You are on page 1of 12

Narrate and Describe: The Problem of Ekphrasis

Author(s): D. P. Fowler
Source: The Journal of Roman Studies, Vol. 81 (1991), pp. 25-35
Published by: Society for the Promotion of Roman Studies
Stable URL: http://www.jstor.org/stable/300486 .
Accessed: 02/01/2014 13:28

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.

Society for the Promotion of Roman Studies is collaborating with JSTOR to digitize, preserve and extend
access to The Journal of Roman Studies.

http://www.jstor.org

This content downloaded from 128.103.149.52 on Thu, 2 Jan 2014 13:28:41 PM


All use subject to JSTOR Terms and Conditions
NARRATE AND DESCRIBE: THE PROBLEM OF EKPHRASIS*
By D. P. FOWLER

The subject of ekphrasis,and in particularof the ekphrasisof worksof art,has recently


begun to receive a great deal of attentionfromclassical scholars.1As will become clear, I
believe thatthe reasonforthisis thatmanyof the theoreticalissues thatare mostpressingin
classical studies - and indeed in culturalstudies in general2- are raised by the study of
ekphrasis.The purposeofthisnoteon theotherhand is modest: I wantto say a littleabout the
narratologicalissues thatare raisedby set-piecedescription(i), and to look at one example in
the Aeneid (II). But even so I have found it impossible not to offersome thoughtsof a
frighteningly general nature (iii). I shall concentrateon the ekphrasisof works of art for
reasonsthatwill again become clear, but some at least ofwhat I shall say will also be relevant
mutatismutandisto theekphrasisof naturalfeaturesand events.

Set-piece descriptionis regularlyseen by narratologistsas the paradigm example of


narrativepause, in the semi-technicalsense of a passage at the level of narrationto which
* A firstversionof this paper was deliveredto the here withekphrasisas an independentgenre,as in the
Universityof Bristol Classics Research Seminar in ekphrastic epigram:on thissee ofcourseFriedlander,op.
October I990. I am most gratefulforthe invitationto cit.,withthefurther bibliography inS. T. Stevens,Image
CharlesMartindaleandDuncan Kennedy,andtoallthose and Insight:EkphrasticEpigramsin theLatin Anthology
who offered comments,especiallyChristopher Gill, John (Diss. Wisconsin-Madison,I983), and SimonGoldhill's
Gould, and Malcolm Heath. Subsequentversionswere 'Reading, seeing, meaning: the poetics of Hellenistic
read to the Cambridge Literature Seminar, at the ekphrasis',forthcoming intheGoldhill/Osborne collection.
University ofWisconsin-Madisonduringan all-too-brief 2I shall referto some of the modernbibliography on
two weeks as BrittinghamVisiting Professor,at the description below:butnotethespecialissuesofYaleFrench
University ofPittsburgh,and at NorthwesternUniversity: Studies6i (I98I), Poetique43 (i98o) and 5i (I982), and
again I owe manythanksto myhostson theseoccasions, Litte'ratureZ8 (I980), and thecollectionsby J.Bessire,
respectivelyRichard Hunter, Alessandro Schiesaro, L'ordredu descriptif (I988), P. Bonnefis, La Description:
Hans-PeterStahl, and Francis Dunn, and to all those Nodier,Sue, Flaubert,Hugo,Verne,Zola, Alexis,Feneon
who offeredcomments,especiallyJohnHendersonand (znd edn., I980), and Y. Wert-Daoust,Description-
Ian DuQuesnay in Cambridge,Barry Powell, Fanny icriture- peintre(i 987). ApartfromGenette,the most
Lemoine, JimMcKeown, and JeffWills in Madison, a important theoretician has been P. Hamon: see 'Qu'est-ce
loyal pupil of Eckard Lefevre in Pittsburgh,and qu'une description?'Po&tique 3(197), 465-85 (trans.R.
BernadetteFort,Daniel Garrison,and JeanHagstrumin Canter in T. Todorov (ed.), French LiteraryTheory
Evanston.In addition,a numberof scholarshave been Today: A Reader (198Z)), and especiallyIntroduction a
kindenoughto commenton writtendrafts:in particular l'analyse du descriptif (I98I: an excerptis translatedas
Alessandro Barchiesi, Irene de Jong, John Elsner, 'Rhetoricalstatus of the descriptive'in Yale French
AndrewLaird,OliverLyne,RobinOsborne,theeditorial Studies 6i (I98I), I-Z6). Note also J. Hagstrum,The
board ofJRS, and the anonymousreaderwho correctly SisterArts(I953); J.Pelc, 'On theconceptofnarration',
divinedthat I want to be loved. The usual disclaimers Semiotica
3(1971), I-I9; M. Barchiesi,
I tempo
e iltesto,
apply,but I hope to be able to takemoreaccountofthe studisu Dante e Flaubert(I987); andJ.vanAppeldoorn,
criticismsI have receivedin an expandedversionofthis Pratiquesde la description(I98Z). Thereis a largemassof
paper to appear in the collectioneditedby J. R. Elsner comparativematerial,whichoftentoucheson issues of
(see n. i below). theory:see e.g. E. L. Bergmann, ArtInscribed:Essays on
' I amthinking especiallyofE. W. Leach, TheRhetoric Ekphrasisin Spanish GoldenAge Poetry(I979);
ofSpace: Literaryand ArtisticRepresentations ofLand- A. Corbineau-Hoffmann, Beschreibungals Verfahren.
scape Republicanand AugustanRome (I988) and S.
in Die Asthetikdes Objektsim WerkMarcel Prousts(I980);
the
Bartsch,Decoding AncientNovel,TheReaderand the P. Dubois, History,RhetoricalDescriptionand theEpic:
Role of Descriptionin Heliodorusand Achilles Tatius fromHomertoSpenser(I98Z); R. L. Flaxman,Victorian
(I989), but thereare currently twovolumesof essayson Word-Painting and Narrative: Toward theBlendingof
the subject in preparationfromCambridge,on Greek Genres (I987); P. Imbert, Semiotiqueet description
textseditedbySimonGoldhilland RobinOsborneandon balzacienne (1978); J. Kurman, 'Ecphrases in epic
Latin editedbyJ. R. Elsner.Bartschoffers a goodgeneral poetry',ComparativeLiteratureZ6 (I974), 1-13;
bibliography:see also those in M. Fusillo, Il Romanzo L. Perrone-Moises,'Balzac et les fleursde l'ecritoire',
Greco (I989), 83-go and S. Richardson,The Homeric Poe'tique ii (I980), 305-Z3; W. H. Race, Classical
Narrator (1990), 5o-69, and add amongstveryrecent Genresand EnglishPoetry(I988), 56-85; J. Ricardou,
worksN. W. Slater,ReadingPetronius
(1990), Z13-30, Problemesdu nouveau roman(I967); L. Spitzer,'The
and D. Rosand, 'Ekphrasis and the Generation of "Ode on a GrecianUrn", or contentvs. metagrammar',
Images',Anion NS i (I990), 6i-iO5. The standardwork ComparativeLiterature 7 (1955), Z03-55; M. Van
remainsthatof P. Friedlander, Johannesvon Gaza und Buuren,'L'essencedes choses',PoetiqueII (I980), 3Z6-46
PaulusSilentiarius
(I9IZ), 1-103; in Latin,thereis an (on Claude Simon); B. Vannier,L'inscriptiondu corps,
excellentsurveyby G. Ravenna,'L'ekphrasispoeticadi pour une semiotique
du portraitbalzacien (197Z);
opere di artein Latino: temie problemi',Quad. Ist. Fil. M. Zink, 'Les toiles Agamanor et les fresques de
Lat. Padova 3 (1974), 1-5z (see also hisarticle'Ekphrasis' Lancelot', Litt&rature 38 (I980), 43-6I. The Tenth
in the Enciclopedia Virgiliana). The most suggestive InternationalColloquium on Poetics held at Columbia
discussionis thatofA. Perutelli,'L'inversionespeculare. Universityin I986 was devoted to 'The Poetics of
Peruna retoricadell'ecphrasis',MD I (I978), 87-98 = La Ekphrasis'(see Rosand, above n. i) but the proceedings
Narrazione Commentata (1979), ch. z. I am notconcerned havenotbeen published.

This content downloaded from 128.103.149.52 on Thu, 2 Jan 2014 13:28:41 PM


All use subject to JSTOR Terms and Conditions
26 D. P. FOWLER

nothingcorrespondsat the level of story.3The plot does not advance, but somethingis
described. There is an obvious sense in which descriptionis more basic - one could
theoretically imaginea narrativewithonlynames in it, and no referring expressions,but it is
practicallyimpossibleforanynarrativeoflengthnotto containdescription.In a deepersense,
however,as Genette noted in his articleon the 'Boundaries of Narrative',4descriptionin
general is secondary, is 'ancilla narrationis, the ever-necessary,ever-submissive,never-
emancipatedslave'. Set-piece descriptionis not even in any real sense necessary. Hence the
controversialnatureofdescriptionand thestrongantipathyto itwhichcriticsfromLessing to
Lukacs have oftenshown.5 This is connectedwith the issue of human interest,put in its
crudestformby Lukacs in 'Narrateor Describe' withitsepigraphfromMarx, 'To be radicalis
to grasp things by the roots. The root of humanity,however, is man himself', and its
declarationthat'objectscome to lifepoeticallyonlyto theextentthattheyare relatedto men's
life,thatis whythe real epic poet does not describeobjects but exposes theirfunctionin the
mesh of human destinies,introducingthingsonlyas theyplay a partin the destinies,actions
and passionsof men'.6Narrativeis about people, descriptiondeals withthings.
to challengethisAristotelianoppositionof 'narrative'and 'description',
It is not difficult
and the exaltationof the tellingof storiesabout human beings over descriptionof things. It
maybe thatwhatwe are interestedin in narrativesis neitherplotnorpicturesbut ideology,the
values inscribedin theworkthroughthemeand imageryas much as by storyand description.
In essence I thinkthisis correct,and thisessay could perhapsstop here. But I wantto go on
talkingoftheproblemofdescriptionbecause thisprimacyofplot and almostmoraldistastefor
description7 has been verydeeplyengrainedin theWesterntradition.Historically,description
has tended to make people nervous. As Riffaterre makes clear,8this lies at the heartof the
traditionallyproblematicstatusof didactic (especiallyscientificdidactic) poetry.To allow a
place fora poem about theworldliketheDe rerumnatura, we mustexplainwhywe are to be
interestedin the blind motions of atoms in the void, to answer Aristotle'sexclusion of
Empedocles as notimitatinghumanaction.9'Loss ofproblems'(in Wittgenstein's phrase) can
be dangerous,because it can blind us to how lingeringare the traces of the beliefsthatwe
dismiss. So I ask thereaderforthemomentto join withme in worryingabout descriptionand
itsrelationto narrative.Ifwe acceptforthemomentthetraditionaloppositionofnarrationand
description,we can isolatethreeapproacheswhichhave triedto deal withtheproblemoftheir
relation.
The firstis to stresstheroleofset-piecedescriptionin 'bringingthescene beforeour eyes'
as traditionalaccountsofenargeia put it,10or to say withBarthes11thatwhat we have is 'the
effectof the real', that what details in a descriptionsignifyis realityitself: 'Flaubert's
barometer,Michelet'slittledoor finallysay nothingbut this: we are thereal'.12Descriptionis
admittedto be narratively(or indeed thematically)redundant,but thisredundancyincreases
our sense of the realityof the scene beforeus. It is just as ifwe were thereourselves. Now,
however,one definesthese'reality'functionsofdescription,theyare undeniable. But theydo
not get us very far. It was an early lesson of old-fashionedstructuralismthat cultural
productionsparticipatein systemsofmeaningindependentoftheconsciousintentionsoftheir
creatorsor users. I maybuy a Nissan car because ofitsreliability(I did) but thesignifications
I See e.g. M. Bal,
Narratology,Introductionto the 7 Not always'almost':as FannyLemoineremindsme,
TheoryofNarrative(i 985), 70, 76-7. some of the antipathyto descriptionmay be more
4 G. Genette, 'Frontiersof narrative',in Figures of explicitlymotivatedby a contemptforthe thingsof this
LiteraryDiscoursetrans.A. Sheridan(I98Z), IZ7-44, at world,whetherfroma Platonicor a Christianstandpoint,
134 (from FiguresII (I969): see also New Literary just as thegrowthofnon-allegorical descriptivepoetryin
History8 (1976), 1-13)- moderntimesis boundup withRomanticpantheism.
5 See especiallyH. C. Buch, Ut Pictura Poesis. Die 8 M. Riffaterre, 'Systeme d'un genre descriptif',
Beschreibungsliteratur und ihreKritikervon Lessingbis Poe'tique3 (I97Z), 15-30.
Lukacs (1972). I shallreturn to Lessing,whohasbeen 9 cf. Tasso, Discourses on the Heroic Poem, trans.
much discussedin recentyears: see the introduction to M. Cavalchiniand I. Samuel (I973), 7-8.
Laocoon trans. A. E. McCormick (I984) with D. E. 10 cf. H. Lausberg, Handbuch der literarischen
Wellberg,Lessing'sLaocoon,Semioticsand Aesthetics in Rhetorik
(I960), 399-407; G. Zanker,'Enargeiain the
theAgeofReason (1 984); G. Gebauer(ed.), Das Laokoon ancientcriticismofpoetry',RhM I Z4 (i 98 I), Z97-3 II.
Projekt(I984: collectionofessays,sometranslated);and 11R. Barthes,'The realityeffect'in The Rustle of
especially T. Todorov, 'Esthetique et semiotique au Language, trans.R. Howard (I986), 14I-8, or trans.R.
XVIIIe siecle',Critique308 (I973), Z6-39 = Theoriesdu Carter in T. Todorov (ed.), French Literary Theory
Symbole(I977), I6I-78 (Germantranslation inGebauer). Today (I98Z), 11-17.
12 idem,148.
6 G. Lukacs,'Narrateor describe?' in Writerand Critic
and otherEssays, trans.A. Kahn (I978), I I0-48, at 137.

This content downloaded from 128.103.149.52 on Thu, 2 Jan 2014 13:28:41 PM


All use subject to JSTOR Terms and Conditions
NARRATE AND DESCRIBE: THE PROBLEM OF EKPHRASIS 27

'boring'and 'unstylish'whichitbearswhencontrastedwitha Porscheare independentofwhat


I thoughtI was doing. Of course an interpreter can take this public signification
and mess it
around,but she cannotignoreit. The ekphrasisofthevillain'scar in a narrativemaycertainly
bringthe scene beforethe reader'seyes, but we can stillask what this car means againstthe
matrixof alternativepossibilities.In fact, it is strikinghow oftenset-piece descriptionin
narrativeis of thingsthat participateparticularlyobviously in social systemsof meaning:
clothingor armour,furniture, architecture, thecultivatedlandscape. These social systemsdo
notdetermineor limitthemeaningofan ekphrasis,but theyalreadytakethereaderbeyondthe
realityeffect,howeverspecified.There is thereforeno real 'solution'to our 'problem'here.
The othertwo approaches are more important.First, we can attemptto deal with the
problemof descriptionby integrating it withthenarrative.A relativelycrude formofthiscan
be foundin Lessing's defenceof the Homeric shield descriptionon the groundsthatnot the
shield but its manufactureis described:13 only slightlyless crude is Lukacs' defenceof the
horse-racein Tolstoy'sAnna Karenina againstthatin Zola'sNana on thegroundsthatitis 'no
mere tableau but a seriesof intenselydramaticscenes which providea turningpointforthe
plot'.14Clearly narrationoftencontinuesthrougha description(there is rarelya complete
pause) and the descriptionmayoccasion reactionsin participantsimportantforthe plot. But
the needs of the plot can usually be satisfiedby a much more exiguous account than we are
offeredin ekphrasis,and it is difficultto see that Lessing's device of describingthe manu-
factureof an object ratherthantheobject itselfis otherthana trick.We could easilyturnany
descriptionof an object intoan accountof itsmaking,but would thisreallyget to theheartof
the problem?Nevertheless,it is an importantpointthatdescriptionis rarely'pure', because
theway thatnarrativeimpurityis introducedis oftenthroughthefigureofan observer.
It is mythirdtactichoweverwhichis mostcommonlyfoundin criticalwritingand which
raises the most importantquestions, that is the relationof descriptionto narrativeon a
psychologicallevel. This can be done in severaldifferent ways. The objectsdescribedmaybe
causallylinkedwitha characteror an action: people choose theirown furniture or wallpaper.
We may have an explicit or implicit observer, through whose eyes the description is
instantiated.We mayhave an instanceof patheticfallacy,howeverwe wishto definethat: the
stormoutsidereflectsthestorminside. Or we mayhave much loosermetaphoricor especially
metonymiclinks with the plot, particularlyones of prefiguration:15the flowerplucked in
chapterone becomes the maidenhoodlost in chapterfour.Much moderncriticalreadingof
ekphrasisin classicalliteraturetakestheformofan attemptto showthatwhatearliercriticshad
seen as 'merely' decorative descriptioncan in fact be integratedwith narrative,indeed
demands to be so integrated.Preciselybecause ekphrasisrepresentsa pause at the level of
narrationand cannotbe readfunctionally, thereaderis possessedbya strongneed to interpret.
As will be seen, I believe thisis the righttactic.But it is obviouslyvulnerableto attack,
and has come underattackfrompost-moderntheorists,forbeing organicistand totalizing.If
we take the dominanttrendsin currentcriticismwith Murray Krieger to be those which
'celebratemarginsratherthan centers,the aporia ratherthan the filledgap, the arbitraryor
even therandomratherthanthenecessary',16 thenit is clearthatfromthosepointsofviewthe
way in which the classical criticismthat I have been describingintegratesekphrasismay be
seen as a minus not a plus. To relatedescriptionin thisway to narrationis to accept its poor
relationstatusbut to give it a limitedformofsocial mobility:the moreradicalmove is to free
descriptionfromthe chains of slaveryand to give it true autonomy.The vanguardof this
approachwas thenouveau roman,particularlyin thetheorizingof Robbe-Grillet,withitscry
that'insteadof thisuniverseof "significations"(psychological,social, functional),one must
tryto constructa worldmoresolid, moreimmediate'17 in whichobjectsaregivena roleoutside
of any metaphoricalor metonymicsystemof reference.Now whetherthenouveau romanin
factachievesthatindependenceoftheobject, and whethereven ifit did itwould be relevantto

13 Lessing,Laocoon, ch. i6, e.g. trans.McCormick, P. Hardie, and M. Whitby(eds), Homo Viator(I987),
op. cit. (n. 5), 84, 'we see inthepoet'sworktheoriginand z43-51I
formation ofthatwhichin thepicturewe can onlybehold 16M4. Krieger,
A ReopeningofClosure(i 989), 3.
as completedand formed'. 17
A. Halsall, "'La Transition", descriptions et
14 op. cit. (n. ambiguites narrativo-discursives dans "Victoire" de
5), III-
15
cf. R. G. M. Nisbet, 'The Oak and the Axe: WilliamFaulkner',in Bessiere,op. cit. (n. z), Z7.
symbolismin HerculesOetaeus i6i8ff.', in M. Whitby,

This content downloaded from 128.103.149.52 on Thu, 2 Jan 2014 13:28:41 PM


All use subject to JSTOR Terms and Conditions
28 D. P. FOWLER

classical criticismare questions I want to avoid, but this anti-organicist,anti-closural


movementin late-modernand post-moderncriticismshould make us be at least a littleself-
conscious about the stresson closure and integrationwhich dominatesclassical criticism.A
generationwhose mottois 'the revengeof the crystal'18 is hardlylikelyto warmto a criticism
whichcelebratesan integrating focuson humansubjectivity.
Moreover, the modernistclassical criticis likelyto findherselfstabbed in the back by
receptiontheoristslikeMalcolm Heath, who would denythatancientreaderswould have felt
thisneed tointerpretwhichis thestandardstarting-point foraccountsof ekphrasis.19 I do not
myselfbelieve thatHeath is right.Basic to his approach is the Hirschianoppositionbetween
'meaning'(identifiedwithconscious intention)and wider'significance'as read in by modern
interpreters,and the exaltation of the formercharacterizedas historicalover the latter
characterizedas 'theforcefulimpositionofalien preconceptionson ancientliterature'.20 It is an
easy challengeto celebrateratherthe latterprocess,and to dramatizeone's criticalpracticeas
Socraticguerillawarfare.But it is moreimportantto deconstructtheopposition.21On theone
hand, the beliefthatone can everfreeoneselffromcontemporaryconcernsis a delusion that
criticshave alwaysto resistiftheyare to avoid self-deception:itis notdifficult to findelements
in thepresentpositionofclassicalstudiesout ofwhicha plausible accountofthefactorswhich
encouragea cool historicismcan be constructed.On the other,Heath himselfadmitsthat a
basic problemwithhis approachis thata culture's'primarypoetic'instantiatedin practicemay
only very imperfectlybe captured by the 'secondary poetic' which is representedin the
conscioustheorizingof rhetoriciansand commentators.22 If one takesa widerview of ancient
semiotics,itis notdifficultto findevidenceofa stronghermeneuticimperativeat all periodsin
phenomenalike divination;and indeed even in literarystudies Heath underplayselements
such as the allegoricaltradition,which begins in the fifthcenturyB.C., not the fifthcentury
A.D.23 Moreover,it is mostimportantnotto accept the characterization of'readingagainstthe
grain as to
necessarilyunhistorical; accept the conscious formulation of its own values by a
culture(or some membersof it) as authoritative looksmorelikea denial of history.One must
both resistthe simplificationof ancientattitudesand accept thatthe criticmayat timesstress
elementsthatmembersof a cultureneglectedpreciselyin the name of history.But Heath's
attackis an importantreminderthatthereis nothingnecessarilynaturalor inevitableabout
modernistintegrationism:it is an aestheticthathas to be defended.We have anotherreason
not simplyto take as an obvious giventhatthe firstthingone does withthe descriptionof an
apple is to finda youngvirginto whomit mightcorrespond.
I wantto returnto thesethemesat the end, but I hope it is alreadyclear how the way in
which we approach ekphrasis is paradigmaticof our attitudesto much wider issues of
interpretation. Let me, however,turnto a question thatis constantlyraised with respectto
ekphrasis,thatof 'point of view'.25I mentionedabove thatone way in which descriptionis
oftenrelated to narrativepsychologicallyis throughthe figure,explicit or implicit,of the
observer.This is put moststronglyby J. Kittay:26

Whenwe reada representation, we also read,or readin, theaccountoftheperception ofthat


representation ... Thereis alwaysa choiceofpercipients. We canreadan actas perceivedbythe
characterwhocarriesitout(ourhero,forexample),bythecharacter whois orwillbe itsobject,or
by anycharacter who mightreactto it, appreciateit, or be confusedby it (e.g. an onlooker,a
confidant, a chorus) ... There are no autonomouslimitson this powerof inferenceand
construction, thisreading-in of subjectivity,
as theremustbe at leastthepossibility of access
(imaginative as well as provided)to a subject-based readingof the represented. Empathyis
available,of one characterforanother,of thenarrator foranycharacter, and of thereaderfor
anyone.

18 J. Baudrillard,
Les strategiesfatales(I983); cf. D. and Medieval Technique (I987), with furtherbibio-
Kellner, Jean Baudrillard, From Marxism to Post- graphy:on TheagenesofRhegium,conventionally made
modernism
andBeyond(I989), 1546Z. theprotosheuretesofallegory,see R. Pfeiffer, Historyof
19 M. Heath,Unity
in GreekPoetics(I989): see also Classical Scholarshipi (I968), 9-I I.
ThePoeticsofGreekTragedy(i 987), 98- II. 24 J.Winkler, The Constraints ofDesire (I990), 126.
20 idem, I55. 25
For myuse hereofGenette's'focalization' (and some
21
cf.D. Kennedy,rev. S. J. Harrison,OxfordStudies ofthe problemswiththeconcept),see 'Deviant focaliza-
in Vergil'sAeneid,Hermathena(forthcoming). tionin Vergil'sAeneid',PCPhS ZI6 (1990), 4Z-63.
22
op. cit. (n-I9), io. 26 'Descriptivelimits',Yale FrenchStudies 6i (I98I),
23
cf.J.Whitman, theDynamicsofan Ancient
Allegory, ZZ5-43, at Z34.

This content downloaded from 128.103.149.52 on Thu, 2 Jan 2014 13:28:41 PM


All use subject to JSTOR Terms and Conditions
NARRATE AND DESCRIBE: THE PROBLEM OF EKPHRASIS 29

The questionoffocalization,of'who sees?', is raisedwithparticufar and obvious forceby


description.Basic to Lessing's famousdistinctionofliteraryand plasticartweretwo notions:
first,that'successionoftimeis theprovinceofthepoetjust as space is thatofthepainter',27 and
second, thatlinguisticsignsare arbitrary,visual ones iconic. A pictureof a cat lookslikea cat,
theword'cat' does not. Againtheseare distinctionswhichinviteand have receiveda greatdeal
of deconstruction,but which I want forthe momentto accept. Both of these aspects of the
literarydescriptionreinforceourfeelingthata literarydescriptionnecessarilyinscribesa point
ofviewmorestronglythana plasticone. On theone hand,thereis thephenomenonofwhatthe
linguistW. J. M. Levelt has referredto as the speaker's linearizationproblem:28 when we
describein wordsa scene,we haveto decide theorderin whichwe areto presentthedetailsand
theduration- whichmaybe zero- ofthedescriptionofeach ofthem.Narratologically, that
is, the visual scene described functionsas storyto the narrationof the verbal description.
There is no neutral, zero-focalizedway of linearizinga visual scene: a point of view is
necessarilyinscribed,thoughtheremaybe accepted ways in a particularcultureof ordering
theelements- asked to describetheirhouse, forinstance,mostpeople willgivea mentaltour
startingat the frontdoor and climaxingaccordingto predilectionin the kitchen,thebedroom
or the study. The focalizationdoes not, of course, have to be thatof the actual observer:as
Kittayremarks,empathyis available, and I can describemy house to a friendin a way that
representsneithertheway I see it northeway she does, but the peculiarinterestsof myaunt.
But the speaker'ssolutionof the linearizationproblemnecessarilyimposes a point of view.
Similarly,the non-iconicnatureofthelinguisticsignmeansthatthereis a muchwidermatrix
ofchoice againstwhicha particularelementis seen. In hiswork19 I2 + i, Sciascia describesa
photographof an Arab being shot amongstthe dunes duringthe Italian imperialistwar in
Libya, and implicitlyhighlightsthecontrastbetweentheemotionalreactionofa modernto the
scene withthepresumedcontemporary reading:29
In Cirenaicala guerriglia pungeva.Tribunalidi guerraassiduamente sedevanopergiudicarei
ribelli:e cioeperpassarliai plotonid'esecuzione.Agliitalianinearrivaqualcheimmagine: schizzi,
fotografie. Conqualsentimento e statoalloraguardataquestafotografia chehosottogliocchi,della
fucilazione di un arabotrale dune?I1 plotoneschierato che staperdareil
su due file,l'ufficiale
segnaledel fuoco,il condannatoche sembralontanissimo dal plotone,come sperdutotra
l'ondulazione delladuna.AgostoI9I3. I9I2 + I.
In Cyrenaicatheguerrilla warwas causingirritation. Courtsmartialwerein constant sessionto
passjudgement ontherebels- thatis,tohandthemovertothefiring squads.A fewimagesofthis
reachedtheItalianpublic:sketches, photographs.Whatwastheirfeeling thenwhentheylookedat
thisphotograph thatI havebeforeme,depicting theshooting ofanarabinthemidstofthedunes?
The platoonlinedup intwolines,theofficer togivetheordertofire,thecondemned
waiting man
lookinga longwayawayfromthefiring squad,almostlostamongst theundulating dunes.August
I9I3. I9I2 + I.

Sciascia attemptsto describethescene neutrally,likea camerawiththe shutteropen. But his


choice of 'fucilazione',likemychoice of 'being shot' mustbe contrastedwithalternativeslike
'being executed','being murdered','being martyred'.The same photographcan be read as a
signoftriumphor an indictmentofcrime,but verbaldescriptionhas to takea stand,however
'objective'it attemptsto be. Again, thereis an obvious sense in whichdescriptionin language
inscribesa pointofview moreforcefully and moreunambiguouslythanplasticart.
Now I stressagain that these oppositionsare in actual factfar more complex: decon-
structionis here as easy and as necessaryas with the otheroppositionsthat I have already
discussed. There are of course various ways in which narrativeart can exist, through
conventionsof placement,the use of panels and frames,the representation of morethanone
momentof timein one picture.30In an architecturalsettingviewpointmaybe 'controlled',as
Robin Osborne has arguedwithregardto theParthenonfrieze.3' The converseofthespeaker's
linearizationproblemis theartist'snon-linearization problem,how to representtimethrough
27 Trans. McCormick,op. cit. (n. 5), 91. Etruscanand RomanArt(I984), withbibliography;on
28 W. J. M. Levelt, 'The speaker's linearization some theoreticalproblems,see N. Goodman, 'Twisted
problem',in H. C. Longuet-Higgins, J.Lyons,and D. E. tales; or, story, study, and symphony',in W. J.T.
Broadbent (eds), The PsychologicalMechanisms of Mitchell(ed.), On Narrative(I98I), 99-I15.
Language (I98I), 305-15. 31 'The viewing and obscuringoftheParthenonfrieze',
29 L. Sciascia,I9I2 + I (I986), 17- 3'HS 107 (i 987), 98-i 05.
in
30 cf. R. Brilliant,Visual Narratives: Story-telling

This content downloaded from 128.103.149.52 on Thu, 2 Jan 2014 13:28:41 PM


All use subject to JSTOR Terms and Conditions
30 D. P. FOWLER

simultaneity, and thereare variouswaysofsolvingthis. In theliterarydescriptionofa workof


art, we may findtracesof both linearizationand non-linearization:the cleverestexample I
know is thatof the ekphrasisin Moschus, where the storyof Argos and To which was non-
linearizedby the artistin spatialtermsis thenrelinearizedby Moschus in a way whichallows
both processes to be seen.32Then too, Lessing has oftenbeen criticizedfor takingas his
paradigm sculpture in the round: there is an even more obvious sense of point of view
inscribedin a panel painting.My example fromSciascia of a photographmightbe said to be
cheating:figurativepaintingis not simplythe accuraterepresentation of reality.Nor even is
photography.Arthoweverrealistis no more'naturally'iconicthanliterature:in boththesigns
are read accordingto systemsof meaningthatare culturalconstructs.In theend, visual artis
notsignificantlydifferent fromliterature.
Nevertheless,Lessing's oppositionscan be used to stress a paradox about visual and
literarydescriptionwhichseems to me ofgreatimportanceforcontemporary classicalcultural
criticism.The tendencyofthetraditionalviewofartthatI have been outliningis thatvisualart
mustbe moreopen, less tiedto a pointofview,less fixedin itsinterpretationthanliteraryartin
whicha pointofview is constantlyimposedby the medium.Yet we are used to literarycritics
beingaddictedto ambiguityand polysemy,whileartcriticsare oftenrigidlyhistoricist.This is
particularlytrue of the criticismof ancient art, where the revoltagainst the 'connoisseur'
traditioninstitutedby Beazley in the footstepsof Giovanni Morelli33has used as its principal
weapon a strictlyfunctionalistmethodology.There are of course exceptionsto this: indeed
some of the criticsof ancientartwhose methodologyhas been mostrigorous(and successful)
in itsfunctionalism have also takenpains to stressthepolysemyoftheartisticscenes theyhave
discussed.34But theconfidenthistoricismofmuchancientartcriticismis in strikingcontrastto
the (by now cliched) rhetoricof 'crisis'thathas grippedart historymore generally.35 In the
studyofRoman artespecially,thestressin recentcriticismhas been on remorselessly showing
how artisticproductionservesthe dominantideology.A good example is E. Lefevre'srecent
discussionof the Porticoof the Danaids in Augustus'temple-complexon the Palatine.36The
interpretationof this monument has always been problematic: Lefevre argues that the
Danaids who murderedthesons ofAegyptusare to be seen as a symbolofthe Romans in their
recenttriumphover Cleopatra,theAegyptiaconiunx.Oftenofcoursethe Danaids are seen as
sinners,but forLefevreit is axiomaticthatsuch an interpretation is impossible.37But how
could artistor patronstop ancientreaderstakinga different view of the Danaids, especially
whenthealternativeviewis so stronglyrepresentedin Augustanliterature?38 Lefevre'sviewof
the monumentleads himto tryto read thefamousscene on thebalteus of Pallas in theAeneid
in similarterms,with Pallas as the worthyDanaids and Turnus the dastardlyEgyptians.39
Once the argumentmoves into literature,its deficiencyis clear: one could not say Lefevre's
viewofthebalteus was impossible,but it is easyto show thatverydifferent viewsare (at least)
equally plausible and thatno amountof evidence foran Augustanreadingcould removethe
32 Moschus, Europa, 37-62. Three scenes are (i987), 41-58, at42: 'nosignhasafixedmeaning... signs
described:Io crossingthe sea (44-), Zeus turningher are polysemic... not all the meaningsproducedby the
back intoa woman(5o-4), and thephoenixarisingfrom signifyingelements in a signifiercontributeto the
theblood ofthedead Argos(SS-6i). That is, theorderof productionof one unifiedcoherentmeaning.Some can
thesceneschronologicallyis ACB (or conceivablyABB if producedifferent perspectives,
warringdiscourses,which
the second and third scenes were taken to be con- deconstruct thedominantone.'
temporaneous). In the last scene, first Hermes is 35 cf.D. Preziosi,RethinkingArt History(I989).
described;then'nearby'Argoswithsleeplesseyes; then `6 E. Lefevre,Das Bild-Programm des Apollo-Tempels
thebirdarising(describedintheimperfect,exanetellen); aufdemPalatin,Xenia Heft 24 (i989).
thenits outspreadwings.The temporalsequence in the 37 ibid.,15: 'Es warezudemabsolutwidersinniggewesen
last scene representsthe spatial arrangementon the wennOktavianim Siegesjubeluber die Agypteran dem
cup, but in such a waythatthetemporalsequencein the Siegesmonument parexcellence
eineDarstellung
zugelassen
storythat the visual representationsupposedlyhad to hattein den Agypterals Ehrentrager erschienen.'
delinearize shows through. Reading narrativeart is 38 cf.P. Hardiereviewing Lefevrein CR NS 4o (1990),
relinearization. 520.
33 On the origins of Morelli's methods, see Carlo op. cit. (n. 36), i6, 'TurnusisteinAngreifer,
39 dessen
Ginzburg's brilliant 'Clues: roots of an evidential Tat negativzu bewertenist: fursie hat er zu bussen. Er
paradigm',inMyths,Emblems,Clues, trans.J.and A. C. kann also nur mit den angreifenden Aegyptus-Sohnen
Tedeschi(I990). vergleichen werden,derenschandlichesHandeln (nefas)
See especiallythe work of ChristianeSourvinou- ein schmahlichesEnde (foede)gefundenhat. Und Pallas
Inwood, e.g. 'Menace and pursuit:differentiation
and ist mitden Danaiden verglichen,insoferner freventlich
the creationof meaning',in C. B6rard,C. Bron and A. angegriffen wird.Aufihnbezogenmussdie Aussageder
Pomari (eds), Images et societe'en Grece ancienne. balteus lauten: Pallas wirdgerahtwerden.Man beachte
L'iconographiecomme methoded'analyse, Actes du auch die Anspielung:Pallas kommtvomPalatin!'
Colloque International,Lausanne 8-II fevrierI984

This content downloaded from 128.103.149.52 on Thu, 2 Jan 2014 13:28:41 PM


All use subject to JSTOR Terms and Conditions
NARRATE AND DESCRIBE: THE PROBLEM OF EKPHRASIS 3I

alternativetracesleftby thecomplexityofthetradition.My argumentis notthatartcriticsare


wrongto argue like this,or thatwe should simplyswitchto seeing artas more open: I think
thereis a genuine paradox in our attitudeto art as both more open and more closed than
literaturethatis notto be resolvedby comingdownunambiguouslyon eitherside. But thereis
a strongcase fora greaterawarenessof thisparadox- which,as we shall see, I believe was
recognizedin antiquity.
Our paradoxicalintuition(if by now we have acquired one) thatartis both moreclosed
and moreopen thanliteraturemakesthephenomenonofekphrasiswheretheymeetofpeculiar
interest.Whateverpositionwe adopt as to the degreeto whichplasticand literaryartinscribe
points of view, in literaryekphrasisthe presenceof the intermediary - usually fictional
visual artistintroducesanotherpotentialfocalizer.This is particularlythe case wherethereis
an underlyingnarrativeelementin thevisualrepresentation beingdescribed.Even leavingout
the more extremepossibilitiesthat Kittaymentioned,we then have a complex hierarchyof
potentialpointsofview,whichcan be summarizedin thefollowingdiagram:
Story

ARTIST (A) VisualRepresentation AUDIENCE I (B)

NARRATOR/VIEWER/
TOPIC (C) AUDIENCE 2 (D)
Description
AUTHOR (D) AUDIENCE 3 (E)

That is, of anyelementin a descriptionwe can ask whetherthefocalizationis thatofthe artist


who made the originalworkof art, or his audience, or the observer,or his audience, or the
author,or his audience: and we have stillnotbroughtin theobserver'sbrother-in-law whom
chapterfourwill revealto be theheroofthe novel.

II

Let me tryto makethisclearerwithmyexample,thefamousekphrasisin Book One ofthe


Aeneid whereAeneas looks at the depictionof eventsfromthe Trojan War in the Temple of
Junoin Carthage.i0Aeneas is herean explicitobserver,whosereactionsto whathe sees are also
explicitlystated: he weeps and groansbecause ofwhathe saw, namque videbat ... Moreover,
withinepisodes, as Eleanor Leach observes,41'the order of presentationcreates confusion
betweenthe visual image and Aeneas' thoughts'.This is clearestin the Troilus panel, where
Troilus is depictedbothfugiens(474) and being draggedby his chariot(476):
amissisTroilusarmis,
partealiafugiens
infelix
pueratqueimparcongressus Achilli,
ferturequiscurruquehaeretresupinus inani,
loratenenstamen;huiccervixcomaequetrahuntur
perterram, hasta.
etversapulvisinscribitur
In anotherpartofthepicturepoorTroilus,a mereboyand no matchforAchilles,had losthis
armourandwas infullflight. His horseshadrunawaywiththechariotandhe wasbeingdragged
alonghelplesson hisbackbehindit,stillholdingon to thereins.His neckand hairweretrailing
alongthegroundandtheendofhisspearwasscoring thedustbehindhim.42
40 Aeneid I.441-93. The bibliographyis predictably ekphrasticcentrepieces',HSCP 87 (I983), I75-84; D.
large: see especiallyR. D. Williams,'The pictureson Clay,'The Archaeology ofthetempleto Junoin Carthage
Dido's temple (Aeneid I.450-93)', CQ NS I0 (I960), (Aen. I, 446-93), CP83 (I988), I95-205; Leach, op. cit.
I45-53; inAeneid
'Ironyandforeshadowing
K. Stanley, (n. i); J.J.O'Hara, Death and theOptimistic Prophecyin
I, AJP86 (I965), 276-77; A. Szantyr,'Bemerkung-
462', Vergil'sAeneid (I990), 35-9.
en zum Aufbau der virgilianischenEkphrasis',MH 27 41 op. cit. (n. 32), 3I4-
(I970), 28-40; 'Dido inthelightofhistory',
N. Horsfall, 42
Trans. D. West (Penguin, I990). I use West'snew
PVS I3 (I973-4), I-I3 = S. J. Harrison(ed.), Oxford versionhereand below because his attemptto makethe
Readings in Vergil'sAeneid (I990), I27-40; W. R. descriptionunambiguousforthe readerunderlineshow
Johnson,Darkness Visible(1976), 99-1I4; C. P. Segal, Vergilleanstheotherway. For somecriticism ofthisas a
'Art and the hero: narrativepointof view in Aeneid i', methodoftranslation, see myforthcoming 'BriefNotice'
Arethusa 14 (I98I), 67-84; R. F. Thomas, 'Virgil's inG&R I99I (2).

This content downloaded from 128.103.149.52 on Thu, 2 Jan 2014 13:28:41 PM


All use subject to JSTOR Terms and Conditions
32 D. P. FOWLER

Similarlythe scene between Priam and Achilles fromIliad xxiv, is introducedwith a


reminiscenceofwhathad happened43beforethescene depicted (11.483-7):
tercircumIliacosraptaverat
Hectoramuros
exanimumque aurocorpusvendebatAchilles.
tumveroingentem gemitum datpectoreab imo,
utspolia,utcurrus,utqueipsumcorpusamici
tendentemque manusPriamumconspexit inermis.
TheretoowasAchilles.He haddraggedHectorthreetimesroundthewallsofTroy,andnowwas
sellinghis dead bodyforgold. Aeneasgroanedfromthedepthsof his heartto see thearmour
strippedoffhim,thechariot,thecorpseof his dearfriendand Priamstretchingout his feeble
hands.

Althoughwe are toldthatAeneas sees Iliacas ex ordinepugnas,44theorderand durationofthe


scenes has oftenbeen seen to representa peculiarchoice,whichthrowsemphasison thetragic
elements like the death of Troilus, the firstof so many dyingyouths in the Aeneid. The
strongestversionof thisview indeed sees Aeneas as misreadingthe scenes. Nicholas Horsfall
and others45 have suggestedthatAeneas got it all wrong.On thisview,
Justas onewouldexpectin a templeofJuno,thechoiceofpicturesillustrates thesuccessofher
whileAeneasis delighted
favourites: tosee thatTroyis notforgotten,
he quitefailsto observe,as
we mustdo, thattheattitudeto Troyshowninthesepicturesis neither norsympathetic.
friendly
justthosequalitieswhichCarthaginians
Theyillustrate mightadmireinthevictorious Greeks-
greedandbrutality,
forwhichtheythemselves hadsucha finereputation.46

On this interpretation, we can see peeping out, especiallyin the finalscene of Penthesilea,
whichclearlyanticipatesDido's entry,hintsof a different way of readingthesescenes, a way
farremovedfromsunt lacrimae rerum.
There are good groundsforrejectingthecrude formofthisthesisas it is put by Horsfall,
who is intenton showingthatAeneas did therightthingin leavingDido. In heropeningwords
to Aeneas she tellsof how Teucer put in at Sidon and toldherofthe'casus ... urbis/Troianae',
and how he 'ipse hostis Teucros insignilaude ferebat/ seque ortumantiqua Teucrorum a
stirpevolebat' (623, 625-6). She associates herselfwith the tragicinterpretation of Trojan
history,and to view heras lyingin so doingwould be an extremesubversionto whichthetext
gives no encouragement.Moreover,thereis clearlya battleof paradigmsin the depictionof
Carthage; is it Phaeacia or the land of the Cyclops? The presenceof artis in itselfpartof the
evidencepushingus towardsScherie; and Diskin Clay pointsout thatwhen Odysseus arrives
in the Cyclops' cave, stressis laid on the factthatPolyphemushas not heardof him,whereas
the fameof Troy has reachedPhaeacia.47And finally,as Clay also stresses,one model forthe
scene in the templeand Aeneas' reactionis the storytold variouslyofAristippusand Plato of
the shipwreckedphilosophercoming upon geometricfiguresin the sand: theretoo what is
discovered is indisputable evidence of civilization.48But while the view that has Aeneas
deceiving himselftotallycannot be right,the question of the focalizationof the ekphrasis
remainsof importance.Take, forinstance,the detail of the descriptionof Achilles 'selling'
Hector's body, 'auro ... vendebat' (484). K. Stanley49pointed out that the reader here
naturallythinksnot of the scene in theIliad but ofthe commonscene in artand elsewherein
literatureof Priam weighingout gold, and Stanley,like Horsfall,saw the presenceof such a
brutalscene in the temple as evidence that Aeneas' readingis fatallyoptimistic,though he
drewverydifferent implications,that'in Vergil'sliteraryand historicalperspective,Achilles
and Aeneas, Greek and Trojan, Roman and Tyrianare bound to thatrealmwheretherolesof

4 On the use of the pluperfect,see Szantyr,op. cit. 4 Horsfall,Stanley,Johnson,Clay, Leach, O'Hara,


(n. 32) and especiallyRavenna, op. cit. (n. I), 34-46, opp. citt.(n. 32) (withverydifferent emphases!).
1.
quotingServiuson 484: 'ingentiarteutiturverbis:nam lHorsfall, op. cit. (n. 32), 138.
hoc loco, quia pingi potuit, praesens tempus posuit, 4 Clay,op. cit. (n. 32), 197.
superius, quia pingi non potuit, sed referri,perfecto 4 idem, 195-6. The storyis told of Aristippusin
exsecutus est tempore dicendo "raptaverat" non Vitruviusvi. i, and of Plato in Cicero, Rep. 1.29: see
"raptabat".' G. Giannantoni,I Cirenaici(1958), 213 (Aristippusfrr.
"On ex ordineand similarexpressions,see Ravenna, 42-3) .c
op. cit. (n. i), I6-17- op. cit. (n. 32), 276-7.

This content downloaded from 128.103.149.52 on Thu, 2 Jan 2014 13:28:41 PM


All use subject to JSTOR Terms and Conditions
NARRATE AND DESCRIBE: THE PROBLEM OF EKPHRASIS 33
the slayer and the slayed are inevitablyunited by the reversalsof time - where, indeed,
''sorrowis implicitin the affairsof men"'. The depictionofAchillesas crueland mercenaryis
ofcourseanywayambiguous: thatcould be a Trojan viewas muchas a Greekone. But can we
be certainthat the scene beforeAeneas was the 'brutal' version?vendebat could represent
Aeneas' interpretation of the scene; one mightsay thatauro pushes the readertowardsseeing
the scene as genuinelyone of ransombecause it looks like one of those details in ekphrasis
which are simultaneouslyabout the painterlysurface and an object in the story being
depicted,S?but theexistenceofthegolden ransomis scarcelyunderplayedin theIliad version
of the meeting."'In this ekphrasiswhere so much is clearly'read-in' ratherthan 'seen' -
insofaras such a distinctionholds - it is not impossiblethatthistoo is Aeneas' view rather
thanwhatis actually'there'.
As oftenwithquestionsoffocalization,thereis morethanone storythatwe can tellhereof
whose points of view the picturesand theirdescriptionsrepresent."2Like other scenes of
ekphrasis,the scene is often- and surelyrightlyin some degree- takenas paradigmaticfor
the interpretationof art,both literaryand visual."3Eleanor Leach,s4forinstance,comments
thattheinteractionbetweenAeneas as readerand theworkofart'willnotappear foreignto the
contemporaryreaderwho understandsthatmeaningis notthe inherentpropertyof a textbut
is insteadcreatedin variantformsthroughvariantexperiencesof reconstructing the workas
text', and that Aeneas' 'deeply sentimentalmisreadingof the frieze shows the process of
perceptionas one ofselection,amplification, and reordering,and thusit casts doubt upon the
reliabilityoffactualcommunicationthroughpictorialnarrative'.But one mighttakethescene
as morenormativeand less aporetic,as enjoiningupon thereaderlikeAeneas to read tragically
ratherthantriumphantly, whateverthe picturethatis offered.This has obvious relevanceto
theinterpretation ofAugustanart.As I mentionedabove, contemporary criticismofAugustan
artis dominatedby functionalismand historicism:it attemptsto show how subjectsthatare
apparentlyaestheticallyneutral actually serve Augustan ideology. If we take the scene in
Aeneid I as paradigmatic,however, it suggeststhat more allowance should be made even
withina historicistframework formorethanone way of readingthe symbols:thatafterall an
observermightbe able to deconstructRoman artas well as Roman literature.The suggestion
would notperhapsbe unparalleled.In a well-knownarticleon Philostratusand Homer whose
importanceBartschhas recentlystressed,Leskyhad suggestedthatsome ofthepassages in the
Imagines were clear 'misreadings'of the underlyingpicture,not throughmisunderstanding
but as a tour-de-force of 'sophistischerDeutungskunst'.55Misreadingand cross-readingare
not necessarilymoderncriticalinventions:and the ekphrasisinAeneid i withits stresson the
complexityof interpretation cannotbe entirelyisolatedfromits time. Too New Historicista
readingofAugustanartbeginsto look unhistorical.

III

In conclusion, I want to returnto the issues with which I began in the light of the
complexitiesof focalizationwhich have emergedin the example fromtheAeneid. I said that
mysympathieswere stillverymuch withthe organicistNew Criticalapproach whichwould
seek linksbetweenekphrasisand the narrativeof which it is part,but I also impliedthatthe
challengeof post-moderndislikeof thisas totalizingand authoritarianneeded to be takenon
board. The politicalmetaphorsare of coursebasic to the assaultof post-moderntheoristslike

50 cf. Ravenna, op. cit. (n. i), I4-I6, R. Debray- reading,butunderstandable froma Trojan pointofview;
'La Pierredescriptive',
Genette Poetiqueii (I980), 293- PriaminAeneid2 providesa counterbalance.'
333 on Heliodorusv. I4, whichself-consciously playswith 52 cf. PCPhS 2 I 6 (I 990), 42-63 .
theconvention. 53 This point might be strengthenedby Richard
51 cf. Iliad xxlv. 76, II9, I37, I46-7, I75-6, I95-6, Thomas' suggestion,op. cit. (n. 32), thatthepresenceof
228-37, 367, 38I-2, 435-6, 502, 555, 579, 594, 685-6. I the peplos at the centre of the ekphrasis (479-82)
owethispointtoAlessandroBarchiesi,whocomments:'It constitutesa sort of mise en abyme in the lightof the
is too easyto forgetthatransom,and gold,playsa rolein tradition ofekphrasticpeploi.
Homer'snarrative too. If Aeneaswas a readeroftheIliad 5
op. cit. (n. 32), 323-3.
(and in a sense he is) he could stillpointout exactlythe 5 Bartsch,op. cit. (n. i); A. Lesky, 'Bildwerkund
same points: cruelty,golden ransom,the gestureof a DeutungbeiPhilostrat undHomer',Hernes75 (I940),
father.This would be a selective, and tendentious, 38-53, at45 = GesammelteSchriften (I966), II-25, at I7.

This content downloaded from 128.103.149.52 on Thu, 2 Jan 2014 13:28:41 PM


All use subject to JSTOR Terms and Conditions
34 D. P. FOWLER

Lyotardon theoriesofinterpretation whichaim fora fixedoveralltruth:thatwayfascismlies.


A carnivalesquedialogylooks radicalagainstthat. But it is not simplyresidualStalinismthat
makesMarxistcriticsin particular56 feelunhappyabout theassumptionthatstresson plurality
is alwaysradical. It can be a wayofevadingthedifficult taskofformulatinga properlycomplex
accountof the relationshipof the individualand society.If we move back frompoliticsto the
text,we can similarlytryto formulatean account of the relationshipof a descriptionto its
narrativewhich takes adequate account of complexitybut does not simply liberate the
ekphrasisto meaninglessness.
A way forwardis perhaps to be found in a neglected contributionby Alessandro
Perutelli.s7He setup an oppositionbetweenthetotalsubordinationofdescriptionto narrative
that he saw in the shield descriptionsin the Seven against Thebes58with the total indepen-
dence of descriptionand narrativerepresentedby theAspis of Hesiod, and contrastedboth
withthe relationshipto be seen in Moschus' Europa, wherethereis an 'inversionespeculare':
lo of the ekphrasiscorrespondsnot to Europa but to Zeus as the bull. This relationshiphe
termed neithernarrative (where descriptionis subordinatedto narrative)nor descriptive
(whereit is set free)but rhetorical,conferringon theekphrasisthestatusofafigure.Although
hisexamplesare limitedones, thisseemsa veryimportantinsight.The mostinteresting recent
workon all typesof 'digression'or narrativepause has been thatwhichviewsthe relationship
withthe main narrativeas a figuredone, in which elementsshiftand are transformed as we
movefromdetailto whole. Perutelli's'specularinversion',forinstance,is clearlyrelatedto the
way thatColin Macleod treatedthe mythologicalexample in his well-knownarticle,'A use of
mythin ancientpoetry'.59 Discussing Catullus 68, he notedthatin theexemplumofLaodamia
and Protesilausboth Catullus and Lesbia are comparedand contrastedwithboth figures:

The myth,then,ofCatullus68 is neither a decorative nordoesitsimply


and learnedirrelevance,
mirror in whichit is set; fortheanalogiesbetweenthetwoarequalifiedby no less
thesituation
significant But theresultof such a complexity
contrasts. is notmereconfusion;themyth,by
anareaoffeeling
indicating beyondthedirectstatements ofthepoem,helpstoexpressa significant
It thusmakesa distinct
ofattitudes.
conflict andcomprehensible tothewhole.
contribution

Similarly,in discussingsimiles,OliverLyne has recentlytriedto movebeyondthealternatives


of multiple-correspondence and decorativeindependence:'

There is thusin mostsimilesa visiblepointof contactwiththe narrative and an illustrative


functiontied to it whichis oftenadvertised;in manysimilesfurther pointsof contactand
functions
illustrative canbe discerned.Butthissortoffunction
is notI maintain theimportant or
mainfunction ofa developedsimileinthehandsofa master.The mainfunction ofa simileisnotto
illustratesomethingalreadymentionedin the narrative,but to add thingswhichare not
mentioned, in a differentmedium:imagery. The poetis switching modes,switching fromdirect
narrativeto'narrative' inthesuggestivemediumofimagery; andhecapitalizes onthefactthatheis
nowoperating notan explicitmedium.An advertised
in a suggestive, illustrativefunctionand
concomitant pointofcontactwiththenarrative mayoftenbe seenas a meansto an end,as little
morethana formal devicetoeffecttheswitchfromdirectnarrativeto'narrative'inimagery.61

Similes and even moreexemplabear, of course, different relationshipsto theircontextsfrom


thatwe mightwish to posit forset-piecedescription.But both these formulationsreflecta
similardesireto Perutelli'sforan accountoftherelationofpartto wholewhichis significantin
a non-reductiveway.

56 See e.g. someofthepiecesin thecollection editedby empiricist'comes close to the termsin whicha modern
M. Krieger,TheAimsofrepresentation (i 987), especially Frenchcritic,L. Perrone-Moises, has discusseddescrip-
D. LaCapra, 'CriticismToday'. tion in Balzac, op. cit. (n. i). Distinguishingbetween
57
op. cit. (n. i). 'static'description,'a fonctionredondante,qualificante,
58 The choice is ironicin the lightof Froma Zeitlin's explicativeou emphatique'and 'dynamic','a fonction
Under theSign of the Shield: Semioticsand Aeschylus' de deplacement,de compensation,de defoulement', she
Seven againstThebes(I982). commentsthatwhereasthefirst'renvoiecirculairement a
59 CQ NS 24 (I974), 82-93, at 88 = CollectedEssays un deja-ditdu recit',thesecondproducesanotherlevelof
(I983), I59-70, at i65. narrative:'la descriptionapparaitici noncommeun arret
6 Wordsand thePoet (I989), 68. du recit(pourrenseigner, reposer,distraireou convaincre
61
It is interestingto observehow Lyne the 'English le lecteur),maiscommele suitdu recita un autreniveau'.

This content downloaded from 128.103.149.52 on Thu, 2 Jan 2014 13:28:41 PM


All use subject to JSTOR Terms and Conditions
NARRATE AND DESCRIBE: THE PROBLEM OF EKPHRASIS 35
Perutelli'suse here of the conceptof 'figure'suggestsan obvious comparisonwith Gian
Biagio Conte's similaruse to explainthephenomenonof allusion:62
Thusallusionworksinjustthesameway,andinthesamesemantic area,as a rhetorical
figure. The
gapinfigurative languagethatopensbetween'letter'and'sense'is alsocreatedinallusionbetween
thatwhichis said (as itfirst
appears),a letter,andthethought evoked,thesense.Andjustas no
figureexistsuntilthereaderbecomesawareofthetwofoldnatureoffigurative language,so too
allusiononlycomesintobeingwhenthereadergraspsthatthereis a gap betweentheimmediate
meaning('afterI havesailedthrough manypeoplesand on manyseas'63)andtheimagethatis its
corollary('as Odysseussailed').In theartofallusion,as ineveryrhetoricalfigure,thepoetry liesin
thesimultaneous presenceof twodifferent realitiesthattryto indicatea singlereality... The
poetryliesin theareacarvedoutbetweentheletterandthesense.It existsbyrefusing tobe only
oneortheother.Thisstillunknown area,thistensionbetweenmeanings, canbe described onlyby
referringtothetwoknownlimitsthatdemarcate it.

Justas withallusion,withany passage wherein any sense we fora moment'stand back' from
the narrativewe have the presenceoftwo realities:thepassage takenin isolationand itswider
context.While as with allusion the extremesset the limitsof meaning,what mattersin the
significanceof the simileor ekphrasisor exemplumis thatwhich in Conte's words 'existsby
refusingto be only one or the other'. Any relationshipwe posit is inevitablyan uneasy one.
Preciselybecause the correspondencesand contrastsare figured,the interpretation of them
cannotbe simpleor clear: thereis roomfordisagreement.We are consciousnotonlyofa desire
forintegrationbut ofa resistanceto it. There is morethana whiffoftheZeitgeisthere. I have
excused my use of political terminologyon the grounds that it is so used by post-modern
critics,but thisis an evasion: I believethattheyare correctto see a connectionbetweentextual
and political integration.And it is temptingto suggestthatthe troubled integrationof the
ekphrasisor exampleor simileor intertextual referenceasfigurerepresentsthesame attemptat
apalintonosharmoniaas thecriticsofLyotardand otherpost-modernists have suggestedas an
alternativeto the simplecelebrationof individualismin the politicalsphere.
We have moved a long way fromthe formalistnarratologywithwhich I began: perhaps
too far.But it is important,I believe,to see thatthismovementis inescapable. It is a common
criticismofnarratology thatitis merelyanothertwentieth-century formalism,a wayoftalking
about textswithoutbringingin ideology. In fact, however,if the issues raised by formal
analysisare pursued,we findwe cannotescape the movementtowardspolitics.The relation-
ship betweenthe aestheticand the politicalis not a simple one: the analogies and contrasts
drawn are preciselythemselvesfiguredin the way that I have suggestedare those between
ekphrasisand narrative.And as figured,theycan of course be interpretedin different ways.
But the relationshipis no more to be denied than ekphraseisare to be separatedfromtheir
contexts- or reducedto them.

Jesus College,Oxford

62 G. B. Conte,TheRhetoric
ofImitation(I986), 38-9. 63 Conteis discussing theallusionto theopeningofthe
Odysseyin Catullusioi.

This content downloaded from 128.103.149.52 on Thu, 2 Jan 2014 13:28:41 PM


All use subject to JSTOR Terms and Conditions

You might also like