182

Micke lial

The collection grew oat of the conviction that the relations between the arts
have something more to offer than the study of each art can offer alone.
Mieke Bal
Rochester and Utrecht, January 1988
t

i^u\
^ &8;

1
$5-1^5

zt

N o rm a n B r y so n

Intertextualily and Visual Poetics*

The phrase Visual Poetics contains a promi.sc: that between poetry and
the visual there is a kinship or affiliation which allows us to cross from one
domain to the other with some kind of ease or sense of natural right of way.
1 have no intention of making the passage from one to the other seem im
possible or unduly difficult, but there are conceptual difficulties in that passage
of which I think one should be aware, I am afraid I will be pointing to more
difficulties than I can resolve, and my aim is rather to raise some questions
than supply answers, I should add that I am restricting my discussion of the
image to representational painting to stop matters from getting completely out
of hand.
One might begin by invoking a classical account of the relation between
poetry and the visual, which, if it were correct, might completely block inter
flow between them: Lessings Laokoon. Lessing argues that painting and poetry
employ different and incompatible systems of signs, the signs of painting being
simultaneous or synchronic, the signs of poetry being diachronic or distributed
in time. An image has but one moment to release its charge; a text, by
contrast, is a progress from beginning to middle to end. Lessing argues that
poetry, being a consecutive or serial form, is singularly unable to deal with
the visual sphere. The more a poetic text tries to describe the visual or to
generate an image, the more it displays its constitutional inability to achieve
the simultaneity which is the images essence. If poetrys visual description is
short, it will lack the detail which characterizes visual experience. If poetrys
visual description tries to pack detail in, the more it adds, the more the pos
sibility of achieving simultaneity recedes. According to Lessing, poetry7should
therefore avoid the visual altogether and leave it to the painters to deal with.
Conversely, painting, which employs static and unmoving forms, is singularly
unable to deal with narrative! the more an image tries to tell a story with
beginning, middle, and end, or with episodes arranged in a series, the less
coherent it becomes. Painters should not try to do the job of poets. At best
the painter can select the central moment which sums up a story' and implies
its beginning and its end; yet the story must be implied, not stated, since it is
not possible for a static or unmoving entity to unfold in time, as all narratives
must.
* T his essay originally appeared in Critical TV.v/.v, volum e
reprinted by permission.

Style: Volume 22, No. 2. Summer 1988

num ber 2 ( HH7). pages-1-6, and is

183

184

Norman Bryson

Let us take from Lessing this notion of the image as atemporal, or allai-once. Obviously it corresponds to a certain truth of ordinary experience.
Paintings stay still; they dont move around the walls; their figures dont slip
out of the frame and gambol about when no one is looking. Sometimes a cult
of ideal form aspires to a shedding o f times contingencies, so that the forms
we see in painting are taken out-of the temporal flux. Or painting may eschew
ideal form and instead set dolvn an unidealized and transient moment. If we
think of the image in terms o f abstraction, and identify painting with the forms
held on the picture plane as surface flatness, time seems notably absent: no
reference to a temporal world or to narrative sequence complicates the pre
vailing stasis. Idealization, realism, and abstraction concur in the timelcssness
of images. Such timelessness allows one to think of a painting as a substance
in Aristotles sense: as an entity having (i) stable location in a single place; (ii)
independent self-existence, requiring nothing but itself in order to exist; (iii)
permanent or enduring form. A metaphysics of substance seems built into the
format of Western painting, into the picture frame. In a sense we can dispense
with frames and regard them as extrinsic to painting; yet even without its
actual frame Western painting is a structure o f framing, and within the frame
substance is held in a state purer than substances in nature. In nature substances
may move, unfold, blend, dissolve, but in the frame substance is held and
displayed in Aristotelian purity: as requiring nothing elseno other paintingin order to exist; as independently self-existent, in a single place, and in per
manent essence.
The first question I want to raise is: what kind of entity is the image, and
if it is not in fact a substance, what is it? Thinking of the image as substance
is obviously untrue to certain aspects of ordinary experience. Images do change:
after five seconds some are completely dead; others get better after the first
ten minutes. Images can also change through an individuals life, and arc one
thing in one decade, something else a decade later. The substance-view of the
image starts to founder the mom ent we shift the point of view of the discussion
to the spectator. In actual experience, looking at an image is a radically tem
poral process, which changes from moment to moment. If we think of the
saccadic movements of the eye, what vision experiences is an image distributed
across discontinuous leaps. Each act of looking attends to a different area of
the image and discloses a partial view, as vision transits through the image in
endless stops and starts. Each view finds a different perch or purchase on the
image, and the successive views are strung together serially, in a flow of time.
One might want to say that gradually these partial views are assembled, like
the pieces of a jigsaw, into a complete image, a total state, which gathers
together and supersedes the work of assemblage. Yet we do not ordinarily
experience vision as ever achieving that totality or saturation. If we could see
that way, then looking at an image would reach a distinct terminus; but vision
seems interminable. It is such that until the moment we drop wc have to go

Imertextuality and Visual Poetics

185

oil looking; we cannot make vision stand still or repeat itself perfectly. The
state of totality is never achieved, is always postponed, is always ahead. Nor
is the idea of accumulating vision quite true to experience: when we look at
a painting we arc constantly revising our previous apprehensions, erasing or
regrouping our findings.
Let us frecze-frame an individual moment of looking and for convenience
call it a glance. The temporality of the glance is, in actual viewing, both re
tentive and protentivc: it brings with it the wake of previous views and even
as it occurs is fell as leading on to the next view. It is an occasion of experience
whose being is its becoming, which perishes immediately on attainment of
completion. The glance is an entity hard to map according to Aristotelian
substance. Since its constitution retains the past and is oriented to the future
it tacks the all-al-onceness of substance. Rather, it is a structure of interpe
netration; past and future interpenetrate it without obstruction.
The interpenetration runs, that is to say, both backwards and forwards.
When wc recognize a Pieta or a Nativity, recognition proceeds by relating the
present occasion back to past occasions when we have seen Pietas and Nativ
ities, back to their monographic conventions. Iconography provides the clearest
example of the retentive structure of the glance, yet it is not only iconographic
forms but all forms that are recognized in this way. And this is true from the
start: even the first form we rectognize is one which is related to a previous
case, and in the logic of inauguiation at work here even the first instance is
already interpenetrated by the past. The glance is a structure without origi
nation. The forward" or protentivc structure of the glance is a function of
the endless mobility or insatiability of vision. When we look at an image, wc
cannot stop our looking: looking goes on, nothing can arrest it; it seeks always
the next view and the next image, and this seeking is built into, is the foun
dation for, each individual glance in the very moment that it occurs.
Now consider what kind of image stands before, or is proposed and
assumed by, the glance. It, too, is structured according to absence of origin,
retentive-protentive temporality, and interpenetration. In a classical concep
tion of the image such as Lessings, the image inhabits a stable location in a
single place; it is a permanent or synchronic form. In the alternative view of
the image which 1 am briefly exploring here, each o f those classical fixities is
unsettled. The image is found to lack all-at-onceness, independent self-exist
ence, permanent form. We can analyze this post-classical or at least other
conception of the image either macrosco pi tally, and consider the individual
image in relation to other images, or microscopically and consider the in
dividual image by itself (though what by itself might mean is exactly the
issue).
In the macroscopic view, the image contains within itself, and is inter
penetrated without obstruction by, the whole suit of prior images, as well as
by the further, as vet empty set of future images, all of which co-exist with or

186

Norman Bryson

inhabit the image in a continuous or framelcss field. In terms of the tradition

of painting, interpenetration by the past corresponds to the massive structure
of aitusion or reference a painting makes, willed or unwilled, to the tradition
which houses it and which it also continues. Again in terms of the tradition
of painting, interpenetration by the future corresponds to the perception of
painting as unfolding in a tradition which runs beyond itself and into future
work: it corresponds, that is, jo the perception of art as embodying a tradition
structured by forward direction rather than mere haphazard accumulation.
Microscopically, in the individual act of looking or glance, interpenetration
by the past corresponds to the repertoire of images previously encountered by
the viewer and now brought to bear upon a particular image in the work of
recognition; while interpenetration by the future corresponds to the awareness,
in the present of viewing, that there is no terminus or saturation point for
viewing. Each moment of the glance is inhabited by the suite of previous acts
of attention which the viewer has brought to bear upon the particular image,
and is inhabited also by a proleptic awareness of the nerd glance which is to
follow and supersede the present one in an endless track of deferral or desire.
What 1 am describing has both a visionary and a common or garden
aspect. Nothing could be more llat than to observe that when we look at
paintings, we typically think of their relation to tradition, or that no sooner
have you enjoyed a particular image or view tliftn you want another one. Yet
the dissolution of the Aristotelian or substance view of images brings with it
some major transformations. Let me prolong this more visionary aspect by
invoking some similes, which maythis is their intention, at any rateconvey
a visceral sense of the changeover from the idea of image according to substance
to the post-classical idea of the image according to unimpeded interpenetration.
Think of the whole span of tradition, both past and future, as rounded : then
each image, interpenetrated by the images of past and future, reflects all other
images from its own standpoint, so that the splendor of tradition is multiplied
ad infinitum in a panoramic spectacle of simultaneous mutual reflection not
unlike Leibnizs monads or perspectival mirrors. Within the boundary of the
individual image, each glance of the viewer across its surface reflects the sum
of previous glances upon the present image, as well as all the images to come;
the image before each glance has the structure of a hologram. Once we shed
the substance-view, we find that the image is subject to wholly different prin
ciples of being. Since it does not disclose itself all at once and does, not stand
in self-existence, it emerges together with the surrounding field of images through
mutual arising. Interpenetrated by past and future images, its frame is dis
solved and crossed through principles of mutual entering, mutual reflection,
mutual containment.
I mentioned earlier that if Lessings idea of the synchronic or timeless
image and the diachronic or temporally unfolding text were correct, it would
be hard to move between the domains of the visual and the literary, as the

Intertextuality and Visual Poetics

187

phrase Visual Poetics promises we might do. The shedding of Lessings

substance-view of the image moves the image into close proximity to the fields
of literature and language. The image regains the fluid temporality denied it
in the substance-view. The idea of the text as an interpenetrative structure is,
of course, important in literary theory in the work of Barthes and Derrida. I
would say that Barthes established the intertextuaf text in two moves. The first
involves the severance of the text from reference to a real world beyond itself'.
Realism is only a rhetorical effect, the result of a text's internal organization,
and what seems like realism is only the repetition within the text of the repeated
discursive forms which in Barthess view construct reality. This first move,
the separation of text from reference to a real world, is completed in the second
move, which disconnects the text from authorial intention. In The Death of
the Author, Barthes wrote that a text is not a line of words releasing a single
theological meaning (the message of the Author-God) but a multi-dimen
sional space in which a variety of texts, none of them original, blend and
dash. Barthess idea of the text is such that, disconnected from authorial
intention on the one hand and from reference to a world beyond the text on
the other, it becomes an intertextual field where no discursive fragment is
bound to any origin whatever. In the field of the text, fragments of other texts
blend, clash, collide; each text is the reflection of other texts in a swirl of mobile
intertextualities. To think of paintings as mutually interpenetrating is to dis
cover in the realm of the image the same phenomenon of mobile intertextuality
made familiar to us by Barthes and Derrida in the field of literary criticism.
The logistics are indeed similar in both domains.
In the case of the image, intertextuality is established by dissolving the
frame around the work. In the case of the text, interpenetration is established
by annulling the point of origin of the work, whether this point is the world
the text refers to or an author in control of its meaning. The phrase Visual
Poetics promises a kinship or affiliation between the literary and visual which
allows one to cross from one domain to the other without a massive change
of gear. In the parallel ideas of the interpenetrative image and the intertextual
text, we seem to have a current mutual ground which criticism might profitably
work with. It seems therefore worth raising at the present time the question:
is there such a thing as a general category of intertextuality or interpenetration
which might be applied in the same way to texts as subject to similar structures
of dissemination and flow? Or are there specificities which apply in the case
of the text but not in the case of the image?
* * *
I would like to keep these questions open; but I would also like to suggest
that the text-flow or intertextuality functions rather differently from the imageflow or image interpenetration. 1 can sketch the difference as follows. In the
flow of textual dissemination described by Derrida, texts flow so fast because

188

Norman Bryson

they are pure information, and information is weightless. To produce meaning

at all the verbal signifier has to travel to another signifier: its essence is move
ment. Moreover, the text is a multiple object; it exists simultaneously in many
places. The life of texts can be easily thought of as inherently mobile, dissem
inating, travelling out in man; directions and colliding, blending, clashing with
other texts similarly mobile, rapid, weightless. For Barthes, the text is just the
clearing of a space in which (we can contemplate aesthetically the general flow
and interflow of signs as these ripple throughout the social organism. Nothing
holds these signs down. They cannot be fixed by authorial intention, since the
multiple text at once flies from the pen of the singular author to a thousand
places and a thousand clashes with other texts. They cannot be held in place
against a world beyond signs, since much of the worldat the very leastis
itself made of signs and so cannot provide anchorage. According to Barthes,
in the most energetic literary texts one finds that the flow of discourse is made
even more volatile:.in Flaubert, Mallarme, or Sollers, the normal forms of
discourse are scrambled, turned and overturned, and the text of jouissance or
bliss bursts out in the splendor o f permanent revolution. According to Der
rida, the disseminating energy of language cannot be contained except when
repressive agencies of authority try to stabilize the flow by holding texts to
fixed interpretations and holding words to the false stabilities of reference.
Dissemination is boundless. Though books pretend to frame themselves with
prefaces and titles, and such things as covers, they are actually part of a seamless
flow of information travelling in all directions and across every individual.
The flow of the verbal signifier may thus be sheer and unimpeded in
terpenetration; but do paintings flow like this? The verbal sign, according to
the theory of dissemination, is.weightless because language is pure information,
and information flies. It has no body to tie it down; its structure is difference,
and difference has no physical dimension. Is painting like this? Of course
painting possesses a discursive aspect. Its narrative dimension, its legible struc
tures, its iconography, its denotations and connotations arc all discursive, and
can also be considered as information. Everything in painting that is like text
can partake of the structures of dissemination and interpenetration; everything
one can read in a painting potentially belongs to the universe of information
and its disseminating flows. But painting may also possess aspects irreducible
to information.
Take the always intractable case: color. Color can convey information;
and if we alter the colors, we alter the information. If Guernica is painted pink,
its vision o f the world may be rosy. But the experience o f color involves an
excess beyond the function o f information, and as color exceeds the power of
discourse to claim it for textuaiity, it touches on a realm o f bodily experience
which diverges radically from the realm o f information. Color can be organized
into information, as in color codes, but the experience o f color stands outside
the code.

Intertcxtuality and Visual Poetics

189

Or think of Cezanne. In a Cezanne landscape, the trace is by no means

identical to information, since Cezannes method bathes each trace in Die luster
of embodied consciousness. The traces do not only exist in the spam' of the
picture plane, but in the time of their application to its surface. The separability
and visibility of the trace correspond to the slowest and most ruminative of
working methods: each implies time prolonged, tune distended. It is important
that they are landscapes of particular places and not the generalized landscapes
of Poussinwe have to take iL as a dam ice that behind the landscape stands
an actual scene. Each trace is compared with its counterpart in the scene,
though not in order to reduce the scene to simultaneous information: Cezanne's
operation is the record of an embodied consciousness moving between land
scape and representation as between distinct but related registers, whose re
lation is precisely the movement of reflexive consciousness between them.
Even the word image may suggest something too simultaneous here: rather,
the canvas collects, gathers into itself moments of duration in which con
sciousness moves among perceptions and among signs.
Such an image is not at all like a text, under dissemination, where texts
race weightless and decontextuaiized, without gravity or density, Ce/anne ex
actly anchors his signs in the actual experience of the body under time and in
a single place. Each trace draws the image deeper into the anchorage of a
consciousness dwelling reilexively between the distance of the landscape and
the proximity of the canvas, and between orders of perception and represen
tation.
Painting thought of in this way is intimately linked to "embodiment,"
yet embodiment has to be one of the most problematic concepts which
painting presents. For the rest of this essay I will be trying to suggest ways of
thinking about embodiment which may free it from the reactionary and un
theorized context in which it remains embedded. My aim is to place this
concept squarely within historical materialist thought and exactly to separate
it from any connotations of the oceanic, of pure Being, of the ground of Being,
of sheer presence, of kingdom, of foundation. I can name three ways of thinking
about embodiment as a particular case of the body, which link it inexorably
with reaction and depoliticization: the body as (i) a pre-social and biological
given, a substance lying undisturbed beneath the socialization of the subject
in culture and outside historical and*social processes; (ii) as unitary; (iii) as a
source o f mystical value. Converging on painting, these notions would claim
any embodiment painting may possess (particularly as this emerges in contrast
to the text) as..the expression of pure Being and of the subject undivided by
the sign; it would not be long before embodiment was claimed as a source
of value and painting as the art which stores ihat value, with the value of
embodiment presented as a haven for troubled and divided subjectivity. When
the idea of embodiment arises, a siren song is at once heard; the body as the
primitive ground of being (in Merleau-Ponly), as the place of blissful, maternal

190

Norman Bry son

refuge from the paternal, coercive empire of signs (in Barthes), as an intrin
sically theological concept (in Heidegger). These are the positions which, for
discussion to advance, must be refused.
To think of the*body as an arena of encukuration rather than as the
substrate to onculuiration is, I would say, easier since Foucault, and harder
since Derrida. If our understanding of signs is thoroughly materialist, then the
material aspect of painting is itaturally thought of as continuous with its status
as sign: since signs are material work, no contradiction arises. The problem
with differun.v is that, by identifying meaning as movement between significrs.
as their mobile inter-relation, the sign is in some sense ^materialized. If the
relation which one signifier has to the rest is posited as synchronicaliy and
diachronically endless, then any local and limited operation of the sign (so
that it acts as x -> y. rather than x n) must seem a descent into contest, a
more or Jess repressive constriction oflheinftrmist expansion ofsigns, a capture
of the sign by matter. A dualism is insrituted between the local and material
capture of the sign by context and the signs inherent expansiveness (such
dualism can be criticized as the unexorcized classicism of Derridas workl. In
terms of painting, this model produces a two-tier account: painting as infor
mation, painting as mute matter. As information, painting can then be thought
to be an infinitely intertcxtual structure, which, overturning the repressive
agency of the frame, Hies out to meet all other paintings under conditions of
mutual dependence and mutual arising. As matter, painting can by the same
token be thought of as the absence of information and as sheer figurality:
as the inert pigment on canvas. One interesting consequence of the extreme
claim for the massive imertextuality of images is that an obverse picture also
presents itself: painting as an art of the material substrate, of pigment below
a threshold o f information, as pre-semiotic or semantic. Such a position ac
tually concurs with the dcpoliticized and reactionary conception of embod
iment in painting as the realm below sign activity and cultural work. All that
is lacking is a rhetoric which then celebrates that substrate as a primitive
expression o f Beingall that is missing is the lyricism of the Siren Song.
One of the most useful texts arguing against such dualism must be Pierre
Bourdieus Outline o f A Theory o f Practice (Cambridge University Press, 1976),
a work which has never, 1 think, been adequately recognized (and has been
eclipsed, like Lyotards Discours, Figure, by later work and later transforma
tions of its author). Bourdieus concept of practice is dose to what is meant
here by embodiment: work with material signs which takes place always (i) in
an actual social and economic arena, and (ii) under conditions of real time.
Bourdieus insistence is that when we look at sign activity we resolutely avoid
reducing it to information, and that we take into particular account its structure
of local and material negotiation, Bourdieus book argues, by example and is
dense: all I can indicate here are some schematic cases. In the case of arranged
marriages (the context is anthropology), while the ceremony of marriage can

Intertextuality and Visual Poetics

191

be viewed as symbolic theater, negotiation towards marriage involves a tem

porality of bargaining in which neither side yet knows the outcome of any of
its moves. The meaning of the moves in the negotiation is determined ret
roactively by the interpretation put on the move hv the other sidethe meaning
does not yetiexist as the move is made. Practice here, as a concept, posits a
temporality o f rea] social lime" and a zone "to one side of meaning,7' where
meaning is being made blit also has not yet arrived. Other examples include
feuding and combat, where each move both counters a previous action and
initiates a further action, whose outcome is unknown; process and performance
arts, where the material dimension itself acts upon each move to influence
outcome; calendrical symbolism, where months are described, not according
to a zodiac or a static order of the whole year, but in relation to the month
before and the month to come.
The Cezanne example possesses embodiment in many of Bourdicu's sen
ses. Cezannes hundredth stroke responds to the previous ninety-nine, and
launches a further move, whose outcome is not yet known; the work acts
recursively and reflexively upon itself in real time, through the agency of its
embodied creator. This real time" of process also occurs within a real social
time : the image reacts to prevailing discursive formations of painting, which
it mutually modifies: it is also placeu :nto a circuit of dealer-merchant trans
actions, where its pricing modifies the field of commodities with unknown
outcome, Embodiment in these cases is a descriptive, not a valorizing term.
Yet our neutral description of embodiment, which docs nothing to praise
embodiment or go into raptures over it. should also include a recognition that,
with actual practical systems, embodiment typically operates in relation to
value. For example: if the potter lacks dexterity, the clay collapses; if the
combatant lacks coordination, he loses: if the negotiator misjudges, the trans
action misfires. Embodiment here is exactly not given by biology or by precultural Being; it is something to be acquired as cultural skill. It is not at all
a quality of the primitive body, as in the phenomenology of Merleau-Ponty,
but of the body cultivated. It is acquired by training in cultural and semiotic
work. A mystical inclination of the argument tends to seize on those aspects
of embodiment which arc outside skill: color is such a potent case for those
who want embodiment to disclose pure Being exactly because color is one
thing the painter does not make (God makes colors). But even in this case,
color is never the idea o f color, even in color-field paintings where color seems
to attain purification and essentiality: color exists only in instances of colors
(plural), individual cases, and the example of color-field painting ought to
indicate, if anything, the high degree of cnculturation involved in working with
colors, the delicacy and skill with which the painter manipulates the cultural
codes and conventions of color and the precision of his task: not some mys
ticism of the azure.

19

Norman Bryson

Perhaps a clearer poini of discussion is provided by calligraphy in the

far Past, Embodiment is calligraphys goal, not its starting' point; the whole
body must be massively enculturated for embodiment to be embodied; and
though calligraphers regard successful work as possessing extreme value, it is
value at the opposite end from a primitive substrate of bodily experience. The
body has, precisely; been trained in material work, of the sign to the point at
which the gross organism ii fully absorbed, leaving little or nothing of the
primitive behind.
Calligraphy is a particularly good counter-example to the "mystical case
of color, since calligraphy turns on the entry of the body into the signs of
languagethe example shows beyond doubt that the subjectivity behind cal
ligraphy has been fully inserted into the symbolic order. We need not, therefore,
think of embodied subjectivity in any way other than as wholly ruled by the
decentering of the subject provoked by the entry into language when we try
to work out what the subject of embodiment is like.
The subject of calligraphy, though embodied, docs notas a result of
embodimentpossess any transcendent unity, which embodiment, in the dis
course of reaction, tends to assume as an automatic guarantee. What has been
said here about calligraphy can be returned to Western painting: here, also,
the painter's embodiment is the product and precondition of cultural skill j :
(he manipulation of material signs.
Such embodiment suggests limits to the concept of paintings intertextuality. I will conclude by stating a formal case, though my statement of it is
provisional and open to objections as a tweedy compromise. I offer it, con
scious of its shortcomings, in the hope that when its shortcomings arc seen,
discussion might move several steps forward:
Texts form a multi-dimensional space in which a variety of other texts,
none of them original, blend and clash. They arc intrinsically intencxtual; that
is, they coexist with other texts according to a logic of mutual origination,
mutual dependence, and mutual interpenetration. They absolutely disobey the
description of Aristotelian substance.
Texts arc like this because they possess no embodiment. They are sheer
information. Paintings are otherwise. Paintings possess embodiment, though
this is a concept absolutely to be purged of mysticism. Embodiment counters
intencxtualily as the unique counters the order of the same. Since the signs
of painting are never sheer information, but in their embodiment are part of
material practice, they are information thoroughly inhabited and interpene
trated by non-information: embodiment is a concept which deconstructs the
opposition between matter and information, upon which the concept of the
text" (as information outside material practice) rightly or wrongly has been
built. By virtue of their embodiment, paintings offer a resistance to intertextuality which texts do not. To this extent they do not behave like texts: they

Intertextuality and Visual Poetics

193

are not appropriately analyzed through models based on texts; they disobey
the formula ut piciura poesis\ and they frustrate the hopes of any easy Visual
Poetics.

-i-j:..;