To these dead forms, came living beauties essence
Able to make them startle with her presence.

—George Chapman, “Ovid's Banquet of Sense”

A fascination with the idea of illusionist representations pervades the history of Western culture. From time immemorial great artists have been endowed with a power to instill a supernatural degree of life into their artifacts. In his classic study, Art and Illusion, Ernst Hans Gombrich takes the Pygmalion fantasy as the starting point of his discussion of the illusionist qualities of art. The famous Greek myth, one recalls, exemplifies the belief in the power of art to give life rather than to represent it. In Ovid's version of the story (Metamorphoses, Book X), “Pygmalion is a sculptor who wants to fashion a woman after his own heart and falls in love with the statue he makes. He prays to Venus for a bride modeled after that image, and the goddess turns the cold marble into a living body” (Gombrich 9). Naturally, this myth has captivated the imagination of many artists—whether idealized by Donatello or satirized by Daumier. The triumph of representational skill is traditionally associated with the power of art to arouse passions. This was common theoretical currency from Antiquity to the Renaissance. A great painter could so enthrall the minds of people that they would even fall in love with a painting that did not represent real human beings. In his Treatise on Painting, Leonardo da Vinci tells the following anecdote:

I made a religious painting which was bought by one who so loved it that he wanted to remove the sacred representations so as to be able to kiss it without suspicion. Finally his conscience prevailed over his sighs and lust, but he had to remove the picture from his house.

(qtd. in Gombrich 94)

If for today's critics illusionist virtuosity is at best suspect as a criterion of artistic worth, it was highly praised by classical authorities. For centuries, Pliny's well-known account of the evolution of painting influenced the criteria used to assess the quality of visual art. In his Natural History Pliny tells us that painting began with the tracing of an outline around a man's shadow, using a single color (“monochrome”); then discovered light and shade, and the contrast of colors; and climaxed with the masterpieces of Apelles, which were so lifelike “they challenged Nature herself” (5.271; 11.283; 36.331). In other words, the progress of the techne was characterized by an evolution from badly representational flat pictures to polychrome “lifelike images which (almost) had the power to deceive the spectator” (Gent 36). In the moral context of Stoic and Christian cultures, the power of art to deceive viewers and arouse their passions was deemed dangerous indeed. (Seneca had already raised this as an ethical issue with regard to Homer's representations [D'Alton 485ff].) Rather, the artist's chief aim should be to instruct—and one way an artist could safeguard his reputation was to interpret his works allegorically. The etymology of the word allegory is revealing: alla agoreuein means to say something different and to say it differently. By introducing allegorical interpretations of their works, painters and poets could side-step the moral objections of their critics, finding an identity whereby moral worth and aesthetic freedom were simultaneously established by the ability to produce marvelous and sometimes uncanny, lifelike illusions (Gent 45).

The aim of this preamble is simply to help establish a broad horizon of expectations for the experience of pictorial aesthetics in the context of so-called fantastic literature, particularly in the modern period. From the Don Juan tradition to the Romantics' “contes fantastiques,” literary works offer numerous examples of narratives dramatizing people's infatuation with pictures, tapestries or sculptures. The moral, political or metaphysical implications of Plato's condemnation of art are of little concern to us here. But the implication that the painter's subject is a lie, because twice or thrice removed from reality (that is, from the realm of Ideal Forms), remains central to any theory of representation. We know that a painting is an illusion; but if we are attracted by it and cannot refrain from gazing at it, if it haunts our thoughts and our dreams, then the sign has become more pregnant and meaningful than its referent (Barkan 9-11; Jakobson 318 ff). As the Latin oxymoron goes, the picture has become a falsa veritas, a false or “fictional” truth (Riffaterre, Fictional Truth).

The history of fantastic literature confirms a posteriori the credibility of Leonardo's account: human beings do fall in love with “supremely lifelike pictures” that do not represent real human beings. Modern fiction is full of examples of characters stepping out of their canvasses and statues descending from their pedestals. Baudelaire's famous line, “Je suis belle, ô mortels! comme un rêve de pierre” (24) is a mere poetic expression of a timeless obsession. Within the French tradition, the theme of the animated work of art, linked with the valorization of the aesthetic experience, is treated compulsively by writers as diverse as Gautier, Dumas, Erckmann-Chatrian, Balzac, Mérimée, Nerval, Nodier, Maupassant, and Villiers de l'Isle-Adam (Bessière, Castex, Jacquemin, Milner). One can really speak of a “Pygmalion complex” in the pervasive Romantic fascination with visual resurrections (Chambers, “Gautier” and “Seduction” 94). This may be due to a reaction (often ironic) against the disorders of modern society or to a new acute perception of the “insignificance of the real” as an artistic motivation (Rosset 25).

My purpose is not to elaborate on these vast problems of interpretation. Tzvetan Todorov, in his Introduction à la littérature fantastique, and Christine Brooke-Rose, in her study entitled A Rhetoric of the Unreal, have tried to account for the existence of the fantastic in many of its forms, viewing even non-fantastic texts as displaced avatars of the fantastic (Brooke-Rose 65). Instead, I would like to focus on the specific genesis of this passion for the unreal—in so far as it relates to the rhetorical description of works of art, the ekphrasis in classical and modern fiction.¹ In spite of the limitations of my scope and selectivity of my corpus, I hope to be able to shed some light on the complex definition of the fantastic as a displaced mode of ekphrastic representation.

A single nineteenth-century fantastic tale will be considered in some detail here for the purpose of discussion: Prosper Mérimée's “La Vénus d'Ille” (“The Venus of Ille”), written in 1837. A quick summary of the plot is in order. As the story begins, the narrator, an archaeologist, visits a provincial antiquarian, Monsieur de Peyrehorade, whose son, Alphonse, is about to marry his beautiful and rich fiancée. M. de Peyrehorade has unearthed a strange-looking statue of Venus that seems to cast strange spells on those around her. During the excavation, the statue had fallen on a worker and had broken his leg so badly that he will hardly be able to walk again. When in a spirit of vengeance another fellow throws a stone at the statue, she apparently throws it back at him as if “to punish the rascal for the outrage done to the goddess” (239) (“puni[r] ce drôle de l'outrage fait à la déesse” [96]).² When Alphonse slips the diamond wedding ring of his bride-to-be on the statue's fourth finger, she clenches her hand and refuses to give it back (264-265). The tragic climax occurs on the wedding night. In a frightful scene reported by the young bride, Alphonse is strangled to death in the very conjugal bed by the vengeful statue herself. Even after the statue has been melted down and made into a church bell, an evil fate is visited on those who possess that piece of bronze: “Depuis que cette cloche sonne à Ille, les vignes ont gelé deux fois” (“Since that bell began to ring in the village of Ille, the vines have twice been frost bitten” [118, 276]).

Much attention has been focused by literary historians on the archetypal topos of the “moving statue” that animates Mérimée's tale and also on its earlier appearances in literature (Gross 115-9; Rousset 109-110; Serres 7-31). Likewise, semiologists and psychoanalysts have proposed various interpretations to account for the origin of the “living statue” fantasy.³ Yet, what strikes this reader in terms of the structure of Mérimée's “La Vénus d'Ille” is the emphasis placed on the descriptive process as a prerequisite for the “fantastic” narration. As Gérard Genette reminds us in his study on literary discourse, “the opposition between narration and description, which was so stressed by academic tradition, is one of the major features of our literary consciousness” (133). Yet, as we shall see, in the fantastic tale (more, perhaps, than in any other genre), the line of demarcation between the descriptive and the narrative tends to be blurred precisely because it is through the actualizing power of situational details that the narrative unfolds.

Let us examine the status of the descriptive in Mérimée's tale. We are first given an account of the archaeological discovery through a dialogue between the narrator and his guide:

“Comment! on ne vous a pas conté, à Perpignan, comment M. de Peyehorade avait trouvé une idole en terre?”

“Vous voulez dire une statue en terre cuite, en argile?”

“Non pas. Oui, bien en cuivre, il y en a de quoi faire des gros sous. Elle vous pèse autant qu'une cloche d'église.” (88)

(“What!” the guide says, “did no one tell you at Perpignan, that M. de Peyrehorade had found a statue in the earth?”

“Did you mean,” remarks the narrator, “a statue in terra-cotta, or clay?”

“Nothing of the kind. It is actually in copper, and there is enough of it to make heaps of coins. It weighs as much as a church-bell.”)

(227)

Detailed, concrete, objective data about the height, weight, matter, color, and shape of the huge Venus are then piled up by the narrator for us to marvel at. We are told that it looks like an idol of pagan times, “perhaps as old as Charlemagne” (228), and, not without irony, that it is “much more beautiful and better finished than the painted plaster bust of Louis-Philippe, which is at the town hall” (229) (“encore plus beau et mieux fini que le buste de Louis-Philippe, qui est à la mairie, en plâtre peint” [89]). Up to this point, we have been given a pure iconic representation in rigorously realistic terms. However, the guide now adds other, more intriguing details that arouse the narrator's curiosity: “Mais avec tout cela, la figure de cette idole ne me revient pas. Elle a l'air méchante … et elle l'est aussi” (89) (“The idol's face is not very nice to look at. She looks wicked … and she is so, too” [229; emphasis added]). Incredulous, the narrator immediately interjects: “Méchante! Quelle méchanceté vous a-t-elle faite?” (89) (“Wicked! What mischief has she done you?” [229]).

This natural reaction is obviously meant to be shared by the amused readers, safely ensconced in their superior isolation. Of course statues can represent wickedness, and the best ones will represent it vividly; but can they be wicked? Only simple folks, innocent victims of superstitious beliefs, can make such a gross error and be deceived by the illusion of appearances because they are quick to imagine fantastic stories about whatever is unfamiliar to them. By contrast, we, as trained archaeologists and reliable scholars, only hear the voice of science and reason. The laws of nature limit us to the safe ground of mere descriptiveness; works of art are meant to be described, analysed, and then (and only then) interpreted. In other words, the ekphrastic code appears to be for the scholar a mandatory hygienic process that must precede any hermeneutic gesture.

To be sure, this attitude is only tenable within the context of an aesthetic based on the Aristotelian principle of verisimilitude—that is, if we expect that, by and large, real human beings behave like human beings and statues like statues, and that we can predict possible developments of their behaviors according to a logical order based on common experience. We all recall, of course, the uncanny tale of the vengeful statue Aristotle tells in his Poetics (9. 12-13). The statue of Mitys at Argos kills the man who had caused Mitys's death by falling on him at a festival. Inasmuch as statues can be expected to sometimes fall from their pedestals and even kill passers-by, this incident still obeys the principle of verisimilitude. True, this particular instance produced much amazement, as it seemed to be more than a mere accident and to involve supernatural powers. But this is what made it particularly fitting for a tragic subject in an otherwise realistic discourse.

In his study “Le discours contraint,” Philippe Hamon lists some fifteen procedures that arise out of the preconceptions of realism, and labels the last one “exhaustiveness of description”:

In realistic discourse, according to Hamon, the world is describable, accessible to denomination, whereas in the fantastic it is unnamable, indescribable. The tenet of realism generates a desire for exhaustiveness, while the assumption of the world's discreteness leads to an aesthetic of the discontinuous, a composition in tableaux, slices of life, scenes, … a synecdochic fragmentation …

(Brooke-Rose 89)

I would propose to rephrase Hamon's fifteenth procedure in light of Aristotelian poetics and to move away from a clear-cut opposition between the “realistic” and “fantastic” poles in terms of descriptive possibilities. The question is not whether verisimilitude welcomes the describable or the indescribable. It is rather whether, within the descriptive mode, it can do anything but obey the readability requirement that says that a statue behaves like a statue (whether it is Aristotle's statue of Mitys, Don Juan's uomo di sasso or Mérimée's “La Vénus d'Ille”).

In uncanny, marvellous or fantastic tales, descriptions can be just as exhaustive as those in realistic stories; they often appear to be structured in various sequences, and function like narrative syntagmas. For instance, in Mérimée's story, we first get a description from the guide, followed by a second one given by the old antiquarian, and finally the narrator's own perception of the statue:

C'était bien une Vénus, et d'une merveilleuse beauté. Elle avait le haut du corps nu, comme les anciens représentaient d'ordinaire les grandes divinités; la main droite, levée à la hauteur du sein, était tournée, la paume en dedans, le pouce et les deux premiers doigts étendus, les deux autres légèrement ployés.

(96; emphasis added)

(It was indeed a Venus of extraordinary beauty. The top part of her body was bare, just as the ancients usually depicted their great deities; her right hand, raised up to her breast, was bent, with the palm inward, the thumb and two first fingers extended, whilst the other two were slightly curved.)

(240)

This apparently “realistic” description of the hand will lose its verisimilitude when we learn about the wedding-ring episode. Through “retroactive reading” we experience the semiosis of these two “slightly curved fingers”: one of them is meant to clench the ring promised to the unfortunate bride. In other words, the descriptive has become “motivated”; it has unlocked itself from the prison of timelessness and yielded to the temporal process of narrative. To use Gustave Guillaume's neo-Latin grammatical jargon, Venus's ring finger belongs to the descriptive only inasmuch as it points to a narrative in fieri or in posse (Guillaume 15, 30).

The same could be said of the hyperbolic tone of the narrator's aesthetic experience when confronted by the work of art:

Quoi qu'il en soit, il est impossible de voir quelque chose de plus parfait que le corps de cette Vénus; rien de plus suave, de plus voluptueux que ses contours; rien de plus élégant et de plus noble que sa draperie. Je m'attendais à quelque ouvrage du Bas-Empire; je voyais un chef-d'œuvre du meilleur temps de la statuaire. Ce qui me frappait surtout, c'était l'exquise vérité des formes, en sorte qu'on aurait pu les croire moulées sur nature, si la nature produisait d'aussi parfaits modèles.

(97; emphasis added)

(However that might be, it is impossible to conceive anything more perfect than the body of this Venus; nothing could be more harmonious or more voluptuous than its outlines, nothing more graceful or dignified than its drapery. I expected some work of Late Antiquity and I beheld a masterpiece of the most perfect period of sculpture. I was especially struck with the exquisite truth of form, which gave the impression that it had been moulded by Nature itself, if Nature ever produces such perfect specimens.)

(241)

This apparently straightforward report by a sensitive art lover takes on an ironic meaning, not only because of its obvious reference to the “Pygmalion complex” topos (what we might call a paradigmatic irony) but because the allusion to the “exquisite truth of form” will soon become true indeed in the syntagmatic unfolding of the story.

In chapter five of his classic study, Tzvetan Todorov stresses the tendency of the fantastic to use figurative language in a literal sense: “Le surnaturel y apparaît comme un prolongement de la figure rhétorique” (82). The word “prolongement” (prolongation) is noteworthy because of its metonymic connotation. For instance, in Villiers de l'Isle-Adam's Vera, the story takes literally the expression “l'amour est plus fort que la mort” (“love is stronger than death”) and develops its plot around it. My contention is that the literalizing process is not actualized—as Todorov seems to believe—in the unfolding of the narrative. On the contrary, the ekphrastic moment is precisely the rhetorical locus where the “event” shines forth. Elsewhere, description may be a contemplative, lyrical luxury, more or less removed from action. In the discourse of the fantastic, description already exhibits its metonymic consequences. It has become the metaphor of a metonymy as well as the metaphor of a literalized metaphor.

The genesis of this process can be traced as far back as Homer. In his Rhetoric, Aristotle remarks that “often Homer, by making use of metaphor, speaks of inanimate things as if they were animate,” and he adds that “it is to creating actuality [energeian poiein] that his popularity is due” (Rhetoric 3; 11. 3; emphasis added). The word energeia, for Aristotle, refers to the paradox of producing a powerful lifelike effect through words. Energeia is usually translated into Latin as actio, and enargeia as illustratio or evidentia (Quintilian 4. 2, 63; 6. 2, 32-36; 8. 3, 61-62). In Roman times, a strange etymological confusion takes place, as the two Greek words energeia and enargeia are semantically conflated. In poetic theory the two meanings combine—as if the artistic power (energy) to represent reality must necessarily be linked with sight, the “noble sense” associated with light and creativity. In Cicero's Orator, the writer's ability to describe inanimate things as if they were animate is expressed in visual, iconic terms (23). The interest in ekphrastic “energy” was reinforced during the Renaissance by the widespread doctrine of ut pictura poesis and the exemplary discourse of classical authors, such as Pliny and Plutarch, on pictorial representation. In his Deffence et Illustration de la Langue Françoise (1549), Joachim du Bellay makes fun of bad translators who lack “ceste energie [meaning both energeia and enargeia], … comme un peintre peut representer l'ame avecques le cors de celuy qu'il entreprent tyrer après le naturel” (40-41; emphasis added) (“this energy, … as a painter can represent the soul of that man he undertakes to paint according to nature, along with the body”).

The pictorial analogy brings body and soul together to give them life; and the portrayed subject regains the “exquisite truth” of Nature itself. The metaphor of the painter appears to be infinitely desirable to any writer whose ultimate goal is hypotyposis, or, in Peacham's words, a “description so vivid that it seemeth rather paynted in tables than expressed in wordes” (Gent 40). In the same vein, Edmund Spenser claims that the pictures of the Shepheardes Calender (1579) are “so singularly set forth, and portrayed, as if Michel Angelo were there, he could (I think) nor amende the best, nor reprehende the worst” (612). But there is more to it. In the Renaissance many artists found material for ekphrasis in dreams, portents, and the marvels of Nature generally. In sixteenth-century visions, strange mythological statues move about with mysterious animation. To quote Chapman's “Ovid's Banquet of Sense” (1595): “To these dead forms, came living beauties essence / Able to make them startle with her presence” (54). Enthusiasm for illusionist representations becomes disturbing as the Reformation develops. This same energeia will prompt the destruction of art, as a radical remedy for idolatry. Iconoclasm is perhaps one of the most extreme historical manifestations of the repressed fascination with the fantastic. Without ekphrastic vigor, literature might have been spared such iconoclastic furor.

In Mérimée's “La Vénus d'Ille,” soon after the tragic episode of her son's death, Madame de Peyrehorade has the statue melted down and made into a church bell, but even in this new form it seems to retain its evil power: “Depuis que cette cloche sonne à Ille, les vignes ont gelé deux fois” (118) (“Since that bell began to ring in Ille the vines have twice been frost-bitten” [276]). These famous last words of the story point to the overdetermination of the descriptive signs as the fictitious embodiment of an animated entity. Venus no longer exhibits her marvellous beauty combined with ferocious looks; she has lost those brilliant eyes which “produced a kind of illusion which recalled lifelike reality” (276) (“produisaient une certaine illusion qui rappelait la réalité, la vie” [98]). Yet she is still empowered with figural meaning. Racine's disquieting line, which Monsieur de Peyrehorade declaimed with naïve enthusiasm at the beginning of the story, functions with paradigmatic rigor to the end: “C'est Vénus tout entière à sa proie attachée!” (97; Phèdre 1.3. 306; “It is Venus herself gloating over her prey” [242]).

The tropological transformation of the text survives the physical transformation of the statue. Or, to paraphrase a now famous dialogue between Paul de Man and Michael Riffaterre, prosopopeia makes all or none of the difference (de Man, “Hypogram”; Riffaterre, “Prosopopeia”). From Quintilian to Lausberg, the rhetorical tradition has dwelt on this conventional figure of animation. In his classic Figures of Speech, Pierre Fontanier gives the following definition:

Prosopopeia, which must not be confused with personification, apostrophe or dialogism … consists in staging, as it were, absent, dead, supernatural or even inanimate beings. They are made to act, speak, answer as is our wont. At the very least, these beings can be made into confidants, witnesses, accusers, avengers, etc.

(404; emphasis added)

At the end of Mérimée's story, the attraction for Evil is no longer personified in the harmonious and voluptuous body of Venus. While it has lost its human shape, it is still powerfully present in the guise of a church bell, as a “supernatural, inanimate being” that plays the role of an “avenger.” The etymology of the trope's name, prosopon poiein, means, as Paul de Man noted, “to confer a mask or a face (prosopon)” (Rhetoric 77-78). The uncanny presence of Venus has lost its beautiful, deceitful face and has now put on a new mask in the very place, the church, where it should normally have been exorcized. The original description thus retains its full ekphrastic power, as anthropomorphism vanishes and a new rhetorical trope, prosopopeia, takes over to ensure the perpetuation of the fantastic to the end.

A significant episode in Mérimée's story concerns the ambiguous meaning of the inscription on the statue's pedestal: “Cave Amantem” (243). The narrator proposes two possible meanings. The Latin words can be translated as “Beware of him who loves thee” (“Prends garde à celui qui t'aime”) or as “Mistrust thy lovers” (“Défie-toi des amants”). However, looking at the lady's wicked expression, the narrator concludes that the most appropriate rendering of the Latin is probably: “Prends garde à toi si elle t'aime” (“Beware if she loves thee”). This third reading, of course, will be borne out by the dénouement of the story. If only Alphonse had read and understood the “caveat” the way the narrator does, he might have been on his guard against the statue's diabolic beauty. Inscriptions are commonly found on works of art, in the form of signatures, addresses, mottos, or dedications, often with an authoritative meaning. They are always epi-graphs, material signs written on the pictorial representation with a special, added significance, powerfully displayed on the surface. In a fantastic tale, every descriptive detail in the text points to the unreal, except for the materiality of the inscription. It is écriture in the Derridian sense, a fragment of discourse, without origin, which absorbs and rejects all hermeneutical readings. The narrator may be aware of this when, commenting on the polysemous compactness of the warning, he observes: “Latin is a dreadful tongue, because of its conciseness” (244) (“C'est une terrible langue que le latin avec sa concision” [99]).

M. de Peyrehorade is skeptical, of course, about the final translation, which the narrator holds to be the true one: “Beware if she loves thee.”

“Humph! [dit-il], oui, c'est un sens admirable; mais, ne vous en déplaise, je préfère la première traduction, que je développerai pourtant. Vous connaissez l'amant de Vénus?”

“Il y en a plusieurs.”

“Oui, mais le premier, c'est Vulcain. N'a-t-on pas voulu dire: ‘Malgré toute ta beauté, ton air dédaigneux, tu auras un forgeron, un vilain boiteux pour amant’? Leçon profonde, monsieur, pour les coquettes!”

Je ne pus m'empêcher de sourire, tant l'explication me parut tirée par les cheveux. (98-9)

(“Yes [he says], that is an admirable interpretation; but, without wishing to displease you, I prefer the first translation, and I will tell you why. Do you know who Venus's lover was?”

“There were several,” the narrator retorts.

“Yes,” continues Peyrehorade, “but the chief one was Vulcan. Should one not rather say, ‘In spite of all thy beauty and thy scornful manner, thou shalt have for thy lover a blacksmith, a hideous cripple’? What a profound moral, monsieur, for flirts!”

I could hardly help smiling at this far-fetched explanation.)

(243-4)

The reader is, of course, tempted to agree with the narrator and dismiss the “hair-splitting” reconstruction laboriously imposed by the old antiquarian, who in this story plays the part of the blind learned fool. Yet the allusion to Vulcan is far from out of place in the context of the story's beautiful bronze statue, wrought by an extraordinary craftsman. Pressed by obvious intertextual yearnings, the reader is invited to respond to M. de Peyrehorade's incorrect translation by recalling, in a synecdochal move, the exemplary work of art, produced by Vulcan at Venus's request: Aeneas's famous shield in the sixth book of the Aeneid.

Obviously, Mérimée has prepared the ground for this literary analogy. For one thing, the text of “The Venus of Ille” is studded with more or less explicit allusions to Virgil's epic poem. In particular, two Latin quotations from Books 4 and 5 of the Aeneid are used to frame the elliptic inscription. Ironically, they are both quoted by Peyrehorade, whose name—although borrowed from a real toponym—carries the derisive connotation of its paronym: “péroraison.” (Pérorer means speechify; Harrap's dictionary gives as a typical usage, “To speechify at the unveiling of a statue: palabrer, laïusser à l'inauguration d'une statue.”) The first reference is to the beginning of Book 4, line 33: “Veneris nec praemia noris” (236). Dido, we recall, burning with love for Aeneas, turns to her sister, Anna, and confesses her dilemma: in spite of her passion, shouldn't she remain faithful to her husband's memory? But Anna reassures her: “Sister, you more dear to me than light itself, / are you to lose all of your youth in dreary loveliness, / and never know sweet children or the soft rewards of Venus? (“nec dulcis natos Veneris nec praemia noris?” [31-33; emphasis added]).⁴ The irony, both in the epic and the fantastic tale, is, of course, that the so-called “soft rewards of Venus” will bring death, both to Dido and Alphonse. Literally speaking they will both “suffer from the wounds inflicted by Venus” (236). The lady's “diabolic expression” (243) is somehow overdetermined by the literalization of the intertextual code.

The second reference to the Aeneid is to the end of Book 6, line 883: “manibus date lilia plenis” (give me lilies with full hands). As Aeneas descends into the Underworld, he meets his father, Anchises, who gives him a long lecture on Roman history that ends with a lament for the death of young Marcellus, nephew and adopted son of Augustus, and designated successor to the Emperor: “No youth born of the seed of Illium / will so excite his Latin ancestors to hope; / the land of Romulus will never boast/ with so much pride of any of its sons” (875-878). Anchises then tearfully begs to perform a devotional rite in honor of the illustrious dead youth:

With full hands, give me lilies; let me scatter
these purple flowers, with these gifts, at least,
be generous to my descendant's spirit,
complete this service, although it be useless.

(883-86)

In Mérimée's story, M. de Peyrehorade is determined to place a wreath of lilies on Venus's head. The intertextual meaning is clear: Alphonse is doomed to endure the fate of Augustus's son. By duplicating Anchises's gesture, the old antiquarian does not realize he is in fact provoking an uncanny twist in the story; but, this time, it will not be a “useless service” (“et fungar inani / munere” 885-86). The “youth born of the seed of Ilium” and the young man of Ille seem to share a uniquely “tragic” destiny. The onomastic pun on Ille and Ilium may have been intended. In the same story, had Mérimée not played on his own first name, giving it in Greek translation—Prosper becoming Eutychès—to the sculptor/lover of his Venus? (245). At any rate, the “soft rewards of Venus” (Aeneid 4.33) and the “pious offering of lilies” (Aeneid 6.883) are both ominous versions of the “love as death” topos. In Mérimée's tale, however, the Latin quotations give an equivocal authority to the scholar's voice, reinforcing the ambiguity as to whether the weird events are supernatural or not.

Returning to the apparently erroneous translation of “Cave Amantem” as “beware of thy lover Vulcan, a hideous crippled blacksmith” (244), we now have ample evidence of the Aeneid's intertextual power in the genesis of this fantastic tale. Yet it is precisely in the descriptive scenes of Virgil's poem that we find the “energy” (energeia again) that will develop the mimetic potentialities of art into pure “fantastic” representations. Recall how in the Aeneid, Venus, alarmed by Aeneas's hard trials, uses her charms on Vulcan to beg arms for her son. In a passage much admired by Montaigne, among others, the goddess,

with snow-white arms outflung around him,
warms him in a soft embrace. At once
he caught the wonted flame; familiar heat
reached into his marrow, running through
his trembling bones …
He gave the embrace that she craved; then
on her breast, outpoured at last,
gave himself up to sleep and rest.

(8.387-404)

Highly conscious of the power of her own beauty, Venus has managed to wring the wanted promise from Vulcan. “Cave Amantem!” or, in the narrator's translation, beware of the one she loves. Montaigne's comments on this scene are well known: “La poësie … represente je ne sçay quel air plus amoureux que l'amour mesme. Venus n'est pas si belle toute nue, et vive, et haletante, comme elle est icy chez Virgile” (Essais 849) (“Poetry … reproduces an indefinable mood that is more amorous than love itself. Venus is not so beautiful all naked, alive and panting, as she is here in Virgil” [Complete Essays 645]). Mimesis, as Montaigne's gloss testifies, is at its climax in this oft-quoted passage of the Aeneid. Both the poet's Venus and the sculptor's goddess have a potentially “fantastic” presence because their unreality is more real than life itself.

This scene is, of course, a prelude to the great ekphrastic scene in which, following Homer's prototype, Virgil describes the “indescribable text” (“non enarrabile textum” line 625) of Aeneas's shield. The significant moments of future Roman history are lavishly represented, in the most vivid detail, on the living symbol of imperialistic power. At the beginning, the primal scene of origins features the mother-wolf reclining in Mars's green cavern:

                                                  and at play beside her,
twin boys were hanging at her dugs; fearless,
they sucked their mother. She, at this, bent back
her tapered neck to lick them each in turn
and shape their bodies with her tongue.

(8.630-34)

Finally, after some one hundred lines of various historical evocations, we witness Caesar's final triumph: “The conquered nations march in long procession, / as varied in their armor and their dress / as in their languages” (722-23).

Motion and speech animate this unreal work of art—so unreal, indeed, that it represents a future that is past and therefore well known to Virgil's readers, but that has not yet unfolded for Aeneas, and is therefore incomprehensible to him. Vulcan, the “Lord of Fire” (414, 628, 710 etc.) is endowed with prophetic genius (“haud vatum ignarus venturique inscius aevi” 627 [not unversed in prophecy or unknowing of the age to come]). In a way, he is the writer of a stupendously fantastic tale. The clues he is willing to give away for the future understanding of the plot can only be grasped by “retroactively reading” the story. By contrast, Aeneas appears to be a naïve reader, marveling at the artful scenes but unable to make sense out of them:

Talia per clipeum Volcani, dona parentis,
miratur rerumque ignarus imagine gaudet.

(8.729-30)

(Such sights he admires on the shield of Vulcan, his
mother's gift, and though he knows not the deeds [res] he
rejoices in their representation [imago].)

(emphasis added)

The opposition between imago and res, sign and thing, is crucial for the genesis of fantastic literature. The reader's “hesitation” between natural and supernatural explanations—Todorov's prime requirement for what he calls the “pure” fantastic—seems to be already present here, at least incipiently.

In the first ekphrastic scene in Virgil's poem, Aeneas has a powerful encounter with a lifelike work of art. Inside the extraordinary shrine Dido is building for Juno at Carthage, he marvels at the frescoes on the temple walls (1. 455 ff). The vivid representations of the Trojan War revive his sorrow and make him halt and weep. Moved by the sight of his family's heroes, he addresses them as if they are alive:

“Achates, where on earth is there a land, a place that
does not know our sorrows? Lo Priam, here too, virtue
has its due rewards; here too, there are tears for
misfortune and mortal sorrows touch the heart. Dismiss
thy fears; this flame will bring you some salvation.”

(1.459-63)

And the narrator adds:

He speaks. With many tears and sighs Aeneas feeds his
soul on what is but an unsubstantial picture.

(464-5)

The Latin words are important: “Sic ait, atque animum pictura pascit inani” (464; emphasis added). By juxtaposing animum and pictura, Virgil suggests a semantic analogy that brings forth the underlying idea of “animated picture.” At the same time, the chiasmic position of “ani/mum” and “inf/ani” serves as a disillusioned commentary on the actual power of this animation. One feels a tension between the forceful emotional sensations elicited by the visual artifact and the rational realization of the evanescent, ungraspable reality behind it. Dido's mural is capable of arousing passions, but in the end it is nothing but a picture.

“Animum pictura pascit inani.” Perhaps this is not the worst definition ever proposed for the perception of the fantastic in literature. Iconic descriptions (i.e. descriptions of which a work of art is the subject) may possess the power—energeia—to set before the reader the very object or scene being described. Yet, in the process of producing enthrallment or astonishment, this power may exceed the limits of verisimilitude. As a result, the very energy that achieves lifelike vividness (enargeia) may also solicit the reader's disbelief. The artist's desire to make the art object breathe, move and speak is somehow allegorized in the sense that it is given multiple levels of meaning (Todorov 69). First, at the literal level, the ekphrastic tradition does prepare the reader to accept the truth of the story. A natural explanation is possible: Pygmalion is swept up in his dream of making his statue alive. Second, at a tropological or “moral” level, it is no longer a question of confusing art with life. Only a supernatural power seems to provide a satisfactory explanation (“poetic justice”) for the end of the story: the gods themselves have turned the cold marble into a living body. Third, at an “anagogic” level, neither the author nor his reader believes what is being narrated, and yet both hold the belief that, no matter how incredible the story may be, it conveys valuable information about the capacity of language to represent the world. In other words, the tradition of ekphrasis is an important testing ground for the “function of disbelief” in literature (Koelb). The shields of Achilles and Aeneas are the objects of incredible descriptions, representing “in tin and gold invisible moral qualities, the sound of music, the rhythms of the dance, the lives of men in society, and even the chiaroscuro and color of the painter” (Hagstrum 21). They are “true” only inasmuch as they give the illusion of the gods' extraordinary achievements and testify to the presence of an intention to communicate a higher order of truth.

A good picture, says Robert Burton, is a “falsa veritas” (233). A false truth is not necessarily a lie. While freeing up the latent powers of the ekphrastic tradition, the “fantastic” remains deeply indebted to its aesthetic origins: the ambiguous rhetoric of “eye-ravishing art.” After all, the fantastic may be nothing more than a particular “illocutionary mode” (Austin 98 ff): a special way of doing something else in the act of describing art and, therefore, a rather limited subversive tendency within the orthodox continuum of rhetoric and literary history.⁵

Notes

1. The literary-historical shift from the more general meaning of the term (i.e. any poetic description of objects, people or places) to the more precise one (i.e. any literary description of a work of art) can be located in the fourth century A.D., with its application to Philostratus Lemnius's Imagines (second century): a series of descriptions of pictures regarded as exemplary exercises in the art of rhetorical description. See Bergmann, Heffernan, Galand-Hallyn, Krieger, and Webb.

2. Quotations of the French text are from Mérimée's Romans et Nouvelles, vol. II, ed. by M. Parturier, pp. 79-125. The English translations are from George Saintsbury's translation, The Writings of Prosper Mérimée, vol. IV, pp. 223-76, with a fair degree of modification. Where no page references are given, the English translations are mine.

3. According to Jakobson, this fantasy is not a projection of mimetic illusion. Rather, it is triggered by the metonymic contagion of images that shape our perception and force us to experience a troubling gap between the statue's signifier and its signified (353 ff). The same fantasy could also be interpreted as a threatening transgressive crossing of the line between the living and the lifeless, and the reader's conflicting desire to open and close the gap between the mobile and the immobile could be analyzed in terms of the Freudian “uncanny” and “death drive” (Gross 128; 233).

4. Translations from the Aeneid are from Allen Mandelbaum, with a fair degree of modification.

5. This article originated as a paper presented, several years ago, at Columbia University. I wish to thank Michael Riffaterre and W. J. T. Mitchell for their suggestions. Since then, a great deal of interest in the rhetorical concept of ekphrasis has been rekindled, thanks especially to the works of Murray Krieger (1992); Ruth Webb (1992); James A. W. Heffernan (1993); and Perrine Galand-Hallyn (1994).

Works Cited

Austin, J. L. How to Do Things with Words. Cambridge, Mass: Harvard University Press, 1962.

Avni, Ora. “Et la chose fut,” Poétique 46 (1981): 156-170.

Barkan, Leonard. The Gods Made Flesh: Metamorphosis and the Pursuit of Paganism. New Haven: Yale University Press, 1986.

Baudelaire, Charles. Les Fleurs du mal. Ed. A. Adam. Paris: Garnier, 1961.

Bergmann, Emilie L. Art Inscribed: Essays on Ekphrasis in Spanish Golden Age Poetry. Cambridge, Mass: Harvard University Press, 1978.

Bessière, Irène. Le Récit fantastique. Paris: Larousse, 1974.

Brooke-Rose, Christine. A Rhetoric of the Unreal: Studies in Narrative and Structure, Especially of the Fantastic. Cambridge, UK: Cambridge University Press, 1981.

Burton, Robert. The Anatomy of Melancholy. Oxford: Henry Cripps, 1624.

Castex, Pierre-Georges. Le Conte fantastique en France, de Nodier à Maupassant. Paris: J. Corti, 1982.

Chambers, Ross. “Gautier et le complexe de Pygmalion,” Revue d'histoire littéraire de la France 72. 4 (1972): 641-58.

———. “Seduction Denied: Sarassine and the Impact of Art.” Story and Situation. Narrative Seduction and the Power of Fiction. Minneapolis: University of Minnesota Press, 1984. 73-96.

Chapman, George. The Poems of George Chapman. Ed. P. Bartlett. New York and London: Oxford University Press, 1941.

Cicero. L'Orateur. Ed. A. Yon. Paris: Belles-Lettres, 1964.

D'Alton, J. F. Roman Literary Theory and Criticism. New York: Russell and Russell, 1962.

de Man, Paul. “Hypogram and Inscription: M. Riffaterre's Poetics of Reading,” Diacritics 11 (1981): 17-35.

———. The Rhetoric of Romanticism. New York: Columbia University Press, 1984.

Fontanier, Pierre. Les Figures du discours. Paris: Flammarion, 1968.

Galand-Hallyn, Perrine. Le Reflet des fleurs: Description et métalangage poétique à la Renaissance. Geneva: Droz, 1994.

Genette, Gérard. “Frontiers of Narrative.” Figures of Literary Discourse. Trans. A. Sheridan. New York: Columbia University Press, 1982. 120-142.

Gent, Lucy. Picture and Poetry 1560-1620. Relations Between Literature and the Visual Arts in the English Renaissance. Leamington Spa: James Hall, 1981.

Gombrich, Ernst Hans. Art and Illusion: A Study in the Psychology of Pictorial Representation. Princeton: Princeton University Press, 1961.

Gross, Kenneth. The Dream of the Moving Statue. Ithaca: Cornell University Press, 1992.

Guillaume, Gustave. Temps et verbe. Théorie des aspects, des modes et des temps. Paris: Champion, 1965.

Hagstrum, Jean H. The Sister Arts: The Tradition of Literary Pictorialism and English Poetry from Dryden to Gray. Chicago: University of Chicago Press, 1958.

Hamon, Philippe. “Le discours contraint,” Poétique 16 (1973): 411-45.

Heffernan, James A. W. Museum of Words: The Poetics of Ekphrasis from Homer to Ashbery. Chicago: University of Chicago Press, 1993.

Jacquemin, Georges. Littérature fantastique. Paris, Fernand Nathan; Brussels: Editions Labor, 1974.

Jakobson, Roman. “The Statue in Puškin's Poetic Mythology.” Language in Literature. Ed. Krystyna Pomorska and Stephen Rudy. Cambridge: Harvard University Press, 1987. 308-367.

Koelb, Clayton. The Incredulous Reader. Ithaca: Cornell University Press, 1984.

Krieger, Murray. Ekphrasis: The Illusion of the Natural Sign. Baltimore: Johns Hopkins University Press, 1992.

Mérimée, Prosper. Romans et nouvelles. Ed. M. Parturier. Vol. 2. Paris: Garnier, 1967. 2 vols.

———. The Writings of Prosper Mérimée. Trans. George Saintsbury. Vol. 4. New York: Croscup and Holby, 1905. 8 vols.

Milner, Max. La fantasmagorie. Essai sur l'optique fantastique. Paris: Presses Universitaires de France, 1982.

Montaigne, Michel de. Essais. Ed. P. Villey. Paris: Presses Universitaires de France, 1968.

———. The Complete Essays of Montaigne. Trans. Donald M. Frame. Stanford: Stanford University Press, 1965.

Pliny. Natural History. Ed. H. Rackham. Loeb Classical Library. Cambridge, Mass.: Harvard University Press, 1952.

Quintilian, Institutio Oratoria. Text with English translation by H. E. Butler. London: W. Heineman; New York: G. P. Putnam's, 1921-22.

Riffaterre, Michael. “Prosopopeia,” Yale French Studies 69 (1985): 107-23.

———. Fictional Truth. Baltimore: Johns Hopkins University Press, 1990.

Rosset, Clément. Le Réel: Traité de l'idiotie. Paris: Editions de Minuit, 1977.

Rousset, Jean. Le Mythe de Don Juan. Paris: Armand Colin, 1978.

Serres, Michel. “Un dieu du stade: Monsieur Alphonse.” Alphonse Juilland: d'une passion l'autre. Ed. Brigitte Cazelles and René Girard. Saratoga: Anma Libri, 1987. 7-31.

Spenser, Edmund. Poetical Works. Ed. J. C. Smith and E. de Selincourt. London: Oxford University Press, 1912.

Todorov, Tzvetan. Introduction à la littérature fantastique. Paris: Editions du Seuil, 1970.

Virgil. The Aeneid. Ed. J. W. Mackail. Oxford: Clarendon Press, 1930.

———. The Aeneid. Trans. Allen Mandelbaum. New York: Bantam Books, 1971.

Webb, Ruth. “The Transmission of the ‘Eikones’ of Philostratos and the Development of Ekphrasis from Antiquity to the Renaissance.” University of London Diss., 1992.