Ekphrasis and the Other

Ekphrasis and the Other


This article reproduced as part of
the Romantic Circles Electronic Edition of Shelley's "Medusa"
by kind permission of the University of Chicago Press.

"Ekphrasis and the Other" by W. J. T. Mitchell from PICTURE THEORY published by The University of Chicago Press, copyright 1994 by The University of Chicago. All rights reserved.

This material may be used and shared in accordance with the fair-use provisions of US copyright law, and it may be archived and redistributed in electronic form, provided that this entire notice, including copyright information, is carried and provided that the University of Chicago Press is notified and no fee is charged for access. Archiving, redistribution, or replication of this text on other terms, in any medium, requires both the consent of the author and the University of Chicago Press.

Adapted for hypertext by Melissa J. Sites.

undying accents
repeated till
the ear and the eye
lie down together
in the same bed
    --William Carlos Williams
    This otherness, this
"Not-being-us" is all there is to look at
In the mirror, though no one can say
How it came to be this way.
    --John Ashbery, "Self-Portrait in a Convex Mirror"

Radio Photographs: Ekphrastic Poetics

1)    A NYONE who grew up in the age of radio will recall a popular comedy duo called "Bob and Ray." One of their favorite bits was a scene in which Bob would show Ray all the photographs of his summer vacation, accompanying them with a deadpan commentary on the interesting places and lovely scenery. Ray would usually respond with some comments on the quality of the pictures and their subject matter, and Bob would invariably say at some point, as an aside to the audience, "I sure wish you folks out there in radioland could see these pictures." Perhaps this line sticks in my memory because it was such a rare break in the intimacy of Bob and Ray's humor: they generally ignored their radio listeners, or (more precisely) pretended as if the listener was sitting with them in the studio, so fully present to their conversation that no special acknowledgment was required. If one can imagine what it would be to wink knowingly at someone over the radio, one can understand the humor of Bob and Ray. One can also, I think, begin to see something of the fascination in the problem of ekphrasis, the verbal representation of visual representation.[1]

2)   This fascination comes to us, I think, in three phases or moments of realization. The first might be called "ekphrastic indifference," and it grows out of a commonsense perception that ekphrasis is impossible. This impossibility is articulated in all sorts of familiar assumptions about the inherent, essential properties of the various media and their proper or appropriate modes of perception. Bob and Ray's photographs can never be made visible over the radio. No amount of description, as Nelson Goodman might put it, adds up to a depiction.[2] A verbal representation cannot represent--that is, make present--its object in the same way a visual representation can. It may refer to an object, describe it, invoke it, but it can never bring its visual presence before us in the way pictures do. Words can "cite," but never "sight" their objects. Ekphrasis, then, is a curiosity: it is the name of a minor and rather obscure literary genre (poems which describe works of visual art) and of a more general topic (the verbal representation of visual representation) that seems about as important as Bob and Ray's radio photographs.

3)   The minority and obscurity of ekphrasis has not, of course, prevented the formation of an enormous literature on the subject that traces it back to the legendary "Shield of Achilles" in the Iliad , locates its theoretical recognition in ancient poetics and rhetoric, and finds instances of it in everything from oral narrative to postmodern poetry.[3] This literature reflects a second phase of fascination with the topic I will call "ekphrastic hope." This is the phase when the impossibility of ekphrasis is overcome in imagination or metaphor, when we discover a "sense" in which language can do what so many writers have wanted it to do: "to make us see."[4] This is the phase in which Bob and Ray's "radio magic" takes effect, and we imagine in full detail the photographs we hear slapping down on the studio table (Sometimes Bob would acknowledge this moment in a variation of his punchline: instead of a wish, an expression of gratified desire-- "I'm sure glad you folks could look at these pictures with us today.") This is like that other moment in radio listening when the "thundering hoofbeats of the great horse Silver" make the giant white stallion with his masked rider gallop into the mind's eye.[5]

4)   It is also the moment when ekphrasis ceases to be a special or exceptional moment in verbal or oral representation and begins to seem paradigmatic of a fundamental tendency in all linguistic expression. This Is the point in rhetorical and poetic theory when the doctrines of ut pictura poesis and the Sister Arts are mobilized to put language at the service of vision. The narrowest meanings of the word ekphrasis as a poetic mode, "giving voice to a mute art object," or offering "a rhetorical description of a work of art,"[6] give way to a more general application that includes any "set description intended to bring person, place, picture, etc. before the mind's eye."[7] Ekphrasis may be even further generalized, as it is by Murray Krieger, into a general "principle" exemplifying the aestheticizing of language in what he calls the "still moment."[8] For Krieger, the visual arts are a metaphor, not just for verbal representation of visual experience, but for the shaping of language into formal patterns that "still" the movement of linguistic temporality into a spatial, formal array. Not just vision, but stasis, shape, closure, and silent presence ("still" in the other sense) are the aims of this more general form of ekphrasis.[9] Once the desire to overcome the "impossibility" of ekphrasis is put into play, the possibilities and the hopes for verbal representation of visual representation become practically endless. "The ear and the eye lie / down together in the same bed," lulled by "undying accents." The estrangement of the image/text division is overcome, and a sutured, synthetic form, a verbal icon or imagetext, arises in its place.[10]

5)   But the "still moment" of ekphrastic hope quickly encounters a third phase, which we might call "ekphrastic fear." This is the moment of resistance or counterdesire that occurs when we sense that the difference between the verbal and visual representation might collapse and the figurative, imaginary desire of ekphrasis might be realized literally and actually. This is the moment when we realize that Bob and Ray's "wish" that we could see the photographs would, if granted, spoil their whole game, the moment when we wish for the photographs to stay invisible.[11] It is the moment in aesthetics when the difference between verbal and visual mediation becomes a moral, aesthetic imperative rather than (as in the first, "indifferent" phase of ekphrasis) a natural fact that can be relied on. The classic expression of ekphrastic fear occurs in Lessing's Laocoon , where it is "prescribed as a law to all poets" that "they should not regard the limitations of painting as beauties in their own art." For poets to "employ the same artistic machinery" as the painter would be to "convert a superior being into a doll." It would make as much sense, argues Lessing, "as if a man, with the power and privilege of speech, were to employ the signs which the mutes in a Turkish seraglio had invented to supply the want of a voice.''[12]

6)   The tongue, of course, was not the only organ that the mutes in the Turkish seraglio were missing. Lessing's fear of literary emulation of the visual arts is not only of muteness or loss of eloquence, but of castration, a threat which is re-echoed in the transformation from "superior being" to "doll," a mere feminine plaything. The obverse of ekphrasis, "giving voice to the mute art object," is similarly denounced by Lessing as an invitation to idolatry: "superstition loaded the [statues of] gods with symbols" (that is, with arbitrary, quasiverbal signs expressing ideas) and made them "objects of worship" rather than what they properly should be--beautiful, mute, spatial objects of visual pleasure.[13] If ekphrastic hope involves what Francoise Meltzer has called a "reciprocity" or free exchange and transference between visual and verbal art,[14] ekphrastic fear perceives this reciprocity as a dangerous promiscuity and tries to regulate the borders with firm distinctions between the senses, modes of representation, and the objects proper to each.[15]

7)   Ekphrastic fear is not some minor curiosity of German idealist aesthetics. It would be easy to show its place in a wide range of literary theorizing, from the Marxist hostility to modernist experiments with literary space, to deconstructionist efforts to overcome "formalism" and "closure," to the anxieties of Protestant poetics with the temptations of "imagery," to the romantic tradition's obsession with a poetics of voice, invisibility, and blindness.[16] All the goals of "ekphrastic hope," of achieving vision, iconicity, or a "still moment" of plastic presence through language become, from this point of view, sinister and dangerous. All the utopian aspirations of ekphrasis--that the mute image be endowed with a voice, or made dynamic and active, or actually come into view, or (conversely) that poetic language might be "stilled," made iconic, or "frozen" into a static, spatial array--all these aspirations begin to look idolatrous and fetishistic. And the utopian figures of the image and its textual rendering as transparent windows onto reality are supplanted by the notion of the image as a deceitful illusion, a magical technique that threatens to fixate the poet and the listener.

8)   The interplay of these three "moments" of ekphrastic fascination--fear, hope, and indifference--produce a pervasive sense of ambivalence, an ambivalence focused in Bob and Ray's photographs: they know you can't see them; they wish you could see them, and are glad that you can; they don't want you to see them, and wouldn't show them if they could. But to describe this ambivalence as I have done is not to explain it. What is it in ekphrasis that makes it an object of utopian speculation, anxious aversion, and studied indifference? How can ekphrasis be the name of a minor poetic genre and a universal principle of poetics? The answer lies in the network of ideological associations embedded in the semiotic, sensory, and metaphysical oppositions that ekphrasis is supposed to overcome. In order to see the force of these oppositions and associations, we need to reexamine the utopian claims of ekphrastic hope and the anxieties of ekphrastic fear in the light of the relatively neutral viewpoint of ekphrastic indifference, the assumption that ekphrasis is, strictly speaking, impossible.

9)   The central goal of ekphrastic hope might be called "the overcoming of otherness." Ekphrastic poetry is the genre in which texts encounter their own semiotic "others," those rival, alien modes of representation called the visual, graphic, plastic, or "spatial" arts. The "scientific" terms of this otherness are the familiar oppositions of semiotics: symbolic and iconic representation; conventional and natural signs; temporal and spatial modes; visual and aural media. These oppositions, as I have argued at length in Iconology , are neither stable nor scientific. They do not line up in fixed columns, with temporality, convention, and aurality in one row, and space, nature, and visuality in the other. They are best understood as what Fredric Jameson has termed "ideologemes," allegories of power and value disguised as a neutral metalanguage.[17] Their engagement with relations of otherness or alterity is, of course, not determined systematically or a priori, but in specific contexts of pragmatic application. The "otherness" of visual representation from the standpoint of textuality may be anything from a professional competition (the paragone of poet and painter) to a relation of political, disciplinary, or cultural domination in which the "self" is understood to be an active, speaking, seeing subject, while the "other" is projected as a passive, seen, and (usually) silent object. Insofar as art history is a verbal representation of visual representation, it is an elevation of ekphrasis to a disciplinary principle.[18] Like the masses, the colonized, the powerless and voiceless everywhere, visual representation cannot represent itself; it must be represented by discourse.

10)   Unlike the encounters of verbal and visual representation in "mixed arts" such as illustrated books, slide lectures,[19] theatrical presentations, film, and shaped poetry, the ekphrastic encounter in language is purely figurative. The image, the space of reference, projection, or formal patterning, cannot literally come into view. If it did, we would have left the genre of ekphrasis for concrete or shaped poetry, and the written signifiers would themselves take on iconic characteristics.[20] This figurative requirement puts a special sort of pressure on the genre of ekphrasis, for it means that the textual other must remain completely alien; it can never be present, but must be conjured up as a potent absence or a fictive, figural present. These acts of verbal "conjuring" are what would seem to be specific to the genre of ekphrastic poetry, and specific to literary art in general, insofar as it obeys what Murray Krieger calls "the ekphrastic principle." Something special and magical is required of language. "The poem," as Krieger puts it, "must convert the transparency of its verbal medium into the physical solidity of the medium of the spatial arts."[21] This "solidity" is exemplified in such features as descriptive vividness and particularity, attention to the "corporeality" of words, and the patterning of verbal artifacts. The ekphrastic image acts, in other words, like a sort of unapproachable and unpresentable "black hole" in the verbal structure, entirely absent from it, but shaping and affecting it in fundamental ways.

11)   Now the skeptic, indifferent to the magic of ekphrastic hope, is likely to point out that this whole argument is a kind of sham, creating difficulties at the level of representation and signs that do not exist. The "genre" of ekphrasis is distinguished, the skeptic might point out, not by any disturbance or dissonance at the level of signifiers and representational media, but by a possible reference to or thematizing of this sort of dissonance. Ekphrastic poetry may speak to, for, or about works of visual art, but there is nothing especially problematic or unique in this speech: no special conjuring acts of language are required, and the visual object of reference does not impinge (except in analogical ways) upon its verbal representation to determine its grammar, control its style, or deform its syntax. Sometimes we talk as if ekphrasis were a peculiar textual feature, something that produced ripples of interference on the surface of the verbal representation. But no special textual features can be assigned to ekphrasis, any more than we can, in grammatical or stylistic terms, distinguish descriptions of paintings, statues, or other visual representations from descriptions of any other kind of object.[22] Even "description" itself, the most general form of ekphrasis, has, as Gerard Genette has demonstrated, only a kind of phantom existence at the level of the signifier: "the differences which separate description and narration are differences of content, which, strictly speaking, have no semiological existence."[23] The distinctions between description and narration, representation of objects and actions, or visual objects and visual representations, are all semantic, all located in differences of intention, reference, and affective response. Ekphrastic poems speak to, for, or about works of visual art in the way that texts in general speak about anything else. There is nothing to distinguish grammatically a description of a painting from a description of a kumquat or a baseball game. An address or apostrophe to a statue could, from the semiotic standpoint, just as well be an address to a tomato or a tomcat. And the projection of a fictive voice into a vase produces no special ripples or disturbances in the grammar that voice employs. When vases talk, they speak our language.

12)   Our confusion with ekphrasis stems, then, from a confusion between differences of medium and differences in meaning. We are continually falling into some version of Marshall McLuhan's dazzling and misleading metaphor, "the medium is the message." In ekphrasis, the "message" or (more precisely) the object of reference is a visual representation; and, therefore, (we suppose) the medium of language must approximate this condition.[24] We think, for instance, that the visual arts are inherently spatial, static, corporeal, and shapely; that they bring these things as a gift to language. We suppose, on the other side, that arguments, addresses, ideas, and narratives are in some sense proper to verbal communication, that language must bring these things as a gift to visual representation. But neither of these "gifts" is really the exclusive property of their donors: paintings can tell stories, make arguments, and signify abstract ideas; words can describe or embody static, spatial states of affairs, and achieve all of the effects of ekphrasis without any deformation of their "natural" vocation (whatever that may be).[25]

13)   The moral here is that, from the semantic point of view, from the standpoint of referring, expressing intentions and producing effects in a viewer/listener, there is no essential difference between texts and images and thus no gap between the media to be overcome by any special ekphrastic strategies. Language can stand in for depiction and depiction can stand in for language because communicative, expressive acts, narration, argument, description, exposition and other so-called "speech acts" are not medium-specific, are not "proper" to some medium or other. I can make a promise or threaten with a visual sign as eloquently as with an utterance. While it's true that Western painting isn't generally used to perform these sorts of speech acts, there is no warrant for concluding that they could never do so, or that pictures more generally cannot be used to say just about anything.[26] It would be very difficult to account for the existence of pictographic writing systems if pictures could not be employed as the medium for complex verbal expressions.

14)   The independence of so-called "speech acts" from phonetic language is illustrated by the subtlety and range of communication available to the Deaf in visual/gestural sign languages. These signs are not, of course, purely pictorial or linguistic (it's hard to imagine what "purity" of this sort would mean in any medium) but they are necessarily visual , and with same sort of necessity that makes Bob and Ray's photographs "auditory." The mistaken assumption that language is medium-specific, that it must be phonetic, is responsible for what Oliver Sacks has called a "militant oralism" in the education of the Deaf, an insistence that they learn spoken language first. Sacks argues that "the congenitally deaf need to acquire, as early as possible, a complete and coherent language--one not demanding understanding or transliteration of a spoken tongue: this has to be a sign language. . . ."[27]

15)   One lesson of a general semiotics, then, is that there is, semantically speaking (that is, in the pragmatics of communication, symbolic behavior, expression, signification) no essential difference between texts and images; the other lesson is that there are important differences between visual and verbal media at the level of sign-types, forms, materials of representation, and institutional traditions. The mystery is why we have this urge to treat the medium as if it were the message, why we make the obvious, practical differences between these two media into metaphysical oppositions which seem to control our communicative acts, and which then have to be overcome with utopian fantasies like ekphrasis. A phenomenological answer would start, I suppose, from the basic relationship of the self (as a speaking and seeing subject) and the other (a seen and silent object).[28] It isn't just that the text/image difference "resembles" the relation of self and other, but that the most basic pictures of epistemological and ethical encounters (knowledge of objects, acknowledgment of subjects) involve optical/discursive figures of knowledge and power that are embedded in essentialized categories like "the visual" and "the verbal." (Panofsky's opening move into the discipline of "iconology" as a discourse on images is, as we've seen, a visual, gestural encounter with another person.)[29] It is as if we have a metapicture of the image/text encounter, in which the word and the image are not abstractions or general classes, but concrete figures, characters in a drama, stereotypes in a Manichean allegory or interlocutors in a complex dialogue.

16)   The "otherness" we attribute to the image-text relationship is, therefore, certainly not exhausted by a phenomenological model (subject/object, spectator/image). It takes on the full range of possible social relations inscribed within the field of verbal and visual representation. "Children should be seen and not heard" is a bit of proverbial wisdom that reinforces a stereotypical relation, not just between adults and children, but between the freedom to speak and see and the injunction to remain silent and available for observation. That is why this kind of wisdom is transferable from children to women to colonized subjects to works of art to characterizations of visual representation itself. Racial otherness (especially in the binarized "black/white" divisions of U.S. culture) is open to precisely this sort of visual/verbal coding. The assumption is that "blackness" is a transparently readable sign of racial identity, a perfectly sutured imagetext. Race is what can be seen (and therefore named) in skin color, facial features, hair, etc. Whiteness, by contrast, is invisible, unmarked; it has no racial identity, but is equated with a normative subjectivity and humanity from which "race" is a visible deviation. It's not merely a question of analogy, then, between social and semiotic stereotypes of the other, but of mutual interarticulation.[30] That is why forms of resistance to these stereotypes so often take the form of disruptions at the level of representation, perception, and semiosis: Ralph Ellison's Invisible Man is not "empirically" unseen; he refuses visibility as an act of insubordination.[31] The emergence of the racialized other into a visibility that accords with their own vision, conversely, is frequently understood as dependent on their attainment of language and literacy. As Henry Louis Gates and Charles Davis put it, "the very face of the race . . . was contingent upon the recording of the black voice ."[32] If a woman is "pretty as a picture" (namely, silent and available to the gaze), it is not surprising that pictures will be treated as feminine objects in their own right and that violations of the stereotype (ugliness, loquaciousness) will be perceived as troublesome. "Western thought," says Wlad Godzich, "has always thematized the other as a threat to be reduced, as a potential same-to-be, a yet-not-same."[33] This sentence (with its totality of "Western thought" and its "always") performs the very operation it criticizes. Perhaps ekphrasis as "literary principle" does the same thing, thematizing "the visual" as other to language, "a threat to be reduced" (ekphrastic fear), "a potential same-to-be" (ekphrastic hope), "a yet-not-same" (ekphrastic indifference).

17)   The ambivalence about ekphrasis, then, is grounded in our ambivalence about other people, regarded as subjects and objects in the field of verbal and visual representation. Ekphrastic hope and fear express our anxieties about merging with others. Ekphrastic indifference maintains itself in the face of disquieting signs that ekphrasis may be far from trivial and that, if it is only a sham or illusion, it is one which, like ideology itself, must be worked through. This "working through" of ekphrastic ambivalence is, I want to suggest, one of the principal themes of ekphrastic poetry, one of the things it does with the problems staged for it by the theoretical and metaphysical assumptions about media, the senses and representation that make up ekphrastic hope, fear, and indifference.

18)   So far I have been treating the social structure of ekphrasis mainly as an affair between a speaking/seeing subject and a seen object. But there is another dimension to the ekphrastic encounter that must be taken into account, the relation of the speaker and the audience or addressee of the ekphrasis [34] The ekphrastic poet typically stands in a middle position between the object described or addressed and a listening subject who (if ekphrastic hope is fulfilled) will be made to "see" the object through the medium of the poet's voice. Ekphrasis is stationed between two "othernesses," and two forms of (apparently) impossible translation and exchange: (1) the conversion of the visual representation into a verbal representation, either by description or ventriloquism; (2) the reconversion of the verbal representation back into the visual object in the reception of the reader. The "working through" of ekphrasis and the other, then, is more like a triangular relationship than a binary one; its social structure cannot be grasped fully as a phenomenological encounter of subject and object, but must be pictured as a menage a trois in which the relations of self and other, text and image, are triply inscribed. If ekphrasis typically expresses a desire for a visual object (whether to possess or praise), it is also typically an offering of this expression as a gift to the reader.[35]

19)   The fullest poetic representation of the ekphrastic triangle is probably to be found in the Greek pastoral or, to use Theocritus's term, "idylls" ("little pictures"). These poems often present singing contests between male shepherds who regale each other with lyric descriptions of beautiful artifacts and women and who exchange material gifts as well. As Joshua Scodel notes:

The art objects used as gifts or prizes may be read as the rewards poets should receive for their productions--honor, fame, money,etc.--and the ekphrastic description asserts poetry's worth by showing that poetry can indeed "capture in and receive such valuable things. But the ekphrastic objects are normally treated as compensatory substitutes for the unfulfilled desire for a female Other.... the ekphrastic object registers both the woman that the poet cannot capture in poetry and the possibility of another, beneficent relationship between a poet and his male audience.[36]

20)   I'm not suggesting that the triangle of ekphrasis invariably places a feminized object "between men" (to echo Eve Sedgwick's memorable study of the "exchange of women"). Bob and Ray's photographs are enjoyed by them in common, in an intimate male friendship that excludes (while pretending to include) their entire audience. Whatever specific shape the ekphrastic triangle may take, it provides a schematic metapicture of ekphrasis as a social practice, an image that can now be tested on a number of texts.

Cups and Shields: Ekphrastic Poetry

21)   The earliest examples of ekphrastic poetry are not, it seems, principally focused on painting, but on utilitarian objects that happen to have ornamental or symbolic visual representations attached to them Goblets, urns, vases, chests, cloaks, girdles, various sorts of weapons and armor, and architectural ornaments like friezes, reliefs, frescos and statues in situ provide the first objects of ekphrastic description probably because the detachment of painting as an isolated, autonomous, and moveable object of aesthetic contemplation is a relatively late development in the visual arts. The functional context of ekphrastic objects is mirrored in the functionality of ekphrastic rhetoric and poetry as a subordinate feature of longer textual structures. Ekphrastic poetry originates, not as an independent verbal set-piece, but as an ornamental and subordinate part of larger textual units like the epic or pastoral. Homer, as Kenneth Atchity points out, develops the significance of his hero "by associating Achilles, in one place or another, with nearly every . . . kind of artifact in the Iliadic world. Like Priam and Hekabe, Achilles stores his treasures in a splendid ["elaborately wrought"] chest; like Nestor, he possesses a remarkably ["wrought"] goblet."[37] Occasionally one of these objects is singled out for special attention and extended description, and these occasions are the putative "origin" of ekphrasis.

22)   I'd like to begin, however, not at the origin, but with a very late and pure example of ekphrastic poetry, Wallace Stevens's "Anecdote of a Jar."

I placed a jar in Tennessee,
And round it was, upon a hill.
It made the slovenly wilderness
Surround that hill.
The wilderness rose up to it,
And sprawled around, no longer wild.
The jar was round upon the ground
And tall and of a port in air.
It took dominion everywhere.
The jar was gray and bare.
It did not give of bird or bush,
Like nothing else in Tennessee.

23)   Stevens's poem provides an allegory and a critique of its own generic identity and might almost be seen as a parody of the classical ekphrastic object. The jar appears, not as a crucial feature of an epic narrative, an ornament to a pastoral drama, or even a focus for elevated lyric meditation. Instead, the jar is the principal actor in a mere "anecdote," literally, an "unpublished," private story that is "not given out" and the tone of which would have to be described as impassive, neutral, and unforthcoming. The jar is stripped, moreover, of all the features that constitute the generic expectations of ekphrasis. There are no "elaborately wrought" ornaments, no pictured scenes or "leaf-fringed legends" of God or man, only the purely functional made object, a simple "commodity" in all its prosaic lack of splendor. Indeed, the apparent lack of visual representation on the urn might lead us to rule this out of the genre of ekphrasis altogether, at least from that central tradition of ekphrasis devoted to the description of aesthetic objects. It is as if Stevens were testing the limits of the genre, offering us a blank space where we expect a picture, a cipher in the place of a striking figure, a piece of refuse or litter where we look for art.

24)   And yet the emphasis on the shape, appearance, and what can only be described as the "activity" of the jar make it clear that this is no "mere" object, but a highly charged form and a representational form at that. If the jar negates the classical expectation of ornamental design or representational illusion, it alludes quite positively to the whole topos of the personified artifact and, specifically, the biblical trope of man as a clay vessel and God as the potter: "we are the clay, and thou our potter; and we are all the work of thy hand" (Isaiah 64:8). The creation of Adam (whose name means clay) out of the dust of the ground is the making of a human "vessel" into which life may be breathed. Stevens simply literalizes the biblical metaphor, or reverses the valence of tenor and vehicle: instead of man-as-jar, jar-as-man--a jar that (like Adam) stands solitary and erect on a hilltop, organizes the wilderness, and "takes dominion everywhere." Stevens also replaces the role of the creating deity, god-as-potter, with something more like a detached observer and "distributor." The narrator does not make the jar; he "places" it; he doesn't show the production of the ekphrastic object (as in Homer's description of Achilles' shield) but its distribution and influence; and he doesn't place it in a new world or a wilderness, but "in Tennessee," a world already named, mapped, and organized. If "Adam" as vessel is reduced to a personified commodity, a barren jar, "God" the speaker/viewer is reduced to a traveling salesman with a sense of irony and order.

25)   The jar's role as a kind of empty relay point or site of commerce and exchange is signaled by its characterization as "a port in air," and Stevens employs it to exemplify three forms of alterity and difference in its relation to its "world" and to the poet: (1) the difference between the human and the natural (the jar is "like nothing else in Tennessee"); (2) between the divine and the human (suggested by the allusion to the biblical creation myth); and (3) between the human and its own artificial productions. And yet each of these forms of otherness is simultaneously overcome: the "slovenly wilderness" (which is already "Tennessee") is made to "surround" the jar in imitation of its roundness; the creaturely subject becomes a sovereign; and the static, spatial image of ekphrastic description is temporalized as the principal actor in a narrative. The poem stages for us the basic project of ekphrastic hope, the transformation of the dead, passive image into a living creature. It does not witness this transformation as a fulfillment of hope, however, but as something more like a spectacle that induces a stunned, laconic ambivalence in the speaker, a kind of suspension between ekphrastic indifference and fear. The jar is, despite all the connotations of orderliness and aesthetic formality that commentators usually celebrate, a kind of idol or fetish, a made object that has appropriated human consciousness. There is something menacing and slightly monstrous about its imperious sterility ("It did not give of bird or bush / Like nothing else in Tennessee.") The double negatives ("did not give" / "Like nothing") perfectly enact the power of the ekphrastic image as intransigent Other to the poetic voice, its role as a "black hole" in the text. Is the jar simply barren and infertile, as empty within as it is "gray and bare" on its surface? Or is it the womb of a more active negation, a "port" that "gives of" an empire of nothingness, fruitfully multiplying its own sterility throughout its dominions? And what, we must ask, does the poet give the reader? Perhaps an "anecdote" that not only refuses to "give itself out," but actively negates every expectation of ekphrastic pleasure.

26)   As the metaphors of barrenness and fertility suggest, there is just a hint of the feminine in Stevens's jar, and the treatment of the ekphrastic image as a female other is a commonplace in the genre. I've already suggested that female otherness is an overdetermined feature in a genre that tends to describe an object of visual pleasure and fascination from a masculine perspective, often to an audience understood to be masculine as well. Ekphrastic poetry as a verbal conjuring up of the female image has overtones, then, of pornographic writing and masturbatory fantasy (the image of the jar as both "round upon the ground" and "tall and of a port in air" lends it an equivocal status as a kind of "phallic womb," as if the jar were both the erection and the visual image that provokes it). Rousseau's Confessions provides the classic account of masturbation and visuality: "This vice . . . has a particular attraction for lively imaginations. It allows them to dispose, so to speak, of the whole female sex at their will, and to make any beauty who tempts them serve their pleasure without the need of first obtaining her consent."[38] Rousseau "fondling her image in [his] secret heart" is like the male poet or reader fondling the mental image of ekphrasis, indulging the pleasures of voyeurism, actual or remembered.

27)   This, at any rate, is one of the mechanisms of ekphrastic hope, but it rarely occurs without some admixture of ekphrastic fear, as we have seen in Stevens's image of the jar. The voyeuristic, masturbatory fondling of the ekphrastic image is a kind of mental rape that may induce a sense of guilt, paralysis, or ambivalence in the observer. A good example of voyeuristic ambivalence is the divided voice in William Carlos Williams's "Portrait of a Lady," an ekphrastic poem that may be an address to a woman who is compared to a picture, or a woman in a picture:

Your thighs are appletrees
whose blossoms touch the sky.
Which sky? The sky
where Watteau hung a lady's slipper.
Your knees are a southern breeze--or
a gust of snow. Agh! what
sort of a man was Fragonard?
--as if that answered
anything. Ah, yes--below
the knees, since the tune
drops that way, it is
one of those white summer days,
the tall grass of your ankles
flickers upon the shore--
Which shore?--
the sand clings to my lips--
Which shore?
Agh, petals maybe. How
should I know?
Which shore? Which shore?
I said petals from an appletree.

28)   The intrusive questioning voice may be that of the addressed woman, or of the poet's unconscious, or of the poem's implied reader, impatiently Interrupting the ekphrasis to demand more clarity and specificity. In either case, the voice resists the smooth, pleasurable fondling of the ekphrastic image, the sensuous contemplation of the woman's body, mediated through the familiar metaphors of fruit, blossoms, petals, wind, and sea. It helps, of course, to know that the poem probably alludes to Fragonard's The Swing , a sensuous rococo pastoral depicting a young swain delightedly looking up the dress of a young woman on a swing. Williams may be projecting his voices into this picture, imagining a kind of dialogue between the swinging maiden and the youthful voyeur. Whether this painting is to be understood as the scene of the poem or only as a passing comparison to a real scene, the ambivalence of voyeurism--the desire to see accompanied by a sense of its prohibition--seems to underlie the dissonance of the two voices. The discreet, metaphorical evasions of "thighs" as "appletrees / whose blossoms touch the sky" is immediately countered by an insistence on explicit literalness: "Which sky?" The answer, of course, can only be the heaven to which the young man wants to ascend, deferred figuratively as a sky in yet another painting, "where Watteau hung a lady's slipper." The speaker can even find a figure for his own ambivalence between hot desire and cold diffidence in his response to her "knees" as a "southern breeze--or / a gust of snow." The movement of the speaker's gaze and of his "tune" follows the line of discretion and diffidence: from thighs whose "blossoms" touch the sky, the "tune drops" to knees, ankles, and the "shore" on which those ankles "flicker." As in Freud's primal scene of fetishism, the boy's gaze finds something threatening in the female image (for Freud, the woman's lack of a penis threatens the boy with the possibility of castration), so he must find metonymic or metaphoric substitutes--a lady's slipper "hung" in the sky; a seashore where the poet seems to fall on his face, the sand clinging to his lips.[39] The more intricate the metaphorical evasions, the more insistent the voice of literal desire, demanding to know "which sky?" "which shore?" and the more fixated and repetitious the poetic voice. The fetishist finally returns to the first substitute image he seized upon, rejecting the traditional fetish-objects of slippers and feet in favor of the initial metaphor of the poem: "I said petals from an appletree."[40]

29)   The most famous instance of ekphrastic ambivalence toward the female image is Keats's "Ode on a Grecian Urn," which presents the ekphrastic object as a potential object of violence and erotic fantasy: the "still unravish'd bride" displays scenes of "mad pursuit" and "struggle to escape." Yet all this "happy, happy love" adds up to a sterile, desolate perfection that threatens the adequacy of the male voice (the urn can "express a flowery tale more sweetly than our rhyme") and "teases us out of thought." Keats may call the urn a "friend to man," but he treats her more like an enemy who only shows up at funerals ("in midst of other woe"), where she pipes the repetitious and comfortless ditty of no tone, "Truth is beauty, beauty truth / That is all ye know on earth and all ye need to know." Perhaps the scholarly controversy over the boundary between what the urn says and what Keats says reflects a kind of ekphrastic disappointment. If the poet is going to make the mute, feminized art object speak, he could at least give her something interesting to say.[41]

30)   A more compelling strategy with the ekphrasis of the female image is suggested by Shelley's manuscript poem, "On the Medusa of Leonardo Da Vinci in the Florentine Gallery." I'm not saying this is a better poem than "Ode on a Grecian Urn," but I do think it lays bare the ekphrastic anxieties that underlie the urn.

31)   If ekphrastic poetry has a "primal scene," this is it. Shelley's Medusa is not given any ventriloquist's lines. She exerts and reverses the power of the ekphrastic gaze, portrayed as herself gazing, her look raking over the world, perhaps even capable of looking back at the poet. Medusa is the image that turns the tables on the spectator and turns the spectator into an image: she must be seen through the mediation of mirrors (Perseus' shield) or paintings or descriptions. If she were actually beheld by the poet, he could not speak or write; if the poet's ekphrastic hopes were fulfilled, the reader would be similarly transfixed, unable to read or hear, but perhaps to be imprinted with alien lineaments, the features of Medusa herself, the monstrous other projected onto the self.

32)   Medusa is the perfect prototype for the image as a dangerous female other who threatens to silence the poet's voice and fixate his observing eye. Both the utopian desire of ekphrasis (that the beautiful image be present to the observer) and its counterdesire or resistance (the fear of paralysis and muteness in the face of the powerful image) are expressed here. All of the distinctions between the sublime and the beautiful, the aesthetics of pain and pleasure, or of the masculine and the feminine, that might allow ekphrasis to confine itself to the contemplation of beauty are subverted by the image of Medusa. Beauty, the very thing which aestheticians like Edmund Burke thought could be viewed from a safe position of superior strength, turns out to be itself the dangerous force: "it is less the horror than the grace" that paralyzes the observer. Medusa fully epitomizes the ambivalence that Keats hints at: instead of "teasing us out of thought" with a paralyzing eternity of perfect desolation, she paralyzes thought itself, first, by turning "the gazer's spirit into stone," and then by engraving the lineaments of the Gorgon onto the beholder's petrified spirit, "till the characters be grown / Into itself, and thought no more can trace."[42] If ekphrasis, as a verbal representation of a visual representation, is an attempt to repress or "take dominion" over language's graphic Other, then Shelley's Medusa is the return of that repressed image, teasing us out of thought with a vengeance.

Figure A. Leonardo's Medusa. Alinari/Art Resource, NY.

33)   Shelley's own voice and text, however, seem designed to deconstruct, not just the repression of Medusa, but the genre of ekphrasis as a verbal strategy for repressing/representing visual representation. The text seems to struggle to efface itself and any other "framing" elements that might intervene between the reader and the image: that this is a painting is mentioned only in the title; after that, Medusa is described as if she were directly present to the poetic observer, and Leonardo's painterly authority vanishes ( in fact, as Shelley may have known, the painting was not by Leonardo).[43] Roles of verbal authority like Stevens's anecdotal narrator, or Keats's apostrophizing poet are eschewed in favor of the purest, most passive sort of description. The speaking and seeing subject of this poem does not speak of or (in a sense) even see Medusa or the painting in which she is represented (figure A): the opening three words appropriate both these roles: "It lieth, gazing." Medusa, the supposed "seen object" of the poem, is presented as herself the active gazer; other possible "gazers" on this spectacle are presented as passive recipients. The voice of the poem is simply a passive recorder, an "ever-shifting mirror" that traces the "unending involutions" of its subject. The pun on "lieth" suggests that the mute ekphrastic object awaiting the ventriloquism of the poetic voice already has a voice of its own.

Medusa is not personified or "given a voice," therefore, to dictate her own story: that would simply amount to a reinscription of poetic authority in the speaker. Instead, the subject of the poem remains irrevocably Other, an "it" that can only "be described" by an anonymous, invisible, and passive poet who has himself been imprinted by Medusa. "It" is the principal agent in the frozen action of the poem; "'Tis" and "Yet 'tis" are its favorite predications. It is as if Being itself were describing itself in and inscribing itself on Shelley's text.

34)   But it is not just the ahistorical and mythic presence of Medusa that Shelley contemplates in this poem. Medusa was a potent image in British cultural politics in the early nineteenth century, deployed as an emblem of the political Other, specifically the "glorious Phantom" of revolution, which Shelley (like many other radical intellectuals) was prophesying in 1819.[44] The use of the female image of revolution was, of course, a commonplace in nineteenth-century iconography, Delacroix's bare-breasted Liberty Leading the People being the most familiar example. This was an image that could be conjured with by both radicals and conservatives: Burke caricatured the revolutionaries of the 1790s as a mob of transvestites and abandoned women, comparing them to "harpies" and "furies" reveling in "Thracian orgies"; Shelley, twenty-five years later, could visualize his revolutionary avenging angel as "some fierce Maenad" whose "bright hair uplifted from the head" is "the locks of the approaching storm," an image that suggests a link between the menacing locks of Medusa and the orgiastic women of Burke's reactionary fantasies.

Figure B. The Contrast, 1792. Emblems of Athena and Medusa, or British Happiness versus French Misery.

35)   But Shelley would not have needed to go to Burke to compose his image; Medusa was, as Neil Hertz has shown, a popular emblem of Jacobinism and was often displayed (figure B)as a figure of "French Liberty" in opposition to "English Liberty," personified by Athena, the mythological adversary of Medusa.[45] The choice of Medusa as a revolutionary emblem seems, in retrospect, quite overdetermined. To conservatives, Medusa was a perfect image of alien, subhuman monstrosity--dangerous, perverse, hideous, and sexually ambiguous: Medusa's serpentine locks made her the perfect type of the castrating, phallic woman, a potent and manageable emblem of the political Other. To radicals like Shelley, Medusa was an "abject hero," a victim of tyranny whose weakness, disfiguration, and monstrous mutilation become in themselves a kind of revolutionary power. The female image of ekphrasis is not an object to be caressed and fondled with contemplative ambivalence like Keats's Urn, Stevens's Jar, or Williams's Lady, but a weapon to be wielded. (Athena's shield or "aegis" is decorated with the head of Medusa, the perfect image to paralyze the enemy.) But this weapon is already latent in the masturbatory fantasies of ekphrastic beauty and shapeliness: it is simply the aggressive, exhibitionist answer to the voyeuristic pleasure staged in ekphrastic urns and jars. Freud's comments on Medusa are worth quoting here:

If Medusa's head takes the place of the representation of the female genitals, or rather if it isolates their horrifying effects from the pleasuregiving ones, it may be recalled that displaying the genitals is familiar in other connections as an apotropaic act. What arouses horror in oneself will produce the same effect upon the enemy against whom one is seeking to defend oneself. We read in Rabelais of how the Devil took to flight when the woman showed him her vulva.[46]

36)   Understood as what Gombrich calls an "apotropaic image," a deadly, monstrous, paralyzing spectacle, the visual image of ekphrasis is properly located, not on an urn or jar, but on a shield which may be displayed to the enemy while protecting its bearer. It seems entirely fitting, then, that the canonical "origin" of classical ekphrasis is the description of Achilles' shield in the 18th book of the Iliad. What is odd about Achilles' shield in this context, however, is that it does not contain an apotropaic image, but an encyclopedic vision of the Homeric world, filled with narrative scenes rather like those we find on Keats's urn. Achilles' shield does not paralyze its beholders with a frightful monstrosity, but overawes them with its impression of divine artifice, an emblem of the irresistible destiny of its bearer. The shield is a perfectly balanced image of fear and hope, serving both as a "beacon" to rally the Achaeans and as a spectacle to dazzle the enemy ("Trembling took hold of all the Myrmidons. None had the courage to look straight at it.") Indeed, it's not clear that anyone in the poem actually examines the shield. Achilles is just happy to have such formidable equipment, and the blind Homer of course can't claim to have seen it. He is just repeating what he has heard from the muses. The image on Achilles' shield is really for us, the readers and listeners who are given time, by virtue of ekphrasis, to look at its production and appearance in detail.[47]

37)   As a generic prototype, Achilles' shield has a capacity to focus ekphrastic fear and hope in the reader as well. For the tradition of literary pictorialism, Homer's set-piece certifies the ancient pedigree of ekphrasis and provides a model for poetic language in the service of pictorial representation. For antipictorialists, the passage is a problem to be explained away. Lessing, for instance, treats the shield not as a prototype of ekphrasis, but as an alternative to it. The distinctive feature of Homer's description is that it is not really a descriptive "arresting of movement" at all, but a continuation of the narrative:

Homer does not paint the shield finished, but in the process of creation. Here again he has made use of the happy device of substituting progression for co-existence, and thus converted the tiresome description of an object into a graphic picture of an action. We see not the shield, but the divine master-workman employed upon it. (p. 114)

38)   An apostle of ekphrastic hope like Jean Hagstrum must, of course, dispute Lessing's argument:

This is not, as Lessing believed, the presentation of an action or a process. The reader's attention is made to concentrate on the object which is described panel by panel, scene by scene, episode by episode. It is remarkable how the eye is held fast to the shield. (p. 19)

39)   Let's look for ourselves at a representative sample of this passage:

He made the earth upon it, and the sky, and the sea's water
and the tireless sun, and the moon waxing into her fullness,
and upon it all the constellations that festoon the heavens,
the Pleiades and the Hyades and the strength of Orion
and the Bear, whom men give also the name of the Wagon,
who turns about in a fixed place and looks at Orion
and she alone is never plunged in the wash of the Ocean.

40)   This passage can, in fact, be mobilized to support either Lessing or Hagstrum, or perhaps neither. If we insist on thinking of this passage as an occasion for visualization, it seems clear that the reader is stationed in a position to move freely from visualization, it seems clear that the reader is stationed in a position to move freely from visualizations of Hephaestos at work to the images on which he works, and equally free to see the images in motion or at rest. The passage equivocates between the categories of time and space, like the image of "the Bear . . . who turns about in a fixed place." The question is one of foreground and background, or scene and frame, or the emphasis imposed by the reader's desire: certainly the god's activity frames each scene, bracketing its descriptive noun phrases with the predicates of making. Just as certainly, the made object seems always already to be completed, even as it comes into being. The same ambiguity arises when we fix our eye on the images themselves. Cedric Whitman notes that "one is not sure whether the pictures on the shield are static or alive. Homer, in fact, is not quite sure just what kind of pictures are made by Hephaestos, whose golden automata have mind and move by themselves."[48] The supposedly "static" images that Lessing wants to temporalize with the verbs of making, and which Hagstrum wants to "hold fast" in the mind's eye, are already in motion, already narrativized, as Hagstrum's own phrase, "episode by episode," should suggest. A similar undecidability characterizes the description of the materials of the shield, and the objects those materials signify:

. . . The oxen turned in the furrows,
straining to come to the end of the new-ploughed land;
all lay black behind them, and looked like ploughed soil,
yet it was gold; indeed, a very wonder was wrought.

41)   Homer's whole point seems to be to undermine the oppositions of movement and stasis, narrative action and descriptive scene, and the false identifications of medium with message, which underwrite the fantasies of ekphrastic hope and fear. The shield is an imagetext that displays rather than concealing its own suturing of space and time, description and narration, materiality and illusionistic representation.[49] The point of the shield from the reader's perspective is thus quite distinct from its function for those who behold it within the fiction. We are stationed at the origin of the work of art, at the side of the working Hephaestos, in a position of perceptual and interpretive freedom. This is a utopian site that is both a space within the narrative, and an ornamented frame around it, a threshold across which the reader may enter and withdraw from the text at will.

42)   The double face of Achilles' shield becomes even more evident when we ask what its imagery represents as a whole, how it functions in the Iliad , and what its status is as the "parent" or prototype of a genre called ekphrasis. Lessing, of course, wants to deny the existence of ekphrasis as a proper genre, a separable, identifiable poetic kind. Achilles' shield cannot be separated from its place in the epic; to emulate this sort of set-piece as an independent literary kind is to produce exactly the sort of visual fixation he fears. Even within an epic context, ekphrasis threatens to separate itself, as it does, according to Lessing, in Virgil's description of the shield of Aeneas. The Aeneid wrongly separates the making of the shield from Aeneas's viewing of it, thus producing an image framed by predicates of vision ("admires"; "are seen") and plain indicatives ("here is"; "there is"; "near by stands").[50] Lessing's rhetoric of blame for this practice is worth quoting:

The shield of Aeneas is therefore, in fact, an interpolation, intended solely to flatter the pride of the Romans; a foreign brook with which the poet seeks to give fresh movement to his stream. The shield of Achilles, on the contrary, is the outgrowth of its own fruitful soil.

43)   The link between ekphrasis and otherness could not be clearer. Ekphrasis is, properly, an ornament to epic, just as Hephaestos's designs are an ornament to Achilles' shield. But ekphrastic ornament is a kind of foreign body within epic that threatens to reverse the natural literary priorities of time over space, narrative over description, and turn the sublimities of epic over to the flattering blandishments of epideictic rhetoric.

44)   If Lessing could have seen the subsequent development of Homeric criticism, he would have found his worst fears justified. Not only did ekphrasis establish itself firmly as a distinct poetic genre, but the great prototype of Achilles' shield seems, in the work of modern classical scholarship imbued with assumptions of formalism, to have established a kind of dominance over the epic of which it is supposed to be a mere ornament. This is not just a question of the popularity of the passage as an object of teaching, research, and analysis. It also has to do with a sense of the function of the shield as an emblem of the entire structure of the Iliad ; the shield is now understood as an image of the entire Homeric world-order, the technique of "ring composition" and geometrical patterning that controls the large order of the narrative, and the microscopic ordering of verse and syntax.[51] Indeed, the shield (and ekphrastic hope along with it) may have even more grandiose aspirations than this sort of synecdochical representation of the whole in the part, for the shield represents much more of Homer's world than the Iliad does.[52] The entire universe is depicted on the shield: nature and man; earth, sky, and ocean; cities at peace and at war; plowing, harvest, and vintage; herding and hunting; marriage, death, and even a scene of litigation, a prosaic alternative to the settlement of disputes by war or blood revenge. Achilles' shield shows us the whole world that is "other" to the epic action of the Iliad , the world of everyday life outside history that Achilles will never know. The relation of epic to ekphrasis is thus turned inside out: the entire action of the Iliad becomes a fragment in the totalizing vision provided by Achilles' shield.

45)   Like Stevens's jar, Achilles' shield illustrates the imperial ambitions of ekphrasis to take "dominion everywhere." These ambitions make it difficult to draw a circle around ekphrasis, to draw any finite conclusions about its nature, scope, or place in the literary universe. Ekphrasis resists "placement" as an ornamental feature of larger textual structures, or as a minor genre. It aims to be all of literature in miniature, as Murray Krieger's reflections on the ekphrastic principle demonstrate. My own claim about ekphrasis would be both more and less sweeping. I don't think ekphrasis is the key to the difference between ordinary and literary language, but merely one of many figures for distinguishing the literary institution (in this case, by associating verbal with visual art). I do think ekphrasis is one of the keys to difference within language (both ordinary and literary) and that it focuses the interarticulation of perceptual, semiotic, and social contradictions within verbal representation. My emphasis on canonical examples of ekphrasis in ancient, modern and romantic poetry has not been aimed at reinforcing the status of this canon or of ekphrasis, but at showing how the "workings" of ekphrasis, even in its classic forms, tends to unravel the conventional suturing of the imagetext and to expose the social structure of representation as an activity and a relationship of power/knowledge/desire--representation as something done to something, with something, by someone, for someone. Stevens's jar, in my view, doesn't fulfill the poetic genre of ekphrasis so much as it implodes it, parodying the whole gesture of the utopian imagetext.

46)   My examples are also canonical in their staging of ekphrasis as a suturing of dominant gender stereotypes into the semiotic structure of the imagetext, the image identified as feminine, the speaking/seeing subject of the text identified as masculine. All this would look quite different, of course, if my emphasis had been on ekphrastic poetry by women. But the difference, I would want to insist, would not be simply readable as a function of the author's gender. The voice and "gaze" of the male, as Williams's "Portrait of a Lady" should make clear, is riddled with its own counter-voices and resistances, and no one is going to blame the Grecian urn for the banalities Keats forces her to utter.

47)   The more important point is to see that gender is not the unique key to the workings of ekphrasis, but only one among many figures of difference that energize the dialectic of the imagetext. The alien visual object of verbal representation can reveal its difference from the speaker (and the reader) in all sorts of ways: the historical distance between archaic and modern (Keats's Urn); the alienation between the human and its own commodities (Stevens's Jar); the conflict between a moribund social order and the monstrous revolutionary "others" that threaten it (Shelley's Medusa); the gap between a historical epic obsessed with war and a vision of the everyday, nonhistorical order of human life that provides a framework for a critique of that historical struggle (Homer's Shield). It is also clear that the otherness of the ekphrastic image is not just defined by the subject matter of the visual representation, but also by the kind of visual representation it is (metal engravings and inlays on a shield; paintings on an urn; a rococo pastoral by Fragonard; an anonymous Renaissance oil painting; a plain, unadorned jar). I have not mentioned the verbal representation of other kinds of visual representation such as photography, maps, diagrams, movies, theatrical spectacles, nor reflected on the possible connotations of different pictorial styles such as realism, allegory, history painting, still-life, portraiture, and landscape, each of which carries its own peculiar sort of textuality into the heart of the visual image. This treatment of ekphrasis, then, like the typical ekphrastic poem, will have to be understood as a fragment or miniature.


1. This definition of ekphrasis as "the verbal representation of visual representation is also the basis for James Heffernan's article, "Ekphrasis and Representation," New Literary History 22, no. 2 (Spring 1991): 297-316. See also Heffernan's The Museum of Words: The Poetics of Ekphrasis from Homer to Ashbery (Chicago University of Chicago Press, 1994).Back

2. Nelson Goodman,Languages of Art (Indianapolis: Hackett, 1976), p. 231. Back

3. For a good survey of this scholarship, see Grant F. Scott, "The Rhetoric of Dilation: Ekphrasis and Ideology," Word & Image 7, no. 4 (October-December 1991): 301-10. Back

4. The pun in "cite"/"sight" above might be cited (or sighted) as an example of a "literal" (that is, conveyed by letters") intrusion of visual representation into verbal representation. Back

5. The iconic character of radio "sound images" is a nonverbal form of ekphrasis. These images (onomatopoeic thundering, studio sound effects) might be said to provoke visual images by metonymy, or customary contiguity. Back

6. Jean Hagstrum, The Sister Arts: The Tradition of Literary Pictorialism and English Poetry from Dryden to Gray (Chicago: University of Chicago Press, 1958), p. 18. See also Wendy Steiner, The Colors of Rhetoric (Chicago: University of Chicago Press, 1982), pp. 42-43, for an account of ekphrasis as literature's nostalgia for the visual arts. Further page references will be cited in the text. Back

7. George Saintsbury, quoted in Hagstrum, The Sister Arts , p. 18. Back

8. Murray Krieger, "The Ekphrastic Principle and the Still Moment of Poetry; or Laokoon Revisited," in The Play and Place of Criticism (Baltimore Johns Hopkins University Press, 1967). This essay, which has without question been the single most influential statement on ekphrasis in American criticism, has now been incorporated by Krieger into a booklength study, Ekphrasis: The Illusion of the Natural Sign (Baltimore, MD: Johns Hopkins University Press, 1992). Back

9. The doctrine can, of course, expand even further to become a general principle of effective rhetoric or even of scientific language, where it appears under the rubric of clear, "perspicuous" representation, modeled on perspectival, rationally constructed imagery. More typical, however, is the use of ekphrasis as a model for the power of literary art to achieve formal, structural patterns and to represent vividly a wide range of perceptual experiences, most notably the experience of vision. The graphic, pictorial, or sculptural models for literary art range from the quasi-scientific claims of perspectival realism, to the grand patterning of architecture, to the focusing of a literary work in a single image, whether an emblem, a hieroglyph, a landscape, or a human figure. Back

10. On the distinction between "image/text," "image-text," and "imagetext," see chapter 3, footnote 9. Back

11. Those who saw Bob and Ray's television debut on "Saturday Night Live" know that their humor loses much of its force when they cease to be invisible voices and are revealed as what we always knew them to be: two very ordinary-looking middle-aged men. Back

12. Gotthold Lessing, Laocoon , translated by Edith Frothingham (1766 New York: Noonday Press, 1969), pp. 68-69. Back

13. For Lessing, arbitrary visual signs (emblems, hieroglyphs, pictographs) such as, for instance, serpents that signify divinity, are well on their way to being a form of writing. See my essay, "Space and Time: Lessing's Laocoon and the Politics of Genre," chapter 4 in Iconology (Chicago: University of Chicago Press, 1986). Back

14. Francoise Meltzer, Salome and the Dance of Writing: Portraits of Mimesis in Literature (Chicago: University of Chicago Press, 1987), p. 21. Back

15. Lessing traces "adulterous fancy" among the ancients--especially women--to the use of serpents as "emblems of divinity" on ancient statues. Not just the phallic shape of the serpent, but its impropriety as an arbitrary sign attached, like language or voice, to a properly "beautiful" and mute statue, is the provocation to adultery. See Lessing, Laocoon , pp. 10-11. Further page references will be cited in the text. Back

16. For further discussion of romantic iconophobia, see chapter 4, pp. 114-20. Back

17. For a more extended account of Jameson on the ideologeme and his use of the categories of space and time, see Jameson's The Political Unconscious (Ithaca, NY: Cornell University Press, 1981), p. 87, and my essay "Space, Ideology, and Literary Representation," in Poetics Today 10, no. 1 (Spring 1989): 91-102. Back

18. As Marcelin Pleynet puts it, "The objective of the text of art criticism . . . is for me to place myself . . . before something that implies another discourse, a discourse that will not be in the text.... " (Painting and System , translated by Sima Godfrey [1977; Chicago: University of Chicago Press 1984], p. v). Back

19. John Hollander proposes a distinction between "notional" ekphrases of "imaginary" or lost works of art and descriptions of visual representation that refer to familiar, widely reproduced, or even present objects of visual representation See "The Poetics of Ekphrasis," Word and Image 4 (1988): 209-19. I want to suggest, however, that in a certain sense all ekphrasis is notional, and seeks to create a specific image that is to be found only in the text as its "resident alien," and is to be found nowhere else. Even those forms of ekphrasis that occur in the presence of the described image disclose a tendency to alienate or displace the object, to make it disappear in favor of the textual image being produced by the ekphrasis. The art history slide lecture is a perfect illustration of this point. A fixed convention of the slide lecture is the declaration that the image projected on the screen is a totally inadequate representation (the colors have faded, the lighting was poor, the texture has all been lost). Even when the lecture is performed in the presence of the object itself, the commentator is never at a loss for strategies of displacement and upstaging, the most obvious being a discourse that removes the object from the museum or gallery and situates it in some other, more authentic or appropriate place (the site of its original display or production, the artist's studio, the artist's mind, or--best of all--the mind of the commentator) . Back

20. The claim that they do take on iconic characteristics, achieving verbal artifacts that "resemble" at some level the visual form they address, is one of the central claims of ekphrastic hope. Back

21. Krieger, Play and Place of Criticism , p. 107. Back

22. Another way to put this is to note that the ekphrasis of a work of visual art need not produce either an "iconic" effect (though this is certainly a possibility) or any other "literary" feature. Ekphrastic prose is an equally available possibility, and the presence or absence of iconism or "literarity" in this prose is not preordained by its reference to a visual representation. Ekphrasis and verbal iconicity are, in short, independent features. Back

23. See Gerard Genette, Figures of Literary Discourse , translated by Alan Sheridan (New York: Columbia University Press, 1984), p. 136. George Saintsbury's account of ekphrasis as a "set description" reflects this desire to "frame" or "set off" description as a separable feature. See footnote 7 above. Back

24. Texts may , of course, achieve spatiality or iconicity, but the visual object invoked does not require or cause these features. Back

25. The strange irreality of these "gifts" does not, of course, prevent us from giving them and from thinking of the whole ekphrastic gesture as a kind of ritual of exchange. One of the most frequent sites of ekphrasis in classical poetry is the singing contest between two shepherds who describe and exchange artifacts as tokens of mutual esteem (see, for instance, Theocritus' Idyll, Virgil's Eclogue V). "The art objects used as gifts or prizes may be read as the rewards poets should receive for their productions . . . and the ekphrastic description asserts poetry's worth by showing that poetry can indeed 'capturew in and receive such valuable things" (Joshua Scodel, correspondence with the author,1989). The Shield of Achilles is a gift from the hero's goddess mother, and the ekphrasis of this shield by Homer is a gift from his muses that is, in turn, given to the reader/listener. I am very grateful to Joshua Scodel for his extended response to early drafts of this essay, and I will be quoting from his letters to me throughout this essay. Back

26. In fact, a moment's reflection suggests innumerable instances where pictures stand in for words and whole speech acts. Linda Seidel has shown, for instance, that Van Eyck's famous Arnolfini portrait is best understood as something like a marriage contract, and the "cave canem" inscription on the fierce dog mosaics at Pompeii (often cited by Gombrich as examples of "apotropaic" imagery) seems redundant in view of the image. See Seidel "The Arnolfini Portrait," Critical Inquiry 16, no. 1 (Fall 1989): 54-86, and Ernst Gombrich, "The Limits of Convention," in Image and Code , edited by Wendy Steiner (Ann Arbor, Michigan, 1981), pp. 11-42. Back

27. Oliver Sacks, New York Review of Books , 23 October 1986, p. 69. Back

28. See Daniel Tiffany's essay, "Cryptesthesia: Visions of the Other," The American Journal of Semiotics 6, nos. 2/3 (1989): 209-19. Back

29. See chapter 1, "The Pictorial Turn." Back

30. I take this to be Jacqueline Rose's point when she says "the link between sexuality and the image produces a particular dialogue which cannot be covered adequately by the familiar opposition between the formal operations of the image and a politics exerted from outside" (Sexuality in the Field of Vision [London: Verso, 1986], p. 231). Needless to say, the positioning of racial otherness within the field of vision would display complex intersections with and differences from the image/text as a gendered or sexual relation See Toni Morrison on "The Alliance between Visually Rendered Ideas and Linguistic Utterances in the Construction of the Color Line," in Playing in the Dark (Cambridge, MA: Harvard University Press, 1992), p. 49. Back

31. Cp. also Toni Morrison's short story, "Recitatif" (1982), with its dehberate confusing of the verbal/visual signs of racial identity, reprinted in Confirmation: An Anthology of African American Women , edited by Amiri Baraka (Leroi Jones) and Amina Baraka (New York: William Morrow,1983) pp. 243-61. Back

32. The Slave's Narrative , edited by Charles T. Davis and Henry Louis Gates, Jr. (Oxford: Oxford University Press, 1985), p. xxvi. For more on this subject, see chapter 6, "Narrative, Memory, and Slavery." Back

33. From the foreword to Michel de Certeau's Heterologies: Discourse on the Other (Minneapolis: University of Minnesota Press, 1989), p. xiii. Back

34. I am grateful, once again, to Joshua Scodel for pointing out that this matter was not sufficiently thematized in early drafts of this essay. Back

35. One might think of the psychoanalytic process of dream interpretation as a staging of the ekphrastic scene in which the manifest visual content of the dream is the ekphrastic object, the analysand is the ekphrastic speaker, and the analyst is the reader/interpreter. Back

36. Joshua Scodel, correspondence with the author. Back

37. Kenneth Atchity, Homer's Iliad: The Shield of Memory (Carbondale, IL: Southern Illinois University Press, 1978), p. 158. See also Marc Eli Blanchard, "In the World of the Seven Cubit Spear: The Semiotic Status of the Object in Ancient Greek Art and Literature," Semiotica 43, nos. 3/4 (1983): 205-44. Back

38. Jean-Jacques Rousseau, Confessions (1781), translated by A. M. Cohen (New York: Penguin Books, 1953), p. 109. See Marcelin Pleynet, Painting and System , p. 153, for a suggestion that Freud's analysis of the relation between narcissism, scopophilia, and auto-eroticism "merits formalization within the field of an approach to painting." Cp. Freud, "Instincts and Their Vicissitudes," in Collected Papers (London: Hogarth, 1949), vol. 4, p. 66. Back

39. "Fetishism" ( 1927) in The Complete Psychological Works of Sigmund Freud , translated by James Strachey (London: Hogarth, 1961), 21: 152-57. Back

40. "The foot or the shoe owes its preference as a fetish--or a part of it--to the circumstance that the inquisitive boy peered at the woman's genitals from below, from her legs up" (Freud, Complete Works , 21:155). Back

41. I've always found Kenneth Burke's rewriting of this line as "Body is turd, turd body" the best antidote to Keats's ending. Back

42. See Carol Jacobs, "On Looking at Shelley's Medusa," Yale French Studies 69 (1985): 163-79, for a good reading of this passage. Back

43. See Neville Rogers, "Shelley and the Visual Arts," Keats-Shelley Memorial Bulletin 12 (1961): 16-17. Back

44. See Jerome McGann, "The Beauty of Medusa: A Study in Romantic Literary Iconology," Studies in Romanticism 11 (l972): 3-25. Back

45. Neil Hertz, "Medusa's Head: Male Hysteria Under Political Pressure," Representations 4 (Fall 1983): 27-54. Back

46. "Medusa's Head," in Sexuality and the Psychology of Love , p. 212. quoted in Hertz, p. 30. Hertz notes the reenactment of female exhibitionism as a revolutionary weapon in the Paris uprisings in June 1848. Back

47. My thanks to Joshua Scodel for pointing this out. Back

48. Cedric Whitman, Homer and the Heroic Tradition (Cambridge, MA: Harvard University Press, 1958), p. 205. Back

49. See chapter 3, footnote 14, for a discussion of image-text suturing. Back

50. The indicatives cited here are the ones singled out by Lessing as "cold and tedious" (Laocoon, p. 116). Back

51. See Whitman, Homer and the Heroic Tradition , and Atchity, Homer's Iliad: The Shield of Memory . Back

52. Cp. Marc Eli Blanchard, "World of the Seven Cubit Spear," p. 224: "The plot of the Iliad , underscored by the manufacture of the shield, has now become a decorative episode on the surface of the metal." Back