©Association ofArt Historians 2010 11
Art History as Ekphrasis
Jas´ Elsner
The history of reflections about what art history is and how it might best be done,
about how art works and what art is, is long and distinguished. I do not propose
to review it here. Nor do I presume to suggest a definition of what art is or how it
may come to be. My concern is more limited. I want to make a proposal about the
nature of art history, which may seem outrageous to those who see themselves in
the tradition of the‘critical historians of art’, or to those who believe art history has
the elevated status of philosophy, or to those who think the discipline is ultimately a
branch of history more generally, but I will make it nonetheless. My proposal is very
simple. Far from being a rigorous pursuit, art history – certainly since its founding
fathers in the modern era,Vasari andWinckelmann, and undoubtedly in the surviving
ancient sources who were their inspiration (Pliny,Vitruvius, Lucian, Philostratus, even
Pausanias) – is nothing other than ekphrasis, or more precisely an extended argument
built on ekphrasis.That is, it represents the tendentious application of rhetorical
description to the work of art (or to several works or even to whole categories of
art) for the purpose of making an argument of some kind to suit the author’s prior
intent. Not everything that results from ekphrasis is art history, but that series of
uses of interpretative description, which attempt to make a coherent argument on
broadly historical or philosophical lines, is definitely art history.The particular rules
governing the making of description and its appropriations have changed radically
over the centuries from (say) Philostratus toVasari to Riegl toT. J. Clark – to suit the
particular intellectual contexts and social aspirations of those writing about art in
these different worlds.And the kinds of results or findings required by different
periods (and sometimes different scholars) from such description have also changed.
Some find that art can reveal artists, others that it can indicate social history or
underlying cultural reflexes, others still that it implies very little at all beyond itself.
But my proposition is that – whatever the particular agenda or argument – art history
is ultimately grounded in a method founded on and inextricable from the description
of objects. I hope this is not controversial.
A number of issues follow methodologically from placing description at the
centre of the enterprise. First, we must be clear about what we mean by description.
Here I take an extremely open-ended view: any account of an object from the
most hardcore formalist analysis (Riegl’s work is exemplary here, and that of his
followers in the SecondVienna School), to a floridly evocative description (what
Panofsky somewhat dismissively called attention to the sensuous aspects of art), to
a highly complex analysis of deeper meanings and symbolic networks (the highDetail
of Michelangelo,
Rondanani Pietà,c.1555–64
(plate 3).
DOI:
10.1111/j.1467-8365.2009.00720.x
Art History | ISSN 0141-6790
33 | 1 | February 2010 | pages 10-27
©Association ofArt Historians 2010 12
Art History as Ekphrasis
point ofWarburgianism), from the mere mention of an object to its dismissal,
from encomiastic praise to vituperative attack – all these and everything in between
constitute ekphrasis, and hence may make up the descriptive basis for the practice of
art history (see plate 1).The reason such accounts are ekphrasis, and hence the bedrock
of art history, is that all these descriptions conspire to translate the visual and sensual
nature of a work of art into a linguistic formulation capable of being voiced in a
discursive argument.The act of translation is central.We conduct it with such ease.
And yet the conceptual apparatus into which the object has been rendered, and its
transformation from a thing that signifies by volume, shape, visual resonance, texture
into one that speaks within the structures of grammar, language, verbal semiotics (call
it what you will) and can be appropriated to numerous kinds of argument or rhetoric,
are quite simply vast. In fact, they are so vast that the truly responsible viewer might
balk at the prospect of so falsifying the object by the act of its verbal rendition.
Or, as in Lucian’s brilliant dialogue‘On the Hall’ (De domo) might rise to the challenge
of creating through verbal artistry a description (descriptive fiction?) which at least
attempts to rival or to emulate the range of emotive, formal and textural resonances
evoked by the object described.
The enormity of the descriptive act cannot be exaggerated or overstated.
It constitutes a movement from art to text, from visual to verbal, that is inevitably a
betrayal. Not everything in the world of the sensual autonomy of the object can be
translated into words, and much that was not there is inevitably added by words.
In other words, description is not merely selective; it is (at its best) a parallel work
of art.To put this another way, however good the approximation in words of the
object described, it can never fully be or fully replace the object. Description may be
seen as a primary interpretative act (like an anthropologist’s account of a different
culture in which unavoidably prior viewpoints born from scholarly training or
cultural background cannot be wholly extricated from‘objective’ ethnography).
In this case, on the bedrock of a verbal interpretation, many further layers of
analysis – each more interpretative than the last, each more authoritative or
speculative than the last – will come to rest. But the difference from anthropology
is that, in most cases, works of art are not cultures or peoples for whom an
unrepeatable ethnography – whatever its partiality, whatever its weaknesses –
may have to stand. Rather, each of us can make our own primary description
(tendentious in that it exists to help me make my specific argument and tendentious
also in that it inevitably emerges within the range of other such descriptions,
differentiating itself rhetorically from them in order to make my point about
the object different and special, but thereby finding its voice not in the direct
inspiration of the non-verbal nature of the visual but rather in the history of other
verbal discourses already floating about the object or objects like it).Yet it is on
this foundation and no other that art history rests. So far as I know no rules have
ever been written for this act of translation and it has hardly been subjected to
analysis. It is the necessary and inevitable move before we start – the only way
we can have objects to discuss at all.Yet the questions it begs are endless and the
fundamental assumption that pictorial or architectural thought operates in parallel
ways to verbal thought, that verbal forms of signification are adequate in any way to
account for visual and material forms of signification, must remain no more than an
assumption taken unexamined as an axiom.
It might be objected that since the end of the nineteenth century description
has become more objective because it can be supplemented with photographic
reproduction – a visual (as opposed to verbal) rendering of the object by visual
©Association ofArt Historians 2010 13
Jas´ Elsner
means and a‘control’ against which the description may be read. If one followed
this line, one might grant the descriptive translation more authority than I have
done, less haphazard partiality and tendentiousness, and a greater objectivity (as
necessary to the foundation block of an argument). But personally I cannot for
a minute entertain the credibility of photographs as anything but tendentious
and personal‘takes’ on whatever they frame in the shot.The photograph is a
visual ekphrasis – interpretative, angled, chosen, made possible by a particular
circumstance, the presence of a photographer in a specific time and place...
The art historian may take his or her own pictures, in which case the‘objective
correlative’of visual proof is merely part of the tendentiousness of the original
ekphrastic formulation. Or photographs may be purchased from an archive or
museum or someone else, in which case the art-historical argument (especially
an argument based on photographs rather than one where the author has recently
been in front of the actual object) will depend in part on other people’s framings
and interpretative views, the appropriation of earlier forms of tendentious (visual)
ekphrasis to one’s own point.This is no different from working with earlier essays
and discussions of the object to hand.
Now, even if my worries about the betrayals of ekphrasis were granted, we cannot
be squeamish about committing the act.Without interpretative description, there
would be no art history. But that is at best an instrumental argument (there may be
someone, somewhere, after all, who can envisage a better world without art history –
as Plato arguably might have done). More to the point, images and objects – insofar as
they are designed to relate to us at all – invite ekphrasis, indeed they require it.
Part of the play of their relations with viewers is to elicit verbal as well as more directly
sensual or visual responses, and in that sense they are themselves the spur to the
range of narratives to which art history belongs.They may be coy about the potential
mistranslation and misrepresentations in this process, but these are themselves part of
the game of soliciting meanings and encouraging often contradictory interpretations.
So the generation of ekphrasis is not only necessary (to art history) but is inevitable in
the viewing of art.
Descriptions often need to be long, for they must entice the non-verbally
responsive object into a state where it is both available as ekphrasis and so angled in its
new descriptive form as to be appropriate to the specific argument being made.
Yet arguments are rarely about one object or even a few.Typically, for them to be
compelling or simply interesting enough to be published and read, they must make a
general point – usually a historical one in the current era. But how can one little object
– one object among millions, which may have survived by pure happenstance and is
known to the art historian by the same chance that makes him or her ignorant of so
many other objects – how can one object carry the weight of a general argument?
It is, in my experience, always the case that the ekphrastic descriptions (often
combined with photographs), that form the basis of our art history books and
articles, are made to carry more weight – both as cultural exemplars of their time or
context and as steps within an argument on which its next stage can rest – than they
can in fact bear. For one aspect of ekphrastic interpretation is to make the particularity
of a work of art more general, by becoming discursively like other objects with which
we may want it to be comparable, than its pre-verbal form actually is.There are many
reasons why art history might wish to suppress its reflex to move from the particular
to the general (not least its disciplinary claims to be a grander philosophical or
historical profession than simply the description of objects) but its unwillingness to
reflect upon the ekphrastic process on which it is founded is one of them.
©Association ofArt Historians 2010 14
Art History as Ekphrasis
Alois Riegl, Late Roman Art Industry (1901), trans.
Rolf Winkes [adapted], Rome, 1985, 139.
As an example we may take the ceremonial picture
representing Justinian and Maximian.A composition
on the plane: centralised; just verticals (contour, folds,
ornaments; the axiality is only slightly reduced in the
figure of Maximian) and horizontals (lines of heads,
feet, garment-seams and arms). Spatial composition:
the figures step frontally out of space in the direction
of the beholder and stare straight at him; even though
the main group shows partial overlaps on the plane,
along with the entourage of five body guards in three
rows, the main group is compressed into one compact
plane-like mass leaving no visible space between the
figures.The floating of the feet repeats the phenomenon
(already observed in the more advanced style early
Christian sarcophagi from Rome) by which the
foreground figures appear to step on the feet of the ones
behind them; this is obvious proof that the artist’s aim
was a complete isolation of the individual figures in
space, even at the price of sacrificing the connection to
the plane (in this case the connection of the feet with
the ground below); linear folds (corresponding to
the engraved folds in sculpture), yet inclined towards
pleating (as can be seen particularly in the double
lines).All this, along with the slim, elongated and
stilted-bodily proportions (together with a reduction
in head size) largely establish a relationship with the
subsequent Byzantine style, which is generally the most
characteristic aspect of the style of these mosaics.
How one can speak of‘decline’in view of works such
as the SanVitale mosaics,is incomprehensible since every
line demonstrates clear planning and a positive will...
Han Sedlmayr,‘Bruegel’s Macchia’ (1934), trans.
Frederick J. Schwartz, in Christopher Wood, ed., The
Second Vienna School Reader, NewYork, 2000, 330.
To facilitate the disintegration of figures, the individual
parts out of which they are composed must be
separated from each other as clearly as possible.
This leads to a preference for objects that are, by their
very nature, piecemeal and patchwork and that tend
to fragment.Thus the clothes of the peasants are not
chosen with a uniform taste but are thrown together
by chance, just as the‘motley’costumes of The Battle
Between Carnival and Lent are improvised out of everything
imaginable.The dappled torsos of the horses and cows
fall by themselves into such patches. In the visions of
madness (and dreams), we see how heterogeneously
composed, fragmented creations –‘condensations’,
to use the specialized term – come together and then
dissolve as in Dulle Griet (Mad Meg).
Alternatively, the positions and movements of the
individual bodies are sought to show the separation and
independence of the limbs: the prisoners’ feet separated
from their bodies by the stocks in Allegory of Hope; the
awkward movements of the dancing peasants in Peasant
Kermis, which have been criticized as‘poorly drawn’;
the contortions of epileptics. Only the legs remain of
the figure who falls into the barrel in Allegory of Gluttony,
only the upper body of the one who falls through the
ice in Skaters in Front of Saint George’s Gate.They all appear
mutilated, and the cripples in fact are.
In its intentions, this disintegration of form
corresponds in the real world to the process of
destruction...
EdgarWind,‘The Significance of Botticelli’s Derelitta’
(1940), in The Eloquence of Symbols, Oxford, 1993, 39.
The painting called La Derelitta, ascribed first to Masaccio,
then to Botticelli, then to that amiable fiction L‘Amico
di Sandro, and recently regarded as part of a series of
cassone panels executed by the young Filippino Lippi
after designs by Botticelli, is a source of discomfort
not only to the connoisseur but also to the student
of iconography.The subject is as enigmatic as the
authorship.A young woman, shut out of a palace, sits
‘derelict’ on the steps before the gate and weeps.This is
the sort of pathetic scene which appealed to nineteenthcentury
novelists by arousing reflections as to what had
happened before and what would happen after.
J. D. Beazley, The Pan Painter (1944), Berlin, 1974, 2.
The Pan Painter likes out-of-the-way subjects; and
the picture on the other side of the vase is unique.
The god Pan is almost unknown inAttic art before
the Persian wars: he had ground for complaining to
Philippides, on the eve of Marathon, that theAthenians
neglected him.After the Persian wars Pan becomes
quite popular atAthens: but not in this context: only
here is he seen pursuing a boy.A young goatherd, in
country garb – goatskin, sheepskin cap, stockings,
whip – is hotly pursued by the goat-god; at the rockseat,
a third, strange person, the wooden image of some
small, Priapus-like deity, views the scene with a round,
bewildered eye.
A word about the painting, before we turn to
other vases by the same painter.A blend of late archaic
©Association ofArt Historians 2010 15
Jas´ Elsner
daintiness and early-classic grandeur; the pathos of
the early-classic period but not its ethos; swift, nay
explosive movement; ravishing elegance; a darting,
fastidious touch; piquant contrasts, deliberate and
amusing disproportions – small things made larger,
big things smaller, than one expects; round heads with
tiny nose and delicate nostril but big chin and bull
neck; wasp waist but sturdy thigh; powerful arms but
tapering fingers; the bow very long, the quiver very
thin; the hounds Lilliputian; Pan’s face small between
long beard and long horns.The forms, even more than
in most vase-painters, approximated to geometrical
shapes, with a special fondness for circle and arc (even
the irregularities of the rock are fully patternised); yet
packed with expression, and tense with life.
E. H. Gombrich,‘Raphael’s Madonna della Sedia’
(1955), in Norm and Form, London, 1966, 64-65.
But can we still see it in isolation? Is not the popularity
it once enjoyed, and our own reaction against it, a
disturbing element? I may confess to you that, when I
approached the Palazzo Pitti this autumn to study the
picture in preparation for this lecture, my heart sank as
I saw the coloured postcards, box lids, and souvenirs
displayed on the stalls in front of the Gallery.
Should I really inflict this on you?A fresh encounter
with the original removed my doubts. My doubts but
not my difficulties. For, after all, you have only my
word for it that the painting looks different from those
baneful reproductions, that the very brushwork shows
a freshness and boldness which banishes all thought of
the sugar-box, and that the colours, under old varnish,
have a mellowness and richness which no print and no
copy can bring out. I remember in particular the warm
golden brown yellow of the Christ-child’s garment, as it
stands out against the deep blue of theVirgin’s skirt, the
dark red of her sleeve and the gold-embroidered back
of the chair, and, most of all, the blending into harmony
of that daring green scarf which so easily brings a cheap
and discordant note into prints.There are some patches
of repair and over-painting over cracks affecting St John
and the fringe on the Christ-child’s face; but by and
large the condition of the picture seems to be good, and
the enamel-like finish of theVirgin’s head, the spirited,
fresco-like treatment of the drapery and the chair with
its impasto highlights all appear to me to betoken the
master’s own handiwork. It is true that even in the Pitti
Gallery it is not easy to come to terms with the picture.
The vast golden eighteenth-century frame produces a
dazzle that all but kills the subtle gradations of tone on
which Raphael relied.As soon as you screen it off with
your hands the picture comes to life.
Michael Baxandall, Patterns of Intention: On the
Historical Explanation of Pictures, New Haven and
London, 1985, 80.
Chardin seems to be doing something strange with
perspective here and there.The chair-back is odd: if
the lady were sitting comfortably on the chair, surely
the chair-back would not be turned to face us and
the picture-plane as much as it does.The tea-pot is
also rather 1910: spout and perhaps also handle are
flattened out on the canvas.Then there are a whole
range of striking colour devices.The most obvious is
the red-lacquered table, assertive but almost unstable.
And if one looks in other pictures by Chardin, one
finds other cases of reds in relation to blues and blacks.
Again, and very conspicuously, there seems something
extraordinarily deliberate and determining about the
differential distinctness and lighting of the picture.
There is a determinate plane of distinctness on the line
of the teapot, hand and arm; and within this plane some
things are more sharply focused than others.What does
all this represent?
Michael Camille, Image on the Edge, London, 1992,
40-41.
In the bas-de-page of the Ormesby Psalter’s 101st
Psalm,
the scandalized look of the gryllus is almost voyeuristic.
Here he stares at a‘bawdy betrothal’, in which another
squirrel-grasping lady accepts a ring from a young
man – an anti-illustration of‘my heart is smitten’ in the
Psalm above. In this marginal masterpiece the margins
include their own meta-marginal parody further out
on the edge. Here, beneath the courtly couple, a fat cat
stalks a mouse, reversing the gender positions above, so
that the mouse in its hole is beneath the knight whose
sword sticks out of his own hole like a phallus.The
complex criticism of their encoded eroticism is further
annotated by the gryllus here who, watching from the
wings, as it were, is an incarnation of scopic obsession
– having a head between his legs instead of a prick. His
look is an ejaculation.
Some ekphrases from a selection of paradigmatic art historians.
A sense of decency has made qualification for inclusion that the
authors be dead,which has the interesting effect that all are men.To
discuss how any one of these,as description, formulates the nature
of the argument of which it is part would take too much space.But
clearly different descriptive choices would in any of these cases have
yielded the possibility for profoundly different results.
©Association ofArt Historians 2010 16
Art History as Ekphrasis
Beyond the issue of generalization, there lies the specific problem of argument.
The object translated into ekphrastic description is available for use in an argument
– indeed, a series of arguable and argumentative assumptions are the inevitable basis
of the descriptive terms chosen to create the ekphrasis in the first place. For much
art history nothing is so compelling in such argument as the adduction of formal
and stylistic observations so as to make the point at issue out of the object’s own
object-hood. Indeed, this was a fundamental aspect of the greatViennese project
in art history from before the end of the nineteenth century until the aftermath of
the 1939–45 war.Yet the problem is that what we adduce as formal is in fact not the
object’s own object-hood and existence as matter but that ekphrastic transformation
which has rendered it into a stylistic terminology. How secure can we be that such
ekphrastic formalism (the closest in art history some might affirm, that an object
can be to its pre-verbal state) is no more than a carefully crafted verbal translation
whose discursive functionings are as far from the actuality of objects as any other
interpretative description? Likewise, in those arguments which turn to questions
of meaning (and are ultimately dependent on the greatWarburgian projects of
investigating symbol and memory as well as the emphasis on meanings in the early
work of Panofsky andWind), to what extent can we be clear that it is the work of art
itself that interrelates with the literary and other cultural artefacts beside which it
has been placed, rather than our tendentious ekphrastic extrapolation of it? In other
words, when the object speaks in art history and when it is heard, what is heard is our
ekphrasis. But we do not stress this point, since the need to elide ekphrasis and object
is essential to the method, if it is to carry the conviction of some empirical validity
(not to speak of objectivity or positivism)...
Let us take an example. Here is a little piece recently composed on Michelangelo’s
Rondanini Pietà (see plates 2–5).
The Rondanini Pietà
Michelangelo is said byVasari to be the apogee of the Renaissance, the point to
which the entire tradition of naturalistic representation has led and at which it has
culminated. If this is true, then his last work – the haunting Rondanini Pietà, which since
1952 has been in the Castello Sforza in Milan – is perhaps his most characteristic
sculpture. For it is not so much the apogee of his art as the emblem for everything
Michelangelo was and for the totality of the tradition ofWestern image-making in
which the Renaissance holds so potent a place.The sculpture’s naturalism is potent
but almost in abeyance: an exquisitely rendered arm belonging to theVirgin from an
earlier, abandoned sculpture of the Deposition but not yet removed; the elongated
forms of Christ’s lifeless legs, breaking into near verisimilitude at the knees before
reverting to sketchy unfinish at the feet and the groin.The very elongation of
forms, stripped down (unfinished?) almost to the point of abstraction, presages the
mannerism to come in Renaissance art.The refusal to finish – the near resistance to
move beyond the pure blocking-out of the faces of the sorrowing mother and her
dead son – foresees every gesture of modernism in later sculpture from Rodin to
Giacometti and Moore.There is, in short, very little, perhaps nothing at all, in the
passions and abstractions of the later development of the tradition ofWestern statuemaking
which the Rondanini Pietà does not foretell.
Yet the sculpture remains naturalistic. Its naturalism is about choosing varieties of
unfinish to pare down representation to its ultimate and simplest. How best to render
the relations of mother and son?With the intense realism of Michelangelo’s earlier
Vatican Pietà? Or with the sketched abstraction of the heads, upper bodies and faces
©Association ofArt Historians 2010 17
Jas´ Elsner
of the Rondanini, into which a viewer may pour all his or her own sense of that grief
and mourning which must surpass all others? However humanist the vision, let us
not forget its piety and genuine devotion: this is not just any old mother and son, it
is the pair of the one mother and son, which means of all and every mother and son.
Here the Rondanini is the summing up of Michelangelo’s long fight with the pietà
as sculptural subject – from the perfection of finish in theVatican sculpture of his
youth via the abandoned and incomplete versions now in the Museo dell’Opera del
Duomo and in theAcademia in Florence, abandoned because they did not satisfy their
maker, to this final meditation.The Rondanini Pietà is Michelangelo’s last statement on
unfinish, and in particular on that moment of (in)completion which comes when
you lay down your creative tools during the unfolding of a process and simply say this
is enough, and on how naturalism to achieve the supreme expression of emotion may
need to resort to incompleteness and abstraction.
In its naturalism, the pietà is in fact entirely a piece in the classical tradition –
looking back beyond Michelangelo’s predecessors like Donatello and the Pisani to the
depths of antiquity, with the same intensity as it looks forward.The very willingness
to tolerate a no longer relevant fragment, like that near-completely rendered arm,
seems to hint at the fragmentary nature of the classical heritage – detached limbs,
torsos, heads, all so perfectly finished yet battered and dismembered.That pin, which
fixes the abandoned arm to the torso of Christ, stands like a deliberate memorial to
all those ancient statues (Roman marble copies of lost Greek bronze originals, we
call them now) which sport such struts to keep the marble limbs from splitting off.
The urge to combine areas of exquisite finish with a roughness which in ancient
survivals is caused by ruin, abrasion and the depredations of time turns this piece into
a modern antique, deliberately poised between finish and incompletion, on the edge
of that state of ruin whose supreme evocation must be the death of Our Lord.
Artistically, what is striking about both the inheritances from the past and the
intimations of the future between which this sculpture is so supremely, magisterially,
poised, are certain very precise gestures of difference.Ancient statues were finished –
their unfinishedness, fragmentation and rough areas are an addition by time to what
was once finished. Noses are smashed and faces made rough where statues once fell
or were broken. But Michelangelo’s unfinish is by design; and it lies before a state
of final finish which he decides never to attain. His work is in progress and for ever
incomplete. But some parts of that unfinish, and perhaps all of it – above all that arm
which looms to the right of the principal group, and the sketchy intimacy of the
faces of theVirgin and Christ – are deliberate, deliberately not worked further upon
or removed. By contrast, the masterworks of modern sculpture that use abstraction,
fragmentation and abrasion, do so with full design.They adopt as a stylistic trait, a
strategy of representation, an aesthetic mode, that which in this sculpture is the point
of unfinish most close to the wear of time on an ancient statue and beyond which
Michelangelo judges that his sculpture will lose more than it can gain.
It is as a work in progress,poised at the zenith of the tradition and encapsulating all
that went before and would follow – positioned between naturalism and abstraction
and engaged with the naturalism of feeling only possible in abstraction – that the
Rondanini Pietà towers above all Michelangelo’s art.It stands as the emblem not only of
the kind of work possible in marble sculpture but also of more philosophical concerns
about the possibilities of representation embedded in marble sculpture.It is as a work
in progress that the Rondanini Pietà becomes the epitome of a tradition of representation
in progress and of a drive to naturalism whose most moving moments come when the
panache of illusionism yields to the pathos and simplicity of pure form.
©Association ofArt Historians 2010 18
Art History as Ekphrasis
2 Detail showing bottom half of
Michelangelo,Rondanani Pietà,
c.1555–64.Milan:Castello
Sforzesca.Photo: Author.
©Association ofArt Historians 2010 19
Jas´ Elsner
3View from right hand side
of Michelangelo,Rondanani
Pietà,c.1555–64.Marble,195
cm (height).Milan:Castello
Sforzesca.Photo:Author.
©Association ofArt Historians 2010 20
Art History as Ekphrasis
4 View from left hand side
of Michelangelo,Rondanani
Pietà,c. 1555–64.Marble,195
cm (height).Milan:Castello
Sforzesca.Photo:Author.
©Association ofArt Historians 2010 21
Jas´ Elsner
5 View withAncient Roman
base of Michelangelo,
Rondanani Pietà,c. 1555–64.
Milan:Castello Sforzesca.
Photo: Author.
©Association ofArt Historians 2010 22
Art History as Ekphrasis
Rather than tarnish another scholar – arbitrarily chosen and cruelly exposed – I have
taken a piece of my own. Let me begin by saying that this little fragment was not
composed for this essay, but beforehand and quite separately – in response to seeing
the object for the first time in the Castello Sforza in Milan. I still stand by the piece
in that it expresses what I think (or feel?) about Michelangelo’s sculpture, if not to
the very best of my conceivable abilities then at any rate as well as I could on the
day. Clearly the account is tendentious in that the sculpture has been reformulated
according to longstanding concerns of my own – the nature of naturalism, the long
process of its rise, fall and rise again in the European artistic tradition, its limitations
(especially in relation to religion). Effectively, the object functions as a springboard for
and embodiment of concerns which neither it nor its maker could fully comprehend
(especially my thoughts about what it presages in later sculpture). Note the big claim
with which I open – reading the object to hand as an emblem of the entireWestern
tradition of image-making (which is implicitly to read Michelangelo inVasarian terms
as a supreme genius). If the piece were cast as history this would perhaps be beyond
the pale; I hope it can stand in the context of a more evocative reflection about the
Western tradition of naturalistic illusionism. But the move to generalization (or rather
the opening frame of generalization which justifies discussion of the particular object
in the first place) could hardly be more strongly made.
The description is deliberately and even insistently formal in the first two
paragraphs – in part because any discussion of naturalism and its departures must
be so. But the assumptions running it (and most other descriptions of naturalistic
images) are large.Why, on the face of it and without recourse to the tradition of
both Renaissance representation and the art-historical discussion of it, should one
impose on pure representation (that is, a block of marble carved in such and such a
way) all these assumptions about finish and unfinish, elongation and verisimilitude,
abstraction and empathy? Or is Michelangelo’s very choice to represent – a man and
a woman, Christ and his mother – not already an opening into the ekphrastic,
embedded in the object before any particular viewer may come along to help it into
a frame of his or her interpretations? One answer to the‘why should we impose’
question, is because we cannot unwrite the traditions of representation and art history
within which we were educated and out of which our thought inevitably must come.
But that is a response from weakness: because we cannot be other than our ideological
conditioning does not greatly strengthen one’s confidence in that conditioning. ...
A stronger answer might be that I do not want to resist all the assumptions of the
tradition because (at any rate in this piece) I am writing specifically about and in
response to the tradition (both that process of making art in which Michelangelo
was so formative a figure and that process of making art history in which naturalism
is a central concern). Fair enough. But still the key question about ekphrasis remains,
which is why the object need be spoken about at all in these terms (terms I need
to make my discussion) rather than any other terms (some of which may not yet
be available to us and others which might just seem utterly eccentric or personal if
expressed in writing). Not only is ekphrasis translating the object into its own terms
but it is dragging it into a network of concepts generated by the ekphrastic tradition
of art history over many centuries, and helping it along down a slippery slope of
interpretative meanings implied already in the choices made in its own manufacture.
A purist would not be wrong to resist this tendency.
And yet, while the tendentiousness is inevitably there in ekphrasis – I mean,
the appropriation of the object for ends that suit the interests of the interpreter, the
counter-question might be why we would want anything else of description (or of
©Association ofArt Historians 2010 23
Jas´ Elsner
art history).At no time – not at the moment of creation, not at the behest of patrons,
not in their receptions in history, have objects been innocent, pure or objective.
The ekphrastic appropriation is in this sense not much different from
Michaelangelo’s own appropriation of a block of Carrara marble and his working
on it to turn it from stone to pietà – and in the case of the Rondanini he has himself
marked his second thoughts (that is, the nature of the changes in his interpretative
intent) by preserving the arm of a different sculpture once to be cut from the same
block. So the worry about appropriation is less about the act of interpreting and
leaping from the described image into one’s own argument than in judging why
my argument may be better, more valid, more justified, more interesting, more
whatever-you-will, than yours or anyone else’s. But my point here is less to judge
which interpretation of any given group is better, than to argue for the need to be
clear about the extent of subjectivity – or shall we call it personal taste, even whimsy?
– in the very formulation of the descriptions from which our apparently rational arthistorical
arguments are generated.
Now it may be that Michelangelo really did think about finish and unfinish,
fragmentation and completeness in statuary in relation to antique survivals – as I
tentatively suggest here. Or even that he did not think about this but was conscious
of it in some part of himself during his handling of the stone or as part of the very
process of sculpting. But clearly it is my own tendentious claim – or that of the
tradition in which I write? – that his sculpture looks forward into modernism
itself. Here the exercise of comparisons – Donatello, the Pisani, ancient statuary,
Michelangelo’s own other pietàs, Rodin, Giacometti, Henry Moore – is itself
interestingly ekphrastic. For though none of these examples is a description of a work
of art, each invokes both a set of characteristic works (theVatican pietà being the most
specific one) and a set of standard accounts in art history. My own interpretation
(interpretative description) builds on these and might be said to be meaningless to
anyone without a sense of the tradition in which they unfold.That is, my ekphrasis is
founded upon and related to a series of other ekphrases (none of which I have read in
the immediate context of writing the piece, some of which I read long ago and some
of which I have never read). Inevitably, though apparently founded on and inspired
by a single object (and written on the same day as seeing it), my ekphrasis is at least as
much about the range and history of other ekphrases of other objects in the tradition.
Indeed, in the sense that it is ultimately about the tradition itself, it may be said to
elide the object, which it professes to discuss centre stage, as merely the emblem
or figment of a tradition in which that object (as‘Michaelangelo’s last sculpture’)
was inevitably a player but not necessarily so savvy and self-conscious a player as my
interpretative impulse has made it.
Now, my piece is illustrated, and by my own pictures. I am proud of my
photographs – for it is not easy to make pictures without a flash or a tripod in the
Castello Sforza (where, as usual in museums, such things are not allowed), and these
were some of the few that came out (from the many I took, despite the harassment
of a guard who believed that five was the maximum any visitor should be allowed).
Nor am I a particularly good photographer (less good certainly than the excellent
digital camera I use). But what do the pictures add? Conventionally, pictures may be
seen to supplement and support a description (or, many might say, to constitute the
prime evidence, the description being there to supplement them by an interpretative
steer). But more potently, reproductions in a text speak the (visual) discourse of the
real, grounding the interpretative flow of ekphrasis (which is itself a verbal appeal
to the real, material, object) with an actual material snapshot of a real thing out
©Association ofArt Historians 2010 24
Art History as Ekphrasis
there, outside the bubble of rhetoric.The photograph is a kind of ekphrasis within
an ekphrasis – a visual interpretative framing within a textual interpretative framing.
The photograph promises that bottom line of‘thereness’ in which the text’s argument
can finally be grounded.That is a powerfully rhetorical role for the visual, and one
particularly suited to ekphrasis, in that interpretative picturing attempts the same
thing as interpretative description in relation to the bigger argument being made.
But it remains no more than a rhetorical role with the photograph (whatever feelings
of achievement my in-focus shots may raise in me) being as tendentious as the rest of
my account.
So much then for an analysis of my own ekphrasis. I have of course been careful to
avoid commenting too much on the content – the tendentiousness – with which the
piece was designed and in relation to which the description functions.That is hardly
my right and is anyway not my purpose in assessing the ekphrastic basis of art history,
although close attention to the rhetorical aspects of interpretative description may
reveal the tendentiousness more acutely than is usual. But it is the tendentiousness
which we cannot control. Nazis like Josef Strzygowski and Hans Sedlmayr were great
art historians whatever one thinks of their wider views or the more horrible turns
which some of their arguments took.We may take a moral stand – agree or disagree
– but then what is at issue (despite appearances sometimes) is not art history but
politics, ethics or ideology.Art history is about constructing compelling and wellfounded
arguments out of objects, and that means out of ekphrasis.Whether my
argument on the Rondanini Pietà is either well-founded or compelling is hardly for me
to say.
Where does this take us? I opened with a proposition about art history as
ekphrasis, because I think it true but simultaneously uncontroversial pragmatically
among art historians (that is, we have no choice but to do interpretative description)
and yet thoroughly controversial ideologically (that is, we find it impossible to admit to
ourselves that this is what our practice and its core methodology consist in).
To say that art history is ekphrasis is to say that it is no less literary and rhetorical as a
discipline than it is philosophical or historical. It is to claim the study of art as playful
and fictional as well as serious and substantial. It is to deny any fundamental difference
between‘art history’ and‘art criticism’ other than the tendency of their ekphrastic
flow and the rules by which they are written. But usually we are embarrassed to be
caught writing fiction with footnotes (why this is so, I have never understood – for
that is what we do, even if the footnotes chart a careful series of‘true’, or at least
generally accepted or empirically attested, parameters within which we steer our
fictive imaginations).Worse, for some reason (another thing I have long failed to
understand) a large number of art historians are insistent on the priority of the object,
as if ultimately art history were a discipline of things and not of words.They are
wrong.Art history is not possible without ekphrasis, just as it is not possible without
objects.That means it is, at the very least, a verbal discipline of the visual, a discipline
of the constant translation and re-translation of art into text, and of the belief in (or
desire for) the potential transfiguration of the visual cast in verbal terms that can make
it more clearly or effectively or essentially grasped.Art history has always turned the
object into text and no amount of cant or bluster can make it otherwise.
This is not to say that the history of art history is the same as the history of
ekphrasis. On the contrary, the history of art history is a great professional chapter in
the history of an ancient descriptive practice that encompasses many turns – playful,
fictional, poetic, metaliterary – that are by no means all art historical. So the claim that
art history is ekphrasis is also a claim for a much longer and grander historiography to
©Association ofArt Historians 2010 25
Jas´ Elsner
the discipline than the usual survey of the usual suspects fromVasari to Panofsky or as
late as Baxandall or even up to Michael Fried and Michael Camille, all of whom have
usually performed as if they were writing prose history about objects or artists.
The ekphrastic passages that deal with art in Homer,Virgil, Dante and a host of epic
poets, in the world of epigram and lyric, in the vast number of novels that describe
works of art from antiquity to the present, are an essential part of the discipline’s
history in that not only do they perform the process of interpretative description
with remarkable panache and self-reflection, but they do so on works of art which
are both real and fictive.That they may turn the discussion to other directions than
just‘art history’ has the potential to add possibilities to our own practice. But beyond
pragmatics and beyond the potential acquisition of descriptive strategies deep in the
tradition (but usually avoided by art historians as just one step too far), a bigger sense
of ekphrastic historiography is a way of sharpening our critical edge.
Interpretative description is always tendentious.When we pay attention to
ekphrasis critically, we isolate the formal and rhetorical elements. So, in my account
of the Rondanini Pietà, I borrow a series of substantive or generic tropes from the arthistorical
tradition – issues of naturalism and abstraction in the question of forms,
comparisons made with art and artists from other periods – and tie them to claims
to generalization both at the opening and at the close.These themes are presented
in ways designed to up the stakes of the importance of the object discussed (and
implicitly in the theme I am using the object as a vehicle for discussing), so that it
can indeed serve as an emblem of something of significance in respect of the whole
tradition. Rhetorical choices include the use of adjectives to support the emotive
force of the piece –‘the haunting Rondanini Pietà’,‘Christ’s lifeless legs’,‘the sorrowing
mother’ – and the repeated terminology of‘apogee’ and‘zenith’ to define a period
both within the chronological history of art and of artistic creativity.The theme of
a work poised between artistic traditions or at the cusp of a shift within them – so
prescient as to presage the future yet also able to encapsulate the past – is verbalized
through the repeated play on‘finish’ and‘unfinish’,‘sketched abstraction’,‘(in)
completeness’, fragmentation, abrasion, those legs‘breaking into near verisimilitude
at the knees before reverting to sketchy unfinish at the feet and the groin’.
In analysing the way ekphrasis attempts to make the object speak, especially by
masking its rhetorical aims through verbalizing the object’s forms and material
characteristics, one may hope to lay bare the tendentiousness – the directing desire
or ideology which fuels the description towards a chosen end (a topic which in the
interests of decency I had better not attempt in my discussion of myself!). In general,
art history takes this desire too much for granted as if it were the result of an argument
(showing a Marxist view of a given set of objects, or their place in ritual, or their
status in an artist’s oeuvre or whatever). But this desire is so often the starting point that
shapes the discussion, informing the ekphrastic opening and ensuring a circularity
by which the argued conclusion is only a restatement of the opening proposition
in different terms.And that, although it may be satisfying ideology, is not good
argument nor good art history.
Worse, how much of that desire is my own? In finding the terms to formulate
my object in description, I not only fix the direction of argument in which it will
lead, but am dependent on the structures, tropes and ground-rules set up by arthistorical
tradition.Through my apparently personal description, the tradition seems
simply to rewrite itself – finding perhaps (one hopes) more subtle meanings, better
analogues, a new document or an old one more precisely assessed, a more nuanced
interpretation in the process.Yet as it rewrites itself through me, what can become
©Association ofArt Historians 2010 26
Art History as Ekphrasis
clear are the very limitations of my particular ekphrastic responses, the questions
those responses allow me to ask, the ground-rules within which my argument
operates.This brings us back up against the object – its glorious resistance to being
fully verbalized, its uncanny ability to be verbalized in a myriad of ways, equally valid
and sometimes mutually exclusive.As description knocks against the object’s objecthood,
the important thing is the chance that is offered to see it afresh in the creative
gap between the visual and our traditions of verbal tropes. In the ancient definitions
of ekphrasis as a rhetorical tool, prime weight was placed on the ability to bring what
is described before the reader’s or the listener’s eyes.The role of ekphrasis – and of art
history itself – is to make the reader or the listener‘see’ more than they saw before,
when they encounter the object next.That search for words to make us‘see’ is at the
heart of the creative struggle against the ways in which what we have learnt can go
stale, and it is an attempt to open to the new.
The formatory nature of the ekphrasist’s desire (in this case the art historian’s) on
both the discussion and its underpinning descriptions is at least partly pre-rational,
and requires both literary methods and a literary sensibility to be diagnosed by the
critical reader and to be given full vent by the descriptive interpreter. Sometimes,
a more fictive and playful form of description may do better justice to our chosen
objects than a recourse to painstaking historicism. Panofsky famously compared
iconology by contrast with iconography to ethnology by contrast with ethnography
and to astrology by contrast with astrography. He wished the first of these
comparisons (that is, the idea of iconology as a parallel to a respectable discipline like
ethnology) to stand as a positive model but not the second (since he was not dealing
in the mystifications of astrology). But with his characteristic acuity Panofsky pointed
to the terrible (or wonderful) truth that art history in fact embraces both extremes,
however ideally we may try to police it. Sometimes, we may need – in other words –
knowingly to allow the intellectually dubious as well as the academically respectable
to inform our making of ekphrasis. Otherwise, it will simply do so without our
acknowledgement: in fact, it frequently does so already.
Notes
This paper is the result of teaching several courses
at the University of Chicago on issues of ekphrasis
and historiography – I am grateful in particular to a
number of students for egging me on down one or
two mad alleyways.In a way it is also my response
to a historiographical turn in my own research that
has had me thinking about and against the likes
of Riegl and Panofsky for several years (here I am
specially grateful to Richard Neer,Peg Olin,Adrian
Rifkin and Joel Snyder for various discussions and the
encouragement not to throw in the towel,as well as
my partner in the folly of attempting to translate the
German of the early Panofsky,Katharina Lorenz).
It also continues two long conversations over many
years with John Henderson on the question of art
and text,and with John Ma on art history and history,
so I thank them especially for having got me here
(even if they are bound to disagree).Insofar as I
have chosen to avoid that great fiction of the real in
academic discourse,namely the footnote,I genuflect
to a venerable tradition in Art History established by
John Onians in a famous piece on imagination in late
antiquity (in 1980) and recently revived in an elegant
paper by John Ma (in 2006).My particular gratitude is
due to audiences in Oxford and at Emory (especially
Jean Campbell andWalter Melion),to whom I have
given different versions of the thoughts contained
here,to David Peters Corbett and the editorial team
at Art History and particularly to an extraordinarily
sharp anonymous reader whose empathetic
response conveyed more telling criticisms and
disagreements more directly than hostile criticism
can ever achieve.
I have been asked to add some bibliographic
suggestions.The ancient prescriptions for ekphrasis as
a‘colour’ of rhetoric are given by a series of authors
–Theon, Hermogenes,Aphthonius, Nicolaus and
John of Sardis – who span the Roman and Byzantine
periods.They are now conveniently translated into
English in G.A. Kennedy, Progymnasmata:GreekTextbooks
of Prose Composition and Rhetoric,Atlanta, 2003, esp. 45–7,
86–7, 117–20, 166–8, 218–21. For discussion of
the ancient techniques and scope of ekphrasis, see
especially RuthWebb, Ekphrasis,Imagination and Persuasion in
Ancient RhetoricalTheory and Practice, Farnham, 2009, with
large further bibliography. Outstanding discussions
of ekphrasis in literary and artistic tradition – from a
variety of different scholarly cultures internationally
– include: Page duBois, History,Rhetorical Description and the
Epic, Cambridge, 1982; Françoise Meltzer, Salome and the
Dance ofWriting:Portraits of Mimesis in Literature, Chicago, IL,
©Association ofArt Historians 2010 27
Jas´ Elsner
1987; Stephen Bann, TheTrueVine:OnVisual Representation and theWesternTradition,
Cambridge, 1989; Lucien Dällenbach, The Mirror in theText, Cambridge,
1989; Murray Krieger, Ekphrasis:The Illusion of the Natural Sign, Baltimore, MD,
1991; JamesA.W. Heffernan, Museum ofWords:The Poetics of Ekphrasis from Homer to
Ashberry, Chicago, IL, 1993;W. J.T. Mitchell, PictureTheory:Essays onVerbal andVisual
Representation, Chicago, IL, 1994; Gottfried Boehm and Helmut Pfotenhauer,
eds, Beschriebungskunst-Kunstbeschriebung, Munich, 1995; John Hollander, The
Gazer‘s Spirit:Poems Speaking to SilentWorks ofArt, Chicago, IL, 1995; PeterWagner,
ed., Icons –Texts – Iconotexts, Berlin, 1996; FrederickA. deArmas, Cervantes,
Raphael and the Classics, Cambridge, 1998; James Elkins, On Pictures and theWords
that FailThem, Cambridge, 1998;Valerie Robillard and Els Jongeneel, eds,
Pictures intoWords,Amsterdam, 1998; Bill Brown, A Sense ofThings:The Object
Matter ofAmerican Literature, Chicago, IL, 2003; Gilbert Sorrentino, Lunar Follies,
Minneapolis, MN, 2005; John Bender and Michael Marrinan, eds, Regimes
of Description:In theArchive of the Eighteenth Century, Stanford, CA, 2005; Gabriele
Rippl, Beschriebungs-Kunst:Zur intermedialen Poetik angloamerikanischer Ikonotexte
(1880–2000), Munich, 2005; the essays collected in a number of special
editions of journals: Yale French Studies, 61, 1981; Word and Image, 15: 1, 1999;
Ramus, 31: 1 and 31: 2, 2002; Classical Philology, 102: 1, 2007.A spectacular
recent and self-conscious example of the extended genre of art-historical
description – constructed in the form of a personal diary – isT. J. Clark‘s The
Sight of Death:An Experiment inArtWriting, New Haven and London, 2006. For
Panofsky‘s discussion of the problems of describing the sensible qualities
of art works, see Erwin Panofsky‘On the relationship of art history and art
theory: towards the possibility for a fundamental system of concepts for a
science of art‘ (1925), in Critical Inquiry, 35,Autumn 2008, esp. 57–63. For
Panofsky‘s comparison of iconology to astrology and ethnology, see Erwin
Panofsky, Meaning in the VisualArts, Chicago, Il, 1982, 32.