Pictorial Models and Narrative Ekphrasis Author(s): Tamar Yacobi

Pictorial Models and Narrative Ekphrasis
Author(s): Tamar Yacobi
Source: Poetics Today, Vol. 16, No. 4, (Winter, 1995), pp. 599-649
Published by: Duke University Press
Stable URL: http://www.jstor.org/stable/1773367
Accessed: 15/04/2008 10:34
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=duke.
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We enable the
scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that
promotes the discovery and use of these resources. For more information about JSTOR, please contact [email protected].
http://www.jstor.org
Pictorial
ModelsandNarrative
Ekphrasis
TamarYacobi
Poeticsand Comparative
TelAviv
Literature,
Abstract Ekphrasisis an umbrella term that subsumes various forms of rendering the visual object in words. This variety is often arbitrarily restricted, largely
because some of the forms, though manifest in artistic practice, are not recognized in criticism. This essay analyzes two such neglected forms: pictorial models
and narrative ekphrasis. Their intersection compounds the ekphrasis of a visual
model (as distinct from a unique artwork) with narrativized (as against descriptive, picturelike) effect, though not only within narrative works. The neglect of
both forms, I point out, relates to theory's doctrinal biases, namely, the insistence on interart reproduction ("mimesis")and so on either-or choices (between
epic and lyric, action and description, narrativity and pictoriality). Instead, I
argue for the centrality and the specifically narrative roles of the pictorial model.
To enhance the evocability and perceptibility of the visual source, literary texts
often allude to a visual common denominator (e.g., the thematic makeup of
"a Madonna with child" or the familiar components of "a Turner seascape").
Furthermore, such ekphrastic models often join forces with narrativity to bring
the visual source into distinctively literary play, not least along the time axis.
Thus, when a visual clich6 is transmitted through the subjectivity of an inside
observer, it enters into narrative patterns such as plot and characterization, as
well as point of view. These and various other interplays between the ekphrastic model and narrativity are illustrated in the second part of the essay, mainly
through the poetics of Isak Dinesen.
Earlierversionsof this paperwere presented at the Fourth IASS Congress(BarcelonaPerpignan, 1989), at the Second International Conference on Word and Image
(Zurich, 1990), and at Indiana University (Bloomington, 1991). The author thanks
the audiences for their comments, and several readers, particularlyMieke Bal and
ClausCliiver.
PoeticsToday16:4 (Winter1995). Copyright? 1996 by The Porter Institutefor Poetics
and Semiotics. CCC0333-5372/95/$2.50.
600
PoeticsToday16:4
1. Source-Target
RelationsbetweenArts:Varietiesof EkphrasisRenumbered
The traffic between visual art and literature has alwaysfeatured the allusive (mimetic, thematic, quotation-like) relations between works in the
different media. The one work's representation of the world then becomes the other's re-presentation, a mimesis in the second degree. Thus
the reworking of biblical and mythological tales, details, moments, and
themes (e.g., the Crucifixion or the birth of Venus) in the spatial arts.
Conversely, we have ekphrasis, where the temporal art of literature alludes to paintings, statues, urns, or, again, their traditional themeswith the difference that texts like William Carlos Williams's "Kermes"or
Auden's "Musee des Beaux Arts" or Dan Pagis's "Portrait,"wholly devoted to a verbal re-creation of visual elements, are comparatively late,
uncommon, and short. Oriented to time, literary discourse will rather
localize the intermedium allusion as an inset motif at some juncture(s).
The earliest and best-known case in point is the rendering of Achilles'
shield in midepic, with numberless counterparts across literature since.
For example, Browning opens his famous dramatic monologue with the
Duke's reference to "my last Duchess painted on the wall, / Looking
as if she were alive" and closes it with the monologuist's pointing to a
statue of Neptune, "taming a sea horse, thought a rarity, / Which Claus
of Innsbruck cast in bronze for me!" Likewise, in the course of Henry
James's Daisy Millerthe portrait of Pope Innocentius X by Velazquez is
observed in a Roman gallery; or we find a reproduction of a Kreutzer
Sonata by Prinet hanging in the Haze house in Nabokov's Lolita. In this
respect, as with the demand for a degree of familiarity with the inset pictorial work, the intermedium allusion in literature operates much like
the intramedium variety.'
Even such a cursory glance reveals the extent of divergence in the
field, for instance, between the allusion to an artwork and to an art
theme, between shorter and longer re-presentational stretches or texts,
between poetry and prose, between description and narration. This goes
to show that ekphrasis,the literary evocation of spatial art, is an umbrella
term that subsumes various forms of rendering the visual object into
words--so various, indeed, that both the reference and the sense of the
term in critical discourse leave a good deal to be desired. The range of
phenomena covered by ekphrasis(and with it the meaning of the con1. Both accordingly fall within a comprehensive theory of "quotation"as secondorder mimesis or re-presentation, exhibiting the set of universals generalized in
Sternberg 1982. The argument there has since gained currency among theorists of
reported discourse in all its linguistic manifestations.But note the reason given there
for the extendibility of the principle to media other than language and to arts other
than the literary,such as "musicaland pictorial"allusiveness. "Twice-removedrepresentation"forms "a qualitative common denominator"and "points the way to an
inter-arttheory of quotation-as-mimesis"(ibid.: 135 n. 14).
Yacobi * PictorialModelsand NarrativeEkphrasis
601
cept and the use of the device) is ill defined or arbitrarily restricted,
largely because some of the forms, though manifested in artistic practice, still need to be recognized in principle and theorized along with
their better-known counterparts.
My concern here is with two such neglected forms, which cofigure in
my title, "Pictorial Models and Narrative Ekphrasis."When these intersect, which they often do across the ages and genres of literary practice, the result is a compounded challenge to theory. We then have the
ekphrasis of a visual model (as distinct from a unique artwork) to narrativized effect (as against descriptive, picturelike, or thematic bearing),
though not only within narrative works proper. The object and the function of re-presentation thus vary together, inverting the common theoretical norm twice over. Modeling and narrativity, it will soon emerge,
are actually two related but independent parameters of ekphrasis, both
central to my argument as well as to the illustrative corpus, especially
Isak Dinesen's artful story world. I will therefore introduce them in turn,
one in this section, the other in the next, exploring their relationship as
we go along.
Let us start with the difference in the re-presented object, between the
artwork and the art model as source. By this shift in level, the allusive
discourse widens and so generalizes its reference from a particular to a
habitual, traditional configuration of elements in the other art or, if you
will, from surface to depth. Take these two examples, which will reappear
later: when Nabokov describes a scene in a Parisian restaurant in terms
of The Last Supper or when Isak Dinesen views two embracing sisters
as "maidenly Laoco6ns," they are not alluding to any specific picture or
statue but to a pictorial model, a common denominator, a generalized
visual image. In what circumstances does such interart dialogue occur,
under what rules, and to what effect, compared with the particularized
variety?
Strangely, this line of inquiry runs not just beyond but against the
common conception of ekphrasis. Leo Spitzer's (1962 [1955]: 72) wellknown definition locates the ekphrastic genre in "the poetic description of a pictorial or sculptured work of art"; it "addresses a particular
visual image," Hollander (1988b: 34) emphasizes to make doubly sure.
And most analysts focus on the one-to-one relation (numerical as well
as mimetic) postulated between the artworks. They will indeed readily
enough admit, at times even examine, the possibility that "a pictorial or
sculptured work" may generate more than a single "poetic description."
Kibedi Varga (1989: 44) thus spells out the rule that underlies many
analyses of otherwise different orientations: "What comes first is necessarily unique; what comes after can be multiplied. One image can be the
source of many texts, and one text can inspire many painters. These secondary series can become the objects of comparative study, which makes
602
PoeticsToday16:4
Table1 Ekphrastic
Relations
VisualSource(representation)
1
2
3
4
one
one
many
many
VerbalTarget(re-presentation)
one
many
one
many
us aware of the fact that illustrations and ekphrasis-in fact, all manifestations of subsequent, secondary relations--are just different modes
Unlimited in number, then, each act of ekphrasis is a
of interpretation."
text's late, "secondary," interpretation of some artwork, "necessarily" a
unique original. But this one-to-many relationship between source and
target, as we may call it, is usually considered the limit of ekphrasisexcluding the reverse, many-to-one possibility, if only by silent yet firm
omission.
Even a promising term like Jean Seznec's (1972: 570) "artistic cliche"
refers in fact to "specific pictures" that "had a sentimental appeal to a
whole generation." The source artwork is again unique, and it remains so
even when worn into a cliche by way of repeated verbal evocation. Likewise, Kranz's (1981: 377ff.) category of "kumulatives Bildgedicht" makes
a welcome change from the norm of the unique art source versus its
ekphrastic reproduction(s). Rather than an ekphrasis of artistic models,
however, his "cumulative" text is one that covers a group (or even the
totality) of works produced by some artist, often as an act of homage.
(See also Cliiver 1989: 57-58, esp. 57 n. 2.)
In this regard, then, the holes left by ekphrasis criticism are conspicuous against the scope of theoretical possibilities, as outlined in Table 1.
All four possibilities are variations, amply realized in literature, of the
image-to-word transfer. Yet the first two categories, sharing a unique
source to be rendered into words, have monopolized the study of ekphrasis. This holds especially for the one-to-one relation, of course. But the
one-to-many variety has also gained some attention in, for example,
Marin 1970, on three literary treatments of a Poussin landscape painting,
or Kranz 1973: 77-87 and 1981: 447, on the dozens of poems inspired
by Breughel's Icarus.This is a suggestive departure from the mainstream,
if only because the presence of multiplicity on one side of the interart (re-presentational, specifically ekphrastic) fence in effect argues for
the inverse perspective as well. When the multiplicity changes sides, it
would rather come to the fore that Auden's (1976: 146-47) "Mus6e"cites
"Breughel'sIcarus,for instance," as a token illustrative of a pictorial type,
"the Old Masters";or that, to bring out "the living quality of / the man's
mind," Williams (1988: 388-89, 505) conflates in "Haymaking" (origi-
Yacobi * PictorialModelsand NarrativeEkphrasis
603
nally titled, for reflexivity and guidance, "Composite") two Breughel
paintings.
The symmetry between the re-presentational pluralities should be evident, as is the status of the latter poems as modern classics of the genre.
For balance, therefore, we need to bring into the picture the last two categories of interart traffic in ekphrasis. Both are defined by the multiple
visual source chosen for verbal (re)modeling. Only, the many-to-one relation consists in a single instance of such modeling; the many-to-many,
in a number of traditional or repeat performances, as when a writer, a
school, or an age revisits a certain image (e.g., a landscape topos) common to various paintings. Of the two, again, I want to concentrate on
the former relation, whose status is theoretically crucial and whose workings may easily be extended to the latter, along comparative or historical
lines. (Speaking of extension, incidentally, the entire table lends itself to
reversal from an ekphrastic network into a map of pictorial imaging of
literary sources.)
2. FromMimeticSpace-imageto Ekphrastic
Dynamism
2.1 WhereHastheModelGone?LessingbetweenAncient
andPresent-Day
Mimeticisms
With the reference of model-oriented ekphrasis delimited, it now remains to establish its sense, that is, its rationale, features, operations,
resources, effects, in short its poetics. We may best approach the question by proceeding to ask whyit has been so neglected in an otherwise
expanding and increasingly sophisticated field of inquiry. Such neglect
certainly does not correspond to the literary facts, the practices of writing (already glimpsed) and of reading. Readers, if anything, are often
more familiar with art models than with the details of specific artworks;
and what we carry around in our heads has both liberating and constraining implications for writers across the entire range of ekphrasis.
(If you take only the generic imperative of making the visual allusion
perceptible, hence readable, consider how much easier it is for a generalized than for a unique reference to meet this need without recourse
to special measures, aids to memory, and the like.) At the transmitting
end, moreover, references to a "van Gogh landscape," as well as to a
"Mona Lisa smile," have found their way into ordinary discourse. What
is it, then, that has so driven analysts toward work-to-work, rather than
work-to-model, relations between word and image? 2
2. The question gains yet sharper point when we compare literary with art criticism,
so attentive to borrowed and inherited thematics. As Panofsky (1982: 40-41) puts it,
iconography is a necessary tool for the analysis of an artwork's "subject matter," while
iconology serves to interpret its "intrinsic meaning or content." On the correlation of
art history and literary study with units of different scope, collective and individual,
respectively, see also Alpers and Alpers 1972 and Steiner 1982: 66-68.
604
PoeticsToday16:4
One answer, which in turn leads to others, concerns the widespread
pressures and preference for strict mimesis in ekphrasis. With the representational thrust so heavily value-laden, the privileging of the strict
and the singular interart transfer appears to form a logical chain: If oneto-one equivalence in re-presentation, then one-to-one equivalence in
number between the re-presented and the re-presenting discourse. This
formula, though never spelled out to my knowledge, has deep roots
in the contemporary approach to the subject and a longer history still
in aesthetics, a rough outline of which is worth tracing. Here Lessing
plays a key role. For he has often been miscast and derogated as a mimeticist by, among others, present-day inquiries into the sister arts (e.g.,
Steiner 1982: 12-14, following Abrams 1953), which themselves betray a
far stronger (inter)mimetic impulse, regarding ekphrasis at least.
A closer inspection makes sense of this incongruity by disclosing the
real point at issue, which could hardly go deeper. In terms of the formula
mentioned above, it bears on the premise ("If one-to-one equivalence in
re-presentation") from which the anti-model directive ("then one-to-one
equivalence in number") virtually follows, and much else besides. The
disagreement, in short, concerns the balance between representation
and communication, mimesis and aesthetics or poetics. Under mimetic
pressure, such latter-day criticism would reverse -not always in an obvious fashion, nor always for the better-Lessing's notoriously restrictive
theory of art. The curious thing is that the notoriety has at times outrun, and the modern opposition hardened as well as relaxed, his actual
(inter)artistic rules.3
Not that Lessing is permissive about representation, any more than he
is a friend to interart re-presentational affairs, including ekphrasis itself.
So far from licensing mimesis, either that of nature by art or that of one
art by another's, the Laocoonsubjects it to regulation by appeal to a set
of higher norms, absolute and comparative. The limits imposed on "the
imitative arts"begin with the object of imitation itself:
Although painting, as the art which reproducesobjects upon flat surfaces,
is now practisedin the broadestsense of that definition,yet the wise Greek
set much narrowerbounds to it. He confined it strictlyto the imitation of
beauty.... The perfectionof the subjectmustcharmin his work.He wastoo
great to requirethe beholders to be satisfiedwith the mere barrenpleasure
arisingfrom a successfullikenessor from considerationof the artist'sskill.
(Lessing1963:8)
3. Compare the reading of the Laocodnin Sternberg 1990: esp. 67ff., where Lessing occupies a position midwaybetween Aristotle's plot-oriented poetics and modernism'smedium- and space-oriented counterpoetics. As will be seen, this helps to
explain the respective attitudes towardekphrasis, notably regarding the question of
narrativity.For a provocativeattempt to refer the dispute to conflicting political and
sexual ideologies, see Mitchell 1986: 95-115 and 1989, also discussed below.
Yacobi * PictorialModelsand NarrativeEkphrasis
605
This statement involves a complex of principled oppositions and hierarchies of value that go against the premium put on mimeticism by
Lessing's successors to this day, as well as by his predecessors. The scope
of mimesis, theoretically extendible to the entire world "in the broadest sense," must narrow in response to the demands of "beauty," the
"supreme law of the imitative arts"and a visual measure in the plastic arts
(ibid.: 11). Imitation stands to beauty as a means to an end, indeed the
end: where aesthetic and factual or realistic value clash, art will opt for
"beauty"and its proper "pleasure," as science will for "truth"(ibid.: 10).
This excludes from pictorial imitation objects such as deformities, monstrosities, facial contortions, rage, and despair, including the ugliness of
the open mouth incurred by Laoco6n's cry (ibid.: 8-14, 153-67). Yet
Virgil "was right in introducing the cry, as the sculptor was in omitting it," because poetry appeals to the ear rather than to the eye (ibid.:
20-21). Medium-sensitive, the rule thus works both ways: precisely that
which places visual beauty out of reach, and so out of bounds for literature, except through oblique evocation, qualifies literature for rendering
the auditory (or the spiritual) world, and vice versa. Note also that the
prescriptive-sounding argument is not necessarily or entirely restrictive,
at least not in the framework of interart transfer. Among other implications, this in effect sets free the re-presentational device that Lessing
officially leaves unfocused and even unnamed: ekphrasis may develop
(rather than at best parallel, at worst attenuate) the original image, by
articulating the cry, for example.
For Lessing, moreover, "beauty" outranks not only "truth" but also
"skill"in mimesis, that is, the drive to conquer or redeem the aesthetic
liabilities of the represented object through the mode of representation.
He has little regard for "manual dexterity, ennobled by no worth in the
subject" (ibid.: 9), for the rise to the challenge known (and elsewhere
admired) as "difficult6 vaincue" (ibid.: 26), particularly the difficulty of
making an exact copy of reality itself or of a prior artwork. This ranking accordingly sets new limits to either of the representational arts and
to the re-presentational traffic between them either way, with historicalevaluative implications to match. Its infringement at the hands of Lessing's contemporaries for the sake of "mere barren pleasure" (ibid.: 8)
largely explains his invidious comparisons of "moderns"with "ancients"
throughout.
By appeal to this value scheme, Lessing carries the theory of artistic
limits even further: from the domain of signifieds per se to their interplay with the signifiers, and from the selection to the combination of
signs proper to the respective arts. Even where roughly the same object
of mimesis is to be presented (or, for that matter, re-presented) with
much the same effect in view, each art form will best operate in terms
of its distinctive signifying conditions. Why, for example, does Laocoon
606
PoeticsToday16:4
wear priestly robes in Virgil's epic and nothing at all in the sculpture?
Because (Lessing argues against the common view) a poetic "robe is no
robe. It conceals nothing. Our imagination sees through it in every part";
and the reminder of priestly status even invests the victim's agony with
a sense of desecration. By contrast, given the artist's visual signifiers, a
robe would divest his rendering of the beauty and expressiveness of the
sufferer's body (ibid.: 39-40). Yet the interart relation between plus and
minus, more representational coverage and less, may also turn round
in other signifying contexts or crosscuts. Lessing thus invokes the difference between "arbitrary"(convention-based) and "natural" (iconic)
signs to explain why the poets fail to describe the muses in the painterly
manner. They can dispense with such portrayal because the name Urania
is enough to perform the necessary reference to the muse of astronomy;
while "in art she can be recognized only by the wand with which she
points to a globe of the heavens" in "dumb language" (ibid.: 67-68).
Accordingly, though neither interartistic comparison made by Lessing deals with ekphrasis proper, both have significant implications for
it, operationally polar yet logically consistent and complementary. In representing the two visual images in question, an ekphrastic writer will
add Laocodn's priestly dress, as well as the priest's scream, and omit
Urania's trappings. Either way, mimetics in and between the arts is a
function of comparative word/image semiotics under aesthetic control.
As with the choice of signs, so with their combination into an artwork. Here the Laocoonmakes its famous plea for harmony between the
arrangements of signifier and signified, medium and object of representation in either art. "Signs arranged side by side can represent only
objects existing side by side," that is, "bodies. Consequently bodies with
their visible properties are the peculiar subjects of painting" as spaceart. By the same token, "consecutive signs can express only objects which
succeed each other ... in time," that is, "actions. Consequently actions
are the peculiar subject of poetry" as time-art (ibid.: 91). Again, this
law of harmony derives not so much from the constraints exerted by
either medium on (re)imaging the world as from the distinctive ends and
options of artistic (re)imaging.
For example, if literature privileges "actions"over "bodies," sequence
over coexistence, narrative over description, this is because art will
always regulate mimesis by the higher value and pleasure of "illusion."
The ability to describe "things as they exist in space," Lessing claims, is
"a property of the signs of language in general, not of those peculiar to
poetry. The [nonliterary] prose writer is satisfied with being intelligible,
and making his representation plain and clear. But this is not enough for
the poet. He desires to present us with images so vivid, that we fancy we
have the things themselves before us, and cease for the moment to be
conscious of his words" (ibid.: 101-2). Thus opposed to mere prosaic in-
Yacobi * PictorialModelsand NarrativeEkphrasis
607
telligibility, Lessing's ideal of poetic transparency so favors narrative as to
rule out of literature straight description (or, for that matter, ekphrastic
re-description). For such writing must draw notice to itself at the expense
of "the things" written about; the part-whole relations that the beholder
takes in at a glance demand laborious processing from the reader. So the
disharmony in arrangement between temporal medium and spatial object would destroy "illusion,"4as Laocoon's cry would pictorial "beauty";
both values count as far superior to representation per se, descriptive or
otherwise, easy or difficult, within this aesthetics.
Given such premises, it is no wonder that Lessing frequently uses the
term Vorbild,"model" in English translation, for the source (of whatever
domain, kind, or number) re-presented by either art. Thus, Virgil having
shaped anew the Laocoon group, ancient sculptors may have taken his
version "as their model" (ibid.: 34-36): numerous graphic returns to a
single literary Vorbild.Conversely, the poet may have worked "after the
model set him by the artists" (ibid.: 42), with the possible (though not
necessary) result of creating some form of ekphrasis (though the treatise will not call it so, let alone name the exact subgeneric formation).
As these examples indicate, Lessing's "model" differs greatly from mine
in that it covers all source-target permutations and is not reserved for
the many-to-one variety.Yet it does point away from strict mimesis, typically and unmistakably so, if we consider the stress laid throughout on
the features changed (added, omitted, played up or down, transformed)
in re-presentation. This is why such mimesis, balancing aesthetic license
and limit, implies no disparagement to the imitators. "On the contrary
the manner of their imitations reflects the greatest credit on their wisdom. ... A model was set them, but the task of transferring it from
one art to another gave them abundant opportunity for independent
thought," for equal novelty and glory by way of deviation (ibid.: 42).
It is not following the Vorbildas subject matter (in effect, the ekphrastic writer's only imperative and hallmark) but copying the manner of
representation that degrades interart re-presentation.
In its overreaction to Lessing, or to some image of him, the modern
study of ekphrasis more often than not reverts to the practices upheld by
his antagonists and the norms behind them. Such reversion particularly
manifests itself in the reversal of the Laocoon'sattitude toward interart
transfer. Rather than a means to higher aesthetic ends and a function of
comparative semiotics, the mimesis of the visual by the verbal then be4. Even without any self-focusing or what we call "reflexive" measures (e.g., fragmentation and apostrophe) adopted by anti-illusionist writing, of the kind Hollander
(1988a) approvingly associates with ekphrastic poetry. To Lessing, they would only
worsen the evil of disharmony. Either way, however, mimesis varies with the poetic
goal, and the extras build or trade on the medium's constant givens vis-a-vis the
object.
608
PoeticsToday16:4
comes again an end in itself: the closer the re-presentation and the more
difficult it is to achieve, the better-across
all differences in signifying
conditions between source and target.
Observe, for example, the frequent occurrence of "reproduction" and
the like as a definitional feature of ekphrasis, or the term's replacement
by "icon(ic)" in Hagstrum's (1958: 18 n. 34 and passim) influential work
The Sister Arts;5 or the insistence on its forming "an exact description
meant, to a certain degree, to evoke and substitute for the painting itself"
(Kibedi Varga 1989: 44). No less suggestive is the need often felt for direct contact between graphic source and literary target in the process
of writing, or for their reconfrontation in the reading. Hagstrum (1958:
42-43) thus finds it necessary to distinguish the poet who "reproduced
what he had seen and responded to firsthand" from the conventionalist
who had not. Others have taken great trouble to identify the original(s)
of, say, Keats's Grecian urn, though Keats himself significantly provided
no clue and no incentive to this quest.6
It is not that either the (inter)aesthetic or the (inter)semiotic disparities that stand in the way of strict mimesis in re-presentation have
escaped the notice of moderns. On the contrary, the contemporary critical insight into such disparities has greatly developed since the age of
Lessing, when these fields were still in their infancy, as has also the art
of interpretation, textual and comparative. To mention only two recent
examples, the array of aesthetic, semiotic, and reading skills brought to
bear by Steiner (1982: 42-47) on E. E. Cummings's "stone children" or
by Cliver (1989: 62-70) on "Starry Night" as translated from van Gogh's
code into Anne Sexton's would be inconceivable in premodern dealings
with ekphrasis. But the more refined the equipment, the more revealing
the all-too-familiar drive toward mimesis against the grain of the target
art, along with the underlying (counter)values. And the very advance in
interpretive skills, best exercised on one artwork at a time, the richer the
better, urges us to compare individual works in or beyond their mimetic
aspect.
Indeed,
considering
the proportion
of reading
to theorizing
in
5. He reserves the former term for a poem in which the artworkbreaksinto speech,
in the traditionof prosopopeia. More recently,Lund (1992: 12ff.) divides "ekphrasis"
(reference to a picture) from "iconicprojection"(picturelikereference to the world),
which in fact arguablyintersects with ekphrastic modeling, because its picturelikeness may attach to a general feature. The one-to-one bond thus defines the field of
ekphrasisagain, persisting acrossvariationsin critical terminology and (given Lund's
resistance to normative bias) in value scheme. (I owe this reference to ClausCliiver.)
6. If with the urn he leaves the original unidentified, elsewhere he generalizesit into
a model of the highest abstraction.Thus in a letter to Fanny:"Ishould like the window
to open onto the lake of Geneva-and there I'd sit and read all day like the picture of
somebody reading"(Keats1953 [1819], 2: 46). (I owe this reference toJoseph Grigely
[pers. com., Conference on Word and Image, 1990].)
Yacobi * PictorialModelsand NarrativeEkphrasis
609
ekphrasis criticism, especially in America, one wonders if the zeal for
interpretation would not be enough to keep the model out of sight,
never mind its offenses against re-presentation. Compared with either
the originals from which it abstracts or their ambitious literary reworkings, the model is liable to appear too flat to qualify for exegesis, as well
as for reality-likeness. Naturally, how these counterforces balance varies
in the field.
Thus Cliiver (ibid.) begins by listing a variety of ekphrastic crosscuts,
and elsewhere he mentions the interest aroused in poets by "an artist's
visual language" as it develops from "workto work"(Cliver 1978: 33). Yet
what he singles out for analysis is "intersemiotic transposition" as "translation," complete with references to its "difficulty" (e.g., Cliiver 1989:
59, 62). For Steiner (1982), whose work is more normative and less wideranging, the difficulty of one-to-one transfer actually dooms ekphrasis
to mimetic failure. But then, the failure itself-rather than commending or at least indicating more feasible alternatives, such as the image as
model - is exclusively promoted to the status of a thematic, even generic,
feature (ibid.: 42, 48). Indeed, where the "definitional failure" is left unthematized in poetic transfer-from, say, Breughel's Returnof theHunters
to Williams's "Hunters in the Snow"-we have instead a case of "structural correspondence," apparently outside the genre of ekphrasis (ibid.:
73-90; cf. Hagstrum's "icon"above and n. 5). Playing with labels to drive
away undesirable elements-whatever their claim to re-presentational
equality-makes a typical and transparent exclusionary ploy.
All this goes to explain why, and to what extent, the reaction to Lessing
falls short of liberation. In the process, it has no doubt helped to redeem
or rehabilitate other modes of transfer than his, other scales of value and
interest operative in (often, indeed, well beyond) the ekphrastic genre
itself. So far so good, except that in reversing Lessing's dogmas, such
opposition in effect locks ekphrasis into another prison, one both older
and newer, yet with symmetrical limits and losses as well as privileges
and gains.
Hence the counterlimits widely imposed on mimesis in re-presentation, as though ekphrasis at its best could not aim to evoke either more
or less than the original image. This is, for example, why we need to bear
in mind Lessing's insight regarding the extra features that are assumed
by the epic Laocoon (the cry, the dress) but denied to his sculptural
analogue. As with the more particularized reworking, so with the less:
on the highest, strategic level, the same counterlimit explains why the
skeletal, abstractive, "modeled" image has suffered neglect as ekphrasis
in favor of the full-bodied image, regardless of availability or frequency
or artistry or any other empirical evidence (also regardless of plain oversight, or the passion for reading). Quite simply, an ideal of one-to-one
correspondence between source and target entails (and privileges) a
610
PoeticsToday16:4
one-to-one numerical relation, an encounter of two individual artworks.
The same holds true if the ideal forced on the device grows a competitive edge, so that the encounter becomes antagonistic: ekphrasis as an
arena of interart rivalry (foredoomed according to some, more openended according to others) over the prize of mimetic fullness. In any
case, the art model would never qualify, because its very form entails
withdrawal (to whatever end) from such fullness and fight; nor would
the less clear-cut arts of abstracting even from a unique visual source,
which complete the repertoire of ekphrastic underre-presentation. By
comparison, overre-presentation might somehow pass muster: though
an equal offense against interart mimesis, and so left untheorized, it at
least makes a bid for outdoing the original's image of the world itself.7
As a result, far from having developed anything like a poetics of the
model in literary transfer, we may fail to identify such a model when
it comes our way. In the absence of a determinate pictorial source, the
ekphrastic image may even be considered fictive rather than deliberately
selective. This befalls Proust's comparison, "One felt the same pleasure
as when one sees in a landscape by Turner or Elstir a traveller in a stagecoach, or a guide, at different degrees of altitude on the slopes of a
mountain pass," with its unmistakable ors, doubled for extra pictorial
coverage and accessibility. "Imaginary ekphrasis," Riffaterre (1990: 51)
nevertheless calls it; a "pseudo-landscape." Here fictionality supposedly
delivers the re-presentation from a threat of lost individuality: better a
unique invented image, as it were, than a composite model grounded on
named historical art sources and alone capable of explaining the text's
features. To escape from the plurality of the object, the critic will if necessary shift the ontology and the meaning, as well as the genre, of the
finished literary product.
Genresor Powers?Hegemonyor Range?
versustheNarrative:
2.2 TheDescriptive
The uneven coverage of the map outlined in the foregoing section is
thus no random matter, nor a passing vogue, nor even the outcome of
a methodological (e.g., interpretive) turn. It is a reflex of critical preferences wide and deep, often hidden but alwaysindefensibly exclusive. The
history of aesthetics, with its alliances across the Laocoondivide, brings
out a principled reason for the unequal attention given to varieties of
transfer that are equally feasible and serviceable in theory, hence equally
manifest in literature, which puts the different forms to different uses.
7. In a relativelydogma-free,corpus-basedaccount, Kenneth Grossargues that texts
featuring ekphrasis aim for more than "mimeticaccuracy"and "willoften reach beyond what is given or visible in the works they 'describe."'But he hastens to qualify
this Lessingesque statement: "Equallyworth attending to, however, are these texts'
often oblique ways of returning to the sculpted object," to "what is given" (Gross
1992: 141-42, 145).
Yacobi * PictorialModelsand NarrativeEkphrasis
611
The art model has simply fallen between two doctrinal stools: that of
Lessing, who would generally minimize interart contact, to the wholesale
(if nominal) neglect of ekphrasis, and that of the longer and stronger
tradition that would mimeticize it into one-to-oneness.
Nor does the model alone show the consequences of doctrinal bias.
For much the same reason, so do the workings of all ekphrasis in literature, as between the extremes of epic and lyric thrust, action and
description, time and space: narrativityversus pictoriality, in brief. The
sharper the focus on interart disparities, Laoco6n-style,the heavier the
pressure on ekphrasis (often along with the rest of literature) to follow the narrative route. Conversely, the more value placed on interart
mimesis-an immemorial routine, whatever its guise-the stronger the
channeling of ekphrasis toward the opposite, static, imagelike pole. By
now it should not be very surprising to find the camps actually divided
along these lines, as if there were again an either-or choice, rather than
a range of ever-available, complementary options.
Lessing (1963 [1766]: xi) thus begins by announcing that under the
name of "poetry" he has allowed himself "sometimes to embrace those
arts, whose imitation is progressive," that is, narrative art at large. In
practice, he goes even further, drawing most of his literary (and all of
his paradigmatic) examples from narrative, especially Homeric and Virgilian epic. This coheres perfectly with his drive to delimit the arts. Of all
genres of "poetry,"narrative is the farthest removed from painting, since
its temporality distinctively covers the object as well as the medium of
representation. And such double coverage involves maximum harmony
and "illusion."
We need not endorse this favoring of narrative, or its grounds, to turn
it to ekphrastic account. Nor, having left the dogma behind, need we
quibble over the question whether Lessing, at narrativizing, addresses
ekphrasis itself. Here as elsewhere, he never refers to it by name, but
he does so often and unconventionally in practice, on any half-generous
reading, and has been given less-than-due credit by interested parties
since.8 (When it comes to a hard-line mimeticist like Hagstrum [1958],
nobody would deny him relevance just because he prefers "iconic" to
"ekphrastic" nomenclature.) The trouble is only that the Laocoonblurs
the line between art's first- and second-order mimesis in its eagerness
to put them both under artistic control, for example, in prescribing the
extension of the literary object (whatever its origin, or Vorbild)in time to
suit the medium.9 Obliquely, but for our purposes importantly, Lessing's
8. The converse also holds true, in other modern arenas, where the Laocoonfigures
all too prominently,as argued by Sternberg(1981, 1990).
9. Where ekphrasis itself forms the object of inquiry, such a blur may indeed lead
to category mistake. For example, to classify under this label the view presented in
612
PoeticsToday16:4
emphasis may nonetheless be extended or converted at several points
from representation to re-presentation in literature-above all, to the
dynamic transformation of pictorial space.
Take the narrativizing of description proper, according to Homer's
method, whereby the elements of the "body"are projected into the flow
of story time, rather than enumerated in spatial coexistence. Thus the
divine chariot comes together piece by piece under Hebe's hands, Agamemnon's dress in the act of dressing, Achilles' shield in the process
of its creation (Lessing 1963 [1766]: 34, 95, 113-25). The last of these
examples, considered by Lessing himself the paradigm of description
turned narration, has in fact been taken by many as the origin of ekphrasis: it is largely a question of whether or not the shield itself (relative
to, say, the chariot) counts as an artwork. However, its generic status or
label vis-a-vis the rest is less principled than their common generic suggestiveness; the dynamism that transforms and integrates the Homeric
objects of description into full literary temporality is always available to
the second-order, ekphrastic transfer of visual space-items.
So, from the pictorial rather than the poetic side of interart transfer, is the "pregnant moment." Originally recommended to space-artists
as an indirect force for temporality, this moment comes just before the
climax in order to "allow free play to the imagination," that is, to the
beholder's story-making imagination. "The more we see the more we
must be able to imagine; and the more we imagine, the more we must
think we see." Hence the need for the artist to choose that moment with
care; and, we might add, the ekphrastic poet's license to follow suit in
his re-presentation. Thus the paradigm case: "When, for instance, Laocoon sighs, imagination can hear him cry; but if he cry, imagination
can neither mount a step higher, nor fall a step lower, without seeing
him in a more endurable, and therefore less interesting, condition. We
hear him merely groaning, or we see him already dead" (ibid.: 16-17).
This interart comparison between literary cry and sculptural sigh, again,
readily widens in principle into a two-way interart movement between
source and target-including the narrativization of ekphrasis via "pregnancy." Nor need such pregnancy remain latent here any longer. Why
should it, considering that the poet enjoys "the liberty to extend his description over that which preceded and that which followed the single
moment represented in the work of art; and the power of showing not
only what the artist shows, but also that which the artist must leave to the
Anna Karenina of Anna's portrait and of Kitty "travelling to her estate, framed by the
window of her carriage" (Mandelker 1991: 16) is to conflate two distinct orders of
mimesis: the re-presentational, alone ekphrastic, and the representational. The latter,
if picturelike, as implied by the window framing, belongs to what Lund (1992) would
call "iconic projection." Any continuities between the orders, or instances of them,
therefore operate on the topmost, shared level: that of (visual-directed) mimesis.
Yacobi * PictorialModelsand NarrativeEkphrasis
613
imagination" (ibid.: 99)? So the ekphrastic writer, as time-artist, is free,
in effect urged, to spell out the narrative implications of the image into
actual narrative discourse with a determinate extension and sequence
and plot.'0
Of course, Lessing himself does not reserve either strategy (the
description-transforming or the pregnant-making and -actualizing) for
ekphrasis, as distinct from space reference in general; nor, we shall see,
does he by any means cover the repertoire of narrativizing options, least
of all those attached to point of view and to inset-frame interplay. But
the possibilities I have already outlined would at least accord with the
constructive, enlarging side of his aesthetics. It is this side that more or
less disappears, and indeed reverses into counterdogma, in most recent
approaches to ekphrasis. Here, poem and poetryusually figure in their
narrowest generic sense, bearing particularly on lyric verse with a strong
(re)descriptive thrust. This apparently serves to tighten the equivalence
10. Alpers (1960) treats the descriptions of paintings in Vasari's sixteenth-century
Lives as ekphrasis with a narrative emphasis, but in a very different, rather peculiar
sense of narrative: the rendering of emotion. She even points out "the absence of any
indication of the development of narrative" (ibid.: 201). On the other hand, she notes
that "Vasari does not differentiate between his characterizations of the same story
as told in different paintings by different artists" (ibid.: 201-3), or, in more positive
terms, he reduces them to what I call a model. Taken together, then, the two features
mean that his ekphrasis specializes in psychologizing (rather than plotting) visual
models; their emphasis does not shift from space to time en route to verbal discourse.
(In another context, Barolsky [1991, 1992] offers a fascinating analysis of Vasari's
Lives as a grand fictionalized history.) Recently Heffernan (1991) has moved into "narrative" as action, but along the lines already suggested by the Laocoon, though the
inevitable return to Achilles' shield (the one-moment picture turned into "a narrative
of successive actions" [ibid.: 301]) soon gives way to illustration from short romantic poetry. Where the analysis departs in essentials from Lessing, to whom it makes
a single passing and uncomplimentary reference (ibid.: 309), the outcome is not
exactly an advance. Thus, Lessing knows better than to associate literature, ekphrastic and otherwise, with the "narrative impulse" and "language by its very nature"
with the "release" of that supposed impulse. On the contrary, never mistaking his
poetic ideal for either the potential or the actual practice of literature, his treatise
expressly challenges the "mania" for description that had overtaken literary narrative itself, against Homer's example. And by the same token, Lessing (1963 [1766]:
chap. 17) explictly points out, language is evenhanded; it can naturally serve either
representational purpose -"description of bodies" as well as development of story
lines-only that literary language should nevertheless confine itself, by poetic fiat, to
rendering what best suits its time-medium. Nor would Lessing commit the error of
elevating a figure like prosopopeia or a theme like interart conflict-both manifestly
omissible, often omitted -to the status of parity with the forces of narration and/or
description, built into ekphrasis as a mode of representing the world. He does not
even waste his ammunition on such free variables. Having meanwhile become poetic
touchstones, though, they have gained a privileged status in ekphrasis criticism (at
least since Hagstrum 1958) at the expense of narrative, and Heffernan (1991) would
now coprivilege them with narrative. To these, as to related issues of principled and
practical variability, across the single constant of re-presentation, I will return.
614
PoeticsToday16:4
between word and image, which converge on a static object of representation with a view to maximizing "spatial form." But the additional
interart convergence entails a multiple reduction in the possibilities of
divergence accessible to and widely realized by verbal art. The loss, to
theory and to the sense made of literary practice, is twofold and often
cumulative, even beyond Lessing's terms of reference.
First of all, like the visual image re-presented, ekphrastic re-presentation then tends to cover the entire work-a short, lyric one at thatrather than form a part inserted by means of quotation into a larger
whole, "quoting" design. Given that such an inset might be no bigger
than a single line or phrase, it negatively resembles the model (even if it
isn't one, which it may be) in its distance from both strict mimesis and
the interpretive, self-contained ideal. Nor are the consequences of marginalizing it less serious. Where work-length ekphrasis has established
itself as the binding or exclusive rule, the orientation to it is liable (if
not anxious) to erase a major feature of arrangement peculiar to literary qua temporal art, namely, that the medium invests its discourse with
extension, direction, and hence also processing force. No matter how
frozen the re-presented object itself-the existent, posture, landscapeits linear re-presentation in ekphrasis enables shifts and turns of understanding denied to the all-spatial visual source, notably in the frame's
movement toward, through, and awayfrom the inset. Even an actionless
world then goes with an eventful discourse. But the norm of overall lyric
ekphrasis would rule out this interplay of (ekphrastic) part with (nonekphrastic or, as in "MyLast Duchess," otherwise ekphrastic) part along
the whole sequence, to the point of playing them off against one another:
undermining or counterpointing the "quoted" image in the sequel, for
instance.
What some contemporary approaches might admit, or value in opposition to others, is the play along the text-length sequence of ekphrasis
itself. Davidson (1983: 69-70) thus introduces his postmodernist corpus
as adverse to "any mimetic function"; in its light, he also finds Lessing
"an Augustan corrective to the excesses of verbal painting," notably to
the modernist rage for self-enclosed spatiality, or "spatialekphrasis." But
"the hermeneutics of existential temporality," which he enlists against
modern criticism, does not carry over to less extensive occurrences, to
more storied uses, to the genre as a whole, or to postmodernist verse
at large (unless, again, circularly defined). One doubts whether it even
applies to the whole ekphrastic repertoire of Ashbery, the parade example, much less to the finest contemporary space-time poet I know,
Dan Pagis (1981), who neither encloses space, pictorial or earthly, nor
meditates on time, but revitalizes the former to retell and deautomatize
the latter's stories (see Yacobi 1976, 1988, 1990a). At most, the "hermeneutic" counterthrust uncovered by Davidson actually operates within
Yacobi * PictorialModelsand NarrativeEkphrasis
615
the severe limits of the postmodernist all-ekphrastic text: to trace the
poet's "encounter with a temporality that had been lost or forgotten,"
say, or his destabilizing a pictorial image into a series of reflections on
textuality, convention, artistic form, historical knowledge. So this argues
more against interart mimeticism than in favor of sequential dynamism,
least of all as triggered by the passing reference to an image or a model by the inset allusion within a discourse that need not even focus on the
interart theme. It still remains to discover how and why ekphrasis ranges
at will from local to central and overall status, from feature to focus,
from specifying to typecasting, from the hermeneutics to the dramatics
of existence.
Second, reversing Lessing's generic bias makes matters worse. For
the common reduction of literature to poetry, and so of ekphrasis to
the short descriptive poem, leaves out narrative-certainly narrativity
across the genres-and results in detemporalizing or "spatializing"the
object of literary re-presentation as well as the medium. The German for
ekphrasis, Bildgedicht,encodes this double bias in the term itself-favoring a text-length piece of verse -elsewhere assumed or conceptualized.
Even Davidson's (1983: 73) postmodernist ekphrasis, in its appeal to time
"lost or forgotten," works to "defeat any sustained narrative" (cf. Lund
1992: 165-66 on a Swedish prose poem). His modernist adversaries go
much further, perhaps none more so than Krieger (1968,1992), followed
closely by Steiner (1982: 42-50).
Here, literature's alleged yearning for imagelike simultaneity, roundedness, permanence, eternal return-in short, deliverance from timefinds its quintessential expression in the "still movement" or "stopped
moment" captured by ekphrasis. Given this all-important object of
(re)description, the anti-Lessing critics tell us, everything else serves to
heighten its effect. Hence, for example, the frequent circularity of the
ekphrastic still object: Keats's urn (Spitzer 1962 [1955]: 73) or Thomas
Browne's (Krieger 1968: 327-28), or the wreath around Cummings's
stone children (Steiner 1982: 44). Each is designed to symbolize the
theme of enclosure, as opposed to temporal closure, and to enable parallels on other levels, such as the circling back of the language to the outset. To carry this ideal to its logical conclusion, the pregnant momentfar from sought, imported, centralized, or elaborated in the transfer to
literature -should be avoided or reduced to stillness, denarrativized.
Accordingly, one can also predict what becomes of narrative: Krieger
(1968: 343) judges "prose fiction" to be handicapped "in proclaiming
itself a rounded object," and Steiner (1982: 48) excludes it altogether for
its inevitable reference to temporal flux: "In the novel, this flow is explicit in the sequence of events depicted; in the lyric poem, its absence
is definitional. Instead, the lyric pretends to represent one now-point ...
a suspended moment." Even in terms of this generic difference, itself
616
PoeticsToday16:4
questionable, the marriage between ekphrasis and the lyric or the still
moment does not follow, nor, correspondingly, does its divorce from
narrative and narrativity.To revert to my earlier counterexample, the
poetry of Dan Pagis (otherwise, e.g., in sound play, nothing like the
prosaic or novelistic mode) falsifies all these generic requirements at
once: the difference, the marriage, and the divorce. In its re-presentation
of graphic artworks, as in first-order representation, this poetry dynamizes the object on every possible level. The world re-presented, the
re-presentational discourse event, the viewpoint, the reading-all spring
into life, and in the service of the mutability theme at that.
"Leafing through an Album" (Pagis 1981: 89-90), for instance, accelerates time so as to synchronize existential with perceptual development:
the running out of the hero's life with the beholder's running through
the photographs commemorating highlights of this life. Apart from the
nonepic framework itself and the novelty of telescoped duration, this
outreaches the Laocoonin at least two strategic respects. One consists
in the appeal to subjective, perspectivized, temporality: the observation
(or literary re-observation) unrolls, indeed mixes, with the (re)observed
birth-to-death action. Another claim to notice lies in the modeling of the
visual source, for the said snapshots (from the baby's onward) have been
so generalized in the transfer that they may apply to anyone's biography, Everyman'stale of mutability. Elsewhere, incredible as it may sound,
the act of visual portraiture itself (apparently anterior by nature to literary ekphrasis, because supposed to generate the "original" object for
re-presentation) gets thrown in for good narrative measure. In "The Portrait," even as the portraitist draws and discourses, the sitter transforms
before his eyes from child to old man to corpse:
The child
Is not sittingstill.
... I drawone line
And the wrinkleson his face multiply;
... his hair whitens.
... He is gone.
(Ibid.:85)
The poem is macabre, yet remarkable for its multifold dynamism in the
smallest compass. Nothing keeps "still."As I have analyzed elsewhere,
no fewer than three processes, each sufficient by itself to despatialize
the "sitting"object, converge on it here: "The speaking barely manages,
and the painting utterly fails, to keep pace with the living" (Yacobi 1976:
19-22; 1988: 95-98). Appropriately, the last-cited analysis of the poems
and their implications has been published in a special issue on narrative
theory.
A complete antipole to the still-moment lyric, whether qua replica or
qua rival of the art source, this dismisses its privileging even within the
Yacobi * PictorialModelsand NarrativeEkphrasis
617
subfield of ekphrasis that may appear to accommodate it best, if not encourage it. Indeed, as with the related attempt to freeze the medium,
artistic practice cuts across all such limits imposed on the object in the
spirit of modernism, allegedly, and often in the name of liberty, too.
Actually, those gratuitous restrictions would fix ekphrasis at the extreme
opposed to the pregnant moment and the like: they deny literature not
just its own peculiar freedoms but even the flexibility that Lessing himself
offers to the artworks that it re-presents.
We also find the two limitations otherwise compounded to the same
effect, as in Kibedi Varga's (1989) word-and-image taxonomy. He starts
by dividing the emblem, where "the beholder is struck by words and
images at the same time," from ekphrasis, where "the reader reads a poem,
without necessarily perceiving" the other part too (ibid.: 33; my emphasis). He then reserves "narration"for the serial as against the single work:
for comics and cartoons versus the emblem or, for that matter, versus
ekphrasis. In this regard, if anything, single verbal objects even supposedly fall below their visual counterparts and may altogether denarrativize
them. After all, painters "have often tried to suggest narration in a single
painting; the whole classical debate from Poussin to Lessing hinges in
part on whether the art of space should compete with the art of time." But
once words appear, narrativity disappears, because they "tend to restrict
the possibilities of interpretation" (ibid.: 35-36)." Hence, for ekphrasis
to comply with such a rule, it must be an overall ("single") descriptive
poem in the first place, rather than an element and so potentially a link
within a narrative. Nor is it surprising, therefore, that this taxonomy also
makes a point of distinguishing Bildgedicht,"a free verbal variation" on
the original image, from ekphrasisas "an exact description meant, to a
certain degree, to evoke and substitute for the painting itself" (ibid.: 44).
This is another of the many terminological attempts to divide the indivisible, to keep out undesirable features which the narrative model (were it
imaginable, here or elsewhere) would compound.
2.3 Ekphrasis,Description,and NarrativeTheory
The breakdown of all such exercises in limitation, Lessing's or his opponents', underlines that ekphrastic form and effect range all the way
between the warring poles. In terms of genesis, if you will, that range
corresponds to the distance between two polar origins of ekphrasis:
the exemplary activation of the device by its originator (the Laocoin's
Homer) and the etymology of the word (the Greek for "description,"
11. Accordingly, this presents the converse extreme to the pairing of language as
such with the narrative impulse, a tie-up that may be found in narratology itself (e.g.,
Scholes 1974: 17) and has recently migrated to one semi-Lessingesque approach to
ekphrasis (see n. 10). But why should either representational string attach anywhere
to the medium?
618
PoeticsToday16:4
first coreferential with the rhetoricians' descriptio,then subsumed under
it). Conceptually, empirically, genetically, ekphrasis makes an assorted
and open-ended bundle of variables, all free except for the constant
minimum of literary reference to visual reference to the world. The descriptive bent (or its inverse) is here no more criterial, or even more
typical, than any of the features that the system accommodates: verse,
figuration (notably prosopopeia), uniqueness, particularity, enclosure,
existential thematics, interart and intersexual power struggle, or their
respective opposites.
But the key term's etymology also reaches beyond origins. "Description" has meanwhile been practiced and studied in fields other than the
interart juncture, among which the narratological has most invested in
the varieties of world making. One would perhaps expect narratology,
therefore, to redress the imbalance caused by the aspiration to imagelike verse within the study of ekphrasis. Yet to date little of the kind
has happened, and not just because ekphrasis usually remains associated
with poetry, in disregard for its abundant manifestations in storytelling.
Were the routine tie-up with one genre of writing broken and the device
carried over at long last to this field of inquiry-if only as a marked,
interart descriptive subtype-the change would in all probability stop
short of essentials. For the extended coverage to make a principled difference, some of the field's larger assumptions (oddly akin to ekphrastic
criticism's) need to change.
Thus hampered, narratology in its present state could not help much
to repair the omissions pertaining to the temporalities of ekphrasisnot even indirectly, by reference to the master category of description.
For narrative theory all too often shares the modern anti-Lessing drive
to "spatial form" while, paradoxically enough, inheriting from Lessing
the bias against literary description, especially in narrative. (The latter
tendency is even stronger than the former; for details see Hamon 1981;
Yacobi 1991.) So, although descriptive writing has in recent years gained
considerable notice, Meir Sternberg is the only theorist who has systematically traced its multifold narrative power: for example, the difference made by preliminary exposition to our sense of the whole (Sternberg 1978: 23-55, 183-235); the sequencing of coexistent reality-items
(Sternberg 1981); the art of the proleptic epithet, state, or stative clause
(Sternberg 1985: 321-64; 1992: 527-28); or, more generally, how, in the
descriptive realm as elsewhere, space and time compose into "spatiotemporal dynamics" (Sternberg 1990, with earlier references). Like every
other discourse element, this realm works under the "ProteusPrinciple,"
whereby "many-to-manycorrespondences" necessarily show between linguistic form (e.g., description vs. narration as given on the surface) and
representational function (e.g., descriptive and/or narrative meaning).
No wonder, then, that when it comes to the Laocoin, Sternberg (1990)
finds its argument all too wanting, rather than excessive, in its orienta-
Yacobi * PictorialModelsand NarrativeEkphrasis
619
tion toward the temporalities of literary discourse and their effect on
spatial design.
But the exception, which does have consequences for ekphrasis as
a time device, only proves the rule: description-the mimesis of spatial objects -is normally theorized as subordinate to the forward-moving
action, rather than as an equal in its own right and sphere, let alone as
a potential ally. Thus Genette (1982: 133), an expert on narrative time,
places this "too well known" inequality among "the major features of
our literary consciousness." "Too well known," doubtless, yet who is in
reality covered by the all-inclusive "our"?Certainly not the exceptional
spatiotemporalist and the entire range of artistic practice (in or out of
ekphrasis, across the genres) that thrives on the interplay between the
two axes, the interpenetration of discourse forces. For very different reasons, "our"would no more include the mainstream of interart analysis
examined above. The promoters of literary (especially lyric) ekphrasis
as description, apparently unknown to the narratologist, assume the reverse hierarchy. But here the polar "literary consciousness[es]" nevertheless meet: in the tendency to associate the elements with opposed
values, to insist on an either-or focus, if not choice, and to withhold from
each the force deemed proper to its opposite. Genre, value, centrality,
power- all supposedly go together, one way or the other.
For Genette (ibid.: 134), though description may be "more indispensable than narration, since it is easier to describe without relating than
it is to relate without describing," it yet remains "a mere auxiliary of
the narrative": "ancilla narrationis,the ever-necessary, ever-submissive,
never-emancipated slave." Of its two "diegetic" functions-the "decorative" and the "explanatory and symbolic"-the first brackets it with the
other figures of style or rhetoric, the second with Lessing's extraliterary
"prose" coherence. By implication, then, ekphrasis must share the limited, one-dimensional role and with it the low status assigned to the rest
of description in "the" or "our" (i.e., this) literary hierarchy; the rendering of its static object becomes a mere "auxiliary"to the narration
of dynamic movement. In fact, by way of mirror image to the interart
camp, the freezing and downgrading of the descriptive in this modern
approach even exceed Lessing's again. For Genette, in his turn, does not
acknowledge so much as the possibility of dynamizing and so integrating spatial objects in the very action, Homer-style. Nor does he explain
why the "decorative" thrust cannot govern a narrative text. And just as
he fails to consider either option for promoting description within the
narrative whole, so he even more curiously denies the existence of selfcontained descriptive genres (ibid.: 134). From all quarters, whether to
mutually or multiply exclusive effect, approaches converge to repress
ekphrastic versatility, in the face of its actual manifestations as well as its
potentials, both "protean."
Within interart study, again, references to ekphrasis in narrative have
620
PoeticsToday16:4
lately begun to (re)appear here and there. Yet the emphasis still falls
on its occurrence as a spatial, indeed antinarrative figure and force in
temporal art. For example, Kurman (1974: 1) begins by defining "ecphrasis" as "the description in verse of an art object," thereby traditionally
prelimiting it on three axes: the functional, perhaps the formal as well
("description"); the generic, especially in its medial aspect ("verse");and
the mimetic ("an object of art," excluding the art model as source). The
generic condition, however, is sufficiently relaxed to admit the cross between "verse"and narrative foregrounded in "Ecphrasisin Epic Poetry,"
from Homer through the Renaissance to Mickiewitz. A promising shift,
given the neglect (duly noted) of the epic manifestations relative to the
lyric. And yet, Kurman does not really supply the omission, far less round
out the generic picture, because he in effect kills the narrative for the
sake of the "verse";he assimilates the newly foregrounded corpus and
elements of ekphrasis, by relentless violence, not just to the old rule
of "description" but to the space-figure prescribed for the lyric ever
since the exemplary analysis of Keats's urn by Spitzer (1962). Ancient
epic, we hear, already expresses "the nostalgia for timelessness that was
to make the device of ecphrasis so attractive to later poets" (Kurman
1974: 3): crossgeneric in essentials, the thrust only rises to panoramic
scale. Throughout literary history, he moreover alleges, a variety of epic
differentia, like similes or, incredibly, dreams and prophecies, cooperate
to reinforce this ekphrastic effect: "to frustrate time," or to "remove the
reader for a time from the main action," or to pass from "a story that
exists primarily in time to an event ontogenetically situated in space"
(ibid.: 3-13). Unsurprisingly, we find the Laocoon"classic,"epic reference
and all, dismissed without explanation in the last note.'2
Similarly with prose fiction. When Steiner (1982: 42-49; 1989: 279)
turns from a poem by Cummings, with its "still-movement topos," to
Edith Wharton's novel The House of Mirth, she carries over the alleged
generic role of ekphrasis: "As the topos of the still, transcendent moment," it "opposes the contingency of plot flow and temporal progression." Thus the novel's heroine, Lily, having gained "ekphrastic power"
from her description, comes to "stand outside time, as part of an 'eternal
harmony' . . . a pure, beautiful visual object, cut off from the world of
causality and contingency" (Steiner 1989: 290).
On a more theoretical level, we find W.J. T. Mitchell (1989: 92) explicitly instancing the ekphrastic genre to counter Genette's sweeping denial
of descriptive autonomy: "[Description] does attain a kind of generic
12. Compare its overt dismissal in Hagstrum 1958: 19-20 as an antidescription treatise; see also Auerbach 1973 [1946]: 3-23 on Homer's eternal present. Contrast the
dynamic reading of such features (simile, prophecy, and dream, inter alia) in Sternberg 1978: 56-128, whereby they all assume narrative power and coherence through
the operations of "gap-filling."
Yacobi * PictorialModelsand NarrativeEkphrasis
621
status in literary traditions such as ekphrasis, topographical poetry, and
art writing." But the hesitant wording, "a kind of generic status," ties
up afresh with the problematics of space and description in literature
and literary theory. Taking up Genette's power ("master-slave")terminology for the relations of time and space in literature, Mitchell regards
ekphrasis as one canonical mode of infiltrating "spatial, pictorial values
into literary forms," where they play "the role of the text's Other, its
negation or death, figured as the object of utopian desire and anxiety"
(ibid.: 92-95; cf. the "poetic Otherness" suggested earlier by Hollander [1985: 15-16]). So below the surface disagreement about generic
autonomy and role, there persist the key assumptions of the mimetic
heritage: that ekphrasis is always descriptive and that its spatial orientation as "the text's Other" goes against the grain of the verbal medium.
Only, Genette's "slave"has become Mitchell's (1989: 97) active antagonist: a rebel against the master, an undercover force for subversion, or
(in the most loaded metaphor) a woman in the land of time-directed
men:
Keats's'Ode on a GrecianUrn,' the canonicalexample,bringsexplicitlyinto
play the multiplicityof [sexual] roles played by literaryspace: as feminine
objectof desire and violence ('thou still unravish'dbride'),as rivaland competitorwiththe poetic voice (the 'sylvanhistorian'can tell 'aflowerytale more
sweetlythan our rhyme'),as a 'coldpastoral'which 'teasesus out of thought'
with its ambiguouseternityof desolation,perfection,and frustration.Keats
maycall the urn a 'friendto man,'but he treatsher like an enemy.
A change is registered here in the ideological force and value (from submissive to subversive, from negative to positive) ascribed to ekphrasis,
rather than in its essential allegiance and thrust (no corresponding shift
from space to time, from statics to dynamics). In essentials, the traditional poetic role shows itself to be constant across the political variations
between Mitchell and Genette. Though meant to privilege and liberate
ekphrasis, moreover, its fixture as "the text's Other" ironically denies it
the narrativizing resources granted to description by Lessing himself,
the supposed conservative in life and art. Finally, is it an accident that,
while Mitchell (ibid.: 97-101, on space in Bronte) sometimes argues his
larger thesis by reference to the novel, his examples of ekphrasis are all
drawn from poetry?13
13. As are even those in Heffernan1991,which shift the focus of the narrativeimpulse
from Lessing'sepic paradigm, the shield in the genesis, with its prose counterparts
since, to the newer favorites, Keats'surn and Shelley's Ozymandias,complete with
their nonepic (far less novelistic) elements, such as the poet's recourse to prosopopeia. But contrast Lund 1992, in which international scope and long historical
perspective (on "iconicprojection")go with a cross-genericrange of illustration.Despite the variance in topic, this wealth of material has indirectly served to retest my
theses against examples, even literaturesunknown to me before, namely,the Scandi-
622
PoeticsToday16:4
In summary, the range of ekphrastic forms and effects, I would argue,
stretch beyond the limits drawn by either polar approach and even beyond the respective latitudes taken together. The preference for interart
mimesis, reinforced by associated modern norms and scales, doctrinally
goes against my two central concerns in what follows: modeling and narrativity in ekphrasis. Where either factor must count as offensive even by
itself, their twinning is beyond the pale. Relative to what it has bred since,
Lessing's approach (with its aesthetic regulation of mimesis, promotion
of literary time, and eye for difference) would in principle encourage
such concerns, especially the inquiry into how the narrativizing of the
image in transmission makes a distinctive crossliterary resource, a variable second to none. Yet the actual aid and tools it could offer for the
purpose remain all too limited. This is not only because Lessing's own
focus of interest lies elsewhere, outside ekphrasis (i.e., in interart comparison rather than in transference from visual to poetic art). Even if it
were otherwise, some holes and counterthrusts would persist at a deeper
level. For Lessing never shows much interest in image models as distinct
from unique artworks; he could not anticipate (what with his nostalgia
for the epic) the refinements of modern storytelling, theorizing, and
exegesis; he brings time to bear on the space-object in a manner both
local and immediate (e.g., the shield transformed into a shield-making
episode, digressive from the main story line) rather than on its points of
contact with the framing tale; and, an aesthetician of harmony, he would
always object to the tension between the descriptive and the narrative
that is built into all ekphrasis (by force of its reference to the world,
albeit the art world) as a mode of what Sternberg calls "spatiotemporal
dynamics." The neoclassicist versus modernist extremes being so exclusive of literary practice and its problematics, as well as of each other,
we therefore need to formulate or reformulate the issues in a larger
perspective.
Thus, where the verbal image is reduced to a common denominator
of the visual and/or incorporated into the temporality of discourse, not
excluding plot itself, some interesting questions arise. How does the representation of a pictorial world in language, time, and movement affect
the respective images, taken singly and together? What are the typical
functions of alluding to an extramedium model rather than to any of
its instances? What is its influence on the (re)cognition and reliability
of the various parties involved: the observer within the fictive world as
against the author and the reader in the frame, the "beholders" of the
literarized picture (and of its dramatic observer) on the rhetorical level?
navian. It is therefore also of interest that my own paradigm case, the English-writing
Danish storyteller Isak Dinesen (as well as, more understandably, the Hebrew poet
Dan Pagis), does not feature in the book.
Yacobi * PictorialModelsand NarrativeEkphrasis
623
Where exactly do the narrativizing potentials of ekphrasis come from?
How do they meet - if necessary, override - the descriptive thrust of the
re-presented original? And again, how does the source model compare
with the singular image in this regard?
In addressing such questions, we must keep the literary repertoire
balanced, for a change. It would therefore be counterproductive (and,
for me, also repetitive) to work with the examples from poems touched
on above or even, as usual, from poetry alone. Instead, I will illustrate
mainly from Isak Dinesen, a born storyteller, a student of painting, and,
across the arts, a master of pictorial modeling in language.
3. Modelsand Spatialitywithinthe StorytellingPoeticsof Dinesen
Dinesen, however, offers far more than illustrative balance. The ekphrastic part, I have been arguing, whether a local inset or a virtually textlength stretch, is inseparable from a complex of wholes that regulate
and interpret it. Among this complex, the framing, "quoting" discourse
and the source, original artwork(s) enter most immediately into the partwhole web of relations. But the immediate structures (textual, crosstextual, intertextual) depend in turn on wider parameters, from the
medium and genre of the respective works; through the discursive forces
in play, especially the ever-shifting balance of the narrative versus the
descriptive; to history- and artist-specific designs, frameworks, conventions, innovations, repatternings. My paradigm case richly exemplifies
both the interrelations at their most complex and their (re)formation
into a determinate poetic unity; or, the other way around, both the hopelessness of the a priori fixture sought by analysts (even compounded in
the "stillmoment" approach) and the penalty of atomism incurred by the
mere anatomizing of the device, taxonomist-fashion. That the ekphrastic part does cohere with its wholes -however intricate the synthesis, or
however it varies from one corpus to another-is not the least of the
lessons that can be derived from this case. Nor is it the least of my reasons for building the general argument around a master practitionera woman at that, and as such alleged by some to reach willy-nilly for the
otherness of space fixities, in opposition to man, the time dweller. We
will find her belying the sexual polarity, no less than the rest, and with a
vengeance.
In Dinesen's poetics, the allusion to visual space-models closely relates
to what is probably her strongest claim to radical originality, as well as her
lifelong enterprise, namely, the upgrading and repatterning of the spatial dimension in literature. Among the set of principles involved, those
most relevant to my argument bear on four aspects: space as meaningful form, as dynamic force, as artistic model, and as interart juncture.14
14. For Dinesen's poetics of space see my previous papers (Yacobi 1989, 1990a, 1991).
624
PoeticsToday16:4
Their bearing on ekphrasis will emerge from the briefest outline and will
later prove, I hope, to extend beyond Dinesen's work. The artist'sunique
space novelties overlie, and their uncovering brings out, a universal of
the literary time-art.
To begin with, space comes to figure prominently as a (if not the) locus
of meaning and design. Thus, in a typical metanarrative comment on the
arena of "SorrowAcre," Dinesen's narrator observes that "a child of the
country would read this open landscape like a book" (WT, 29, also 30,
37, 60-61).15 Such analogies between nature and book, between physical
setting and covert yet readable writing, draw notice to the operations
of space as a semiotic system throughout Dinesen's art. Nor does this
system provide a mere addition to or substitute for others. As the same
narrator comments a few lines before, "The country breathed a timeless
life, to which language was inadequate": where words fail, the semiotics
of space comes to the rescue. Itself a dimension of the fictional world, in
short, space works to organize and interpret that world into a pattern of
significance along various lines.
One major line is implied by the very sentence "The country breathed
a timeless life." An apparent personification, this states a literal fact
within the Dinesen universe. According to the grand design of her world,
space and the objects traditionally immobilized in it (sea, land, forest, artwork) not only assume a life of their own but may even secretly manipulate human life, for good or for ill. In startling, ideological opposition to
anthropomorphism old and new, things thus rise to the status of animacy,
agency, not excluding determinate personality. Their rise, far from being
a figure of prosopopeiacal (etymologically, "person-making") speech,
generates a new world-picture. What is more, speech is the only feature of
personhood that Dinesen withholds from extrahuman reality-not least
from artworks, for example, statues qua re-presented objects/subjectsas if to make the cleanest break with the tradition of prosopopeia. Having
discussed the details in my "Plots of Space" (Yacobi 1991), I shall now
generalize their relation to the strategy of ekphrasis. For Dinesen, crossing the traditional value-laden lines between space and time, still and
mobile existence, description and narration, is a matter of high realism
as well as experimental art.
Again, with space as with other patterns, Dinesen is acutely aware of
the models through which reality is mediated, transmitted, perceived
in literature and art at large. Her staging of professional tellers, her
allusions to Scheherazade, her play on conventions of the most diverse
kinds and ages, her stylized plots, figures, setups, language--all flaunt
this awareness in regard to literary manner and matter. But the same
15. For simplicity and interlinkage, all references to Dinesen's works include an abbreviated title (e.g., WT= WinterTales).
Yacobi * PictorialModelsand NarrativeEkphrasis
625
consciousness of models-including their force, traditionality, range of
inevitability, and hence problematics for representation-arises with respect to the visual, spatial media as well. Thus the shrewd dialectics
voiced by a fictional art connoisseur, Count Augustus in "The Poet":
I have learnt that it is not possibleto paint any definiteobject, say,a rose, so
that I, or anyother intelligentcritic,shallnot be able to decide, withintwenty
years, at what period it was painted, or, more or less, at what place on the
earth.The artisthas meant to createeither a pictureof a rose in the abstract,
or the portraitof a particularrose;it is neverin the least his intentionto give
us a Chinese, Persian,or Dutch, or, accordingto the period, a rococo or a
pure Empirerose. If I told him that this waswhathe had done, he wouldnot
understandme. He might be angrywith me. He wouldsay:'I have painteda
rose.'Stillhe cannothelp it. I am thus so far superiorto the artist.... At the
sametime I could not paint,and hardlysee or conceive,a rose myself.I might
imitateanyof theircreations.I mightsay:'I willpainta rose in the Chineseor
Dutch or in the rococo manner.'But I shouldneverhavethe courageto paint
a rose as it looks. Forhow does a rose look? (SGT,382-83)
The speaker divides the producer from the consumer of art. On the productive side, not only the artist but also his art mediate between the real
object and the beholder. For consciously (like Dinesen) or not (like the
naive painter envisaged here), the artist's rendering is always bound by
conventions: a school, a period, a style, a culture, "Chinese, Persian, or
Dutch, or... rococo." Yet the artist, no matter how blind or even resistant to the forces of convention, will alwayssummon up against them "the
courage to paint a rose as it looks"; and so he will, to some extent, free
his "abstract"or "particular"image of a rose from the constraints of the
given rose-model. His very blindness-like the acute self-consciousness
of a Dinesen in literature, or the equivalent in painting-may radicalize
the omnipresent play of mediate and immediate world imaging, pictorial cliche and novelty, tradition and the individual talent. On the other
hand, lacking the "courage"as well as the gift for immediate vision, however partial, even the most knowledgeable connoisseur falls below the
most naive creator. Were he tempted into shifting sides, he would be
reduced to imitating particular "creations"or the traditional "manner."
For how does a rose look?
That this credo appears in a tale entitled, and focused on, "The Poet"
suggests literary analogues. In Dinesen, however, the arts involved also
enter into much closer relations than meta-artistic or comparative analysis. Not satisfied with exploring their theoretical common ground, she
draws them together into an actual common life, setting one within the
other, by way of ekphrasis, if not one beside the other, in syncretic multimedia form. Her ekphrastic allusions from literature to models of spatial
art (much like Pagis's on a poem-length scale and a different worldview)
then become part of the operations of space in the discourse as a whole,
626
PoeticsToday16:4
themselves uncanny, with the result of compounding the problems and
inspiring original syntheses.
Here, as though representing such coexistence at work in or on literature's world were not challenging enough, space comes to double in representation as interspace: a multiple, interart semiotic system, whereby
the signs ("languages") of different media get superimposed on narrative
in words. Superimposed, I say, because the elements brought together
appear to be anything but harmonious, even less so than within the
genre of lyric or descriptive poetry (where critics usually find ekphrasis).
For the built-in tension escalates toward diametric opposition between
Dinesen's narrative art and the arts it draws on and alludes to-between
target and source media. Nothing would be more deplorable to Lessing, nothing more desirable to the anti-illusionists, especially in a writer
who casts her interart net beyond the pole opposed to her own. Neither
value judgment applies here, for reasons to emerge soon, but the givens
and choices on which either would pounce do come to the fore. Considering the multiple temporality of her target medium, which unfolds
story lines in time sequences, it is remarkable that her main extraliterary
source media for allusion should turn out to be the spatial arts. (The theater, a more likely source for re-presentation because it crosses the two
polar harmonies into a time-space art, comes only second.) Oriented by
definition to space, painting and sculpture as sign systems distinctively
arrange signifiers and signifieds alike side by side rather than in a line:
they render states through sign configurations, not (or not primarily)
actions through sign sequences.
This immediately raises the question of their integrability in temporal
context. To be sure, this question arises wherever the art forms brought
together pull against each other because they extend in different ways.
Yet nowhere (e.g., in language's descriptive re-imaging of a still life, or in
the transfer from the midway art of theater to either extreme) does the
tension present such a challenge to overall integrity as in narrativized
ekphrasis, with its doubled double extension: the original's versus the reworking's. Narrativized, I insist, because the difficulty of synthesis varies
not according to the text's genre by itself but according to the shape
and function with which the text invests the ekphrasis. Prose narrative
may and often does opt for the descriptive variety, least inharmonious
due to the object-to-object correspondence across the media; conversely,
poetry, the lyric included, is free to radicalize matters by charging the
interart allusion with narrativity (single-line or, we recall, even multiple)
on top of sequentiality. In this regard, though not exclusively, the functions involved outrank or crosscut the genres of literature at ekphrasis,
and, crossgenerically, the descriptive type makes a simpler case than the
narrative turn. In integrability, the former even reduces to the latter-
Yacobi * PictorialModelsand NarrativeEkphrasis
627
which subsumes its problem of assimilating an import, of linearizing a
graphic configuration -but not vice versa.
As it happens, Dinesen combined the source and the target art forms
in her life: she studied painting but practiced writing. (Her affair with
the visual arts actually came early and remained lifelong.) So her twofold expertise makes it especially instructive to trace the meeting and
interplay of these polar opposites in her fiction, where one assimilates
the other. Since the imaging of the world in painting and sculpture apparently runs counter to narrative in both the object and the medium of
representation, how are their spatialities indicated and incorporated in
the most time-bound (because storied) of all mimesis?
The generality of reference in ekphrasis is, among its other services,
a central means to this end: it helps the writer exploit, and the reader
grasp, objects imported from pictorial space, so that they gain coherence both in and beyond their spatiality. As a rule, Dinesen refers the
particular narrative situation not to an equally particular artwork but
to some traditional model or theme of spatial art: Lacoon, Diana and
Acteon, the adoration of the Magi, the Madonna with child, and so on.
This rule of inequality in specification (target vs. source, word vs. image,
plot vs. state) has various reasons and effects. Three of them stand out:
evocability, perceptibility (or accessibility), and assimilability (or maneuverability), especially in the framework of narrative re-presentation. All
of these factors being essential not only to Dinesen's poetics but also to
model-oriented ekphrasis as such, in relation to work-directed varieties,
let me introduce them in turn.
For one thing, Dinesen's expertise in both arts combines with her literary interests to point away from the mimetic tradition, not least from its
objective regarding ekphrasis. To her, such one-to-one interart mimesis
is not much more appealing than feasible. A portrait or statue is unreproducible in words, owing to the variance between the two media,
so that any attempt to detail, let alone rival, such an object would not
only court failure but distract attention. And to what purpose beyond
itself, Dinesen might wonder? Instead of incurring such loss, then, why
not generalize the reference (or the re-presentation) in the first place
and in the service of less traditional as well as more viable ends? In the
critical, instrumental approach to mimesis, if in nothing else, including
the values and priorities of literature, Dinesen would side with Lessing
against the copy-making tradition.
Indeed, the second reason for generalizing the interart allusion into
model range already brings us to a constructive role played by the device,
namely, as a force for perceptibility in ekphrasis. The question whether
or not the reference from word to image is perceptible has far-reaching
interpretive and even theoretical consequences. (So, for that matter, has
628
PoeticsToday16:4
the pointedness of inverse transfer.) What hangs in the balance is nothing less than the viability (communicability, accessibility, readability) of
ekphrasis, which a misconception like Kibedi Varga's (1989) seriously
and gratuitously threatens.
When "the artist is inspired by a preexisting image and writes ekphrasis," this critic states, "the reader reads a poem, without necessarily
perceiving the other part too." Hence, "this distinction implies that we
must argue from the point of view of reception rather than production"
(ibid.: 33). Whatever its taxonomic use, the statement is odd to the point
of undoing the genre. Of course, readers (though not "the reader") are
alwaysliable to miss "the other part," the source image. In such readings,
however, ekphrasis as a word-image relation simply vanishes to leave an
ordinary "poem";and they are accordingly not just partial but incompetent readings by generic rule.'6Therefore the argument for an approach
through "reception rather than production" does not follow, either. On
the contrary, were the choice between them necessary or generically
imaginable at all, then the receiver's possible ignorance and blindness
would force us to approach ekphrasis from the well-informed viewpoint,
the producer's--except that "us" would then denote just another receiver, himself perforce in the know, to another self-contradictory effect.
Instead of a binary taxonomy, production and reception necessarily
make two sides of a communicative affair. Within a theory of ekphrasis as such, whatever the performances of individual readers/receivers
commit or omit, "the reader/receiver" as well as the producer must by
definition be aware of the ekphrastic bond of word-to-image reference.
The only question left open- and it makes a variable of high importance,
as well as of great explanatory power-is what happens in (inter)artistic
practice. Given the variations in pictorial expertise among readers, how
can we establish a common ground that will define "the reader"? What
measures does the (or a certain) literary producer contrive to ensure,
standardize, sharpen, channel (etc.) our awareness of the cross-reference
necessary for the working of ekphrasis, for the desired generic communication between the parties to take place and effect? Where, in short,
does the key to generic perceptibility lie in this or that text, corpus,
school, manner, or any other crosscut of the field?
Here, all other things being equal, model-directed ekphrasis compares
favorably with the alternative variety. Dinesen's avoidance of definite in
favor of generalized reference to artistic practice works as a safety measure, because it virtually ensures the reader's familiarity with the referent
(topos, posture, setup, theme) invoked. In all that regards ease of spotting and deciphering and patterning, it has the advantages of a stereo16. Cf. Cliiver 1989 on the difference between reading an ekphrastic text as an
"original" and as a "translation."
Yacobi * PictorialModelsand NarrativeEkphrasis
629
typed allusion, a pictorial cliche. Catching the reference to such a model
presupposes nothing like the expert's encyclopedic knowledge or like
Count Augustus's ability, in the face of any given artwork, "to decide,
within twenty years, at what period it was painted, or, more or less, at
what place on the earth." Instead, if the reader has not viewed one particular artistic image of Laocoon, or Diana, or the Adoration, then he
has viewed another. And even if he has encountered none, or no longer
recalls any with confidence, the chances are that his repertoire of mental
schemata includes the appropriate topoi for the discourse to activate in
the process of reading.
As we advance from mimesis in re-presentation to accessibility in communication, then, the negative reason for Dinesen's method joins forces
with a major positive; the dismissal of the unreproducible comes with
perceptual gain. In Ehrengard,typically, both reasons motivate a fictional
painter's evasion of the request to verbalize spatial reality: "You ask me
for a description of Schloss Rosenbad. Imagine to yourself that you be
quietly stepping into a painting by Claude Lorrain" (E, 31). Like his creator, herself with a foot in either art, the character does not so much
depict as recall the object in bare outline; nor does his recall point to
any specific Claude Lorrain but to the model of nature underlying many
of the pictures, and hence far more amenable in transfer to the reading
directive "Imagine to yourself."17
In certain quarters, such provision for salience, access, and contact
with the audience may be regarded as a concession unworthy of high
literature -a flattening of ekphrasis to the lowest common denominator with a view to the widest appeal. Let me therefore emphasize that
Dinesen's recourse to the model, complete with the poetics (or, if you
like, the problematics) of readability behind it, has its equivalents among
other sophisticated writers, classical and modern.
Here are a few telegraphic cross-references. "The taper, sensual fingers," R. L. Stevenson (1952: 205) writes about Alain, the old Sire
de Maletroit, "were like those of one of Leonardo's women." Rossetti
(1985: 130), looking through the train window, registers scenes "byHans
Hemmling andJohn van Eyck."Or witness howJohn Ashbery (1986: 26,
95, 235) brings a model to bear on a scene from "a long novel," where
"each snowflake seems a Piranesi"; or on being left "alone and skinless,
a drawing by Vesalius"; or on poetry ('And," the title reminds us, "Ut
Pictura Poesisis her name"), whose writing throws the mind's "extreme
17. And further, a model institutionalized in literary pictorialism since the neoclassical age; hence a model with a whole tradition of modeling behind it, as interart topos,
so to speak. For a short overview of Lorrain's fortunes see Lund 1992: esp. 90ff., and,
for contrast, see ibid.: 36-39 on the markers needed to identify specific references
across the arts.
630
PoeticsToday16:4
austerity" into collision "with the lush, Rousseau-like foliage of its desire
to communicate."
Less telegraphically, George Eliot, commenting on the realism of her
"simple story," Adam Bede, multiplies pictorial analogues, contrastive and
like-minded. She expresses her delight in "many Dutch paintings," where
I turn, without shrinking, from cloud-borne angels, from prophets, sibyls, and
heroic warriors, to an old woman bending over her flower-pot, or eating her
solitary dinner ... or I turn to that village wedding, kept between four brown
walls, where an awkwardbridegroom opens the dance with a high-shouldered,
broad-faced bride, while elderly and middle-aged friends look on, with very
irregular noses and lips. (Eliot 1956 [1859]: 173)
In defense of her novelistic art of "truthfulness," Eliot thus appeals not
to masterpieces but to models from the sister art: to what "an old woman
bending . . . or eating" and "an awkward bridegroom . . . bride . . .
friends" share (or equally lack) vis-a-vis "cloud-borne angels" and their
genre. The distinctive low verisimilitude is now pronounced interartistic
on top of intrapictorial, a model of models, in fact, not unlike Dinesen's
Lorrain, only more abstractive still. But if "lofty-minded people despise"
this common denominator, they will know it as well as the admirers.
Auden's "Musee des Beaux Arts" may appear to be a strange companion piece, if only because it is so celebrated for its (modernistic and
"particular" [Hollander 1988b: 34]) ekphrasis of Breughel's Fall ofIcarus.
However, this re-presentation of the painting stands anywhere but alone,
or even first, in literary history; and its traditional bearing on the original finds its complement in the reverse one-to-many relation. Before the
close-up on the one named artwork, Auden (1976: 146-47) pays collective tribute to the insight into "suffering" displayed by
The Old Masters: how well they understood
Its human position; how it takes place
While someone else is eating or opening a window or just walking
dully along,
thus leading up to "Breughel's Icarus, for instance" in the next stanza.
Compared with Adam Bede, and with the later Proust hillscape, everything has changed in the transfer except the procession of "or's" (here
underlined by the climactic "for instance") unique to language and incrementally exemplifying a shared pictorial quality. This procession also
recalls the serialization of the snapshots in Dan Pagis's poem, except that
his series coheres along a different illustrative line ("modeling") -temporal rather than existential. Where would one find the ordered story
of humanity's mutability if not in an album, and a crosscutting model of
existence if not in a museum?
Again, between the studied anonymity of "many Dutch paintings" or
"the Old Masters" and the titling of Icarus, as between the distilled visual
Yacobi * PictorialModelsand NarrativeEkphrasis
631
attribute ("truthfulness," "suffering") and the singular case in point,
there lies the topos identifiable from its relative, foregrounded or encoded detail. Thus, when Nabokov describes a scene in a Parisian restaurant by reference to The Last Supper-just
as when Dinesen (WT,
evokes
the
in
the
embrace
of
two
Laocoin
142-43)
tragic sisters-he alludes not to any specific picture or statue but, again, to a pictorial model,
a generalized visual image, with one difference. The referent has long
become a stock figure. So, in "Spring in Fialta," Nabokov does not, and
need not, explicitly identify the model. His narrator watches a novelist
"presiding" at a long table and comments:
For a moment his whole attitude, the position of his parted hands, and the
faces of his table companions all turned towards him reminded me in a grotesque, nightmarish way of something I did not quite grasp, but when I did so
in retrospect, the suggested comparison struck me as hardly less sacrilegious
than the nature of his art itself. He wore a white turtle-neck sweater under a
tweed coat; his glossy hair was combed back from the temples, and above it
cigarette smoke hung like a halo. (Nabokov 1967: 16)
The narrator dramatizes the theme of perceptibility in his own belated
recognition of "the suggested comparison," due to the failure of his
memory at the time. Yet he counts on his reader to reconstruct the
unnamed model or field of allusion from the given and interrelated ingredients: a Christ figure, signified by the halo above him; his central
position emphasized by his parted hands and by the disciples all gazing
at him. And the narrator can safely assume our reconstruction, though
he himself "did not quite grasp" the origin of the pattern, except "in
retrospect." For his momentary lapse alerts and triggers our memory;
and we are also in a better position to make the interart connection
("comparison") because he mediates and pinpoints it for us in the telling.
Himself confronted at the time with a dense and heterogeneous reality the modern profane scene that encloses, crisscrosses, and so blurs the
ancient sacred image "in a grotesque, nightmarish way"-the narrator
does not expose us to, let alone trick us into, anything like the same bewildering fullness at second hand. On the contrary, his discourse picks
out from the original scene a minimal cluster of identifying items that
runs through many specific works, styles, periods, and schools of visual
art.'8 For good measure, the alluder even throws in such loaded giveaways as the evaluative "sacrilegious" or the interpretive, metaphorical
"halo." His selective modeling does duty for naming The Last Supper.
18. Indeed, their own imagings, however distinctive, recognizably select from their
common literaryoriginal, so that we encounter a mimesis in the third degree and on
two mutually reinforcing levels of abstraction: the interart chain goes from the New
Testament to the paintings (one-to-many) to a work of literature that focuses them in
turn (many-to-one) on a writer of novelistic literature.
632
PoeticsToday16:4
Nor does perceptibility exhaust or even head the model's constructive recommendations in ekphrasis, certainly not in Dinesen's. Rather,
most important of all (and, accordingly, my chief concern in what follows) is a third reason for the generalized reference from literary tale
to visual topos: the effect on our sense of priorities and plenitude. The
borrowed sign (Laocodn, the Magi, etc.) is not just left unparticularized,
under the pressure of verbal re-presentation or communication or both,
but is intentionally departicularized in their favor. The very disparity in
the scale of representation between the source and the target arts is
enough to establish a generic scale of importance. This twofold scaling
characteristic of the model does more than rule out the tensions that
loom so large in criticism, namely, the interart struggle for the honors of
mimesis or stillness or just dominance, sibling or sexual. Conflict gives
way to a means-end nexus of the inset-frame variety.The ekphrasis here
forms, not an entire work, much less a poem, least of all one emulating
the original artwork in its rage for description-as theory would have
us expect-but a part within the storytelling whole, and a drastically reduced part by any comparative standard at that. The less particular the
visual image, the more evident its function as an aid, not a rival (Other,
enemy, counterforce), to the particularized narrative and the richer its
contribution to narrativity as to everything else in time.
Thus, instead of settling for one variant or treatment or reading of the
spatial model in art history, as with the ekphrasis of unique artworks,
Dinesen exploits them all; the entire tradition (possibly including works
yet unpainted at the time) comes to bear on the individual tale, with
enormous gains in the range of reference and suggestiveness. Nor do all
those gains follow automatically: the principle or potential is one thing,
the use made of it another, variable in thrust and extent. By the logic of
departicularization, much the same rescaling affects the equivalents just
cited-from the novel, from poetry, from the short story-except that
the ends there are not primarily narrative, not even within the narrative
works among them. Rather, they pursue ends that belong to discourse
in general, descriptivity included. George Eliot's modeling is less narrative than metanarrative, designed to advocate a certain poetics by artistic
analogy; Auden's serves to extrapolate a valued idea of life, Nabokov's to
draw a portrait, Proust's a landscape. None would therefore constitute a
tale by and for itself, or dynamize the march of the framing tale, if any,
or even our expectations about it. Observe how the closest approach to
such maneuvering stops well short of the developmental extreme: Nabokov's hide-and-seek game operates on an existent (who's being drawn as
what?) rather than an event, unfolds along the sequence of linguistic telling rather than of lifelike happening, and is articulated by a teller who
has already experienced the enigma and now guides the reader through
it with the wisdom of hindsight, as might any puzzle maker.
Yacobi * PictorialModelsand NarrativeEkphrasis
633
It is not at all (pace Lessing) that such uses are inferior, untypical,
or unartful in modeling, or that the users are incapable of changing
or widening their means-end combinations; rather, and significantly,
Dinesen puts the narrative turn first. As we shall see, the departicularized allusion then sheds its original descriptive thrust to gain narrative
maneuverability, in the service of plot dynamics and/or point of view.
For example, such allusion may retain enough accord with ongoing developments to foreshadow the sequel-possibly behind the characters'
backs and against their expectations-or enough discord to ironize the
hero's view ("modeling") of himself in its terms.
As with the describers above, so with the fellow story makers here: observe the distinctiveness of both kinds of crossart maneuver, together or
apart, from alternative narrativizing options, of both the Lessing and the
one-to-one variety. On the one hand, neither Dinesen-type maneuver
overlaps, or so much as intersects, with those adumbrated in the Laocoon
and sufficiently elucidated by now. In fact, the respective logics strategically contrast, complementing each other to make up for our benefit an
ekphrastic repertoire of transformation. Lessing's interest focuses on the
set piece, the local and virtually detachable unit of representation: how
a bundle of singular descriptive features can get narrativized (the dress
into a ceremony of dressing, the shield into a tale of its genesis) to comply with "illusion" and pass literary muster, without regard for the work
as a whole. (So Kurman [1974] can proceed to detemporalize, in effect
to redescriptivize, the unit by treating it as a retardatory moment along,
or against, the overall epic sequence. Likewise see Mandelker 1991: 2 on
local novelistic suspension via ekphrasis.) But Dinesen's is an art of relations, and the ekphrasis springs to life within and for it alone, often to
decisive effect. For ekphrasis to destabilize the plot ahead of time, especially if otherwise stable-looking, it must integrate into the overall plot;
for perspectival impact and subjectivity, let alone incongruity, the art
object it re-presents (in whatever shorthand) must enter into the character's field of vision, self-vision, re-vision, which we ourselves observe as
part of the represented world at large. However small-scale, unmoving
by itself as well as unspecific, in contrast to, say, Achilles' shield in the
making, the structured pictorial inset here comes to dynamize the larger
frame, as the shield does not: to affect in little the fulfillment of the only
necessary and sufficient conditions for verbal storytelling. These, on the
other hand, might indeed draw equal energy from the reworking of a
particular artwork (think again of "The Portrait" multilinearized), but
not with equal economy. The difference (whether judged an aesthetic
and operational plus, with Dinesen, or, with the admirers of dificultevaincue,a minus) persists and matters, as does the still more clear-cut one in
our aliveness to the original in art history. So, above all, Dinesen's intertextual or, rather, interart play of meaning in ekphrasis operates under
634
PoeticsToday16:4
this rule: minimum allusion from word to image for maximum inclusion
and integration of imagery.
This bird's-eye view will for the moment suffice to indicate not only
the roles of modeling and the ways to story making in ekphrasis but
also how and where they come together. Among the various possibilities, as I likewise hope to have established by now, those favored by
Dinesen are the least known to theory, though not to fellow practitioners
across literature, and would repay more detailed tracing and comparison.
Let us therefore explore the two, often converging lines of transformation into narrativity:via plot and perspective, the movement of the tale
itself and the management of the telling (especially where the authorial
teller clashes with some dramatized, self-deceiving or otherwise fallible
reflector).
4. PictorialModelsin NarrativeTransformation;
or, Dynamizingthe Static
A common denominator of visual and verbal structure, perspective yet
acts on the inset models as a powerful narrativizing factor, due to the
superior perspectival resources of the framing text. To begin with a fairly
simple (or simple-looking) transfer, many of the allusions, though short,
include the factor of gaze. In one death scene, for instance, the survivors
remainedquite still aroundhim [the dead man].The figureof the old Prince,
lying immovablyon the ground, still held the center of the pictureas much
as if he had been slowlyascendingto heaven,and they his disciples,left behind, gazingup towardhim. OnlyNino, like one of those figureswhichwere
put into sacredpicturesas the portraitof the man on whose order they were
painted,kept somehowhis own direction.("RoadsroundPisa,"SGT,208)
The dead Prince appears in a double focus. Within the imagined ("as
if... like") reality of the "sacred picture," he as a Christ figure ("slowly
ascending") literally holds the center of the other participants' gaze (the
one exception among the "disciples," Nino, emphasizing the unity of
the rest). And he likewise dominates the show within the verbal report,
where the narrator organizes around him the space of her own fictional
arena by casting it in terms of visual point of view and direction built
into a familiar model, so familiar as to offer a ready-made descriptive
metaphor.
Not a special case, except in detail, this exemplifies the law of ekphrasis. For ekphrastic re-presenting (as a subcategory of quoting in general)
entails not just re- but multiperspectivizing, if only into a twofold view,
namely, the original's (now inset in transfer) and the frame's; for example, TheLast Supper'sperspective on Christ and its Nabokovian modeler's on his novelist as Christ figure, or that embodied in the bronze
statue of Neptune, "taming a sea horse," and that smuggled into its
mention by the Duke as wife tamer, self-made widower, and prospective
Yacobi * PictorialModelsand NarrativeEkphrasis
635
bridegroom.19Under this law, the only differences are in the number, the
congruity, and the role of the viewpoints brought together. In regard to
viewpoint, even the two examples from Nabokov and Browning-neither
of them harmonious compared with, say, Rossetti's "van Eyck"-like Belgian landscape-significantly part ways. The former shows, if not much
the greater, then the more localized and coexistent harmony in disharmony, as befits the portrait's descriptive thrust; while the latter plays on
the disharmony between the sculptor and the monologuist's intent to
generate dramatic tension, a sense of conflict past and newly approaching with the remarriage: narrativity,in short. So, to return to Dinesen's
old Prince, does the noncoincidence in "gaze"between Nino and the rest
of the "disciples,"which typifies Dinesen's strategy at its least complex.
More often, and yet more integral to narrative technique -as distinct
from the rule of ekphrasis as such -we find the viewpoint on or through
the art form mediated, split, problematized, and so dynamized to match
in interart transfer.The invocation of ekphrastic models then comes not
or not only in the narrator's own name but through the perspective of
fictional observers, dramatized and less than reliable, who themselves
view their reality in such artistic terms. We must then distinguish between two typically narrative and often incompatible viewpoints on the
relevant model: from within the fictive world and, simultaneously and
more intricately, from without, where the storyteller communicates with
the reader alone.
On the rhetorical level, when the narrator of "The Caryatids: An Unfinished Gothic Tale"wishes to transmit the fearless spirit of the heroine's
children, she compares them in her own voice (or so it appears) to cherubs of old Relievi, "who are represented riding on lions, spurring the
mighty lord of the desert with their little rosy heels" (LT, 114). Reliable,
because authorial, this imaging of the children in outline functions primarily to "make the reader see" them through the visual vehicle, with
a view to portraying both their character and their appearance. Despite
the attendant danger to the "little rosy" things, the impression given is
one of static tranquility, for it is the habit of such fearless cherubs to
19. Likewisewith entire subsets of ekphrasis,usually discussed under another, more
unique-looking rubric. For example, Hollander's (1988a) tradition of reflexive ekphrasis, culminating in modernist poetry, not only makes an extreme variety of
anti-illusionism,as suggested in note 4 above. It also thematizes a certain clash of
perspectivesbetween the arts, or their instances, qua representations:the first-order
and world-directedmimesis of the picture as against its inset and discourse-focusing
mimesis within the poem, or, in short, objective versus reflexive, hence "subjective,"
artistry.So (re)perspectivized,this style shows itself to be no more, and no less, than
a well-definedvariantof the all-ekphrasticlaw, one choice among many perspectival
interplays available-not excluding the bid for "illusionist,"objective-to-objective
accord, a la Lessing.
636
PoeticsToday16:4
tame "the mighty lord of the desert." As one of the pictorial models
utilized early in this Gothic tale, moreover, the "cherubs of old Relievi"
convey the seemingly idyllic nature of the world, where children might
ride lions.
On the other hand, the interart reference loses its stability along with
its authority (or, in other words, modulates into narrativityproper) when
Childerique, the heroine, frames herself within an ekphrastic model.
Here, the ironies of narrative versus visual perspective, as of narratorial
versus figural vision, become a dynamic and complex tool. For the heroine's unreliability in her act of self-framing and self-imaging sharply contrasts with the narrator's sure hand, insight, foreknowledge. Accusing
her half-brother of dishonorable treachery, Childerique complains: "Is
it for ever, then, the task of the women to hold up the houses, like those
stone figures which they call caryatids?And are you now... going to pull
down all the stones of our great house, upon your own head, and upon
mine, and the heads of all of us?" (LT, 132-33). Notice that in importing
the sculptural model and willingly situating herself within it by a kind
of self-identification or role playing, Childerique, the fictional agent, expresses her sense of self-satisfaction, if not, or no longer, well-being and
harmony. She is caryatid-like, as it were, and the caryatid (alleged "to
hold up the houses") represents stability. Yet behind her back, on the
rhetorical level, the static image of stability bristles with tension, the invoked space-art with clues to time and hence with potential narrativity.
Unlike the heroine, the reader already knows that she is incestuously
married to her brother. So the threat to the honor and integrity of a
great house actually arises from her side. Our superior knowledge enables us to discern the narrative instability of the static caryatid model,
and even to predict events: once Childerique takes the initiative, in the
belief that she is safeguarding her ancestral "house," she will probably
bring it down. And in retrospect, its downfall comes to threaten even the
idyll that the earlier ekphrasis spun around the children of the house.
Regardless of any intermediate changes in the world, the interekphrastic
chain reaction is enough to remold our understanding and expectancies,
to reperspectivize the model on the entire story front, in brief.
Therefore, if the glance at carved lion-riding cherubs bore some resemblance at the time to Cummings's "stone children" as analyzed in
Steiner (1982: 42-48), minus the mimetics and sense of failure vis-a-vis
the original universalized there by analytic fiat, the allusion to the stone
caryatids inverts the parallel into contrast. Again, Childerique does not
at all resemble, not even for a moment, Steiner's (1989: 290) Lily in The
House of Mirth,because her two-edged "ekphrastic power" has an effect
antithetical to fixture "outside time, as part of an 'eternal harmony'... a
pure, beautiful visual object, cut off from the world of causality and contingency." Indeed, Lily herself might be profitably reviewed in the light
Yacobi * PictorialModelsand NarrativeEkphrasis
637
of her Dinesen counterpart, whereby she, her ordeal, and Wharton's art
would all be released from the stasis imposed on them against the grain
of novelistic discourse.
By a further turn of the narrative screw, such transformation may
occur in closure or in retrospect. Thus, since Childerique's self-image
is manifest early in the plot, and its falsity exposed in advance, it leads
us to envisage her painful route from ignorance to knowledge and selfknowledge. In "Roads round Pisa," however, Carlotta's unreliable selfportraiture, charged with the tension of incongruity, comes as late as the
end. When she accepts her fate and makes her peace with her granddaughter, she is described from within as "having taken for herself the
part of Joseph," while the real father is left with "no greater part in
the picture than the youngest Magus of the adoration" (SGT, 215). On
the face of it, the ekphrastic allusion to the well-known pictorial model
signals that the dynamics of one art, Dinesen's, has now given way to the
statics of the other; the conflict has apparently been settled and a point
of rest and closure achieved. But the divergence between the heroine's
(or inset reflector's) and the author's (or the text's) perspective on the
model complicates the effect. While believing that she has found new
harmony, Carlotta now identifies herself with the male role of a pseudofather, Joseph: a sign that her problematic sexual identity, dramatized
throughout, remains unresolved. This is hardly the closure, far less an enclosure, circular or otherwise, so valued by the Spitzer group. Given the
balance of power within this twofold view, the cross-reference to Joseph
does not even perpetuate the friction between outcomes, as does Keats's
to the "still unravish'd bride," caught forever between virginity eternally
preserved and imminently lost, between art's moment and life's movement. Instead of freezing the end in space, the double-edged pictorial
analogy launched by Carlotta ironically heightens our sense of ongoing
temporality, of the narrativity of narrative.
In both cases, the self's appeal to an ekphrastic model generates a tension much like that between tenor and vehicle in a far-fetched, jarring
metaphor, except that its incongruity points inward (to character) and
forward (to reversal). At times, indeed, the model even takes the surface form of an overt linguistic metaphor, or simile, as when Childerique
in her blindness reserves for women the duty "to hold up the houses,
like those stone figures which they call caryatids." No matter how it is
conveyed, the interart similitude drawn by the hero turns into ominous
dissimilitude (the caryatid into a destroyer, the placid Joseph figure into
a restless sexualJanus) in the higher, more informed narrative outlook.
But such tension between source and target may become so extreme
as to force itself on the visualizer's own eyes in midadventure, with active
consequences for his character, drama, even fate, as well as for our
understanding and expectations of all these from the outside. The cross-
638
PoeticsToday16:4
reference enters the arena, the mind's and the world's. In this stronger
variation peculiar to narrative, the hero changes his image of himself
and/or others in the course of the plot; the irony surfaces through the
character's movement from casting himself in some pictorial role to
working out a very different model, usually the hard way. As the artdirected perspective evolves, or revolves, so does the plot in (fictional)
life; the two axes become indissoluble. Lessing, though he neither treats
ekphrasis as such nor dreams of mental, subjective images in art, literary
or otherwise, much less of their replacing one another in psychodramatic
sequence, would yet find the dynamics highly appropriate. For "nothing
obliges the poet to concentrate his picture into a single moment...
Every change, which would require from the painter a separate picture,
costs him but a single touch; a touch, perhaps, which, taken by itself,
might offend the imagination, but which, anticipated, as it had been,
by what preceded, and softened and atoned for by what follows, loses
its individual effect in the admirable result of the whole" (Lessing 1963
[1766]: 22, also 58-61). Over and above carrying the principle to an
extreme, Dinesen reshapes it to suit her own existential, narrative, and
aesthetic concerns.
A plotted chain of visual (self-)images featuring the Laocodnnaturally
recommends itself as an example. In "The Invincible Slave Owners,"
Axel Leth first senses the rapport between a young girl he loves and her
duenna, expressing it in terms of traditional pictorial harmony: "The
elder woman's austerely plaited hair had in it a faded reflection of the red
within the girl's floating locks. It was as if the artist had found a little of
the colour left upon his pallet and had been loath to waste such a glorious mixture" (WT, 131). Later on, coming to feel at one with his world,
Axel also integrates himself into such pictorial harmony: "The afternoon
was so perfectly still, so golden, that he felt as if he had found his way
into a picture, some classical Italian painting, that suited him well" (WT,
139). This is, of course, an incongruous place for a nineteenth-century
Danish nobleman in Baden-Baden to project himself into, as he himself
learns once his beloved proves to be an impostor and her duenna her
sister and accomplice.
With the breakdown of his romantic illusion comes a switch in the
nature, the distance, and the reliability of his spatial allusions. No longer
blinded and constrained by the initial, wishful image for his own position
vis-a-vis the others (as if their color harmony merged with the classical
Italian painting), Axel realizes the complexity of the situation from a distance: "The two tragic sisters in the wood, their red locks emblazed by
the sun, in their very contortions had been so harmonious that he saw
them as a classic group, two maidenly Laocodns, locked in one another's
arms, and in the deadly coils of the serpent" (WT, 142-43). Through the
Laocoon model, Axel expresses in miniature not only the tragic insepa-
Yacobi * PictorialModelsand NarrativeEkphrasis
639
rability of the sisters but also its destructive effect on his love; hence the
metamorphosis of the redness vis-a-visthe overall composition in the two
images of the red-haired sisters. Initially he tries to fuse two pictorial harmonies, one purely formal (the "glorious"red mixture of paint) and the
other thematic as well (the "classical Italian painting," with its "perfectly
still, . . . golden" afternoon). But the continuity through classical color
norms proves unsuitable. Instead, the "red locks" compose with "the
deadly coils of the serpent": harmony in terms not of color but of round
form and pattern. At the same time, repetition cuts across variation; the
classicism of the different pictorial models and techniques implies the
anachronism of the ideal for which the two sisters are willing to sacrifice both present and future. In narrative terms, the Danish nobleman
thus finds himself excluded not only from the spatial configuration they
project but also from the temporal zone they would occupy.20
In turn, the future-the main narrative potential of the allusion-recalls but goes beyond the Laocoon'sfamous use by Lessing, namely, as
an exemplar of the "pregnant moment," chosen by artists in order to
circumvent the static nature of spatial art, while articulated by writers
(at original work or at reworking) into the movement suitable to their
medium. In Dinesen's tale, the interpenetration of arts and media, as
well as setups, goes not to neutralize or transcend either component
but again to reinforce the features of the narrative system. By an apparent paradox, the storyteller (against literature's capability, or Lessing's
advice) leaves the sculptor's moment itself intact in transfer and, if anything, immobilizes it in the modeling process; that the Laocoon family
dies, and will go on dying forever, only underlines by way of contrast
how Axel's adventure soon moves into another new phase (which, again,
owes much to the merging of media). At the same time, the ekphrastic
model, perceived from the outside, operates to defamiliarize and aestheticize the charged psychological situation (cf. Langbaum 1965: 16467) without simplifying it. On the contrary, the narrative gains depth and
complexity from the tense combination of the formal pictorial elements
with the thematic model of the Laoco6n.For the impression of static
calm given by two analogous female figures, done in the same color and
implying "classical"harmony, is undercut by the tremendous narrative
power of the Laocoon model; and that power stretches from external
to inner drama. The possibly simple portrayal of the sisters is enriched
by the (positive, at least pathetic) implications traditionally associated
with the Laocoon group; for example, the fatal embrace, the estrange20. It is interesting that, genetically, this tale was inspired by Courbet'spainting of
two women on a balcony (see Hannah 1971: 20). In the narrativizingprocess, however, the unique genetic art source gave wayto a sequence of conflicting models from
variousperiods and places, in line with the poetic rule.
640
PoeticsToday16:4
ment from society, and the agony of the death scene carry over from
the ancient to the modern figures. Notwithstanding both their origin in
a static surface structure and their compounded literary fixture there,
generally speaking, the ekphrastic models bring out (deepen, suggest,
forward, interpret) the emotional and psychological life of the characters
in Dinesen's stories.
Applied to the sisters, the shifting visual images make for "roundness"
or complexity in the process of their disclosure. With regard to Axel,
the image maker himself, the transition from the first to the last model
traces character dynamism as well. His perspective, moreover, develops
along with his personality; we witness his modulation from an involved
agent, who encounters and interprets the world (events, people, circumstances) in the subjective manner of the experiencing-I, to the relatively
detached, more objective, and so narrator-like observer of others. In
terms of the ongoing action, again, the change of pictorial models attends and reflects the hero's moment of discovery (anagnorisis), which
in turn rechannels the subsequent flow of the plot. Indeed, the visual
space-metaphors in Dinesen are as a rule located, and certainly replaced,
at central plot junctures. They may heighten or complicate our expectations at the outset, as when Childerique models herself on a caryatid;
they may lay bare the threat to and below the apparent stasis of the end,
as with Carlotta identifying herself with Joseph at the Adoration. Most
energetically, as with Axel, they come at the pivotal moment of recognition that leads through a change of model, attitude, and direction (all at
once) to the unexpected end: anagnorisis followed by peripety, modernstyle, with ekphrasis projected into a key role throughout the narrative
sequence.
5. SpecificWorksas Models:FromMetaphoricalto Literaland IconicImages
Turning now to another part of my introduction, let me further develop
the argument that rendering unique versus generalized images and investing them with narrative versus descriptive force are two independent
parameters of ekphrasis. I have already briefly illustrated the point from
such one-to-one re-presentations as Dan Pagis's three-track "Portrait,"
suggestively contrasting with the accelerated but humanity-wide tale told
by his "Album,"which in turn nicely contrasts with the descriptive key
to Auden's "Musee" of Breughel ("Icarus,for instance") and other "Old
Masters"wise in the ways of "suffering,"all poem-length. In light of the
foregoing argument, however, it will be hard to adduce a more revealing
test case than the practice of Dinesen, who sometimes does allude (in
midtale, again) to specific artworks as well.
Of special interest are those, most often statues, that exist as part of
the fictional world: on the same ontological level as the characters who
refer to them. Having real, literal existence within Dinesen's fiction, these
Yacobi * PictorialModelsand NarrativeEkphrasis
641
artworkswould seem to contrast with the model-images discussed so far.
For example, mention of theLaocoonwill bear on a genuine, historical art
object; even mention of a nonhistorical statue or picture will at least create an imaginary equivalent, a particular fact within the fiction. On the
other hand, Axel's reference to "aclassic group, two maidenly Laoco6ns"
evokes an image that has no existence in the real or artistic world (except
as a type, topos, or schema, subjectified at that) and only metaphorical
existence within the fiction by way of allusive analogy to fictional entities,
the two sisters. It is a figure's, or elsewhere the frame's, figure of thought
about the world. (Recall the explicit markers of figuration: "as if," "like,"
in the comparison of the old Prince's death to the Ascension or of Childerique to a caryatid.) Yet the distance between the "literal" and the
"metaphorical" occurrences of visual art shrinks remarkablyin Dinesen's
poetics. These meet both in the ways they aid perceptibility and in their
more substantive roles (plot movement, character portrayal, psychological insight, clues to interpersonal relations, etc.), not to mention their
intersemiotic twists and turns.21
Consider such literal statues as Psychewith Lamp, or a trio of equestrian figures in clay ("The Cloak," LT, 29; "Of Secret Thoughts and of
Heaven," LT, 54, 62). Rhetorically speaking, they all prove to be modeldependent (as variations on a familiar artistic theme and/or on one
another), verbalized in outline rather than in exhaustive detail, and
hence also easily integrated. From the reader's viewpoint, then, their
existence within the fictional world does not yet render them inaccessible and so break or even limit the Dinesen rule of "perceptibility."
Furthermore, they too serve to link two media or languages in Dinesen's
semiotic montage, with a notable difference in the mode of interrelation
and signification. For while the "metaphorical" model-images signify as
such primarily by virtue of epitomizing recurrent cultural objects and/or
themes, the artworks existing in the tale's world draw their significance
mainly from their similarity to unique individuals in it. Psyche is, genetically, a replication of the sculptor's beautiful wife-though not an official
representation, as the painting in the Browning dramatic monologue is
of the Duchess-while the success of the three childish equestrian figures
depends on whether their future owners recognize themselves in them.
The former captures the woman's image only to signify the goddess; the
latter both image and signify the children who have been portrayed in
21. The differences between two- and three-dimensional art as allusive or modeling
systems within this corpus fall beyond the scope of this essay. A clue to the difference
possibly appears in "The Monkey," where the Prioress maintains that "it was the devil
who invented a third dimension. Thus are the words 'straight,' 'square,' and 'flat' the
words of noblemen, but the apple was an orb, and the sin of our first parents, the
attempt at getting around God. I myself much prefer the art of painting to sculpture"
(SGT, 115).
642
PoeticsToday16:4
clay; yet both equivalence links manifestly involve a relation other than
that between the Christ figure and the Prince in death or the caryatids
and Childerique in house propping. So the difference between the art
source's metaphorical and literal existence in the re-presenting fiction
correlates with a shift of emphasis from metaphorical to iconic (or at
least near-iconic, quasi-iconic) resemblance to fictional creatures.
In turn, this difference extends from our semiotic process of making
sense with the aid of the respective visual images to their dramatic force
of making or motivating plot. If the interart play of imagery affects the
reader, then the iconicity of the fictionalized statues comes to be recognized as such by the dramatis personae, leading them to action, reaction,
and perception. The icon's very genesis (even if previously hidden, marginalized, unseen, far from dramatically enacted, Homer-style) forces
its way to notice, to meaning, to the center of storyhood. If you will,
the original mimesis (in the iconic sense, i.e., as visual representation,
not directly as literary re-presentation) turns dynamic. For instance, the
seemingly accidental fact that the statue of Psyche has been patterned
upon the sculptor's wife becomes, by force of likeness, the Aristotelian
"moving element" that initiates and sustains the action in "The Cloak."
Or, in a final move toward self-recognition, the three equestrian figures suggest themselves to the artist-protagonist as an image for his own
"triple" nature, pulling him different ways throughout life in various
roles and periods. The course of events, as well as our understanding of
it, thus turns on the artwork'surge, and the fictional observer's eye, for
resemblance.
As agents of plot, moreover, the statues existing within the fictional
world will not be reduced to objects: not even to objects used by agents
for propulsive ends, as when an art object serves to encode the Duke's
admonitory message to his future bride; or as the murder weapon in a
detective story; or, more widely, as a bone of contention. (On his very
deathbed, Browning's "bishop orders his tomb" so that "Gandolph ...
shall see and burst.") Even at their most earthbound and instrumental, I believe, such roles must have their place in any comprehensive
theory of ekphrasis; but our corpus gives them a strategic, existential
twist, revolutionizing their plot value and much else besides. In perfect
accord with the ontology of Dinesen's story-world, unparalleled for its
all-embracing dynamism in the modern or even the Western tradition,
the statues rather acquire the powers and honors of full-fledged existence, including life in time. Traditionally regarded as spatial, inanimate
things, fit only for human contemplation and description, or contention,
analytic or acquisitive, such artworks are narrativized by Dinesen to the
extent that they uncannily ascend to the top of the scale of being; they
claim nothing less than equality with the human cast, whose fortunes
they affect by their similarity, along with their contiguity, to its members.
Yacobi * PictorialModelsand NarrativeEkphrasis
643
Animacy, agency, subjecthood, neighborship, iconicity: to assume this
complex of statuses and roles, the art objects singled out for ekphrasis
must of course also lead individual lives, parallel to their human counterparts. Each features as an existent in the world itself, the story's or history's own, not as a model abstracted by the storyteller's discourse from
pictorial imagings of the world. Only thus can the ontological fellowship
escalate from the distinctive but unremarkable coexistence on the same
level (as reality-items, comparable to the Duke and the late Duchess's
portrait, "looking as ifshe were alive") all the way to the uncanny network
of correlatedness (as dramatis personae on a reality-wide stage).
In "The Cardinal'sThird Tale," for instance, the statue of St. Peter in
the Vatican directly changes the life and character of Lady Flora; it mars
her health and looks, by infecting her with syphilis, but reconciles her
through illness to the human condition. This reversal goes with a multiple analogy between the reversing and the reversed agents. To begin
with, the statue bears a physical and a mental correspondence to the
tall, "immovable," "denying" heroine (LT, 89-90). Again, as in the case
of the Laoco6n,the history of the saint (denial followed by dedication)
plays an important role in the analogy with the woman who follows the
same path. As Peter to Christ, so Lady Flora to St. Peter. The history of
the actual statue (transformed from the Roman Jove into the Christian
saint) even foreshadows her direction of change, by force of precedent.
Nor is the strange analogy wholly reserved for us outsiders. Overtaken
by a sense of kinship from their first encounter, rather, the lady is mysteriously attracted to St. Peter's statue, visiting it again and again. As always
in this pattern, the beholder's own eye for similitude (the thematic selfrecognition in the icon of the other) clinches matters; it wears down her
resistance to the point that she kisses the statue, which in turn leads
to her sickness and metamorphosis. So the icon's multiple reference in
story and history-to Jove, St. Peter, and Lady Flora herself-elevates it
from inanimacy to a humanlike state, if not to superhuman action.
In "Peter and Rosa,"we even witness how a purely metaphorical model
reappears as a statue-agent within the fictional world. It all starts when
the adolescent Peter suddenly likens his cousin Rosa to a ship's figurehead:
Now Rosa [standingon the windowsill], in her stockingedfeet, with the skirt
of her blue frockcaughtbackby the cross-barof the window,wasso like the
figureheadof a big, fine ship that for an instant he did, so to say, see his
own soul face to face. Life and death, the adventuresof the seafarer,destiny
herself,here stood straightup in a girl'sform. (WT,259)
As with Lady Flora, only by way of simile rather than icon, the observer
finds his own self ("soul") in the other's three-dimensional image. The
network of analogy unfolds as a chain of equivalences, moving from the
644
PoeticsToday16:4
human to the inanimate-sculptural to the psychic domain. First the girl,
"her blue frock caught back by the cross-bar of the window," is cast as a
wooden effigy, typically blue and also undetailed, owing to the tautness
of her frock. Then the model of a stylized wooden effigy reveals itself
as a figure for the immaterial self. From the viewpoint of a boy who has
decided to run away to sea, the figurehead analogy is a way of giving a
physical, tangible existence to a yearning. At the same time, it also reveals Peter's love for the girl. Whether she, who at once betrays his secret
dream to the grown-ups, deserves or repays this love is another matter,
which the narrative will soon bring to the fore. At the moment, "seen"
through Peter's imaging eyes, Rosa is the anima to his masculinity; the
spirit (placed, literally and figuratively, high above him) to be admired;
the realization of the moving power that impels him to abandon the
confines of his present existence for the ocean's endless space. No wonder, then, that on their next meeting, "the sea had become [for Peter] a
female deity, and Rosa herself as powerful, foamy, salt and universal as
the sea" (WT, 267). He can hardly distinguish between the two.
The figurehead simile having done its work, a genuine iconic counterpart follows to complicate and complete the effect. Peter tells Rosa
a story "of a skipper who named his ship after his wife. He had the
figurehead of it beautifully carved, just like her" (WT, 274). This does
not simply double or even merely realize the image, but animates it,
too. Hitherto leading a figurative existence, the wooden figurehead now
transforms into the humanlike third side of a love triangle, since the
wife mistakes the iconicity ("just like her") for substitution, as if the icon
really replaced her in her husband's heart. This counterreading of the art
sign, performed against its visible iconicity and its owner's intent, thereupon leads to a countermovement that annihilates the entire triangle,
appropriately by visual violence. Stealing the figure's "eyes" (a couple
of precious blue stones that echo Rosa's "blue frock"), the jealous wife
blinds the ship, which sinks with her husband, while she herself literally
turns blind. Misunderstood, the iconic relation proves destructive within
the inset tale. In the frame, however, as Peter recounts the wife'sjealousy,
treachery, and final punishment, Rosa (herself figurehead and sea rolled
into one) perceives with horror the analogy between her own betrayal of
Peter's secret plan to escape and the wife's betrayal of the skipper. This
insight brings about her decision to die with Peter rather than let him
live and discover her guilt.
Here, then, the figurehead is first evoked as a pure visual metaphor by
a hero who, like Axel or Childerique, consciously draws the comparison.
But he remains unconscious of its effect on his emotions, on his actions
(confiding his secret to the girl, telling her about her female analogue),
and, through her reaction, on his fate. In the process, the figurehead
Yacobi * PictorialModelsand NarrativeEkphrasis
645
simile assumes real existence, agenthood, even pathos within the inset
tale (the fiction within the fiction), leading to self-discovery and catastrophe in the frame as well. From the reader's vantage point, the twofold
or two-level role of this "moving object" radicalizes the actional force
exerted by similarity links in Dinesen's universe. Where does (visual) art
end and (story) life begin? All this gives a new twist to the relations between the dynamics of plot and the statics of ekphrastic objects within
the narrative arena: the two interpenetrate as never before.
The fact, mentioned earlier, that of all human traits this interpenetration excludes voice-the art objects always keep silent, along with
the rest of their order-may therefore be duly appreciated now as well.
This conspicuous absence pulls the artworksall the more forcefully apart
from the immemorial line of prosopopeia, inside ekphrasis and outside
it, as voice transfer to the domain of muteness: they are figures in life,
not figures of literary speech. Here again, Dinesen's work reads like a
silent commentary on ekphrastic practice and theory, framed within the
overall realm of mimesis.
One way of refining the difference would be to consider her writing
against the background of the "Moving Statue" tradition recently explored by Kenneth Gross. For now, let me quote his reference to two
cases in point: "Something in their tropes is worth pausing over, for each
struggles to find a voice and a mode of consciousness for the statue that
is fitted to its peculiar history, ontology, and use as a statue, rather than
simply as the image of a human being. The statue speaks but it speaks
(impossibly) as a statue" (Gross 1992: 145). Dinesen reverses matters
almost point by point, from the troping to the utterance to the statuelikeness multiply divorced from humanlikeness. In terms of origins, she thus
rejects both of the etymologies from which critics have often derived,
selectively orjointly, their privileged features of ekphrasis: (mimetic) descriptiveness and/or (fantastic, if figurative) voice-lending. My argument
about the variability of those features goes to prove her, in systematic
reversal, the better theorist, the all-around counterbalancer.
6. FalseSexualArrest
Let me end by giving the argument a socioartistic turn. Unhappily, the
age-old critical bent for imprisoning the protean dynamism of literature's forms in preconceived formulas shows signs of extending its jurisdiction from the representation to the representer. Both lines of arrest,
I would therefore point out, are countered by the above analysis. It
not only illustrates the rich play among dimensions of genre (ekphrasis/poetry/narrative), structure (space, time, space-time) and interart
mimesis (text/image/model). As the work of a woman, Dinesen's art also
questions the more recent, fashionable drive to literary fixture: that of
646
PoeticsToday16:4
marrying theoretical and sexual categories. Thus Mitchell's (1989: 97)
bid for linking ekphrastic (and related) poetry to sex, within what he
calls the "politics"of literary space:
Whathappens,we mustask,when the strategiesof literaryspacingare appropriatedby a femalewriter?I suggesttwo hypotheses:that the genre in which
womenconquerliteraryspace is principallythe novel, and that this conquest
is manifestedin the frequentoccurrenceof the heroine or female narratoras
subject,
painteror keen-sightedviewer,a "seeing"as opposed to a "speaking"
a dwellerin space ratherthan in time.
This is well intentioned, no doubt, yet it reduces women ("female writer"
of literature or "heroine or female narrator as painter or keen-sighted
viewer" within literature) to one-dimensionality, reminiscent of that attributed by a philosopher of yesterday to "Man."Even beyond the dynamization of ekphrastic images, neither hypothesis, about the female's
novelistic writing or about her space-dwelling, applies to our corpus
(or to others like it). A full demonstration of how and why they break
down would involve rehearsing much of my argument in "Plots of Space"
(Yacobi 1991). Here I can only touch on some relevant facts.
For one thing, Dinesen rejected the novelistic genre in favor of the
"story,"the "sublime"or "divine art";precisely because, among other reasons, she would restore the primacy of the "speaking"subject, the telling
voice. "Stories have been told as long as speech has existed" (LT, 23ff.).
With prosopopeia categorically ruled out, moreover, this primacy also
takes on existential significance. Far from dividing women from men,
"speech" unifies humanity vis-a-vis the nonhuman order, speechless at
its most humanlike.
For another thing, in the world of Dinesen (though herself a painter)
the roles of pictorial artists are reserved for men, while women prominently figure among her archetypal, as well as occasional, storytellers.
They range from her ArabianNightsancestress, the much-admired Scheherazade, to the old woman who makes a parable of storytelling in "The
Blank Page." No wonder, then, that Dinesen's practice of ekphrasis follows her general poetic rule. The tale would not "conquer" space but
bring it to life in new ways; it aims not to deliver the heroines (any more
than the heroes) from the prison house of their sex-much less at the
cost of reimprisoning them forever elsewhere -but to deliver the world
of coexistent entities from stasis. In ideological as well as in artistic terms,
this looks like the more liberated and liberating strategy. The idea of
segregating the "literary"sexes proves false, at any rate, another myth of
ekphrasis. But if we must bring sexual variables into the field of poetic
representation, the choice is clear. Either we keep our correlations flexible and empirical -so as to accommodate the spectrum of arts wielded
by "female writers," among the rest-or we consider Isak Dinesen, if
Yacobi * PictorialModelsand NarrativeEkphrasis
647
not a freak of nature, then perhaps a male alter ego of Baroness Karen
Blixen.
References
Abrams, M. H.
1953 The Mirror and the Lamp (New York: Oxford University Press).
Alpers, Svetlana
1960 "Ekphrasis and Aesthetic Attitudes in Vasari's Lives,"Journal of the Warburgand
CourtlandInstitutes 23: 190-215.
Alpers, Svetlana, and Paul Alpers
1972 "Ut Pictura Noesis? Criticism in Literary Studies and Art History," New Literary
History 3: 437-58.
Ashbery, John
1986 SelectedPoems (Harmondsworth: Penguin).
Auden, W. H.
1976 CollectedPoems, edited by Edward Mendelson (New York: Random House).
Auerbach, Erich
1973 [1946] Mimesis: The Representationof Reality in WesternLiterature,translated by
Willard R. Trask (Princeton: Princeton University Press).
Barolsky, Paul
1991 WhyMona Lisa Smiles and Other Tales by Vasari (University Park: Pennsylvania
State University Press).
1992 Giotto'sFatherand theFamilyof Vasari'sLives (University Park: Pennsylvania State
University Press).
Cliver, Claus
1978 "Painting into Poetry," Yearbookof Comparativeand GeneralLiterature27: 19-34.
1989 "On Intersemiotic Transposition," Poetics Today10: 55-90.
Davidson, Michael
1983 "Ekphrasis and the Postmodern Painter Poem," Journal of Aestheticsand Art
Criticism42: 69-79.
Dinesen, Isak
1934 Seven GothicTales(New York: Smith and Haas).
1942 WinterTales(New York: Random House).
1957 Last Tales(New York: Random House).
1963 Ehrengard(New York: Random House).
Eliot, George
1956 [1859] Adam Bede (New York: Pocket Library).
Genette, Gerard
1982 "Frontiers of Narrative," in Figures of Literary Discourse, translated by Alan
Sheridan, 127-44 (New York: Columbia University Press).
Gross, Kenneth
1992 TheDream of the Moving Statue (Ithaca: Cornell University Press).
Hagstrum, Jean H.
1958 The SisterArts (Chicago: University of Chicago Press).
Hamon, Philippe
1981 "Rhetorical Status of the Descriptive," YaleFrenchStudies 61: 1-26.
Hannah, Donald
1971 "IsakDinesen" and Karen Blixen: The Mask and the Reality (London: Putnam
and Co.).
Heffernan, James A. W.
1991 "Ekphrasis and Representation," New LiteraryHistory 22: 297-316.
648
PoeticsToday16:4
Hollander, John
1985 "Introduction," in Rossetti 1985: 3-16.
1988a "The Poetics of Ekphrasis,"Wordand Image 4: 209-19.
1988b Melodious Guile: Fictive Pattern in Poetic Language (New Haven: Yale University Press).
Keats, John
1953 [1819] The LettersofJohn Keats, edited by Hyder Edward Rollins (Cambridge,
MA.: Harvard University Press).
Kibedi Varga, A.
1989 "Criteria for Describing Word-and-Image Relations," Poetics Today10:31-53.
Kranz, Gisbert
1973 Das Bildgedicht in Europa: Zur Theorieund Geschichteeiner literarischenGattung
(Paderborn: Schoningh).
1981 Das Bildgedicht:Theorie,Lexikon, Bibliographie,3 vols. Literatur und Leben, n.5.,
23 (Cologne: Bohlau).
Krieger, Murray
1968 "Ekphrasisand the Still Moment of Poetry; or, Laokoin Revisited," in Perspectives
on Poetry, edited by James L. Calderwood and Harold E. Toliver, 323-48 (New
York: Oxford University Press).
1992 Ekphrasis: The Illusion of the Natural Sign (Baltimore: Johns Hopkins University Press).
Kurman, George
1974 "Ecphrasis in Epic Poetry," ComparativeLiterature26: 1-13.
Langbaum, Robert
1965 The Gayetyof Vision:A Study of Isak Dinesen's Art (New York: Random House).
Lessing, Gotthold Ephraim
1963 [1766] Laoco6n: An Essay upon the Limits of Painting and Poetry, translated by
Ellen Frothingham (New York: Noonday).
Lund, Hans
1992 Text as Picture: Studies in the Literary Transformationsof Pictures, translated by
Kacke Gotrick (Lewiston, NY: Edwin Mellen).
Mandelker, Amy
1991 "APainted Lady: Ekphrasisin Anna Karenina," ComparativeLiterature43: 1-19.
Marin, Louis
1970 "La description de l'image: A propos d'un paysage de Poussin," Communications 15: 186-209.
Mitchell,W.J. T.
1986 Iconology:Image, Text,Ideology(Chicago: University of Chicago Press).
1989 "Space, Ideology, and Literary Representation," Poetics Today10: 91-102.
Nabokov, Vladimir
1967 Nabokov'sDozen (Harmondsworth: Penguin).
Pagis, Dan
1981 Points of Departure,translated by Stephen Mitchell (Philadelphia: Jewish Publication Society of America).
Panofsky, Erwin
1982 Meaning in the Visual Arts (Chicago: University of Chicago Press).
Riffaterre, Michael
1990 Fictional Truth (Baltimore: Johns Hopkins University Press).
Rossetti, Dante Gabriel
1985 TheEssential Rossetti(New York: Ecco).
Scholes, Robert
1974 Structuralismin Literature(New Haven: Yale University Press).
Yacobi * PictorialModelsand NarrativeEkphrasis
649
SeznecJean
1972 "Art and Literature: A Plea for Humility," New LiteraryHistory 3: 569-74.
Spitzer, Leo
1962 [1955] "The 'Ode on a Grecian Urn'; or, Content vs. Metagrammar," in Essays
on English and American Literature, edited by Anna Hatcher, 67-97 (Princeton:
Princeton University Press).
Steiner, Wendy
1982 The Colors of Rhetoric: Problems in the Relation betweenModern Literature and
Painting (Chicago: University of Chicago Press).
1989 "The Causes of Effect: Edith Wharton and the Economics of Ekphrasis,"
Poetics Today10: 279-97.
Sternberg, Meir
1978 Expositional Modes and TemporalOrderingin Fiction (Baltimore: Johns Hopkins
University Press).
1981 "Ordering the Unordered: Time, Space, and Descriptive Coherence," Yale
French Studies 61: 60-88.
1982 "Proteus in Quotation-Land: Mimesis and the Forms of Reported Discourse,"
Poetics Today3(2): 107-56.
1985 The Poetics of Biblical Narrative: Ideological Literature and the Drama of Reading
(Bloomington: Indiana University Press).
1990 "Time and Reader," in The Uses of Adversity:Failure and Accommodation in
ReaderResponse,edited by Ellen Spolsky, 49-89 (Lewisburg, PA: Bucknell University Press).
1992 "Telling in Time (II): Chronology, Teleology, Narrativity," Poetics Today 13:
463-541.
Stevenson, Robert Louis
1952 "The Sire de Maletroits Door," in The GreatShort Stories,199-220 (New York:
Pocket Books).
Williams, William Carlos
1988 The CollectedPoems,Vol. 2, edited by Christopher MacGowan (New York: New
Directions).
Yacobi, Tamar
1976 "The Manipulation of Time as an Anti-Realistic Factor in the Poetry of Dan
Pagis," Hasifrut 22: 18-37.
1988 "Time Denatured into Meaning: New Worlds and Renewed Themes in the
Poetry of Dan Pagis," Style22: 93-115.
1989 "Space as Interspace: Multiple Sign-Systems in the Poetics of Isak Dinesen."
Paper presented at the IV IASS Congress, Barcelona-Perpignan. Abridged version in Signs of Humanity, edited by Michel Balat and Janice Deledalle-Rhodes,
2: 897-902 (The Hague: Mouton de Gruyter, 1992).
1990a "Dimensions of Space: Isak Dinesen," in Proceedings of the Twelfth Congress
of the International ComparativeLiteratureAssociation, edited by Roger Bauer and
Douwe Fokkema, 3: 79-85 (Munich: Iudicium).
1990b "Space as a Figure of Time: Direction Locatives and World Pictures in Pagis's
Poetry," HebrewLinguistics 28-30: 123-36.
1991 "Plots of Space: World and Story in Isak Dinesen," Poetics Today12: 447-93.