More often than not, books are the children of discontent or dissent.

The idea for the essays

in this book (and the conference where they were first delivered) was engendered by my
personal dissatisfaction with the state of visual poetics and the research in word-image
relations. In the late 1980s, while attending international meetings and the annual ASECS
conferences in the United States and acquainting myself with the rhetoric of experts in
narratology, art history, philosophy, and literary theory, I began to realize that the

fields of ekphrasis and intermediality (whose changing boundaries I will try to outline below)
had apparendy seen litde progress or change. Since I was and continue to be personally
concerned, in a manner of speaking, in a study of the intertextual and intermedial relations in
Hogarth's graphic art, I deemed it high time in the early '90s for a meeting that would deal
with the problems at hand and, if possible, get us ahead. The critical and ekphrastic
malaises I had diagnosed were the following. Structuralist and poststructuralist theoreticians
- from Bakhtin, Barthes and Kristeva, to Foucault, Lacan, and Derrida - had published a
series of studies that should have shaken the foundations of a number of disciplines,
including humanist art history and literary criticism, and put into question such cherished
beliefs as the "mutual illumination of the arts." To be sure, "French theory," as it came to be
called in the 1980s1 , did have an international effect. Roland Barthes, for instance, taught
the world that everything from painting to objects, to practices, and to people, can be studied
as "texts." Barthes, Kristeva and, in their wake, Umberto Eco made us see what semiology
can do for the understanding of cultures and social practices and their expression in images.
Foucault and Althusser opened our eyes to the social production of meaning and its
inscription through power. And the Lacanian psycho-analysis demonstrated how the human
subject is formed in the play of gender difference. Finally, Jacques Derrida's still hotly
contested, and frequently misunderstood, statement "il n'y a pas the hors texte," is, if we
remain clear-headed, rather conventional by modern philosophical standards; but what it
does tell us, ruthlessly and honestly (which is the real reason why it has found so much
objection), is that there is no Archimedean point outside of language from which the truth
claims of language itself can be verified. Where, I had to ask myself in the early '90s, are the
studies in ekphrasis or visual poetics that engage with or profit from Derrida's concepts of
the "supplement" and the "différance" or his claim that in art it is "the represented and not the
représenter, the expressed and not the expression" which matters? One had to look very
hard and for a long time to find such critical works.

One of them saw the light of day in 1988, when Norman Bryson published a collection
entitled Calligram, which assembled important essays in "the New Art History from France"
(including articles by Kristeva, Baudrillard, Marin, Foucault, and Barthes). Bryson intended to
contrast the AngloAmerican approach in art history, and notably the continuing "Perceptualist
procedure" as represented by Gombrich (in his Art and Illusion), with the French insistence
on the sign, its social production and formation, and the problems of power and control. If
Bryson's ultimate aim was to prove, like the Sternian traveller in A Sentimental Journey, that
they "order this matter ... better in France," he was honest enough to concede that "as
elsewhere, in France too a conservative art history dominates the major institutions" which
tend to "consolidate the most positivistic aspects of art historical writing." Excepting some
challenging contributions from absolute outsiders like Régis Debray or members of the
Grandes Ecoles, this still holds true today.3 Apart from Calligram, Bryson has produced a
number of interesting studies suggesting new ways in visual poetics. Like Mieke Bal, another
important theoretician, Bryson orginally comes from what he terms the "Literary Field"; and it
seems to me that it is precisely the crossing of disciplinary borders (Bryson from English
Literature to art history; and Bal from narratology to visual rhetoric) which makes their works
innovative. Bal's Reading 'Rembrandt' is, in my opinion, the most advanced and sustained
attempt to date to get us beyond the word-image opposition. It is precisely because she has
a feminist axe to grind and thus reflects on ideology, gendering and vision within the more
immediate art historical and rhetorical concerns of the book, that her approach is both
persuasive and timely, given the preceding decades of gender-biased criticism. Bal may not
really deliver what she promises in the subtitle of her leviathan monograph; and yet Reading
'Rembrandt' (the inverted commas serve to deconstruct the author concept) remains a
powerful statement that urges us to consider pictures as rhetoric or encoded signs that must
and can be 'read' with the tools provided by narratology and poststructural theories, including
feminism

But the voices of Bryson, Bal, and Barbara Stafford5 , vociferous as they may be, remain
those of enlightened singers or prophets in the desert - art history in America and Britain has
yet to respond to their challenges.6 In many ways the missing echo is comparable to the
lacking response in art history and literary criticism to the poststructuralist re-thinking of the
author concept and of authorial intention. Beginning in the late 1960s, theoreticians as
diverse as Lacan, Althusser, Barthes and Foucault all questioned the status of the
homogeneous, independent, individual that was now being described as an ideological
construct. The concept of the author was one of the first casualties of this attack. The
advantage of the new view (sometimes ridiculed as the denial of the existence of biological
individuals) is that the human being, including the former "genial" (god-like) author7 , is now
located within a system of conventions, practices, norms and so forth, in short: of discourses
that actually articulate him/her - the author comes into being as someone "spoken." This is
the important notion behind what is often called "the death of the author."8 For both art
history and literary criticism, one could have imagined farreaching and fruitful consequences
of such challenging of traditionally cherished hermeneutic concepts. But even the most
superficial survey will show that critics in both fields continue, unperturbed, inquiring into the
possible intention of say Hogarth or Swift.9 My last point in this brief litany of complaints
about the lacking response to recent theory is the discussion of one of the darling concepts
in the relations between words and images. This is "utpictura poesis," which in art criticism
and history has led to what is often termed the "correspondences of the arts" or "the sister
arts" or even "the mutual illumination of the arts." It is a notion that has seeped over into the
practice of ekphrasis. Indeed, Horace's famous dictum, which can be translated in several,
and to my mind different, ways (e.g., "as is painting so is poetry," or "as in painting so in
poetry"), has cast its long shadow over the Western tradition. The simile, however
interpreted10 , asserted

the likeness, if not the identity, of painting and poetry, engendering an extensive body of
aesthetic speculation. In the Renaissance and Enlightenment periods, it brought about the
beginning of a theory of art and art history. Lessing's objections in his Laokoon: oder über
die Grenzen der Mahlerey und Poesie (1766) were concerned less with the wish to do away
with the Horation assumption (or demand) than with the purity and delimitation of the arts,
which in itself reflects neoclassical and bourgeois aesthetic preoccupations. His insistence
on the spatiality of visual art and the temporality of verbal art does not, however, produce a
clash in the postmodern sense; paradoxically, it puts into practice yet another détente
between the arts, a friendly alliance which, as Tom Mitchell has demonstrated, masks
Lessing's fundamental fear of the irrational power embodied in images and icons.11 Vestiges
of the Laokoon mark the literature of the sister arts, from the most historical of accounts (R.
Lee, Hagstrum) to the most theoretical (Krieger, Steiner, Bryson); and one must indeed
conclude with Scott that "the great devide between painting and poetry — at least in critical
circles - remains as formidable today as it was in the eighteenth century."12 It seems that
these critical circles are not afraid of "the risky presumption that the visual work of art can be
translated into the terms of verbal discourse without remainder."13 The ineradicable belief in
a correspondence appears to be suspiciously neat and fitting. Is there no room in ekphrasis,
we may ask with Stephen Bann, "for the modern conviction (deeply rooted at least since the
time of Baudelaire) that a good proportion of what is experienced in looking at a work of art
simply cannot be expressed in verbal terms?"14 If poststructuralism has taught us anything it
is the knowledge that making meaning depends on the fickle nature of the sign, which is
subject to personal and social determinants. When, it may be asked, will this undeniable fact
be recognized and considered by our more reluctant confrères who are writing on visual
poetics and the relation of verbal and visual art? These, in a few words, were the reasons
that persuaded me to organize a symposium on ekphrasis and intermediality and to publish
the proceedings in Walter Pape's distinguished series.

Let me now turn to this book. As the reader will notice immediately, it cannot serve as a
remedy for the malaise described above - nor was it designed as such. The main target is a
survey and description of the status quo in the writing about art and the relations between
texts and images. This is in itself a first important move before one approaches new shores.
But a large part of this pioneering work of adventurous discovery and (as discoveries go)
appropriation is still to be done. We, the contributors to this volume, are glad to provide
bases from which departures are possible in many postmodern directions. If the collection
also gets us a few steps ahead, it is all the better for all concerned, even if that progress is
only minimal. I was lucky to find contributors from several countries, and I think this is
important (not merely because of what the traveller says at the opening of Sterne's Λ
Sentimental Journey)·, the reader will be able to recognize different approaches here; to
some extent, these differences are due to national outlooks, but overall they reflect more of
the personal critical interests of the contributors. I have also insisted on the possibility that
we can publish in two languages, which guarantees "freedom of enunciation" (at least for the
greater part of the contributors), for anyone writing or publishing in a foreign language knows
how hard it is to find "le mot juste." From the very beginning, I wanted to limit the period
under survey to the eighteenth and nineteenth centuries because a number of studies and
proceedings of conferences have already been published on the twentieth century or the
entire modern period, and because even more critical literature is available on the time
before 1600.15 With the exception of Walter Pape's contribution (which does return, by way
of Crumb's images, to the eighteenth century too), the essays in this book focus on the
Enlightenment and the nineteenth century. My aim was to unite art historians, philosophers,
literary scholars (from French, German, and English departments), and comparatists. This
aim has been largely achieved, although we would have welcomed more colleagues from art
history if they had they been available (Norman Bryson, Barbara Stafford, Klaus Herding,
and Werner Busch were, unfortunately, busy elsewhere). The book is in four parts. The
essays in the first section deal with the theory of reading and describing visual
representation, including the works of such arch-ekphrastics as Diderot and Gautier. Part II
is concerned with theory and practice in the eighteenth century, as we are introduced to
discussions of the readings of works by Watteau, Fragonard, Hogarth, and visual
representations of Raynal and Humboldt. Part III presents analyses of iconotexts from the
nineteenth century, and a special, final, section deals with the popular genre of caricature.

Since so far I have not even explained how I define and understand the key terms in the title
of this book (ekphrasis, iconotexts, and intermediality), I must ask the reader to forgive me
while reminding her, however, that one of my favourite authors, Laurence Sterne, practices
the same procedure in his Tristram Shandy. Whereas Sterne delivers as it were only in
Chapter 20 of volume III, I feel obliged to satisfy the reader's curiosity forthwith. In the
following sections of the introduction I will address three issues. Firstly, the definition and
discussion of the important terms and concepts in their various

1. Definitions and Concepts In Herman Melville's mighty Moby-Dick or, The Whale (1851),
art and artists constantly interfere in the text — and so do the fictional readers or observers
of art works. A great concern with the production and understanding of painting as a visual
text to be decoded seems to lie at the heart of the novel, constituting as it does one
particular form of a general epistemological questioning. From Chapter I, where "an artist [...]
desires to paint you the dreamiest, shadiest, quietest, most enchanting bit of romantic
landscape in all the valley of the Saco" (p. 4), to Chapter 118, in which the Pequod's three
masts are compared to (visual representations of) "the three Horatii pirouetting on one
sufficient steed" (p. 545), the text repeatedly turns to ambiguous paintings and art objects
placed in front of helpless observers - and the readers of Melville's novel. One of the more
intriguing examples occurs when, in Chapter 36, Ahab first nails the golden doubloon to the
mast, and then, as the English of Shakespeare and the Elizabethan revenge tragedy blends
with the powerful prose of the Bible and Milton's lofty verse (and many other pre-texts),
forces his crew to unite in an "indissoluble league." Ahab calls out to his harpooneers:

"Advance, ye mates! Cross your lances full before me. Well done! Let me touch the axis." So
saying, with extended arm, he grasped the three level, radiating lances at their crossed
centre; while so doing, suddenly and nervously twitched them; meanwhile, glancing intendy
from Starbuck to Stubb; from Stubb to Flask. It seemed as though, by some nameless,
interior volition, he would fain have shocked them into the same fiery emotion accumulated
within the Leyden jar of his own magnetic life.

The scene is obviously inspired by two major pretexts (or subtexts), one verbal and the other
visual: Shakespeare's Hamlet and Jacques Louis David's Le serment des Horaces (colour
plate I). David's painting — and its highly ambiguous, contradictory, socio-political
implications - are integrated into Melville's text. The dramatic scene relies for its meaning on
the reading of David's canvas and all the sources behind it, including the classical story of
the Horatius family. Understanding Melville's text thus becomes an intertextual or (as I shall
argue below) intermedial venture that should take into account the indeterminacy of the
painting

It seems to me that such appropriations of images by texts need to be reassessed in the

light of recent theory which rejects essential assumptions of the sister-arts approach (e.g.,
the translatability, the correspondence, and the unity of text and image17). In an analysis of
John Braine's use of Edouard Manet's Olympia in the novel Room at the Top (colour plate II)
I have shown that fiction can be as iconotextual as, say, an engraving: Braine misreads (i. e.,
limits the meaning of) Manet's painting while subjecting it to the strategic aim of portraying
the change of his hero's male gaze. Trying to understand Braine's novel and its appropriation
of a picture means trying to understand, firsdy, Manet's Olympia and, secondly, the particular
way in which the text represents the image in what amounts to a misrepresentation.18
Melville's and Braine's use of or implicit references to David's The Oath of the Horatii and
Manet's Olympia are examples of ekphrasis, although they could of course also be
considered and studied as allusions, as important knots in the large fabric established by
such intertextual/intermedial weaving of the textual fabric.19 The latest definition of
ekphrasis seems to accomodate my reading. Tom Mitchell, Grant F. Scott, and James
Heffernan, for instance, define the term as "the verbal representation of visual
representation."20 But do they mean the same thing as David Carrier who also describes
ekphrasis as "verbal re-creations of the visual artwork" but then associates it with what he
terms "Renaissance artwriting," though he concedes later on in his study that ekphrasis can
be "considered a kind of interpretation"?21 Obviously, as one of the most distinguished
specialists on the subject has recendy argued, "a study that parades under the name
ekphrasis can be many things."22 It seems that what we need at present is a clarification of
extant concepts. We may begin, in the good old German philological fashion, with the origin
and development of the word and the concepts it has designated. If critics agree at all about
ekphrasis, they stress the fact that it has been variously defined and variously used and that
the definition ultimately depends on the particular argument to be deployed.23 Ekphrasis
is/was a poetical and rhetorical device and a literary genre, and if we can trust Raymond A.
Macdonald it has quite recendy contrived to become "a charismatic panjandrum of literary
critics and art historians," a popular new star in the
poetic and art historical firmament.24 Consisting of the prefix 'ek' (or 'ec' and even 'ex')
meaning 'from' or 'out of,' and the root term 'phrasis,' a synonym for the Greek lexis or
hermeneia, as well as for the Latin dictio and elocutio (the verb phra^ein denotes 'to tell,
declare, pronounce"), ekphrasis originally meant "a full or vivid description." It first appears in
rhetorical writings attributed to Dionysius of Halicarnassus and then became a school
exercise in rhetoric. It seems that in the fifth century A. D., there was a tendency to limit the
basically rhetorical term to (poetic or literary) descriptions of works of art, and in this limited
sense ekphrasis has had a long and complex history that stretches from the description of
Achilles' shield in Homer's Iliad, to the ekphrases of the tapestries of Minerva and Arachne in
Ovid's Metamorphoses, and down to Shakespeare's The Rape of Lucrece, Keats's "Ode on
a Grecian Urn," W.H. Auden's "Musée des Beaux Arts," William Carlos Williams's Pictures
from Brueghel, and Lawrence Ferlinghetti's When I Look at Pictures (1990). This poetic
description of works of art (mostly paintings) in a literary mode is the subject of James
Heffernan's recent perceptive survey of a body of literature (from Homer to Ashbery) that
evinces the paragonai struggle between word and image.25 Some of these works of art
engendering ekphrastic description are of course not real (e.g., the painting and the urn in
the texts of Shakespeare and Keats). John Hollander therefore proposes to call the texts
based on them "notional" ekphrases, but I think WJ.T. Mitchell has a point when he argues
that "all ekphrasis is notional, and seeks to create a specific image that is to be found only in
the text as its 'resident alien'."

Looking at the development and use of the word in the English language, one is immediately
struck by what Grant F. Scott has aptly termed "the split personality" of ekphrasis.27 The first
entry in the Oxford English Dictionary dates from 1715, when it is defined as a specifically
rhetorical term, "a plain declaration or interpretation of a thing"; whereas the second entry
(1814) associates it with "florid effeminacies of style." The implications (in terms semantics,
ideology, and gender) are, of course, myriad and fertile.28 Ekphrasis, then, has a Janus
face: as a form of mimesis29 , it stages a paradoxical performance, promising to give voice
to the allegedly silent image even while attempting to overcome the power of the image by
transforming and inscribing it.30 In the final part of the introduction, I shall return to this
central paradox of the ekphrastic enterprise, a paradox that is closely linked with the verbal,
rhetorical, nature of writing (whether it is literary or critical) and the assumption that there is
an essential difference between image and text.

Ekphrasis, then, originated in the field of rhetoric and has been appropriated by literary
critics and art historians. Whether this development is a recent phenomenon, as Raymond
A. Macdonald argues, remains to be determined; but it is true that the present use of the
term as "coterminal with the description of the subject-matter of an art object" disencumbers
it of its rhetorical restriction.31 Beginning with Svetlana Alpers's seminal article on
"Ekphrasis and Aesthetic Attitudes in Vasari's Lives" (1960), the term has taken on a new
conception. While this is a normal process in criticism, there is a danger that is perhaps best
exemplified in Murray Krieger's latest claim that "by removing the italics," he "has brought
'ekphrasis' into the language."32 The danger is precisely that the term is now treated not as
a neologism but, in Macdonald's words, one of "antique usage with an extended pedigree"
which thus falsifies the rhetorical and poetical bases of classical criticism of the visual arts.33
What Bernadette Fort distinguishes in her contribution as "critical ekphrasis" extends the
meaning and usage of the term into criticism and art history. This form of ekphrasis has been
with us at least since the time of Diderot, Heinze, and Goethe

Surveying what the contributors to this book have said about it, I should like to suggest that
we extend the use of ekphrasis (as a poetic, literary, mode) to encompass "verbal
representation" in its widest sense, including critical writing. David Carrier and James
Heffernan have both argued that ekphrasis may and should open itself to the vast body of
writing about pictures "which is commonly known as art criticism."35 We should drop, once
and for all, the tacit assumption that the verbal representation of an image must be "literary"
to qualify as ekphrasis - in our age of the arbitrary sign it has become extremely difficult to
distinguish between a "literary" and a "critical" text.36 If ekphrasis is "the verbal
representation of visual representation," a definition most experts now seem to accept, the
first part of that definition can only mean: all verbal commentary/writing (poems, critical
asssessments, art historical accounts) on images. All such writing is essentially ekphrastic:
the difference between the critical and the literary versions is one of degree, not one of mode
or kind. Another extension of the term that seems urgent at the moment concerns the public
(non-specialist) reaction to pictures. Ekphrasis has been and continues to be the province of
specialists who, like art historians generally, compartmentalize out those who do not respond
in the way they do: "madmen, women, children, and less educated people generally
(especially the primitive and the illiterate)." David Freedberg's description of the art historical
control of response to pictures has also been true for ekphrastic writing - what remains to be
(re)discovered is what Wolfgang Kemp terms "the ars conversationis" of the eightenth
century, i.e., the public discussion (in coffee houses and taverns) of art works and the
mentalités that informed it.37 Tom Mitchell's
thoughtful analysis of "pictures and the public sphere," conducted in the light of Habermas's
notion of a "democratic public," is a timely and most useful step in the right direction.38 Let
us now leave ekphrasis for a while and turn to the second provocative term, iconotext, which
I applied to the implicit allusion to David's Le serment des Horaces in Chapter 36 of
Moby-Dick. 39 By iconotext I mean the use of (by way of reference or allusion, in an explicit
or implicit way) an image in a text or vice versa. I have borrowed the term from Michael
Nerlich, who introduced it to designate a work of art made up of visual and verbal signs,
such as Evelyne Sinnasamy's novel with photographs, La femme se découvre, in which text
and image form a whole (or union) that cannot be dissolved. Alain Montandon then adopted
the expression in his publications to define works of art in which writing and the plastic
element present themselves in an inseparable totality.4

While Nerlich and Montandon apply the expression to modern and postmodern works of art,
I have extended iconotext both chronologically and semantically. Chronologically, in that in
my various publications on Hogarth's use of writing as a visual constituent of his engravings
(e.g., the Bible, newspapers, pamphlets, ballads), I have repeatedly referred to such prints
as iconotexts (Nerlich and Montandon would prefer to restrict it to twentieth-century collages
and montages). Like some of Gillray's caricatures discussed by David Bindman, Hogarth's
engravings showing some form of eighteenth-century discourse in writing are iconotextual
precisely because they integrate the semantic (denotative and connotative) meaning of the
written texts that are iconically depicted, urging the 'reader' to make sense with both verbal
and iconic signs in one artifact.41 In this sense, Hogarth's prints can be compared to the
traditional emblems of the Renaissance (on which Hogarth also drew), a classical example
of iconotexts which are, however, pre-determined in that the reader was expected to
recognize and accept commonplace assumptions. However, Hogarth's graphic art also
subverts and burlesques this tradition.42 I have extended the term semantically by applying
it not only to works which really show the interpénétration of words and images in a concrete
sense (e.g., in Hogarth's Beer Street, or Gillray's French Liberty. British Slavery) but also to
such art works in which one medium is only implied (e.g., the reference to a painting in a
fictional text). Unlike 'iconicism,' a term coined by Richard Wendorf to designate iconic
biography and iconic portraits in a descriptive system that maintains the separation between
spatial and temporal forms, iconotext refers to an artifact in which the verbal and the visual
signs mingle to produce rhetoric that depends on the co-presence of words and images.43
In his latest discussion of verbal-visual relations, Tom Mitchell refers to such works as
imagetext, distinguishing between such composite works and "'image/text' as a problematic
gap, cleavage, or rupture in representation," and 'image-text' as designating relations of the
visual and verbal.44 It seems to me that iconotext, in the sense defined above, is an
appropriate and less cumbersome term we

can apply to pictures showing words or writing, but also to texts that work with images.
Ottmar Ette, in his essay in this collection, employs iconotext in this sense.45 Some of the
contributors to this volume seem to like the term, and if what I have said about it is
convincing it will perhaps gain greater currency in the future. Finally, I must say a few words
about the second neologism in the tide of the introduction. What is intermediality? A short
answer would be: a sadly neglected but vasdy important subdivision of intertextuality. In a
perceptive article on intertextuality and visual poetics, Norman Bryson has broached the
question whether, from an intertextual point of view that takes into consideration the work of
Barthes and Derrida, images can be 'read' like texts. He concedes that one may think of
paintings as mutually interpenetrating (his term for my intermediality), that, in other words,
there is mobile intertextuality in the realm of the image too; but he finally shies away from
treating texts and images alike (in intertextual terms): Bryson claims that paintings, unlike
texts, "possess embodiment," a concept which in his view deconstructs the opposition
between matter and information and which finally accounts for the fact that paintings offer a
resistance to intertextuality.46 Even though this is a vastly helpful essay, Bryson is mistaken
in assuming an essential difference between texts and images when it comes to semantic
and rhetorical relations in the intertextual field 47 Arguing that since images, like texts, are
rhetorical and must use signs to express meaning (signs, however, that are verbal, iconic or
both at the same time), I will return to this issue too in the final part of the introduction.

Suffice it to say here that it may be useful to revive that obsolete word, intermedial, which the
Oxford English Dictionary lists as a synonym for "intermediate." The passage from
Moby-Dick quoted above is a prime example of intermediality, of the "intertextual" use of a
medium (painting) in another medium (prose fiction). While intertextual studies are at present
the rage in English departments (at least in Europe48), the use of pictorial works as an
essential constituent in a text is a field where much remains to be done. The main reason for
the absence of studies in intermediality is, of course, the reluctance of many scholars to
work in what they still consider to be alien territory. Three years ago, when I gave a lecture at
the Universität Tübingen on intermediality in Victorian poetry, one of the colleagues in the
local English department thanked me afterwards for what he termed a fascinating paper
"moving close to the border of art history." The wording is telling. We need more studies,
then, in both visual poetics and ekphrasis to explore those fascinating works that combine
visual art and prose fiction (or music and fiction49), urging the reader not to give preference
to one medium but to consider both.50 Mieke Bal has made a great step forward in this
direction with her admittedly radical Reading 'Rembrandt'. But if we are to overcome the
"correspondence-of-thearts approach" that assumes a "mutual illumination" (which still rules
the fields of both literary criticism and art history51) in the study of iconotexts while turning a
blind eye to the warfare involving difference and différance in each medium and between the
media, we should move from the theoretical ground prepared by Bal, Bryson, Hansen-Löve,
Zander, Reader, Plett, and Fowler, to practical interpretations that take into account the
intermedial nature of iconotexts. I propose a name or tag for such enterprises - let us call
them studies in intermediality

The Essays

In the section concerned with theories of reading visual representations, JeanPierre Dubost
opens the batting with an assessment of iconolatry and iconoclasm in the French
Enlightenment, the age of eloquence (l'âge de l'éloquence) that saw the meeting,
confrontation, and mingling of religious rhetoric and libertine discourse. For a book inquiring
into verbal and visual representation, i.e., into rhetoric in all its forms, there could be no
better starting point than the libertine discourse on love. Dubost is a specialist in this field.
He has written one of the most perceptive and intelligent post-doctoral studies on "literature
and libertinism" in the Enlightenment.53 In the present article, he marshals the forces of
poststructuralist theory and philosophy to tackle the aporias of a discourse that has come to
be recognized as a hallmark of the French eighteenth century. Tracing this form of "écriture"
back to earlier religious and rhetorical models, Dubost argues that libertine writing (an
apposite term indeed, for it is not confined to literature but absorbs different discursive forms,
from the confession to the autobiography and the novel) can be distinguished from other
modes of erotic expression, and that it can be understood both in the context of the century
but also in terms of the history of hypotyposis and ekphrasis. What is important, however, is
not analogy over the ages, but the mimetic status of libertine writing as such.
The article contains two major contentions. Firstly, that the rhetorical strategies of libertine
writing were based on a conscious, and frequently even a deliberate, de-finalization of
sacred icons, broadly conceived; secondly, that these strategies are not a mere reversal of
positive images/pictures (in the widest sense of the term) but that they constitute and
foreground the great potential of ambiguity and ambivalence contained in the images to be
subverted. This logic of exhaustion (of representation) is carried to an extreme point by the
Marquis de Sade. It is in his libertine writings that one finds, according to Dubost, a
permanent hypotyposis (one of Dubost's key terms) of evil which prevents, once and for all,
the pictorial control of evil (i.e., the body) through the text. But this reckless "emptying,"
radical though it may be, also creates the possibility of thinking without images. Despite all
appearances, this thinking (in its discursive form) is surprisingly close to Kant's iconoclastic
aesthetics. In fact, one could maintain that the total disillusionment through images/pictures
is the blind illustration as it were of aesthetic ideas, an illustration that is both necessary and
impossible. Exploring the semantic, linguistic, and philosophical relations between forms of
representation that only seem to be different at first glance, Dubost takes us straight into the
heart of our business: representing in words and images, and the representation of images
in words. His key term, hypotyposis, also used by other contributors (see, for instance,
Bernadette Fort's article), can be translated as a "vivid description of a scene, event, or
situation, bringing it, as it were, before the eyes of the hearer or reader"; to this definition in
terms of rhetoric listed in the Oxford English Dicitionary we may add significant additions as
recorded in Webster's New Twentieth-Century Dictionary·, hypotyposis has general,
rhetorical and scientific denotations; it also means "wordpicturing," while scientists use it in
the sense of "a concise outline of a subject."

As Bernadette Fort shows, it is only a short step from hypotyposis, a performative strategy,
to what she defines as "critical ekphrasis," the cradle and fountain of modern art criticism.
Trying to answer the question whether ekphrasis and interpretation are two
incommensurable modes of writing (David Carrier's contention pronounced in 1987), Fort
explores the ways in which Diderot, in his famous commentary on Fragonard's Le
Grand-prêtre Corésus s'immole pour sauver Callirhoe (written for the Salon of 1765),
reworked the conventions and codes of ekphrasis. In a closely argued and wholly persuasive
contribution, she demonstrates how Denis Diderot once again moves to the forefront of
literary and critical innovation by drawing on a literary mode and subordinating it
systematically to the reconstruction of representational strategies that are at work in
Fragonard's grand history painting. Diderot's ekphrasis, then, manages to be several things
at the same time: a poetic piece in its own right, and a seminal example of interpretive art
criticism. What makes Fort's article a daring and innovative enterprise is the fact that she
uncovers not merely Diderot's sophisticated rhetorical and critical
manoeuvres. In a most impressive manner she demonstrates what research on ekphrasis
can do for us and the culture of the past: on the one hand, ekphrastic texts reveal crucial
insights about the art of their time (insights, to be sure, we may want to modify and adapt to
our "eyes"), and, on the other hand, they also unveil strategies of writing, the powerlines of
critical discourse and aesthetic response. Together with Bernard Dieterle's recent
monograph on "narrated pictures" and Osterkamp's postdoctoral study of Goethe's
ekphrastic criticism (which, as can be expected, pardy draws on Diderot), Fort's essay does
a power of good to our understanding of ekphrasis as practiced in the age of
Enlightenment.54 Frederick Burwick takes us from the ekphrastic enterprise of the
Enlightenment into that of the early nineteenth century, investigating as he does the mimetic
crisis of Romanticism. Like Dubost, Burwick pursues the less obvious, subterranean,
connections between mimesis, ekphrasis, and aesthetic illusion.55 Burwick starts with the
recognition of a crisis in representation, a crisis that seems to have affected the Romantic
writers as they strove to abandon the ruins of imitation and began to worship at the shrine of
subjectivism. At this point of disjuncture, Burwick shows how they made a virtue of necessity,
developing a number of strategies for coming to terms with what was indeed a crisis of
representation. Focusing mainly on texts by De Quincey, with insightful brief excursions to
writings by Coleridge, Wordsworth and other Romantics, Burwick sketches the manner(s) in
which confession is metamorphosed into ekphrasis as the self becomes the major
preoccupation of the writers. Burwick demonstrates how the province and techniques of
mimesis were being redefined in Romantic literature, a process that involved a great concern
with perception and the reliability of mimesis in the representation
of reality. Following contemporary critical emphasis on metonymy and difference, he looks at
the aesthetics of skepticism and incredulity as an alternative to illusionism and "willing
suspension of disbelief." In such ekphrastic poems as Keats's "Ode on a Grecian Urn" or
Wordsworth's "Elegiac Stanzas" the competing verbal and visual challenges of
representation or, in Bloom's terminology, misprision56 , involve the same self-reflexive
strategies that inform Coleridge's poems about writing (or not writing) poems, or Blake's
pictures about pictures, and books about books. Burwick's valuable contribution delineates in
great detail how the skeptical repudiation of the mimetic and ekphrastic endeavour is not the
break-down but the breaking-out-of the reflexive entrapment as art imitates art that imitates
art. After Frederick Burwick's "archeological" inquiry into the epistemological and ekphrastic
backgrounds of representation in the English Romantic period, Alain Montandon presents
his critical account of a French approach, a few decades later, in the writings of Théophile
Gautier, surely one of the most influential figures of his time (as I have demonstrated in my
contribution on Oscar Wilde's poem, "Impression du Matin"). Montandon argues
persuasively that Gautier's effort as an ekphrastic has been vastly underestimated. A
specialist in what may be termed the "transposition of art," Gautier developed a kind of
narrative discourse which helps unmask the "dispositif libidinal," which is Lyotard's coinage
for an erotic-libidinous dimension of the work of art.57 As Montandon proves, Gautier
develops a new ekphrastic perspective that is postmodern avant la lettre, moving as it does
from the ideas of correspondence and transposition to a more radical insistence on the act of
creation. In one of his last works, Spirite, Gautier even puts into question the power of
language when it is faced with the indescribable (art); he thus restores both the importance
of the image as representation and its connotative realm beyond denotation.

In paying particular attention to Gautier's repeated concern with two special iconotextual
forms - the hieroglyph and the arabesque - Montandon explains the particular fascination
these codes of drawing in writing (and vice versa) held for a writer who was attracted by the
paradox of describing and thus catching the inexpressable, seductive, aspect of art.
Montandon finds the same aesthetic principle at work when Gautier engages in ekphrases of
paintings by Rubens and Leonardo da Vinci or more classical examples (fictive works of art
described in prose) in Mile de Maupin. For Gautier, describing a painting becomes a kind of
intimate archeology that must uncover and recover the process of creation and the mystery
of artistic attraction. Seduction is one of the key words in this ekphrastic enterprise, and it
turns up time and again in Gautier's prose which probes the limits of verbal representation
even as it seeks refuge in it. For Gautier, the essence of real beauty, which always
evokes a feeling of both strangeness and familiarity, remains something mysterious and
inaccessible. Part II assembles four contributions on particular iconotexts and their verbal
representation in literary and critical discourse. The survey starts with Watteau, an enigmatic
painter at the beginning of the century. His person and his paintings have generated a true
critical industry, and his work continues to be the focal point of ekphrastic attempts
investigating what is invariably described as the charm, the mystery, and the pleasure of his
canvases. Catherine Cusset focuses on the two versions of Watteau's Pilgrimage to the
Island of Cythera, which are arguably his best known paintings. Considering the texts which
surround them, she examines the relation of images to narrative and meaning as depicted in
the paintings themselves. Cusset diagnoses narrative inconsistency in the paintings and
establishes a useful link between Montandone discussion of the "arabesque" and Derrida's
exploration of the parergon: they both have particular significance for Watteau's rococo
treatment of pleasure depicted in what postmodern scholarship has described as a semantic
vacuum. Foregrounding the changes made for the Berlin version, Cusset penetrates to the
very heart of Watteau's highly ironical representation of pleasure in the framed space of
"nothingness." Watteau, it seems, stages the experience of aesthetic pleasure even while
opening a space of reflection that, ultimately, defies both definiton and delimitation.

From Watteau it is only a short step to Fragonard and his contemporary, Sterne. Both tried to
describe or transcribe the intensity of those "sentimental" (in French: "sensibles") moments
which constitute the flesh of human relationships. In his analysis of the analogies in the two
artists' use of eroticism - an investigation that happily steers clear of the
"correspondences-of-thearts-approach" (see Ogée's note 5) - Frédéric Ogée goes to the
deep structures of verbal and visual representation in Fragonard's Les débuts du modèle
(The New Model), painted soon after the publication of Sterne's A Sentimental Journey
(1768). Ogée argues that Fragonard comes close in spirit to the novelist. Like Sterne's
narrator, Yorick, when he tries to capture the magic of the "exquisite" moments spent with
the "fair fille de chambre," Fragonard seems to delight in the representation of the ecstasy of
restraint, of the suspended moment. The still untouched, blank, canvas ("toile vierge" is
perhaps a better, more ambiguous, term in French) in the middle of the artist's studio is
Fragonard's ironic and erotic comment on the flourish of the artist's cane delicately lifting the
model's petticoat. Ogée reads the painting as a remarkable celebration of that "sentimental
commerce" which, in the same period, Sterne was trying to evoke in his "quiet journey of the
heart in pursuit of Nature." Ogée is thus concerned with the deep currents running between
ekphrasis and ecstasy, between representation (in words and pictures) and eroticism. His
thesis is that Fragonard and Sterne employed the depiction of eroticism as a way of
suspending time and of frustating the reader's or beholder's narrative urge. Far from being
limited to mere sexual/textual teasing, as critics have often dismissed it, this eroticism can be
seen as an invitation to join in its essential suspension, an invitation to "sympathize" with the
vibrations it provokes, and to enjoy the shimmering quality of its transcience.

Laurence Sterne seems to have been the great mediator between painterly and fictional
representation — and the ekphrastic attempts to describe and dominate both. In Tristram
Shandy, he was himself very much concerned with the possibilities of representation as
such, constantly exploring its epistemologica! and semantic limits. Ronald Paulson sees him
as an influential source for Lichtenberg whose ekphrasis of Hogarth's graphic art, Paulson
argues, was shaped in decisive ways by the German's reading of Sterne. In his essay on the
critical reception of Hogarth's engraved works in the eighteenth century, Ronald Paulson has
selected three important Enlightenment ekphrastics whose works reflect the special
concerns of the age. Like Bernadette Fort in the first section, Paulson lays bare the
power-lines of ekphrastic discourse and the ways they shaped the contemporary
understanding of Hogarth's œuvre. Paulson's candidates for this archeological search are
three important commentators on Hogarth: the Reverend John Trusler, the Swiss-born Jean
André Rouquet, and the German "Universalgenie," Georg Christoph Lichtenberg. While the
first two wrote with Hogarth's (and Jane Hogarth's) approval, Lichtenberg, the first
semiotician, developed a new kind of ekphrasis which, according to Paulson, must be seen
in the light of Sterne' influence. What is interesting here is the fact that the three
commentators wrote for English, French, and German audiences and with essentially
different aims. To explain the iconotextual character and the palimpsest nature of the
Hogarthian prints, they had recourse to other discursive models and literary genres - which
were also Hogarth's 'architextes' in the Genettian sense (e.g., the novel, the drama, the
"progress") — thus anticipating Rodolphe Toepffer's apposite comparison, pronounced in the
nineteenth century, that Hogarth's engravings are "littérature en estampes" whose richness
in detail could scarcely be matched even by two novels from the pen of Samuel
Richardson.58 One of the many interesting points that emerges from Paulson's perceptive
study of the critical, explanatory, commentary on Hogarth's prints is the impression that
ekphrasis (drawing as it does on contemporary discursive patterns and mentalités) inscribes
itself in the visual representation - to the point of domesticating if not disfiguring the image
itself. Paulson's close reading of his predecessors (for his analysis could be extended to his
own stupendous work on Hogarth59) demonstrates that we are in dire need of a 'Critical
Heritage' volume on the verbal response to Hogarth's engraved iconotexts. For the recent
critical debate on "reading Hogarth," an ekphrasis that has been rekindled by the publication
of Paulson's catalogue raisonné and the revised three volumes entitled Hogarth,60 has
proved Hogarth's graphic works to be highly complicated and deeply ambiguous iconotexts -
the very opposite of what the Reverend Trusler considered them to be.

Ottmar Ette's contribution takes us from the ekphrasis of the Hogarthian iconotexts to a
fascinating reading of the images accompanying the works of Raynal and Humboldt. Ette's
inquiry into the semantics of the portraits of Guillaume-Thomas Raynal and Alexander von
Humboldt, as they appear in various texts of these writers published in the eighteenth and
nineteenth centuries, proves, once and for all, the rhetorical nature of "graven images" and
their proximity to both inscription and description. Made up of linguistic and iconic elements,
the engraved portraits of Raynal as well as the engraved and painted pictures of Humboldt
are prime examples of intermedial iconotexts that claim presence, authority, and, ultimately,
reality for what in essence is fictive representation. What links the iconotextual depictions of
the two writers is, according to Ette, the representation of the authorial gaze suggesting the
great thinker. This visual and rhetorical manoeuvre was, of course, partly desired and
intended by the represented themselves. Ette manages to tease out of the iconography of
the portraits the ideas and concepts that shaped contemporary scientific writing ("écriture")
and portraiture. The act of writing is caught in another iconic pregnant moment, staging the
place (e.g., the work table) where writing apparently takes shape and urging the observer to
consider the great authors, their painters or engravers, and the visual narration of all these
elements. Ette's valuable reading of important iconotexts from the eighteenth and nineteenth
centuries should take an important place in the research on the rhetoric of the body which
has been underway for some time.61 Chapter III is entirely devoted to the nineteenth
century. Selecting Jane Austen's novels straddling the turn of the century as his hunting
ground, Peter Sabor assesses Austen's handling of ekphrasis. As the Austen buff knows,
each of her six completed novels, as well as several of her early writings and Sanditon,
which remained unfinished at Austen's death in 1817, contains at least one scene in which
pictures play an important part. Sabor is concerned not with the theory of ekphrasis in these
cases, but with the narcological and semantic functions of images described or discussed in
words. Far from serving merely decorative purposes, these pictures, Sabor argues, are used
to reveal psychological nuances in the novels' characters and to explore the dynamics of
their relationships. Austen's fictive images — such as the portrait of Darcy in Pride and
Prejudice or the print of a boat in Persuasion - are interesting less for their representational
value than for the fact that they create debates about their subjects. Sabor shows that in this
respect the pictures function in the same endlessly engaging and provocative manner as the
texts of which they are a part. Austen's invented images are, in Sabor's terms, "speaking
pictures," discussed within the novels by the characters and designed to be discussed
outside her fiction by her readers.
Sabor's exemplary investigation of the function of pictures in novels can serve as a starting
point for what is surely one of the most underdeveloped fields of research. We may pursue
Sabor's line of inquiry (extending it, for instance, to the sophisticated handling of fictive
images in Emily Bronte's Jane Eyre), or we may venture into the discussion of "real" pictures
(e.g., well known paintings) in fiction. Both directions of research should try to lay bare not
the "correspondences of the arts" but the paragonai juxtaposition, conflation and/or
confrontation of the rhetoric of the image and the rhetoric of the text incorporating the image.

Wolfgang Lottes returns to the area of ekphrasis Bernadette Fort has tilled in the opening
section: his subject is what some critics now term "art writing," and his focus is on the
Victorian English reception of Botdcelli. Lottes's survey of that variety of ekphrasis stretches
from Joseph Archer Crowe and Giovanni Battista Cavalcaselle ("the new Vasari") to
Swinburne, Pater, Ruskin, Symonds, and the Pre-Raphaelites. Lottes shows how each of
these critics appropriates Botticelli, adapting the quattrocento painter's work to prevalent
English aesthetic principles while reacting both to the original paintings and the discourse
about them that had already been established. Swinburne, Pater, and Ruskin in particular
practically made Botticelli the artistic star of the finde-siècle cult of beauty and ennui. It
seems that Botticelli's paintings are sufficiently ambiguous (in the sense of Eco's "open
work" of art) to allow such appropriation or - in Lottes's terms - usurpation by ekphrastics of
the most heterogeneous dispositions. Lottes's detailed survey is a case study of the verbal
reaction to and, indeed, usurpation of paintings in the very act of ekphrasis. He also
comments on the role the mentalité and taste of the ekphrastic (the personal disposition)
play in this process marked by interference, displacement, and différance. James Heffernan
examines another traditional department of ekphrasis: the poetic representation (which is
always, as in the cases studied by Lottes, both an appropriation and misprision) of pictures.
These works of art can be imagined, as in Browning's poem, or real. One of Heffernan's
arguments is that a transition in ekphrasis took place from one to the other, as the museum
became the place where the experience of art is regulated. Beginning with a close reading of
Browning's "My Last Duchess," Heffernan first addresses the powerfully gendered rivalry
between word and image. Browning, he maintains, pursues a double strategy in this
ekphrastic poem: on the one hand, the verbal representation of a fictive image shows that no
pictorial work of art can ever be wholly mastered by the authority of the word; on the other
hand, in representing the Duke of Ferrara as the prototype of the museum curator, Browning
prefigures the way that twentieth-century ekphrasis enters the museum and salutes the
power of its words. Typifying this move, W. H. Auden's "Musée des Beaux Arts" verbally
reconstructs Breughel's Landscape with the Fall of Icarus, a painting whose meaning is
initially determined - for all who see it in the Brussels museum — by the words of the title
affixed to it there. It is Heffernan's contention that by viewing Breughel's picture in light of
these words, and of the various texts that stand behind Breughel's paintings, Auden remakes
Icarus in words as a museum-class specimen of how the Old Masters could represent
suffering. Always attentive to the paragonai aspect of ekphrasis, to the conflict and
competition it stages while pursuing its own rhetorical

and artistic aims, Heffernan's essay is a thoughtful close reading of important canonical
texts. In my "explication d'iconotexte," which follows Heffernan's analysis, I also provide a
close reading of a poem which is made up of allusions to verbal and visual texts. My aim is
to test the theories of intertextuality and discourse analysis (as advanced by Barthes,
Kristeva, Foucault, and, to some extent, Derrida) in a practical, "archeological," case study
uncovering the discursive field of late nineteenth-century art and poetry as re- and deflected
(i.e., represented) in a single poem by Oscar Wilde. It is my contention that such intertextual
or intermedial readings have more to offer than authororiented approaches, precisely
because they consider the author as part, perhaps even as a victim, of the discursive field of
his/her age. The intermedial approach also yields an evaluation of the iconotext discussed in
that one can determine to what extent an author or artist was able to recognize and deal with
the mentalités that dominated contemporary perception and thinking. The final section of the
book unites two articles concerned with caricature, providing examples of iconotexts par
excellence. David Bindman focuses on text as part of the visual image in the caricature of
James Gillray. Such text-image constructs need more critical attention. There is, to begin
with, a great variety of iconotexts, reaching from book illustrations to individual engravings
(with titles or words inscribed) and social and political caricature.62 John Dixon Hunt has
recendy opened a new avenue in the discussion of book illustrations by examining what
happens when images appear in different textual surroundings. He finds that the imagery is
indeed subdy changed by its altered verbal context.63 David Bindman does a similarly
important job in exploring the functions of text within some visual images produced by James
Gillray. If Hogarth reduced the amount of words in his paintings and engravings as he
matured as an artist, his sucessor Gillray relied on text either as an integral part of the image
or (more often) as what Genette would term the "seuil" (paratext) of the entire construct: the
space beneath the image that constitutes an access to (and an ekphrasis of) the "main" text
and its meaning.64 Bindman demonstrates that Gillray's caricatures without text inside the
image are ironically concerned with "High Art" in an intertextual and intermedial game with
Royal Academicians and their values. Printed text (or "speech-text" as Bindman prefers to
call it) as part of the central image usually appears only in specifically political caricatures
which are studded with allusions to verbal and visual pretexts. What emerges from
Bindman's discussion is that in the case of Gillray's caricatures, "les mots dans la peinture"
(to use Michel Butor's apposite phrase from his book tide) assume not only a status as visual
facts in themselves, they also go beyond their verbal meaning and create a space inhabited
by word and image. Together, they establish a rhetoric whose shared verbal and iconic
dimensions deserve more of our attention. Walter Pape brings up the rear with an
examination of the paragone of signs in Robert Crumb's visual reading of a famous
Boswellian text, the London Journal. Crumb's "abbreviations" (a favourite term of
Lichtenberg's in his characterization of Hogarthian iconotexts) may be considered as iconic
ekphrasis (which paintings, based on texts, are too): instead of words trying to contain and
re-create images, we are faced with images "reading" (in the sense of interpreting) words,
with a visual rendering offering a specific interpretation of an eighteenth-century text. This
article too provides new perspectives on the research in text-image relations, and I can only
hope that Pape finds successors in the exploration of a fascinating variety of ekphrasis.

3. Perspectives
One of the many paradoxes surrounding the image in Western culture is the fact that, from
the very beginning, it has had a bad press precisely because commentators (ekphrastics, for
instance, or philosophers) have felt its great irrational power. In his recent fascinating study
of "the life and death of the image," Régis Debray links the genesis of pictures with death,
and more precisely with the human wish to overcome the fear of nothingness (le néant) and
to preserve what must necessarily decay while domesticating the basic horror presented by
the idea of the void. Images, Debray argues, have been powerful and attractive all along
because they create the illusion that the visible is also readable, accessible, and —
present.66 Victorian writers and artists particularly the Pre-Raphaelites, exploited this
potential of the image in representations of women that combine necrophilia with the wish to
overcome the male fear of the femme fatale·, women are invariably depicted as conflations
of saints and sinners or as powerless sleeping or dead beauties.67 Apparently, we have to
live with the dilemma that we hold images in high regard but cannot really grasp them. Wim
Wenders's movie Bis ans Ende der Welt is a wonderful artistic comment on the state and
future of a world terrorized by pictures, of images that are helpless and render their human
observers as helpless as Narcissus, whose story could serve as a motto for the movie.
Another example that comes to mind are the truly marvellous and pregnant "scenic images"
in Robert Wilson's plays relying on the charm and power of pictures. Always dividing the
audience, they are both fascinating and difficult precisely because they demand a decoding
of their rich, arcane, rhetoric that replaces the words in traditional plays.68 Summing up and
echoing more than three decades of research on the paradoxical nature of pictures, Umberto
Eco argued in 1993 that the image possesses an irresitable force. It produces an effect of
reality, even when it is false. It cannot say by itself that it does not exist or that it is false,
whereas the text can do that. Without text, the image lies or gives way to a multitude of
interpretations.

Michel Foucault and Roland Barthes had expressed the same opinion many years earlier.69
The image, then, is both powerful and helpless in the sense that, as Wittgenstein wrote, it
cannot represent its form of representation but merely contains it as a trace that is frequently
overlooked by the observer.70 To a reader, it will initially always be more attractive than a
text; and yet in order to mean something it needs mendacious and/or distorting words: a tide,
an epigraph, a signature, an ekphrasis.71 Given the undeniable power of visual
representations, it should be no surprise to learn that Western culture has sought to limit that
power through an exegetical and, indeed, ekphrastic tradition which translates the pictorial
into the readable, thus controlling and encircling it with words.72 The human terror or fear in
front of and of the image has found expression, on the one hand, in the ekphrastic fear cast
in words73 and, on the other hand, in both philosophical discourse and the response of
iconoclastic observers destroying or disfiguring what they consider dangerous visual
representations. Plato extends his distrust of artists to the use of colour in general, whether it
is paint on canvas or the use of colour in rhetoric or make-up: cosmetics, rhetoric and
paint(ing) are potentially dangerous (because they serve as means of seduction) and must
be held in check. Jacqueline Lichtenstein has shown how this attitude, something one could
also term a mentalité, informed Western writing about rhetoric and painting into the
seventeenth century and beyond.74 And in his study of the history and theory of response to
images, David Freedberg has explored the psychodynamics behind the power of images,
the reasons why they arouse us religiously or sexually, and why we want to adore or destroy
them
Since this seems to be well known and documented, I should now like to return to a couple
of threads I left dangling while weaving the first section above. They concern the nature (or
concept) and paradoxical work of ekphrasis itself, a performance that promises to make the
silent image speak even while silencing the unspoken (and, perhaps, unspeakable) or
imposing verbal rhetoric (a kind of painting, in Plato's view) upon the image. There is no
denying the fact that ekphrasis continues to be a thoroughly paradoxical venture. But I think
there is a way out of the present dilemma. A part of that way has been sketched by Bryan
Wolfs seminal contribution on what he mockingly terms the "unnatural relations" between
literature and painting.76 If we marshal forces, supplementing Wolfs suggestions with some
important ideas provided by Mitchell, Bryson and Derrida, the present stalemate can be
overcome. I shall begin with Mitchell, move on to Bryson and then join Wolf to find the way
out of the cave. Tom Mitchell describes Western culture as one that can be characterised by
its woven fabric of signs in which the dialectic of word and image seems to have been a
constant.77 Although Mitchell is very much concerned with the struggle between verbal and
visual representation (the paragone), arguing convincingly that the image has always been
policed by the word, he makes a point that does not warrant his eventual pessimism. Mitchell
is of the opinion that "there is no essential difference [his italics] between poetry and
painting, no difference, that is, that is given for all time by the inherent natures of the media,
the objects they represent, or the laws of the human mind."78 What Mitchell says here is that
art and literature (and writing as such) cohabit within the same representational space; he
then goes on to stress, and study, the paragone, the fight for supremacy, between what he
terms "two kinds of signs." But instead of stressing the difference, as indeed we have done
since Lessing, we can also look at the representational space as a rhetorical field in which,
though a struggle may be going on, sign systems are at work. What is important, then - and
this constitutes the escape route - is the fact that representation (in its visual and verbal
varieties) has two common denominators: rhetoric and the sign. We may regret, with Mitchell
and other critics who have written in his wake, the marginalization and suppression of the
visual79 , and that we "have not gone far enough in our exploration of text-image
relations."80 But if images are rhetorical and "impure," if they always incorporate texts in one
way or another,81 we simply cannot sustain the assumption that there is a line separating
the visible from the readable. Derrida has recently argued that the demarcation does not run
between painting and words — it traverses both the pictorial and the lexical field in "the
labyrinthine way of an idiom."82 Nor is the cherished distinction between the 'spatial arts'
and writing a useful concept when we consider Derrida's lucid observations on spacing as
shared by texts and images and on the problems of mastering space by the look/gaze.83
We may and should therefore move on to an exploration of the common ground shared by
the image and the word, e.g., rhetoric, spacing, 'inhabitation,' and the partisan
re-presentation of mentalities.84 It is at this point that the poststructuralist critical studies of
the rhetoricity and semiotics of the visual arts (works published by Bryson, Moxey, and
Mermoz85, to name just three examples) come in handy. Both Mieke Bal and Bryan Wolf
(and, more reluctandy, Tom Mitchell) have recendy argued that the visual is as rhetorical as
the verbal. Our great mistake is to be seen in the fact that we have "committed an act of
historical suppression" by considering visual art as silent and surrounded by an aura of
nonverbal immediacy. We should, therefore, begin to read images, that is to study, not words
in paintings (the province of investigations by Svetlana Alpers and Michel Butor86) but their
semiotic and rhetorical structure.87 Art historians and critics from related fields seem to be
agreed that, so far, almost all attempts to establish a new "visual poetics" (a term suggested
by Mieke Bai in 1988) have failed.88 The reason is certainly to be found in the continuing
assumption of the differ - ence between what are essentially sign systems. In the wake of a
new and promising interest in the socio-political aspect of art, of its cultural setting and
background89, what is needed at the moment is, in Bryson's words, "a form
of analysis in art history dialectical enough, and subtle enough, to comprehend as interaction
the relationship among discursive, economic, and political practices."90 Inspired by
semiotics and discourse analysis, a new ekphrasis will have to consider the visual image as
sign, as discursive work that comes from and returns into society. Writing about art, then, will
involve the consideration within poststructuralist perspectives of what Bryson calls the
"archival" method (tracing the work of art back to its original context of production); and it will
reconsider the hermeneutics of interpretation, which is an act of recognition, reception, and
"reading" (hence Bal's title, Reading 'Rembrandt). In such an approach that combines the
best of what poststructuralism has to offer us (from semiotics to, yes - why not! -
deconstruction91) a visual work of art ceases to be treated as a "homogeneous site of
ideological and semiological coherence, and becomes a space of dispersion and
sedimentation in which conflicting possibilities work in parallel with - or, in certain cases,
against - authorial cl/aims and objectives."92 I agree with Bryan Wolf that "all art, whether
verbal or visual, must be anchored within a historical semiotics"93 that explores a poem or a
piece of canvas with pigment as what they are: signifiying systems and rhetorical constructs
that we try to appropriate.94 Insisting on the unity of the sister arts,

and not on their difference, we can study their common rhetorical structure.95 In this new
approach, history becomes rhetoric, and art (in its verbal and visual varieties) a system of
representation working with signs, codes, and frames that are neither "natural" (as Krieger
would have it96) nor silent or inaccessible. The need, then, is for what Wolf terms "cultural
ekphrasis," a reunification of writing and painting under "the common banner of
representation,"97 and for an intermedial exploration of the working of both linguistic and
pictorial signs in one medium.98 Intertextuality in art thus becomes a possibility in research,
not in the Brysonian, exclusive, way that again attributes a special status to paintings
("embodiment"), but in a more extensive variety of ekphrasis, broadly understood, which
considers visual art as a space serving as both meeting ground and battle ground for
encoded, rhetorical, sign systems that refer to texts and images

But the introduction and application of intermediality is not all that remains to be done in
ekphrasis and visual poetics. There are other untilled fields awaiting courageous visitors
(non-specialists) from neighbouring areas. For instance, the study of images as distorted
re-activations of collective memories, of what some scholars call mentalités. Together with
the oral tradition100 , texts and rites, images thematise and preserve what is virtually
present in the (subconsciousness and memory of people, precisely because visual art (of the
painted or engraved varieties) is also rhetorical and thus embodies unspoken attitudes.
Images throw light on a "latent" memory that is always in danger of being obscured, hidden
and displaced. Recently, Pierre Bourdieu has argued that we tend to overlook the large
invisible basis that informs the history of ideas ("l'immense socle invisible des grandes
pensées"), mainly because the varieties of "habitus" or doxa they display were so
self-evident for the contemporaries that they never bothered to record them in documents. I
would add that images can serve as what Bourdieu (quoting Satie) terms "memories of an
amnesiac" ("mémoires d'un amnésiaque") - for those who care to look and resist the
temptation of reading for reality, they display the silent codes much more obviously than
writing.101 Extending my suggestions above, I hold that pictures can be studied
meaningfully in an approach combining a historical semiotics with discourse analysis and a
shot of deconstruction. They can be decoded, not as the product of the genial artist (whose
intention must be unveiled) but as partisan representations of discursive and pictorial
traditions and mentalités.

One cannot, however, do such a step and then stop walking. For the pursuit of the power
lines of discourse in intermedial representations will eventually cast doubt on other concepts
involved in such an enterprise. One of them, my final point in this survey, is the role of the
observer/reader. Although we have already come a long way in our understanding of the role
of the reader (of texts and pictures), a way marked by the allegedly "open" concepts (the
implied reader) developed and introduced by Iser and Eco103 , the notion of a
predetermined art work that must be re-established by the "receiver" still prevails. Umberto
Eco's seminal The Open Work (Cambridge, 1989) at long last suggests a participating
reading agent who actively formulates the meaning of a work - not necessarily that intended
by the author. As Eco writes, the open fictional text or the painting is a "work to be
completed," offering a field of interpretive possibilities, a configuration of substantially
indeterminate stimuli which the recipient employs for his "readings

In his recent devastating attack on the "pure" eye of the beholder and the implied (educated)
reader (often a mask of the educated critic), Pierre Bourdieu suggests substituting such
concepts of the cultivated reader in hermeneutics and art history (as advanced, for instance,
by Gadamer, Panofsky, and Iser) with a social history of the observer. For the eye of the
beholder is not a given constant; it is the product of institutional settings and social forces
constituting what Bourdieu labels the "habitus."105 It is by historicizing the categories of
thinking and perceiving in the observer's experience, not by dehistoricizing them in the
construction of a transhistorical ("pure") eye106 , that we can arrive at an adequate
understanding of understanding ("comprendre le comprendre"107), which must of course
include ekphrasis: Bernadette Fort's article in this collection is in part a giant, and timely, step
in this exploration of the conditions of writing about art. Unlike Bourdieu, I do not think that
only sociology or social history can achieve this aim.108 For behind the eye there is always
a he or she: vision and looking is gendered and subject to both social and biological
influence. There is, admittedly, a male gaze, but we must be careful with this concept in
criticism lest it

become what Edward Snow apdy terms "a fixed and almost entirely negative term." The
critical analysis of looking and of visual representation has become a difficult business: as
Michael Ann Holly reminds us, "representational practices encoded in works of art continue
to be encoded in their commentaries." Traditional (humanist) scholars may find it hard to
accept the fact that both vision and 'reality' do not exist per se — we always produce vision,
and realism (that of a photograph or a painting) is just one function of this process. While we
have much to learn from feminist (film) criticism about the tyranny of the (male) gaze, we
must not underestimate the equally powerful attraction of art objects, especially if they
represent women.109 What we need is less disciplining by traditional fields and more
interdisciplinary analysis. I can only hope that this need will provide the subject of another
conference - and, perhaps, another book.