin all majorritualcontexts,
in contrast, was exclusively exercisedbypersonsofhighrankandtapu,thegreatest author-
ityinreligiousmatters beingheld,regardless ofpersonalabilityormana,bythoseatthetop
ofthegenealogicaltree- providedtheyhadnotlosttheirstatusforsomereason(Bowden
1979,1984).
Gell's failureto understand thattheMarquesasstatussystemhad two quitedistinct
partsto it is reflected in his inability to explainaspectsof theart.He notesat one point
(Gell 1998:202ff) that certain Marquesanutilitarian objects,suchas fanhandles,wereorna-
mentedwithimagesofdeities(etua) andthattheelaborateness oftheornamentation varied
directly withwhathe assumeswas thepoliticalpoweroftheirowners.The mostelaborately
ornamented wereownedby 'themostpowerful chiefsand chiefesses'(p.202; myempha-
sis). He also notes that the motifs on these objectshad apotropaicrolesdesignedto protect
theirownersagainststatus-degradation. Gell believesthatthepoliticalpoweroftheowners
oftheserichlyornamented objectswas directly correlated withthenumber of 'supporters' a
personhad.As he putsit,'A chief'spowerwas a function ofhowmanysupporters he had',
and the 'multitudinous' imageson thehandlesof thefanssuch a personownedreferred
'symbolically to thesalience of 'numbers'in Marquesanpoliticallife' (p.202). Buttheevi-
dencedoes notsupportthisinterpretation. Gell himselfnotesthatsomeof themostelabo-
ratelydecoratedfan-handles wereownedby highrankingwomen.Followingmanyother
writers on Polynesiahe refersto thesewomenas 'chiefesses'.WhatGell failsto realise,
however,is thatin Polynesiawomenformerly, whatever theirrank,neverexercisedeffec-
tiveleadershipin intergroup affairsand thattheyconsequently neverhad 'followers'after
themannerof military leaders.The womenhe refersto as 'chiefesses'werein factnot
politicalor military leadersat all, butthehighestrankingwomengenealogically in their
communities, persons whose marriages were of in
frequently greatsignificance intergroup
politics,and whosepersonaldignity and honourbothreflected andredounded to thecredit
of thegroupsto whichtheybelonged.Because Gell failsto distinguish betweenstatus
basedon rankandpoliticalpowerbasedon personalachievement, especiallyin warfare, his
attempt to correlatethe elaborateness with which fan handles and other utilitarian objects
weredecoratedwithapotropaicimageryand thesize of a person'sfollowingin factional
politicsfails,fortheownersof some of themostelaboratelydecoratedof theseobjects
neverhad 'followers'orledpoliticalfactions.
Fortunately, a betterexplanation is available.On thebasis of theevidencewhichGell
himself citesitis clearthatthedegreeofornamentation on suchobjectsas fan-handles was
directly correlated not with the political power of their owners but theirrank genealogically
defined,themostlavishlydecoratedbeingownedby themenand womenof thehighest
rank(and status).Giventhatstatuscouldbe lost(e.g. through enslavement) itis notsurpris-
ing that the images on such objects were thought to have a protective role. Protection
against the loss of status would be a matter of great concern to persons of high rank, andit

317

This content downloaded from 146.96.128.36 on Mon, 1 Sep 2014 21:29:12 PM


All use subject to JSTOR Terms and Conditions
ofAlfred
Critique GellonArtandAgency

wouldmakesenseforthemto own suchapotropaicdevices.Loss of status,and hencethe


needforprotectionagainstit,wouldnothavebeenof thesameconcernto peopleof lower
rank.Not surprisingly, objectstheyownedwerenotas elaborately
theutilitarian decorated
withapotropaic
imagery.

THE EXTENDED MIND


The concludingchapteris devotedto whatGell calls the'extendedmind'.Specifically, he
arguesthata collectivity of artworks unifiedby a commonstyle,such as theMarquesan
objectsconsideredin thepreviouschapter, constitute an externalised or objectified partof
theirmaker'smindorconsciousness. 'The pithofmyargument is thatthereis isomorphy of
structure betweenthecognitiveprocesseswe know(fromtheinside)as "consciousness"
andthespatio-temporal structuresof distributedobjectsin theartef actualrealm'(p.222) -
suchas the'oeuvre' of a singleartistor a stylistically unifiedbodyof workproducedby
different peoplefroma singlesociety.
Takinga cue fromHusserl(see pp.232-42) Gell arguesthatsubjectively experienced
consciousness is composed,amongotherthings, of 'retentions' or memories of thepast,as
well as 'protentions' relatingto futureevents(p.236; the latteris misprinted in several
places as 'portentions'). Retentions and protentions are notstaticor stableaspectsof con-
sciousnessbutunstableanddynamic, fortheyarecontinually beingmodified in thelightof
new experiences.Retentions are continuallybeing modified as the events to whichthey
relateare seenin thelightof laterexperiences, andprotentions arecontinually modifiedas
thefuture eventsto whichtheyrelatebecomeprogressively nearer.Gell contendsthatthe
features
stylistic thatlinkartworks exhibitingthesamestyleareanalogousto thoselinking
the different partsof subjectivelyexperiencedconsciousness.Individualartworks, for
instance,displaycomplexretentions from(or resemblances to) earlierworkscomposedin
thesame style,and equallycomplexprotentions in relationto workscreatedlater.These
complexesof stylistic inter-relationships 'recapitulate... theprocessesof [individual] cog-
nitionor consciousness'(p.254). Stylistically unifiedartworks, furthermore, constitute a
'single,coherent objectdistributed in space and time'and forman extendedor objectified
partof theircreator'smind- or mindsiftheyhad multipleauthors.Gell attempts to illus-
tratehis argument through detaileddiscussionsin turnof New Irelandmalangansculptures
(pp.223-228),New GuineaKula canoes (pp.228-32),Maorimeetinghouses(pp.251-58),
andtheworkofthe20thcentury Frenchartist MarcelDuchamp(pp.242-51).
Superficially, Gell's contention thatthelinksbetweenthemembers of a setof stylisti-
callyrelatedartworks 'recapitulate'theprocessesof individualconsciousnessis plausible.
Everyarthistorian is accustomedto documenting thecomplexsimilarities and differences
thatconnectearlierandlaterworksby an individualartistor thosefromthesameperiodin
thehistory of a singlesociety.But Gelljumpsfromthisrelatively innocuousfactto a con-
clusionwhichis notentailedby it andwhichis in factincorrect. He infersthatthefeatures
thatconnecta set of stylistically unifiedartworks, like theretentions and protensions of
individualconsciousness,are necessarily'dynamic'and 'unstable'and thatanalysesof
thosestylistic interrelationships can therefore neverbe absolutebutare at bestprovisional
(p.256). In thecase oftheworkofa singleartist, forinstance,'we cannottotalize[the]oeu-
vreas a temporalobjectwhichcan be regardedsub specie aeternitatis. All we can do is
compile a 'file' ofdifferent temporal perspectiveson the oeuvre as a whole' (pp.241-2).
Gell's argument wouldhavesomeplausibility ifitappliedto theworkofa livingartist
or a societywherethesetofobjectsunderdiscussionwas periodically beingaddedto.Then
it wouldmakesenseto say thatthestylistic interrelationships betweenthemembers of the
setweredynamicandunstable, forlaterworksmightbe createdwhichsignificantly modify
theanalyst'sinterpretation ofthestylisticlinksbetweenthemandearlierworks.Butwhena
setof objectswhichis closed,as in thecase of theworkof a deceasedartist, thenabsolute

318

This content downloaded from 146.96.128.36 on Mon, 1 Sep 2014 21:29:12 PM


All use subject to JSTOR Terms and Conditions
Bowden

andnotjustprovisional of thestylistic
interpretations features thatconnectthemembers of
thesetarepossible,forno newobjectsarebeingaddedtotheset.
It is true,of course,thateven in thecase of a deceasedartist, his or herworkis not
finiteor boundedsincehitherto unknownworksby thesame personcould come to light,
and thesepotentially could lead to a re-evaluation of the stylisticlinksthatconnectthe
worksthatwerealreadyknown.Butthisis a relatively minorpoint.For thereis no reason
whyan analystshouldnottaketheworksthatare known,or evena smallsub-setof them,
suchas thoseproducedin whatmightbe consideredtheartist'smaturestyle,and on the
basis of thatboundedset offeran interpretation of theirstylisticinterrelationships in
absoluteterms.This,in fact,is preciselywhatGell himselfdoes in his analysisoftheMar-
quesanartstylein Chapter8. In thatchapterhe bases his analysisof theMarquesanstyle
noton all knownworksbuton thoseillustrated in Steinen'sbook.Furthermore, Gell does
notprefacehis analysisin thatchapterby sayingthatthestylistic relationsbetweenthe
motifs he is considering areinherentlydynamicandunstableandthathis analysistherefore
is provisional;on thecontrary, he offersan analysisof whatconnectsthesemotifsstylisti-
cally in absoluteterms.It is perfectly truethatany interpretation of a bodyof empirical
data,whatever the academic is
discipline, alwaysprovisional in thesensethata better(e.g.
moreconsistent or simpler)interpretation of thesame bodyof datacan alwaysin principle
be advanced;butthisis notwhatGell is arguing.

AN OVERVIEW
Artand Agencyraisesmanyotherissueson whichI could commenton at lengthif space
permitted. butconfusednotionofthe'distributed
One is thefashionable person'whichGell
discussesat lengthin chapters7 and 9. In thisconcludingsection,however,I restrict
my
comments tothreemoregeneralissues.

4 studiesofart
Anthropological9
As alreadynoted,Gell beginshis book witha no-holds-barred denunciationof earlier
anthropological analyses of artwhich focus on what he calls 'aesthetic'
systems andthecul-
tural'meanings'thatartworks supposedly encode.He denouncesthesenotbecausetheyare
illegitimateintellectuallybutbecausetheyareinherently unanthropological. Theyareunan-
thropological because they do not focus on the transactional contexts in which artworks
occur,andin particular on thewayartworks mediatesocialagency.If studiesof symbolism
and 'aesthetics'areto be undertaken at all, Gell considers, theyshouldbe carriedoutbyart
historiansand 'critics'(p.2). Gell partlywithdraws fromthispositionat theendof Chapter
7 butitnevertheless remainsthebasicpositionadoptedthroughout thebook.
No seriousanthropologist woulddenythatartworks typically, evenuniversally, serve
as indexesof socialagency.In theSepikRiversocietyin whichthepresentwriter conduct-
ed fieldwork themembers ofeach clandeliberately andquiteconsciously decoratethecere-
monialbuildingsthey own in slightly different ways from all otherspreciselyforthepur-
of
pose distinguishing themselves as members of a distinctsocial group(see Bowden1992).
Furthermore, thesize of thesebuildingsand thelavishnesswithwhichtheyare decorated
withbarkpaintings andpolychrome sculptures areuniversally interpretedas indexesofthe
size,politicalstrengthanddepthof artistic skillof theclansthatownthem.Entirely legiti-
mately, Gell exploressimilarissuesin Chapter9 in relationto theceremonialarchitecture
of 19thcentury Maori.
But as he does so oftenin thisbook Gell startswitha valid pointand drawsinvalid
conclusionsfromit. In thiscase, he infersfromthefactthatartworks frequently serveas
mediatorsof social agencythatthisis theprimary role of artcross-culturally (see Gell
1998:251) and even the only one that should be of interest to anthropologists.But restrict-

319

This content downloaded from 146.96.128.36 on Mon, 1 Sep 2014 21:29:12 PM


All use subject to JSTOR Terms and Conditions
CritiqueofAlfredGell on ArtandAgency

inganthropological analysesof artto thewayobjectsmediateagencyhas theeffect ofradi-


callyimpoverishing bothartas a culturalphenomenon and theanthropology of artas an
intellectualdiscipline.It impoverishes theanthropology of artas an intellectualdiscipline
sinceitprevents as
anthropologists anthropologists from exploring a wholerangeof other
issuesrelatingto thesocial role of art.These includetheway artworks of all kindsgive
expression to thegenuinely sharedbeliefsandvalueson whichdifferent societiesarebased
- suchas beliefsabouthow societiesare structured, howpeopleshouldinteract withoth-
ersas individuals and as members of groups,whatmenandwomencan legitimately aspire
to,and how humansrelateto thephysicalenvironments in whichtheylive: plantand ani-
mal. By insisting thatanthropological studiesshouldfocusexclusivelyon agencyGell is
also preventing thedisciplinefromsolvingmanyof themostinteresting problemsin the
study of artcross-culturally,such as how peopleunderstand intellectual
creativity,and why
thereare no close parallelsof themodernWesternconceptof 'art' in theindigenouslan-
guagesofSub-Saharan Africa,theAmericasorOceania(e.g. Appiah1995;Kaeppler1989).
Tellingly,Gell does notevenpracticewhathe preaches.He denouncesstudiesof symbol-
ism as inherently unanthropological at thebeginning of thebook,yetin Chapter8 under-
takeswhatcan onlybe describedas a classic(ifflawed)symbolicanalysisof a bodyof art.
There,as noted,he attempts to showthattheformalprinciplesthatunderliethestylistic
unityofMarquesanvisualimagery also underliethestatussystem, andthattheartas a con-
sequencethematizes widerMarquesancultural parameters (Gell 1998:157).

Aestheticvalues
A secondmajorproblemwiththebookis thatit providesno coherent accountof aesthetic
values,i.e. thecriteria peopleuse in to
anysociety judgequality in art.
The studyofaesthet-
ic valuesis nota trivialor secondary matter anthropologically,as Gell implies(p.6), butis
centralto thestudyofartcross-culturally, forhowthemembers ofa societyjudgequalityin
artreflects theway theyconceptualise artas a culturalphenomenon, as well as how they
understand artistic Gell
creativity. does not directly addresstheissue of aestheticvaluebut
he does alludeto it.The problem, as alreadynoted,is thatwhathe saysis bothinconsistent
and,at times,astonishingly uninformed. His primary argument, outlinedin Chapter5, is
thatwhatgivesartworks theirvalue aesthetically is theirtechnicalcomplexity - thesame
feature,in his view,thatmakesthemeffective as mediatorsof agency.He also implies,
thoughwithout providing anyevidence,thataestheticvaluesare thesamecross-culturally.
Gell advancedan identicalargument in his well-known1992 articleThe technology of
enchantment ... ', and thatarticleactuallyprovidesthemostmemorablestatement of it.
Therehe recallshow his parentstookhimas an elevenyearold childon his firstvisitto
SalisburyCathedral.He recallsbeingutterly captivatednotby thecathedralitselfbutby a
matchstick modelofitwhichthechurchauthorities hadon displayto advertisetheirrestora-
tionappeal. He reportsthathis strongest impression whencontemplating thismiracleof
model-making was his totalinabilityto understandhow its makerhad constructed it.
Indeed,thismodel,ratherthanthecathedral,epitomisedforhimas a boy 'dexterity in
objectified form'(Gell 1992:47).

At one levelI had perfectinsightintothetechnicalproblemsfacedby thegenius


who had made themodel,havingmyselfoftenhandledmatchesand glue,sepa-
ratelyandin variouscombinations, whileremaining utterlyat a loss to imaginethe
degreeofmanipulative skillandsheerpatienceneededto completethefinalwork.
Froma smallboy's pointof view thiswas theultimateworkof art,muchmore
entrancingin factthanthecathedralitself,and so too,I suspect,fora significant
proportionof theadultvisitorsas well. Here thetechnology of enchantment and
theenchantment of technologycometogether. The matchstick model,functioning

320

This content downloaded from 146.96.128.36 on Mon, 1 Sep 2014 21:29:12 PM


All use subject to JSTOR Terms and Conditions
Bowden

[fortheCathedral'sFabricFund],is partof a tech-


as an advertisement
essentially
nologyof enchantment, butit achievesits effectvia theenchantmentcast by its
technicalmeans,themanner ofitscomingintobeing...(1992:47).

The factthatGell shouldgivesuchprominence in histheory ofartto a childhoodexpe-


rience(or at leastto his memoryof it) is highlyrevealingforit suggeststhatdespitethe
muchgreater knowledgeof arthe acquiredas an adult,andespeciallyas an anthropologist,
he neverchangedhis viewsaboutwhatgave artitsaestheticpower(and henceculturalsig-
nificance)fromthosehe held as a boy.More to thepoint,thereare two majorproblems
withGell's approachto aestheticvalues.First,and contrary to whathe assumes,aesthetic
valuesdo varycross-culturally,andvaryradically.Thismeansthatthereasonsa Marquesan
or a SepikRivervillagerwillgiveforvaluinga sculpture as a sculpture,or a paintingas a
painting - such as thefactthatit exactlyreplicatesan ancestralprototype - will differ
fromandevencontradict thoseoffered in,say,a modernWestern society.Gell seemsto be
unawareof this,andhenceof theneedto explainthesedifferences anthropologically - as
they can be.

Fig. 3. Claude Mellan (1598-1688), Head of Christon the Sudarium(1649).


Engraving,43.2 x 32.4. Courtesyof The Victoriaand AlbertMuseum, London (cat-
alogue V&A: E.2546-1960)

321

This content downloaded from 146.96.128.36 on Mon, 1 Sep 2014 21:29:12 PM


All use subject to JSTOR Terms and Conditions
CritiqueofAlfredGell on Artand Agency

Fig. 4. Claude Mellan(1598-1688), Head of Christon the Sudarium(1649) (detail).

Second, even if was concededthatthe mostadmiredartworkscross-culturally are


invariably objectsof greattechnicalcomplexity (something thatis notself-evidently true;
cf.Anderson1979:9-12)itdoes notfollowthattechnicalcomplexity counts,in anysociety,
as a reasonforvaluingsomethingas art. Put slightlydifferently, technicalcomplexity
mightbe a necessarycondition of something beinghighlyvaluedas artbutthereis no evi-
dencethatiteverconstitutes a sufficientcondition.The factthattechnicalcomplexity is not
directly related to the way objects are evaluated in
aesthetically, any society, needs no
detaileddemonstration here.One illustrationof thepoint,however,mightbe given.This
derivesfromWestern artand relatesto an etchingentitledHead ofChriston theSudarium
(1649) bythe 17thcentury FrenchartistClaude Mellan(Figs 3 and4). This work,likethe
artist'setchingsas a whole,is famousnotbecauseitis a masterpiece aesthetically- indeed
itis not- butbecauseoftheastonishing technicalvirtuositytheartisthas displayedin cre-
atingit.The entirefigureofChristhas beenformed froma singlespirallinewhichbeginsat
thetipofthenose(see Fig. 4). Mellan'stechnicalvirtuosity was suchthathe was evenable
to use thissameswellingand tapering line,brokenonlyat thebordersof theplate,to form
thelettering at thebottomof thefigure(see also Lambert2001:47). Accordingto Gell's
argument, themiraculoustechnicalvirtuosity embodiedin thisimageoughtto entailthat
theetchingis a masterpiece of art.But historians rightlydo notrateit particularlyhighly,
forlikeMellan'sworkas a wholeitlacksotherfeatures whichin modernWestern societies
are consideredmoreimportant to theway artworks arejudgedaesthetically.The twomost
important ofthese,ofwhichGell makesno mention, area work'sconceptualoriginality and

322

This content downloaded from 146.96.128.36 on Mon, 1 Sep 2014 21:29:12 PM


All use subject to JSTOR Terms and Conditions
Bowden

theway a workrevealsin new and penetrating wayssome aspectof theworld,eitherthe


physicalor socialworld,orboth(see Bowden1999;Bowden2002). Indeed,Mellan'swork
forplacingtoo greatan emphasison bravuratechnicalskillsat
has actuallybeencriticised
theexpenseoftheseotherproperties. In hisbriefentryon theartistin TheOxfordCompan-
iontoArt,HaroldOsbornewrites:

After12 years'trainingin Romehe [Mellan]wonconsiderable popularityinParis,


for
particularly his of
engravings compositions after Vouet. His use of longsinu-
ous lineswithno cross-hatchinggave his workat itsbest . . . directness
great and
Excessiveskillturnedmuchofhis workintoa sheerdisplayofvirtuos-
clarity.
ity,as in his Head of Christdone in one spiralline (Osborne 1978:710; my
emphasis).

ofart
Definition
A thirdmajorshortcoming of Gell's bookis thatit failsto providea coherent definition of
artandhencefailsto specifywhatthebookis ostensibly about.This is a significantfailing
in a workthatpurports to offerbotha theoryof artand a modelof how anthropological
analysesof artshouldproceedcross-culturally. Gell was clearlywell awarethathe had not
adequatelyaddressedtheissue of how 'art' is to be defined,forin theopeningchapterhe
attempts to sidesteptheissuebydismissing itas irrelevant methodologically (Gell 1998:7).
Butthisstrategy, to use thevernacular, is a cop-out.Gell knewit,furthermore, forimmedi-
atelyafterbrushing theissue aside he attempts to justifyhis indifferenceto how 'art' is to
be definedby saying- highlymisleadingly as it happens- thathe will limithis discus-
siontoobjects which no-one would dispute were art,suchas theMona Lisa (p.7).
Gell also attempts to sidesteptheproblemof how to define'art' by sayingthatthis
conceptis onlyfoundin certaintypesof societies,notablythosewith'institutions' suchas
galleries,artschoolsandcritics;anydefinition, consequently, wouldhaveno valuein cross-
culturalresearch(p.12). Gell is correctwhenhe statesthatthemodernWestern conceptof
artis onlyfoundin certaintypesof societies,notablyEuropeansocietiesand theircultural
offshoots (Appiah1995;Kaeppler1989). Butthefactthata conceptis notfounduniversal-
ly does notmeanthatit has no value in cross-cultural research.If he werecorrectexactly
the same could be said of almostall the analyticalcategoriesthatanthropologists use,
including'polities','economies','religion',even 'kinship'and 'family'.Rather,thefact
thattermscloselyparallelling artare notfoundin theformerly nonliteratesocietieswhich
anthropologists havetraditionally studiedshouldprompt theanalystfirstto offeran ethno-
graphically informed accountofhowthetermis usedin thosesocietiesin whichitis found,
and thenofferan explanation forwhyclose equivalentsare notfoundin manyothercul-
tures,even when are
they greatart-producing societiesin Western terms.If Gell had done
thisit wouldhavetakenhimclose to theheartof manyof themajorissuesin thestudyof
artcross-culturally.
AlthoughGell rejectstheneedto define'art' he nevertheless revealsin passingwhat
he believesthecharacteristics of artworks to be. Artworks display'technicalexpertise',
of a
'imagination high order', and 'exploit the intrinsicmechanisms ofvisualcognition with
subtlepsychologicalinsight'(Gell 1998:68). This is certainlybetterthannothing,but
unfortunately thethreefeatures he mentionsare notlimitedto artworks and consequently
do notdistinguish thisclass of objectsfroma wholerangeof otherentities,suchas any
well-designed producton a supermarket shelf,thebesttelevisionadvertisements, and any
number ofcomputer games.
Paradoxically, Gell's failureto offera methodologically usefuldefinition of 'art'reflects
thefactthathisbook,despiteitstitle,is notprimarily aboutartatall.Theprimary focusofthe
bookis indexesofagency, ofwhichartworks areonlyonevariety. Indeed,forGell'spurposes

323

This content downloaded from 146.96.128.36 on Mon, 1 Sep 2014 21:29:12 PM


All use subject to JSTOR Terms and Conditions
CritiqueofAlfredGell onArtandAgency

it is completely irrelevantwhether theindexeshe is discussingcomeundertheheadingof


'art'or not.The factthathisbookis centrally concerned withagencyrather thanartis made
clearintheveryfirst chapter whereGell definestheanthropology ofartas an academicdisci-
plineas thestudyof 'socialrelations in thevicinity
ofobjectsmediating socialagency'(p.7;
emphasissupplied).Notethathe says'objects',not'artworks'. A fewpageslaterhe modifies
thisdefinitionslightly by sayingthatanthropological analysesof artfocuson objectswhich
relateto social agentsin 'art-likeways',or occurin 'art-likesituations'. But thesephrases
clarifynothing, for all he means by them is thatsuch objects'permit...abduction[s]of
agency'(p.13) - something thatis truebydefinition ofindexesofagencyofall kinds,even
natural objectssuchas smokethatis takenas a signoffire(p.13). Laterhe againmodifies his
definitionof theanthropology of artas an intellectualdisciplineby statingthatit dealswith
indexesthathavebeenmadeintentionally art-
(pp.16,17). But thisalso failsto distinguish
worksfromanynumber ofotherobjectswhicharemadeintentionally, andpotentially serveas
indexesof agency,butwouldnotbe classifiedas 'art'.Theseincludeuniversity examination
papers,parking ticketsandeventheboiledeggsthatGellapparently enjoyedforbreakfast and
arethesubjectofoneofhismorememorable meditations (p.101).

NOTES
1. One oftheanonymous reviewersofthisarticlesuggested
thatcriticismofGell's bookshouldbe temperedby
thefactthatit was publishedposthumously.Butthisis questionable.If Gell had notwantedthemanuscript
or had notwanteditpublishedin theformin whichhe leftit,he wouldhaveinstructed
published, hisexecu-
torsaccordingly.
The factthathe didnot,andthathisexecutors publishedthebookoverhisownnamerather
thanthatofan editor,impliesthatthetextrepresents
Gell's consideredviewson itssubject-matter.

REFERENCES
ANDERSON, RICHARD L. 1989.ArtinSmall-scaleSocieties.EnglewoodCliffs,New Jersey: PrenticeHall.
APPIAH,KWAMEANTHONY.1995.WhyAfrica?Whyart?RA: theRoyalAcademy No. 48,pp.4O-41.
magazine,
BOWDEN, R.D. 1979. Tapu and Mana: RitualAuthority and PoliticalPowerin Traditional Maori Society.The
.JournalnfPacific,Hixtnrv 14:50-61
1984. MaoriCannibalism: An Interoretation.
Oceania 55:81-99.
1992. Art,Architecture and CollectiveRepresentations in A New Guinea Society.In Coote, Jeremyand
Anthony Shelton(Eds),Anthropology and Aesthetics . Oxford:ClarendonPress,pp.67-93.
1999. Whatis WrongwithAn ArtForgery? An Anthropological Perspective.TheJournalofAesthetics and
ArtCriticism 57 (3):333-43.
2002. Whatis 'Authentic' AboriginalArt?PacificArts,Nos 22 and23, pp.1-10.
CONKEY, MARGARETW. andCHRISTINE A. HASTOFF (eds). 1990.The Uses ofStyleinArchaeology. Cam-
bridge:CUP.
GELL, ALFRED. 1992.The Technology of Enchantment and theEnchantment of Technology.In Coote,Jeremy
andAnthony Shelton(Eds). Anthropology and Aesthetics. Oxford:ClarendonPress,pp.40-63.
1998. ArtandAgency:AnAnthropological Theory.Oxford:ClarendonPress.
KAEPPLER, ADRIENNE L. 1989. Artand Aesthetics. In Alan Howardand RobertBorofsky. Developments in
PolynesianEthnology. Honolulu:Univ.ofHawaiiPress,pp.211-240.
LAMBERT,SUSAN. 2001. Printsand Techniaues. London:V&A Publications.
LAYTON, ROBERT. 2003. ArtandAgency:A Reassessment. TheJournaloftheRoyalAnthropological Institute.
(N.S.) 9, 447-464.
LOUNSBURY, F.G. 1964. A FormalAccountof Crow-and Omaha-TypeTerminologies. In Goodenough,W.
(Ed.), Explorationsin CulturalAnthropology . New York:McGrawHill.
OLIVER, DOUGLAS L. 1974.AncientTahitianSociety.3 vols.Honolulu:University PressofHawaii.
OSBORNE, HAROLD. 1978.TheOxfordCompaniontoArt.Oxford:ClarendonPress.
SAHLINS, M. D.1966. Poor Man, Rich Man, Big-man,Chief:PoliticalTypesin Melanesiaand Polynesia.In I.
HogbinandL.R. Hiatt(eds). ReadingsinAustralian and PacificAnthropology. Melbourne:Melbourne
University Press,pp.159-79.
SMITH,BERNARD. 1989.EuropeanVisionand theSouthPacific.(2nded.). Melbourne: OxfordUniversity Press.
THOMAS, NICHOLAS. 1998. Foreword,to Geli, A. Artand Agency:An Anthropological Theory.Oxford:
ClarendonPress,pp.vii-xiii.
WEINER,ANNETTE. 1992.InalienablePossessions:theParadoxofKeeping-while- giving.Berkeley:University
ofCalifornia Press.
WOLLHEIM, RICHARD. 1979. Pictorial
Style:Two Views. In B. Lang (ed.) The Conceptof Style.[Philadel-
phia?]:University ofPennsylvania Press.

324

This content downloaded from 146.96.128.36 on Mon, 1 Sep 2014 21:29:12 PM


All use subject to JSTOR Terms and Conditions

S-ar putea să vă placă și