The Shield of Perseus

The Flower of All the Arts

Alberti, near the start of the second book of On Painting, reminds us
that the allegorical founder of the art was the son of the river god
Cephisus and the nymph Liriope: ‘I used to tell my friends that the
inventor of painting, according to the poets, was Narcissus, who was
turned into a flower.’ He explains why the famous episode is germane
to his subject – ‘What is a painting but the act of embracing by means
of art the surface of the pool?’1
Embracing the surface of the pool is precisely what Narcissus
cannot do. Ovid has the enchanted boy plunge his arms into the water,
try to kiss its shimmering surface, and weep in grief and confusion at
his failure: ‘His eyes are deceived, but the strange illusion excites his
senses. Trusting fool, how futile to woo a fleeting phantom! You’ll
never grasp it.’2 The image he sees there is ‘fraudulent’, an ‘empty
hope’; it is ‘a reflection consisting of nothing’.3 From these glittering
waters, where the love-lorn Narcissus is forever frustrated by the
untouchable likeness of his beloved, Alberti feels that painting can
succeed in delivering to us the objects of sight and desire; painting is
not being likened to the water’s reflective surface, but presented as a
way of avoiding Narcissus’s fate. Its ‘truly divine power’ is to ‘make
the absent present’; Alberti makes so bold as to ‘venture to assert that
whatever beauty there is in things has been derived from painting’.4
It is an interesting position to take, and the effective opposite of the
Platonic. Plato, while banishing poetry from his Republic, dismisses
painting as mere imitative play: the copying of copies, incapable of
drawing us nearer to truth and knowledge about the world; in fact,
making us even further removed from it.5 Alberti, however, awards
painting a privileged ability to take us closer to truth, and toward the
world. He is thus in loose, and most probably coincidental, agreement
with Plato about the illusory or faulty nature of the perceptible world,
but unlike Plato he believes painting is an art which can assist us in
4 Black Light

avoiding its deceptions. It seems unlikely that Alberti was familiar

with the Athenian, given the partial and scarce translations available
at the time On Painting was written,6 but it seems likely he would have
found himself at odds with Plato’s views in any case.
Nor is his view by any means a conventional ‘modern’ or rationalist
position, since he considers the world which painting takes as its
model to be as essentially false and ungraspable as the reflected image
of the original narcissist. In this brief but telling aside, painting seems
to be seen as a path toward truth, able to illuminate and enliven the
world as it simultaneously satisfies our desire to lay hold of it. It can
prevent us being deceived by phantoms through making them
material, and thus allow us to have what we want; truly showing us
the world and revealing its otherwise latent beauty, its transformative
power is the analogue of Narcissus’s passion, not his reflection.
Painting, ‘the flower of all the arts’,7 is Narcissus triumphant.
The boy’s image, gleaming there in the ‘clear, unmuddied pool of
silvery, shimmering water’,8 is his punishment for mocking the advances
of nymphs and men alike, the tragic Echo first among them. His self-
regard is rewarded with its fitting prize, and his perfect reflection in the
shining water deludes him into the only love he cannot disdain. It is one
of the two most famous reflected images found in the antique mytho-
poetic tradition. The other has a quite different function.

The Protections of Perseus

The element of Perseus’s story which seems to have most interested the
ancient mind is the rescue of Andromeda, and the subsequent intrigue
and drama surrounding the hero’s marriage to her. This episode was the
subject of lost plays by both Euripedes and Sophocles (fragments of both
survive); Ovid, after a description of Perseus’s defeat of Atlas, devotes
most of his attention first to the rescue of the Ethiopian princess and then
to an extended and very bloody account of the fight with the allies of
Phineus, Andromeda’s original betrothed.9 (Apollodorus presents the
entire story with equal emphasis; but then Apollodorus presents
absolutely everything with equal emphasis, which is to say, none.)10
These latter parts of the story are of less importance here than the deed
and prize which underwrites Perseus’s later successes, the slaying of the
Medusa and the taking of her head (a deed which seems to have formed
the subject of a lost play by Aeschylus, The Children of Phorcys). As with
many Greek myths, much of the story is rather vague: sources are
fragmentary and contradict one another, and there are gaps and
 The Shield of Perseus 5

embellishments. The version given below is composed mostly according

to Ovid and Apollodorus.11
A son of Zeus, Perseus had been conceived while his mother Danae
was locked in a dungeon. Her father Acrisius, king of Argos, had
imprisoned her, fearing she would fall pregnant with a boy: an
oracular prophecy had predicted his death at the hand of his
grandson. Zeus, untroubled by Acrisius’s sequestration of the girl,
visited her in a shower of gold that poured into her lap, and she bore
him Perseus. Unconvinced that the child was of a god and in fear of
the oracle’s words, Acrisius locked both Danae and her infant son in a
wooden chest, and cast them out to sea. Their ark washed ashore on
the island of Seriphos, where its living cargo was discovered by one
Dictys. They were taken in, and Perseus was raised either by Dictys
himself, or possibly by his brother Polydectes, king of Seriphos.
In time, Polydectes fell in love with Danae and demanded her hand in
marriage. Perseus, ‘grown to man’s estate’12 and thus his mother’s
guardian and protector of her virtue, refused permission. Seeing the
need to be rid of the young man, Polydectes made a pretence of making
ready to marry another woman, Hippodameia. Perseus, glad to sponsor
a wedding which would protect his mother from Polydectes’ advances,
rashly promised to provide any present to the king, even the head of the
Gorgon Medusa. Polydectes, who had received horses from his other
friends, took Perseus up on his offer, and demanded the Gorgon’s
severed head.
We come now to the parts of the myth that concern us more directly.
Both Hermes and Athena resolved to help Perseus in his task. (Athena
evidently held a grudge against the Gorgon sister, since according to
Ovid it was she who transformed her from a beautiful woman into a
snake-locked monster – the virgin goddess had been outraged at the
sight of Poseidon raping Medusa inside one of her shrines.)13 From
Hermes, Perseus obtained an adamantine sickle; from Athena some
say he received a polished bronze shield and the advice needed to kill
Medusa. He then visited the Graiae, sisters of the Gorgons who shared
between them a single eye and tooth, in order to discover the
whereabouts of the nymphs who possessed other items he needed.
Stealing their lone eye, he forced them to divulge the location of these
nymphs, from whom he then collected the cap of Hades, a magical
wallet to put the Gorgon’s head in, and some winged sandals. He then
flew to the remote lair of the Gorgon sisters.
The Gorgons – Medusa, Stheno and Euryale – were the daughters of
the sea by Phorcys. Only Medusa was mortal, having been
6 Black Light

transformed by Athena into a hideous monster, whose hair was a mass

of snakes and whose gaze turned all who looked on it to stone. Perseus
found them asleep; under the guidance of Athena, he avoided looking
at the deadly being by viewing her reflected image in his polished
shield, and swiftly beheaded her with the sickle of Hermes. In order to
escape the other sisters he donned the cap of Hades, an item which
conferred invisibility on the wearer, and, being careful to avoid
looking at the still powerful severed head of Medusa, he slipped his
prize into the magic wallet and began his flight back to Seriphos to
confront Polydectes. We will leave the victorious Perseus suspended
here, and look in more detail at the moment of the narrative which is
specifically important to us: the beheading of Medusa, and the ruse
through which the hero is able achieve this feat.
We may infer that Perseus received his polished shield from Athena,
though neither Ovid nor Apollodorus is actually explicit about this.
(Given the goddess’s grudge against the Gorgon, her guidance of
Perseus’s hand and her later placing of the severed head on either her
aegis or shield, it seems likely that it is so.) In its shining surface, he sees
Medusa reflected; thus can he kill her, being able to do so because he can
safely look at what would usually kill him. He disarms her, having the
means to avoid the threat she poses. His shield, which is not there to
protect him from arrows or blows, provides him with safe sight. It
protects him from the danger of seeing her.
But it also seems that her sight is also perilous. Though the texts do
not indicate decisively that the Gorgon’s look alone kills, Ovid certainly
indicates that it is her gaze particularly that is fatal when he has Phineus,
Perseus’s foe in the violent latter stages of the story, plead with the hero
to put away the head after two hundred men have been turned to stone
by ‘the stare that kills’; the poet also specifically mentions ‘the terrible
eyes’ of the Gorgon.14 However, Perseus comes upon the Gorgon sisters
while they sleep, and it seems it would still be fatal to look upon
Medusa: the polished bronze shield must be used. In Apollodorus it is
fatal to behold the Gorgon sisters (‘they turned to stone such as beheld
them’)15 and though Ovid mentions Medusa’s gaze, as noted above, he
also says that to set eyes on or catch sight of her is enough.16
It is this visual aspect, whether in the form of looking or being
looked at, which finds emphasis in other versions or appearances of
the myth. Benvenuto Cellini’s extraordinary bronze Perseus, in which
the hero lifts Medusa’s newly severed head to display it out across the
Piazza della Signoria, has the eyes of the Gorgon closed, and the hero’s
face averted downwards; here he seems to be indicating that it is
 The Shield of Perseus 7

looking on the head (and his sculpture) which petrifies – her eyes need
not be open. Freud, identifying Medusa’s head with the mother’s
genitals, also lays stress on seeing the head rather than being seen.17
Presenting decapitation as cognate with castration, and thus the fear of
Medusa’s severed head as ‘a terror of castration that is linked to the
sight of something’,18 he thus elides a further connection that the mad
and violent eroticism of Cellini’s sculpture makes explicit – that the
decapitation of the female monster by the male hero is also a protection
against the threatened castration. (The closed eyes also make sense in
this context, since having them closed functions to deprive the Gorgon
of even residual power.)19
The psychoanalyst Sándor Ferenczi, a long-time friend and corre-
spondent of Freud, gives a similar reading, speaking of the ‘frightful
impression’ made by the ‘terrible symbol of the female genital region’.
Unlike Freud he also mentions the ‘fearful and alarming starting [sic]
eyes of the Medusa head’, suggesting they have ‘the secondary meaning
of erection’.20 By this he presumably indicates that some sort of power
emanates from the Gorgon’s eyes themselves and since he, like Freud,
identifies the presence of symbolic penises (the snakes of her hair) as
further symbols of the castrative danger (‘the multiplication of penis
symbols signifies castration’, writes Freud),21 we can read this as a
similar case, i.e. if the eyes symbolize penile erection, the signs of danger
can be read in these ‘alarming starting’ eyes too.
Roger Caillois, in The Mask of Medusa,22 identifies various legendary
creatures, found in bestiaries and traveller’s tales from Pliny the Elder
onwards, as rationalized or corrupted versions of the Medusa. The
notorious Basilisk, as well as the less well known Catoblepas and
Gorgo (the name seems telling here), all destroy what they touch or see
by means of baleful emanations, often from the eyes. The sheep-like
Gorgo’s heavy mane screens off its deadly vision; should it shake its
locks from its eyes, the results are fatal for those it beholds; the
Catoblepas also has poisonous eyes, and a head so heavy it can barely
lift it from the ground. The corrupting influence of the Basilisk is well
known: scorching the earth before it, killing vegetation, and splitting
rocks with its malignity, Caillois reports that it is nevertheless ‘killed
by its own reflection’ – ‘the surface of calm water is quite fatal to it.’23
That all these creatures are said to be found in the deserts of Libya or
Ethiopia adds credence to his interpretation, since much of the final
sequence of the myth takes place in Ethiopia, and Ovid attributes the
fact that Libya is ‘infested with poisonous reptiles’ to a few drops of
the Gorgon’s blood dripping down on the country as Perseus flew
8 Black Light

homewards – ‘the earth received them and gave them life as a medley
of poisonous serpents’.24
Further evidence that the danger is not merely from seeing her, but
stems from an actual emanation, comes from the fable contained
within the story about the origin of coral – having dispatched the sea
monster sent to consume Andromeda, Perseus places the severed head
on a bed of seaweed while he ritually cleanses himself. Ovid tells us
that the seaweed ‘absorbed the force of the Gorgon’, and was turned
into coral.25
Other versions and transformations could no doubt be found.
Caillois is almost certainly correct to link the motif in its entirety to a
very generalized fear of being looked at which would cover not only
all similar myths, but also various masquerades and folk beliefs
concerning the evil eye. (Caillois being Caillois, he is even prepared to
argue that it seems to extend in a similar form into the non-human
world, where it is evidenced by the proliferation of eye-spots in insects
as a protective measure against predators.) ‘Almost everywhere,’ he
writes, ‘we see in man this tenacious, almost ineradicable, fear of the
eye whose gaze paralyses, roots him to the spot, suddenly deprives
him of thought, movement and will.’26 We can be additionally sure
that it is this danger which Perseus must avoid, as well as the danger
of seeing Medusa, through the fact that the cap of Hades is a magical
source of invisibility – its presence as a significant accoutrement of the
hero is in fact unintelligible unless there is a real need not to be seen.
Hesiod mentions it in context of his flight from the Gorgons, saying
that it ‘had the awful gloom of night’ about it;27 Hyginus, in the most
explicit explanation of its purpose, describes it as ‘a helmet which kept
its wearer from being seen by an enemy’.28 It is the rhyme for the
shield which allows the bearer to see, and points decisively to the fact
that the sight of the Gorgon is deadly, not merely the act of setting eyes
on her. The texts do not suggest this meaning for the cap –
Apollodorus merely says it helps him escape after he has killed
Medusa, and Ovid does not mention it at all – but it seems obvious
that to see without being seen is the condition for success.
Here we must return again to the matter of reflection. Perseus’s
safety is ensured provided he does not look upon the deadly sight of
the real being but only upon the reflected image: the image on the
shield cannot see anything, nothing dangerous emanates from it. In the
shield’s mirror, transformed to an image, the Gorgon is disarmed: an
image does not see anything, and so it is safe to look at. With its help,
the real danger, the deadly aspect of the living monster, can be
 The Shield of Perseus 9

defeated; Medusa is undone by the impotence of her own image

(Caillois somewhat perversely prefers to think she is killed by it, rather
than by Perseus, whatever the sources might say).29
The ambiguous difference between seeing and being seen is thus
exploited to the fullest point – for Perseus, it seems that being seen is
made safe since he manages to see by proxy, as it were. He is seen only
by a picture, which has no power to look; thus he can look on it with
impunity. It also seems, then, that the shield-mirror protects him from
his own look as it meets hers: for it is not merely being seen, but seeing
the face of Medusa which is the danger. His shield allows him to see her
seeing him.
Here, then, is a parable about the power of the image: it does not
look, but it can show, and therefore it can be looked at. It can protect its
bearer (just as the image of an eye gives protection against the evil eye
of folklore) against the danger that comes from without. It shields –
with the literalism that so many Greek myths possess, it is actually on
a shield – from the look that kills, and enables its bearer to defeat the
monster whose look is certain death. The image – the perfect mirror
image, the representational image – protects those under its sign, and
gives them safety. It provides the freedom to look, the freedom to act,
the freedom to be victorious against a deadly, horrendous enemy.
(Later, Perseus presses the Gorgon’s gaze itself into his service: with
it he slays his enemies in the house of Cepheus and the house of
Polydectes. Here he takes advantage of those who do not have the wile
of representation to protect them: they are forced to see the full deadly
horror of being looked at. But he knows to avert his eyes, to avoid
looking at being seen.)
For Perseus, reducing the world to an image can deliver him from
its danger; for the Florentine Alberti, the painted image wrests some
truth from the world’s deceptions. Either way, the import is the same –
in Alberti we simply see the task assigned to the shield transferred to
painting. The oblique formal relation between the myths of Narcissus
and Perseus seems to be illustrated (perhaps unconsciously, perhaps
not) in the work of certain antique artists – there exist Etruscan
mirrors, and vases from southern Italy, which are decorated with a
scene of Perseus and Athena looking at the reflection of the Gorgon’s
severed head in a pool of water.30 What gulled Narcissus is as
beguiling as ever, only now it is attainable in painting, and what
makes Perseus’s enemy strong is the very thing that, once reduced to
an image, seals her fate. Picture the world, and you control it; but do
not look directly on it, and do not let it look at you.
10 Black Light

The Advice of the Mirror

Here is Alberti again, assuring us now that reflection gives the clue to
the improvement of our paintings, those beautiful fragments marvel-
lously grasped from Narcissus’s own impossible, untouchable mirror:
‘I do not know how it is that paintings that are without fault look
beautiful in a mirror; and it is remarkable how every defect in a picture
appears more unsightly in a mirror. So the things that are taken from
Nature should be emended with the advice of the mirror.’31
Reflect again the truth you have captured, and its own beauty and
defects alike will become apparent; it will itself become clear and truly
visible, and those errors it might have inherited from Nature herself
will be seen and may be ‘emended’. The mirror’s advice will be
objective; improving, even. The painting, coming into form as a
worldly object, may begin to suffer a transformation into as faulty an
object as its model: the mirror will remedy this – as paint is applied,
and effects are created, the mirror is ‘an excellent guide’ to let you
know when you have ‘arrived at what is required’.32 Backwards and
forwards: first the world as illusive reflection, then reflection as your
guide. Drawn off from the surface of the glistering pool into the fixities
of paint, the mirror – parody and fragment of the greater reflection –
will inform you just how to render these images truthfully and
without fault. There are, however, mirrors with their own faults.
Edouard Manet’s A Bar at the Folies-Bergère (1881-2; Figure 1) is a
painting about which rivers of ink have already been spilled. Manet
himself, the world in which he lived, and his specific place in it have
also, and to an extraordinary degree, become objects of scholarship.
The fierce academic intensity of the study devoted to this man and his
world is forbidding; long awarded a variety of historical stations, all of
the gravest significance (the founder of modern painting; the first truly
avant-garde painter; he in whom the first stirrings of painterly
modernism itself are clearly visible; the living avatar of 19th century
Paris, and so on), Manet is the object of a kind of cultic veneration.
With idolatry comes anxiety, in this case manifested in the obsessive
marshalling of sources. A world has been produced for the painter and
his paintings which is saturated with reference and implication, a
world so detailed that it does not in fact really correspond to any
world we can truly imagine being inhabited. It is a total world where
the historical significance of all things is somehow already present and
engaged with, their historical functions somehow on display. But it is
the condition of living societies that their internal structure is
 The Shield of Perseus 11

mysterious to them, like the structures of language in the act of

speaking: this is precisely why the words of contemporary critics, the
presentation of which is an almost unvarying feature of writings about
Manet, are so important – they display something their authors could
not imagine was on show, something that at best they only half-knew.
Cantaloube, Claretie, Chesneau, Stop: how many other second-rate
19th century art critics and bad cartoonists are as familiar to us as these
and their like, whose negative reviews lashed Manet at all those
notorious Salons and whose bizarre and telling parodies poked fun in
the newspapers and magazines? The words of these critics appear to
us now as so much unselfconscious testimony, given without
hesitation, like dreams from the analysand reclining on the couch.
The same can be said for the posters, the advertisements, the café-
concert playbills and all the rest – an endless parade of ephemera, all
of which now seems pregnant with meaning, innocently confessing
the secrets of their epoch to Manet’s diligent historians.
These critical accounts, though, essential as they are, seem in fact to
take us away from both Manet’s Paris and the Manet of our age, and
into an art historian’s Paris, where we are guided by desk-light
through the darkness of time past an impressive and intimidating
mass of materials and sources: paintings famous and obscure, popular
prints and reviews, letters and posters and titbits of information about
what and who was celebrated, decried, famous, unknown, and so on.
This materia prima almost constitutes the very study of Manet, as
though this meticulous historical trawl was necessary before it was
possible to look at the work; invocation of it is a rite, as if it was
inadmissible to speak about Manet’s paintings without the pious re-
iteration of what has previously been said, all the way back to the
famous salon of 1865 and the Olympia furore.
None of this has ever served to make Manet’s paintings any less
mysterious. In fact, the paintings themselves tend to make the mass of
writings look ever more irrelevant and strange; it is almost as though
those who have written on him are forced to examine everything about
his life and time except his paintings, as though the intense
examination of his world will finally yield a way of looking at his
work that removes its unnerving power and slippery elusiveness. The
scandal of Manet’s critical reception is the locus of the great power
shift in criticism and judgement, the moment when the ability to
recognize the necessary qualities in art shifted away from establish-
ment critics to artists themselves and their maverick champions – the
Zolas and Baudelaires, and later the Schukins and Vollards – who got
12 Black Light

it right from the start: his dismissal by his contemporaries and their
persistent failure to recognize his greatness marks the moment where
art criticism made probably its greatest and most notorious blunder.
Critics and art historians have returned again and again to this trauma,
determined every time to correct it anew; it is the ultimate failure, in
permanent need of remedy, and every fear that an assessment of an
artist might prove incorrect in the historical long-term is referable to it.
Bataille is right, in his book on Manet, when he writes that the
transfixed, hilarious public and the savage criticism which greeted
Olympia indicates a mystery in itself – they had all been ‘led by some
unerring instinct into the lair of danger’, and the violence of the
response is the most telling proof of the power in Manet’s painting.33 It
is a danger that art historians and writers have understandably and
belatedly tried to protect themselves against, building up around the
artist and his work an encasing edifice of historical matter. Manet is
the limit case for error, the blank and alien sign of unintelligibility and
mistaking, the permanent reminder that the current of history is swift,
cold and unpredictable.
What, then, of the Bar, Manet’s endlessly strange final major work,
completed a year or so before his death? In addition to the enormous
literature on Manet (in which it invariably features), there is an entire
book of theoretical essays by various authors devoted solely to its
impenetrable mysteries, a critical laying-siege to the work which is –
predictably – in great part concerned with prior art historical
approaches to the painting.34 Despite the thick patination of study, it
remains variable and multiple; faced with the painting, the various
interpretations and theories offered fall away, and when recalled
appear distant and extraneous, as if they were talking about something
else entirely. In any case, nearly all of them are different. Other major
works such as Olympia or Dejeuner sur l’Herbe have been more or less
stilled by the academy’s unremitting gaze, their meaning and
significance largely fixed in place. For these great paintings the litany
of sources is known by heart (Titian’s Urbino Venus, an engraving after
Raphael’s Judgement of Paris, Goya’s Naked Maja, Ingres, Velasquez,
and the rest); they are by no means any less troubling than the Bar, but
they have been comprehensively accommodated. Their reception by
the public is a famous legend, as noted above; the oft-parroted list of
important features in the paintings (unacceptable nudity, contempor-
ary dress, lack of justification for scene depicted, spectre of prostitu-
tion, gaze directed at viewer, flatness, etc.) is also well known. Their
place in the history, and how this history is thought to move with and
 The Shield of Perseus 13

through them, is steady. The received truths about Manet guard them
like sentries, and keep the order of play correct: first Manet, then
Impressionism, and so it goes.
Not that there is not some truth in these asseverations; it is not for
nothing that Bataille so memorably describes the advent of Manet’s art
as the coming of ‘that moulting time’35 for painting, its cracked and rigid
casing dramatically sloughing off in his work. That was, in general, how
the matter was seen by sympathetic contemporaries, as well as now. But
the Bar doesn’t feature in this story in the same way – despite its
importance and majesty being generally acknowledged, it does not have
the recognizable place and narrative historical function assigned to it that
the earlier paintings do, or even an agreed meaning. It does not because
it will not; it holds out, obdurately, as the theories and source materials
proliferate. There are real and enduring difficulties in A Bar at the Folies-
Bergère; they arise because the genuinely perturbing forces harnessed in
the painting are directed outwards and seem specifically and intention-
ally to damage and limit any attempt to understand it. It refuses to have
a meaning, an explanation or a solution. These problematic forces, forces
that continue to reflect away all attempts at definitive interpretation, are
made manifest in the painted surface of a mirror.
Stubbornly and wilfully incomprehensible, the famously erroneous
mirror that takes up the entire background of the Bar simply will not be
made good. No wrangling with viewpoints or angles will resolve its
contradictions or relieve the frustration it generates. It refuses to be made
sense of; Bartleby-like, it would prefer not to, and its recalcitrance cannot
be argued against. We know it should reflect the reverse side of what we
are being told is immediately before us: a marble bar counter, a selection
of bottles, and the barmaid herself. Instead, it sends back a dizzying
profusion of jumbled elements, not one of which really corresponds to
anything we can see; but since all of them are sufficiently similar to what
we expect to see, we are sent round again in a maddeningly fruitless cycle
of identification and misidentification. Through the mirror’s work, the
picture exists in a condition of dynamic irresolution, permanently
oscillating back and forth between what we think happens at one
moment, but doesn’t quite, and what we think should not happen at the
next, but always does.
The details of the painting are familiar. Described and pored over
innumerable times, the multiplying accounts are themselves a parodic
echo of the way the picture forces the eye to shuttle again and again
over uncomfortable terrain, without permitting it any rest. The
counter, a cold and dark-veined grey marble, itself gently reflective
14 Black Light

(we seem to see a suggestion of the barmaid’s presence in it), runs the
length of the canvas along its base. At its left topmost extremity there
is the merest hint of a corner, cutting down at an obtuse angle and so
linking the bar with its reflection, which duly indicates that the
counter effectively finishes with the left edge of canvas, although the
perspective here is already rather muddling. At this left end, half a
dozen unopened bottles; aperitif, champagne, beer. On the right side, a
similar number. Between them, incongruously, a glass containing two
roses, and a cut glass compotier containing a bright rubble of ripe
tangerines, their gleaming skins a gentle rhyme with the golden foil of
the champagne bottles on the left.
The barmaid herself stands centrally, leaning on the bar, deceptively
flat until the lace jutting crisply forward from the picture-left side of
her jacket’s low, squared-off neckline suddenly gives her bosom an
unexpected volume. In fact, her body is not quite squarely on to the
viewer; her right shoulder (that is, picture-left) is slightly dropped, she
appears to lean slightly heavier on her right hand, and the irregular
outline of her hips and waist suggests her weight is on her right leg.
Her head gently leans picture-right, the briefest suggestion of
contrapposto; around her neck is a black ribbon necklace, with a cameo
or locket hanging down. Her jacket is a midnight blue-black, as infinite
as the cloak of the Virgin in Duccio’s Maesta, chased with a scumbled
white lace at the neck and cuffs, and studded with a dead-straight
constellation of perfunctory, flower-like white buttons down the front,
which is also the dead centre of the painting. Her skirts are silver grey;
at her breast a posy of flowers chimes with those on the bar-top below,
and Manet has emphasized them with thick daub of yellow-cream
paint, which shines in relief. She looks out, delicately but clearly made-
up, the powdered rose-pink of her cheeks in touching contrast to the
colder white flesh of her chest. The sheen on her neatly combed fringe
shines gold through hazelnut. The light, flat and source-less, seems to
be emanating from somewhere outside the painting, behind us and to
the upper left (this is, we can note in passing, effectively the traditional
source of the light of the annunciation).
The space behind the bar is indeterminate; the distance between her,
the bar-top, and the mirror appears by turns shallow and relatively
deep. The lower edge of the mirror runs parallel with the counter,
separated from it by a variably thick reddish band which reads as a
strip of wall visible beneath its gilded frame. It could be two feet away;
it could be five. This constitutes everything that at least pretends to be
‘real’; these few clues, condensed into the shallow space of the bar and
 The Shield of Perseus 15

the limited space inhabited by the barmaid, are all we have to go on.
The mirror’s surface takes up everything else.
In this surface we are presented with an entirely invented world,
apparently an immaterial double to the one that we are told is
physically before us; a world which, we must suppose, is occurring
behind us. It is thinly but liberally scuffed with silver-grey patches,
which perhaps signify the cold, glassy material existence of the mirror
(this seems certain at the point between her right forearm and the
champagne bottles, for instance), and which might, if that were the
case, allow us to grasp at least the even flatness of its physical
presence; but since elsewhere these patches might equally represent
the blue haze of tobacco smoke inside the reflected establishment, the
possible comfort that proof of the mirror’s real existence might bring is
also called into doubt.
We do not have any choice about reading it as a mirror, though; there
is too much to say that it is. The bar is there, the bottles are there, the
barmaid is there, all in what appear to be approximately the right places.
In fact, they are in nothing like the right places, but the illusion is
incredibly forceful, and has to be in order to fight off the equally
confusing sense of deep space generated by the distant balcony and the
apparently infinite recession, perhaps into further large mirrors, of
crowding top-hatted figures. Manet’s unmistakable choppy and irregular
brushstrokes eventually dice this host into an abstract, blocky hatching
which is finally melded into areas of vertically dragged indeterminacy,
areas which again seem to indicate the mirror’s surface. The mirror is in
general painted with looser work, with occasional passages of thick paint
(for instance, on the reflected bottles’ tops), which pull the recessed space
abruptly forward to the picture plane. This effect not only contradicts the
depth and distance created by the mirror as it reflects the room, but also
denies the ‘real’ space, anterior to the mirror, behind the bar, within
which the mirror should hold its place on the wall behind the barmaid:
the viewer is brought up sharply to the surface of the canvas, and the
already confused spatial effects are made even more complex.
Along the surface of the reflected bar are several dramatic
discontinuities, in large part stemming from the entirely gratuitous
introduction of a recessive single-point perspectival view along the
counter’s reflected left-hand edge. This 45 degree angle – whose
ostensible vanishing point is exactly behind the centre of the barmaids
face – seems responsible for pushing everything up- and right-wards
in a most peculiar way which superficially appears to respond to an
imagined viewing position, but in fact cannot really do so. As well as
16 Black Light

being up to the right, the group of bottles at left are on the far side of
the reflected bar, when they should be at the near side; in the gap
between the barmaid’s arm and waist there appears to be the pinkish
suggestion of a seemingly nonexistent bottle and what one can only
presume is the reflection of one of the champagne bottles – its location
is completely baffling. Passing over the reflection of the barmaid
herself, the cut roses are reflected at the right hand edge of the
painting; the distance between them and the left-hand group of bottles
seems to have telescoped wildly and they are now an immense
distance away from these bottles, as is the barmaid’s reflected right
hand (her reflection-left). Excepting the roses, nothing on the right side
of the bar is reflected at all; the organizing illogic of the mirror decrees
that they must presumably be somewhere off to the right-hand side of
the picture, though more precisely they simply do not exist.
And then there is the matter of the barmaid. She herself seems
physically transformed by the mirror; her back seems rounder than it
ought, and she appears generally more massive. (Manet here, as ever, is
a master of economy; in the blackness of her jacket, he opens the space
between her left arm and body with a single dry smear of light yellow).
Her posture in the mirror does not properly correspond either; she
seems to slump slightly, and to lean further over the bar than her
original, who is upright and graceful by contrast. Most worrying of all
though, looming in from the right, is her interlocutor. Top-hatted, with a
trimmed beard and moustache, his indistinct features defined by the
sharp white line of his far shirt-collar, he stands at the bar before her. His
cane is held in a hand that is a purplish, brusquely smeared paw, and a
dabbed touch of Manet’s brush lights and models his nose; his hat
extends upwards off the canvas. He appears to be very close to the
barmaid, despite the width of the bar top.
How are we to understand his presence? His relation to the barmaid
as it is shown in the glass famously indicates that, if he was really
there, he would be standing in exactly the position held by the viewer
of the painting. Jack Flam, in an excellent essay on the painting, has
noted his physical similarity to the man at the left-hand end of the
reflected balcony, a man who also appears to look directly towards the
place occupied by the barmaid (which is to say, directly at the viewer).
A possible reading is thus that we are seeing one and the same man,
and a psychological narrative of fantasy and desire is being played out
from left to right: his look, the object of his desire (the barmaid), and
the ‘imaginary encounter’ between them; the desires could equally be
hers, and perhaps we are even seeing ‘a kind of shared reverie, based
 The Shield of Perseus 17

on their mutual awareness of each other’.36 There is even no need for it

to be imaginary: there is nothing to say that each stage of the encounter
is not real, the painting presenting a simultaneous narrative in the
manner of the trecento Sienese masters. Whether or not we accept this,
the problem of the right-hand figure’s location is a conundrum which
will not go away: if we demand that a ‘real’ body is being reflected, his
relation to the girl means he can only be placed precisely where we think
we are. On the other hand, we also know that in truth he is not real or
present at all, no more so than the absent oranges are real for the mirror.
The account given here, like all accounts of all paintings, is partial.
There is more that could be said about the formal properties and internal
dynamics of the picture, and much more that has already been said
elsewhere; indeed, as noted above, the Bar’s internal contradictions
provoke infinite and unresolvable interpretative extension – the process
of describing and teasing out meaning could, and no doubt will,
continue as long as the picture has viewers. This seems to indicate that
the real problem with the painting is elsewhere. No solution will ever be
found within the fractured, restless scene: at every moment the viewer is
systematically disappointed and frustrated, and since no amount of
explanation will knit together Manet’s frayed and paradoxical weave,
any amount is possible, and all falls equally short, only managing to
prompt more – more historical study, more speculation, more
interpretation. The problem, it seems, is not what is in the picture, and
does not have to do with the complicated matter how the image works –
it doesn’t ‘work’, and never will. The problem is how the picture works:
how it is that it operates so successfully to elude the ascription of even a
relatively final meaning. Naturally this agility stems from the image
itself, and especially the programmatic confusions in the mirror (though
not them alone); but the question must be how we account for the
function of these disjunctions beyond their internal, pictorial ones. It is not
enough that we note the refusal of the mirror to make sense, and leave it
at that; in itself that is nothing more than paradox or irritation which
neither makes the picture nor makes the picture great. The question is
how it is that the entire ensemble, including but not limited to the mirror,
provokes a higher order of operation to begin, an historical operation
whose effect has been permanently to bind attempts at critical
interpretation into an exit-less maze of description, re-description, and
an ever increasing mountain of scholasticism and exegetic history? How
is it that this painting creates an impasse at which the false progress of
judgement and interpretation is paralysed, its usual assertive confidence
descending into nervous echolalia?
18 Black Light

Nature Mort

Something that cannot be ignored is that every single distinct item in the
Bar is readable at some level as a momento mori: the ripe fruit; the
unopened bottles; the attractive girl; the location itself, a place of
transient entertainments and pleasures; and of course the flowers. Flam
perceptively notes that in this context we may also see the reflected man
as a figuration of Death, to go with Maiden of the bar.37 The history of
iconology makes inevitable the conclusion that the painting must in part
be seen as a huge, complicated vanitas, painted by a dying man.
Manet, it seems certain, died from complications arising from
syphilis, probably contracted in his youth. The final, tertiary stage
appears to have set in during 1876; by 1878 he could no longer bend his
right leg. The nervous and muscular degeneration symptomatic of the
tabes dorsalis form of the disease had set in; he suffered the ‘lightning
pains’ associated with the illness, and walking became very difficult.38
At the time the Bar was painted, he was suffering gravely; effectively
unable to paint, his work was restricted to smaller pieces, and he had
even sought advice on the painting of miniatures. The Bar was the
exception, but it sapped his energy, and he worked in fits and starts.
In the final years between 1881 and his death in 1883, he produced a
series of small, extraordinary flower paintings; Gordon and Forge, in a
book devoted solely to them, reproduce 16, every one a jewel-like tour de
force.39 Of flowers, always a reminder of death, and always a touching
affirmation of life’s brief splendour and happiness, the dying painter
was said to remark, ‘I would like to paint them all’.40 Gordon and Forge
note that these flowers are given contrast by their vases: that the
exactitude and honesty of the bouquets have a counterpoint in the
splintered glitter of the glass containers.41 Working in concert with the
consummate depiction of the flowers themselves (which Manet paints
astonishingly, impeccably: he is the still-life painter non-pareil, perhaps
absolutely) there is an equal, possibly greater, and at least as breath-
taking effort to attend to the effects generated when light is bent and
baffled through glass, refracted and torqued through clear water where
the standing stalks are shivered into those familiar, angular brush marks.
Is this not, in some sort, the same opposition as that which we see in
the Bar between the mirror and its model? These crystalline crucibles
are as perfect and captivating as the flowers they contain. The
precision with which they are rendered is a moment of oscillation that
shows the other side of the visible blooms reaching up out of the glass:
their submerged selves, whose appearance is quite as visible yet
 The Shield of Perseus 19

impenetrably distorted. The blooms are summoned with decisive

power; but with them comes their shade, an image of the duality of
things and their representation; the flowers of Narcissus and their
unknowable counterpart in the limpid water.
It is unavoidable that the roses in their glass on the bar top are seen
as part of this series of flower paintings (the marble slab of the bar is
known to have been in his studio: many of the small flowers are
evidently painted on it too). On examination, it does indeed begin to
look separate from the rest of the larger picture, as if it were a painting
on its own. In fact it appears not to belong to the scene at all. The
perspectival treatment of the foot of the glass contradicts everything
else on the bar top; its curve does not fit with the bases of all the other
objects present, and indicates a much lower viewing position, one
which seems to be at close to 90 degrees from the picture plane. We
thus see the glass of roses very near to exactly side-on. It is certainly
presented in a very different line of sight to that indicated by the
decisive angle of the ‘real’ bar, and this an inconsistency which is every
bit as strange and worrying as the failure of the mirror to fit with its
model. Perhaps more so, since the painting is openly concerned with
that latter, more gross, failure and does not appear to be about the
random introduction of incompatible elements to spaces which should,
at least discretely, be whole. Even if the mirror and the ‘real world’ do
not fit with each other, they at least make some sense internally – indeed
this is what the picture depends on for its effect, since it makes its capital
from the disjunction between two discrete worlds which should be one.
But the flowers simply look like a different painting, as though Manet
has used the surface of the Bar – not the surface of the bar in the painting,
but the actual surface of the painted canvas – to support a different
painting. This troubling effect is made worse by the perfunctory nod to
the flowers’ existence which is present in the reflection: by contrast to the
rest of the mirror’s world, the assurance that they are there, in more or
less the right place, becomes disturbing. Their odd isolation is further
exacerbated by the fact that they exist entirely within the black void of
the barmaid’s jacket. Manet ensures they touch nothing within the scene
– the marble reaches to the glass’s thick clear belly, while the flowers rise
up against a completely black background; the curved grey wedge of the
barmaid’s skirt and the lace of her cuff are brought excruciatingly close,
but do not touch, and the modest bouquet is isolated.
The perspectival shift at the foot of the glass is extremely peculiar,
and forces a reassessment of the viewer’s relation to the picture. The
depicted scene itself indicates our position relative to it; the
20 Black Light

perspective is robustly, if oddly, directed to us by the angled edge of

the bar, and we are then fixed in position by the barmaid’s face and the
feet of the bottles. The flowers, isolated and side-on, redirect us to a
different position, one which is in more obvious relation to the picture
surface. Just how major this shift is can be seen by looking from the
glass back to the reflection of the bar top: the disjunction is severe,
making the reflected bar tip upwards radically, and this has a knock on
effect on the ‘real’ bar, which then dances between two possible angles
of incline, depending on whether we look at the flowers or the bottles.
Acknowledging the flowers as the anomaly, our relation to the picture
starts to flicker back and forth between the flattening plane of the flowers
and the distances and angles of the bar and its world. As an interruption
of the regular distortion of form that characterizes mathematical
perspective, it serves to interfere with the psychological singularity of
the viewer, that stationary, observing individual, locked in a fixed
position, which such perspective has both generated and presupposed
since the triumph of mathematical models in the Florentine renaissance.
(Viewed as a memento mori it also calls to mind the anamorphic skull in
Holbein’s Ambassadors, another huge and complicated painting, which
deals with the ephemeral nature of earthly power and knowledge, and
also features its subjects face on to their audience). It calls into being a
second order of looking: a looking at the picture surface that sits uneasily
alongside the view into the space of the scene.
Here we are perhaps close to the Bar’s secret. We cannot ignore the
fact that it is, perhaps primarily, a vanitas painting. The most obvious
statement of this theme – obvious because it is the most clearly
separated and most incongruous element in the scene, and because
historically it also finds a place in a series of final flower paintings
made by a sick and dying man – is found in the flowers. These flowers
introduce to that theme a level of direct communication with the
viewer that goes beyond the comfortable and learned world of
allegories and symbols. Though it is far less extreme than Holbein’s
famous skull, Manet nevertheless emphasizes the significance of the
vanitas aspect of the work in a similarly decisive moment of disregard
for internal pictorial coherence. By fracturing the picture’s already
fragile consistency (fragile: yet we never think the bar and mirror do
not belong to each other) with an effectively alien element, Manet
undoes the visual knot by which the viewer is connected to the scene
depicted; no longer bounded by the perspectival convention which
makes the space of the picture an extension of the space of the viewer,
the flowers announce their meaning without an intermediary. Their
 The Shield of Perseus 21

message is delivered raw, so to speak, to a viewer required to re-

register the picture as an object in a discontinuous world. Their
meaning – life, and death – is not being communicated to this
disorientated viewer by what is in the painting, but by the painting.
In Holbein’s great masterwork, the awesome interruption of the
scene by the huge elliptical skull makes the overall meaning of the
picture completely explicit. There is nothing uncertain about its
message: the trappings of power, knowledge and glory are limited,
and all the achievements of man will be overtaken by death. Behind
the sumptuous green curtain hangs the cross, just visible at top left;
when the curtain falls on life, a different kind of power and different
kind of reckoning to those of the sciences and arts whose symbols are
laid out between the two young masters will be revealed, the final
judgement of an observant and ever-present God.
But what is the message in Manet’s painting? Holbein’s skull calls our
attention with direct force to the allegorical content of The Ambassadors,
but Manet’s flowers are less certain. For if they are calling our attention
to something which is carried by the rest of the painting, then we are
brought back before the central problem again, a problem which is far
more ambiguous than the broken lute strings and rich fabrics of the
Holbein. This problem is the mirror and its rejection of the world it
should show in favour of a similar but persistently different one.

Mirror and Model

What can a representational image achieve? We have seen what it could
achieve for Alberti: his vision of the painter’s art, aided in its steps
toward perfection by that most perfect of representations, the mirror,
was of a practice capable of taking us closer to an uncertain world.
Painting can show us truth and beauty where we might otherwise
remain deceived, and give to us what we might otherwise be denied. On
the polished shield of Perseus, the image is protective: symbolized in the
glare of Medusa, the dangers of the world, represented in reflection, are
incapable of harming us. The representational image is capable of
truthfulness; it is capable too of illustrating those things which are
themselves mortally dangerous, and of making them safe. Indeed, life
and death – all the world, and even its ordering principles – can be
captured, displayed, acted upon and manipulated by the painter as he
reflects them, and reflects on them, in the canvas’s painted mirror.
It does not appear from the Bar that Manet shares this kind of
positivist optimism, and here we must look carefully at the mirror
22 Black Light

again and ask: what is happening here to the representations we

expect? For the overall function of the disappointments and decep-
tions of the mirror in the Bar is to provoke a continually failing search
for representational resolution. Manet has intentionally put any such
resolution definitively beyond reach; in the impossible space of the
Bar, the representative imagination, which is so adept at recognizing
objects and spaces in images and colours, begins to seize up. The
mirror cannot show the bar but, even more than that, it cannot not.
Manet ensures there is no alternative to connecting them: the mirror is
not released from its moorings to the bar entirely, but has been
permitted to drift away from them. The situation is thus more serious
than a complete dissociation between bar and mirror would be:
instead of denial or dismissal, he introduces uncertainty. The unclear
relations between the bar and its reflection uncover the mysterious
link between the idea of representing things through images (under
which lies hidden the very idea of all signification) and the painted
representation itself. In the space between the mirror and the bar there
occurs an alarming disturbance between the sign and the thing
signified. The discomfort and uncertainty that is inevitable when
looking at the Bar comes from a disruption of the conventions of
representation, a disruption that displays the vulnerability of any
construction of reality. Representation – and by extension, the sign
itself, and all of language – is brought into question, and when the
arbitrariness of its operating principle is exposed, the fragility of the
world of signs becomes obvious.
The gap between our knowledge of the world and the world itself
usually remains hidden behind the signs and representations that
figure and refigure perception into the artefacts of language and
image. Manet allows the imagined signs in the mirror to slide gently
away from what they signify on the bar, and in doing he shows us the
arbitrariness of the sign; what we compulsively search for, what we
automatically – logically – expect to find, simply may not be there. The
idea that painting should or could show something – what sort of idea
is this? It is an idea that depends on everyone playing by the rules, and
no-one saying that the game itself is a precarious fantasy; but it is only
a matter of refusing these rules in order to cause the whole house of
cards to come tumbling down.
This, in fact, was always Manet’s practice, as we are so often told in
the literature: to strip the accreted layers of fiction from representation,
to present and represent things somehow ‘just as they were’, whether
the model on the studio bed whose false existence as a mythological
 The Shield of Perseus 23

goddess was cast off, or the young gents in ordinary dress lunching on
the grass. It is not that Manet does not believe in the possibility of ‘good’
representations, pictures which really can show the world as it is; if he
did not, then finally to declare against, through the creation of a painting
which functions as a negation of this possibility, would have no power. If
anything, he has pushed too hard at the problem of representation:
because in the end the search for the true nature of the image reveals
nothing at all, a mask behind which there is no face, a merely arbitrary
language of expectations and conventions: ‘bottle’ for bottle, ‘barmaid’
for barmaid, and so on, without the secret heart of it being revealed. This
is what Manet runs up against: and so, at the last, he pushes beyond like-
for-like representation and on into the problem at its core: the final
arbitrariness of the signs that are used to denote things.
This arbitrariness indicates something further and more frighten-
ing: for if the representative sign is arbitrary, then what it denotes
cannot be grasped, and is finally unknowable. Real, alien, and active,
the world is not capable of being known or controlled by representa-
tions, however perfectly they reflect what we assume we see, or what
we seem to know. A mirror reflects the world perfectly; the invocation
of this perfection is what gives the painting its strange charge, for if we
are presented with the things we seem to be shown on the bar, then we
expect the mirror to do the work it should do ‘in real life’ and show
them in reverse. A touching expectation, Manet seems to say, but one
based on nothing except the mere faith that I will do what I am
supposed to do, now I have shown you a picture of a mirror. But he
has not done what he was supposed to do; instead he shows us exactly
how misplaced our faith in images is, exactly how little control we
have over what they do and how they are made (and also displayed
here, of course, is the extent of the control that the makers of images
possess over the viewers of images). By extension, we are shown how
arbitrary and limited our understanding of the world is, and how little
power we have to protect ourselves from the moment when it calls
time on us, and death returns us to mere matter.
The mirror of the Bar is the antithesis and negation of the shield of
Perseus; in the sliding away of the image we see a final retort to the
happy dream that the mortal gaze of the world itself can be happily
deflected by a clever ruse, set down on canvas and held fast. Instead,
Manet shows us something else: a glimpse of the gap between our
world of signs and representations and the vast cosmos, which they
feign to deliver. Paltry mutterings and meaningless daubs in the face
of a universe as indifferent as the barmaid. ‘Know yourself’ was the
24 Black Light

motto of the oracle at Delphi. Manet responds dryly: Know yourself?

You do not know yourselves, or the world, and you cannot. What
made you think painting would help you? Everything you saw, you
can be made to see again, wrongly; these bottles are not where you
thought they were, this bar and this girl and these oranges are not
where they ought to be, and you? You have been replaced by this man
in a top hat; you are not the person you imagine you are, not in the
picture, and not in real life either. (One is reminded of Lacan’s remark
concerning the gaze, and the disappointment of its desires: ‘The
subject is presented as other than he is, and what one shows him is not
what he wishes to see.’)42 Contra Alberti, painting will not give you
everything you want, and it will not protect you: not from deception,
not from the world, and not from death. And remember, says the
dying man as he paints his final big picture, his great vanitas,
remember that you too will die.

Watch out for your eyes

In 1881, Manet was awarded his first academic honour, a second-class
medal at the Salon; it meant he no longer had to submit works to the
jury, and no longer had to risk rejection and exclusion. After a career
marked by misapprehension, vilification and scorn, he could now
show the public and the critics whatever he liked. At the Salon of 1882,
he showed them the Bar.

Polydectes . . . who ruled the tiny island of

Seriphos,
failed to be moved by the dangers that Perseus had faced
and the courage
he’d shown in so many exploits. He still held on to his
stubborn,
relentless hatred and never ceased from his unjust anger.
He even belittled the hero’s glory and claimed that
Medusa’s
death was a fiction. ‘Here is the proof that my story is
true,’
cried Perseus to all. ‘Watch out for your eyes!’ And the
face of the king
was turned at once to bloodless stone by the face of
Medusa.43

It would be possible to identify the barmaid, with her glazed,

indifferent stare, as a kind of Medusa. In this connection it would
also be possible to invoke Freud and Ferenczi, with their ‘terrible
symbol’ of the female genitals, and note, as Flam does, the clear
 The Shield of Perseus 25

suggestion of the female sex in the grey triangle of the barmaid’s skirt.44
Recalling Manet’s syphilis, the link from the barmaid-as-Gorgon to the
old misogynist mythology of the vagina dentata would then be an easy
leap. Too easy, and also inaccurate – it would not accord with the picture,
which in truth suggests none of those things. If anything, it is the
barmaid that seems statue-like, petrified; she does not transfix us with a
terrible stare – indeed, she is more being looked at than looking.
While the mirror of the background delivers its various riddles to the
viewer standing before the famous canvas, another kind of mirror effect
is also created: one which identifies the barmaid with the observer. Their
positions are set in a mirror-like relation: facing each other frontally, with
the canvas’s lower edge approximately equal to the bar’s lower limit, the
viewer and the barmaid look at each other through the surface of the
picture as though one were the reflection of the other.
Once noticed, it is quite hard to shake this impression: looking at
the picture begins to feel like looking in a mirror, its invisible glass
along the front of the bar. This effect is accentuated by the fact that
most of the picture is a mirror, which is indeed showing the reflected
image of a scene apparently behind us. It receives a further
reinforcement via the reflection that is ostensibly ours – the top-hatted
man is in the ‘wrong place’, while the space in which we would expect
our reflection to be found is occupied by the barmaid alone. A further
element of the painting encourages us to identify with the barmaid,
rather than the man: the similarity, noted by Flam, between him and
the man on the balcony at far left. The possible encounter between
them has been remarked on above; but while the right-hand version of
this man is clearly looking at the barmaid’s reflection, the man at left
only appears to be looking at her through the fact that we project a line
of sight for him that would have him looking straight at her. But in
point of fact he is looking directly at us, thus setting up an uneasy
equivalence between ourselves and the barmaid. Perhaps the most
convincing argument for the notion of the viewer-as-barmaid is found
in the bottles of the reflection: for we have noted they are not
accurately reflected, but on inspection they are more or less accurately
reproduced. That is, the reflected barmaid’s relation to the reflected
bottles – which bears no resemblance to the ‘real’ barmaid’s relation to
the ‘real’ bottles – is in fact a reproduction of the viewer’s relationship
with the ‘real’ bottles that stand on the bar.
Mirrors, which appear to reverse left and right, in fact reverse front
and back. (This is what Magritte, the great logician of Surrealism,
recognizes in the famous Not To Be Reproduced of 1939.) Let us suppose
26 Black Light

for a moment that we are actually looking into a mirror, a possibility that
Manet does indeed seem to leave open, if only by implication; if the
barmaid is taken as a reflection of the viewer, then the painting redoubles
into an endless hall of mirrors, and the mirror of the background shows
this viewer identified not with man, but with the barmaid: it thus shows
an image of the viewer looking at the painting itself.
And what then is this painting? How is the act of looking at Manet’s
great work presented? What, in other words, does Manet tell us we are
looking at? He tells us that we look into the face of the mysterious, top-
hatted man, identified already as a figuration of death, a disembodied
presence, a reflection without a model. There in the mirror, invisibly
waiting for the viewer to look in and try to fathom its depths, is the
Gorgon’s face, a black-hatted Mephistopheles, the cold glare of Death
himself. It is symbolized by the man, not by the barmaid; and it is the
secret message of the painting, an image of the magic mechanism which
keeps the picture forever slipping away from understanding. Look at the
picture, and you yourself are looked at: it is you, the viewer, who is
transfixed by the painting, for death is looking back at you.
Finally, is this man not in some sense Manet himself? A Parisian
dandy, well-dressed and well-mannered, familiar with the girls and
bars and social life that the Folies-Bergère exemplified; a portrait, if we
are correctly informed, of another painter, Manet’s friend Gaston
Latouche? And is it not finally the artist who turns the living to stone,
whose gaze fixes them in their places with mixtures of pigment mined
out of the earth itself (the lapis blues and malachite greens) or, like
Medusa, renders them into cold marble? At his final salon, the dying
Manet holds up Medusa’s head to the crowds and critics that had
mocked him for twenty long years; it is a proof of his power, proof that
his story was true, as with Perseus in the house of Polydectes – Watch
out for your eyes!

You want to see? Well, take a look at this!45

I must, to begin with, insist on the following: in the scopic field, the gaze
is outside, I am looked at, that is to say, I am a picture.46

Who is looking at who in Manet’s baffling, frightening hall of mirrors?

We look at the painting; but stare too long and it seems the painting
looks at us. We might say this look is symbolized by the man; but a
picture of a man cannot look at us. It is the painting itself which looks.
Because the image in Perseus’s shield cannot not look, it can show.
 The Shield of Perseus 27

Manet’s image does not show anything definite, he has made sure of
that. Instead, it can look.
Caillois, as we have seen, notes the prevalence in the insect world of
‘ocelli’, eye-like markings which function to frighten or distract a
predator, and links them to the transfixing gaze of Medusa, the various
mythical beasts whose eyes paralyse and kill, and the plethora of
folktales and beliefs which feature the evil eye. The ocelli of insects
work because they give predators the illusion of being looked at by a
different, more dangerous creature than the one which they actually
belong to; the ocelli of myth really are deadly.
Manet’s picture itself has something resembling Caillois’ ocelli: the
white orbs at the top, nominally two round electric lights but which
are in fact simply pure white unmodelled circles. Viewed in this way
they become disconcerting: the images of the painting dissolve, and
the object, suddenly indeterminate, is in possession of two piercing
eye-spots. Perhaps this the most radical moment of all in Manet’s
extraordinary painting: the moment it escapes being a painting and
becomes a mask, charged with a power that goes beyond its image.
This is not the same as the hoary ‘acknowledgment of materiality’, or
drawing attention to the picture plane, or any of art theory’s old saws.
No, here Manet uncovers a most secret mechanism, the moment of the
superstitious engagement and identification with inanimate matter –
the very creation of meaning from raw material, the moment real
significance is forged from configurations of mere paint.
The uncertainty of the mirror, the indeterminacy of the viewers
identity, the interruption of the flowers, and finally these strange eye-
spots – in all this unstable flickering we catch glimpses of a concealed
principle, which can never be looked on directly, but might perhaps be
intimated, or momentarily caught sight of in a mirror. Manet discloses
for us the moment when painted patterns – mere swirls – are
compulsively identified with the objects they seem to stand for, and
then displays this process as arbitrary, faulty and fragile. In the
transfixing eyes of the picture Manet shows to us in all its majesty and
drama the birth of the representative dyad – viewer and image, the one
demanding the other and holding them in place long enough for pattern
to become picture, and for the play of arbitrary forms to become imbued
with all the gravity of meaning and significance. And this secret moment
is made briefly visible only because Manet immediately takes away what
he has given: first the image of the bar, then the not-image of the mirror.
The wild and uncontrollable force that is at play in A Bar at the Folies-
Bergère – the force that makes it ever more confusing and that has
28 Black Light

prompted mountains of study without the picture ever having been

tamed – is the involuntary animating force on which representation is
founded. It is the secret agency which is at work in the creation of all
images, the unknown compulsion which makes line and colour and
shape suddenly appear to have depth and weight and form: the ancient
magic of Lascaux that causes mere marks to represent something else,
something real and material. And finally, of course, it is a manifestation
of the greater mystery that underpins the creation of all human meaning:
the mystery of the birth of language and the sign. This is the eternal
question, the Sphinx’s riddle; and the vertiginous impossibility of the Bar
is the endless desire to unravel the mystery of that birth.

Notes
1 Leon Battista Alberti, On Painting [1435] trans. Cecil Grayson (London:
Penguin, 1991), 61.
2 Ovid, Metamorphoses trans. David Raeburn (London: Penguin, 2004): book
3, lines 341–3 (henceforth Ovid, 3: 341–3).
3 Ovid, 3: in order, 439, 417, 434–5.
4 Alberti, op. cit. note 1: 60; 61.
5 This famous discussion occurs in Republic, book 10.
6 There had been various Latin renderings of Plato made in Italy before
Marsilio Ficino’s major work of translation, at the behest of Cosimo de’
Medici, in the 1460s. The Republic had received a translation in Milan, by
Uberto Decembrio, as early as 1402, but there is nothing whatsoever in On
Painting to indicate Alberti had read it, or indeed had ever read anything
at all by Plato. For a detailed study on the Renaissance reception of Plato,
see James Hankins’s Plato in the Italian Renaissance (2 vols., New York: E. J.
Brill, 1990).
7 Alberti, op. cit. note 1: 61.
8 Ovid, 3: 407.
9 Ovid, 4: 604–803 and 5: 1–249.
10 From a literary point of view the Library is, as its translator Sir James
Frazer remarks, ‘the dull compilation of a commonplace man, who relates
without one touch of imagination or one spark of enthusiasm the long
series of fables and legends which inspired the immortal productions of
Greek poetry and the splendid creations of Greek art.’ J. G. Frazer,
Introduction to Apollodorus The Library, trans. J. G. Frazer, (London:
Heinemann, 2001 [1921]), xxxiii.
11 For full sources, including artistic sources, see the Timothy Gantz’s
invaluable Early Greek Myth, vol. 1, (New York: Johns Hopkins, 1993)
pp.303–7, and Jenny March’s equally useful Cassell Dictionary of Classical
Mythology (London: Cassell, 1998). Here are to be found detailed the
variations, discrepancies, visual sources etc.
12 Apollodorus ibid.: book II, ch. iv, paras. 1–2 (henceforth II.iv.1–2).
 The Shield of Perseus 29

13 Ovid, 4: 796–803. Hesiod however says they lay together in a ‘soft

meadow amid spring flowers’. Hesiod, ‘Theogony’, pp.78–154 in Hesiod;
Homeric Hymns; Epic Cycle; Homerica trans. Hugh G. Evelyn-White
(London: Harvard, 2002 [1914]), 99.
14 Ovid, 5: 217; 241.
15 Apollodorus, II.iv.2.
16 For instance Ovid, 4: 780–1 or 5: 200–3.
17 Sigmund Freud, ‘Medusa’s Head’ [1922], trans. James Strachey, in
Standard Edition. . . vol. XVIII; Beyond the Pleasure Principle, Group
Psychology and Other Works, (London: Hogarth, 1955), 273–4.
18 Ibid.: 273.
19 Cellini’s Perseus, with its huge and vicious upward-curving sword
stationed horizontally near his waist, is thus virile Man victorious over
castration, that is, over the danger of a desire that may result in castration
– a victory achieved through the destruction of the object of desire. This in
turn seems related to the meaning of the chaste Athena’s hatred of
Medusa; as Freud notes, her positioning of the Gorgon’s head on her aegis
‘repels all sexual desires’ (ibid.).
20 Sándor Ferenczi ‘On the Symbolism of the Head of Medusa’ [1923] Further
Contributions to the Theory and Technique of Psycho-analysis compiled. John
Rickman, trans. Jane Isabel Suttie et al. (London: Karnac, 1980 [1926]), 360.
The piece is very short: a mere three sentences.
21 Freud, op. cit. note 17.
22 See footnote 14 for bibliographic details.
23 Roger Caillois The Mask of Medusa [1960] trans. George Ordish (London:
Victor Gollancz, 1964), 101. Again, Caillois may well be taking some
liberty with his sources – Pliny says nothing of the sort, in fact he indicates
that only weasels are capable of killing it.
24 Ovid, 4: 616–20.
25 Ovid, 4: 745.
26 Caillois, op. cit. note 23: 101.
27 Hesiod ‘The Shield of Heracles’ pp.220–54 in Hesiod etc. op. cit. note 13:
237.
28 The Myths of Hyginus trans. and ed. Mary Grant, (Lawrence: University of
Kansas, 1960). See p.194 for Hyginus’s account of Perseus and the
Gorgons.
29 ‘I suspect rather he used the mirror to reflect back her own fatal visage.’
(Ibid.: 99.)
30 Gantz, op. cit. note 11: 306.
31 Alberti, op. cit. note 1: 83.
32 Ibid.
33 Georges Bataille, Manet tr. A. Wainhouse & J. Emmons (New York: Skira,
1955), 61.
34 12 Views of Manet’s Bar, ed. Bradford R. Collins, (Princeton: Princeton
University Press, 1996).
35 Bataille, op. cit. note 33: 69; italics in original.
36 Jack Flam ‘Looking into the Abyss: The Poetics of Manet’s A Bar at the
Folies-Bergère’, pp.164–88 in Collins, ed., op. cit. note 34: 182.
30 Black Light

37 Ibid.: 172–3.
38 Beth Archer Brombert’s excellent and original biography Edouard Manet:
Rebel in a Frock Coat (Boston: Little, Brown & co, 1996) contains a detailed
section on Manet’s illness and death. See pp. 398–9 and pp. 452–3. The
book also contains an excellent account of the Bar, pp. 439–43.
39 The Last Flowers of Manet Robert Gordon and Andrew Forge (London:
Thames and Hudson, 1986).
40 Ibid.: 5.
41 Ibid.: 14.
42 Jacques Lacan, The Four Fundamental Concepts of Psychoanalysis [1973] tr.
Alan Sheridan (London: Penguin,1994), 104.
43 Ovid, 5: 242–9.
44 Flam, op. cit. note 36: 167.
45 The phrase is taken from Lacan (Lacan op. cit. note 42: 101). In passages
which draw heavily on the work of Roger Caillois, the psychoanalyst is
speaking about the message that the painter sends through the painting to
the viewer: ‘The painter gives something to the viewer who must stand in
front of his painting which, in part, at least, of the painting, might be
summed up thus – You want to see? Well, take a look at this!’
46 Lacan, ibid.: 106.