Abstract

This essay suggests that the minimal 1966 exchange between Jacques Lacan and Michel Foucault in
Lacan’s seminar actually stood in for a much fuller debate about modernity, psychoanalysis and art than
its brevity would indicate. Using their contrasting interpretations of Velázquez’s painting, Las Meninas,
as its fulcrum, ‘‘The Other Side of the Canvas’’ discovers a Lacanian critique of Foucault’s history of
modernity, circa The Order of Things. The effort here is to insert the interpretation of Velázquez into the
context of both Lacan’s ‘‘Science and Truth’’ (originally the first session of the 1966 seminar) and
Foucault’s recently published book. Our interpretation develops above all from Lacan’s contrast
between the definition of a painting as a ‘‘window’’ and Foucault’s implicit understanding of it as a kind
of ‘‘mirror’’—a distinction in which Lacan discovers his seminal concept of ‘‘object a.’’ Pursuing the
understanding of object a as the ‘‘surface’’ of the perspectival window allows us to understand why
Lacan expands the discussion of Velázquez both into an understanding of twentieth-century paintings
(Magritte, Balthus) and an implicit interpretation of the difference between philosophical and
psychoanalytic approaches to science and history.

The other side of the canvas

In May 1966, Michel Foucault made a visit to Jacques Lacan’s seminar. Foucault had heard about Lacan’s
reference in a previous session to his recent book, Les mots et les choses and, therein, to Foucault’s
interpretation of the painting of the Meninas by Diego Velázquez. The non-discussion that followed—
Foucault’s contribution to the session was limited to about 10 words—should not be mistaken for a non-
event. Quite the contrary, Lacan’s ‘‘friendly’’ reference to Foucault’s interpretation of this painting hides
a profound dispute about the character of the contemporary world and who we are in it.1 With their
disagreement about Las Meninas, Lacan and Foucault in fact reveal contrasting approaches to
subjectivity and history.

Las Meninas painting by Diego Velazquez

Diego Velázquez, Las Meninas

Las Meninas (detail)

Las Meninas (detail)

Click on images to see larger.

What was in dispute between Lacan and Foucault in 1966 is the character in a painting of what’s called
the picture plane, a site that is the very theme of Velázquez great pictorial essay. Certainly Foucault and
Lacan agree that the key to Las Meninas lies in the large canvas—of which viewers see the back alone—
taking up the ‘‘picture’s’’ left-hand edge and from which the ‘‘painter’’, Velázquez himself, has
apparently just stepped back.2 Furthermore, as does Lacan, Foucault understands the way this canvas
stands in for the picture plane itself, anchoring the perspectival and compositional tricks by which the
painter has achieved his ‘‘baroque’’ effects. He sees how this canvas anchors a ‘‘ceaseless exchange,’’
one which allows an ‘‘infinite’’ play between the positions of viewer and scene, subject and object.3
That is, for both Lacan and Foucault, the turned canvas thematizes the very possibility of a play of
representations in which, for example, the painter is both ‘‘in’’ the scene and painting it from ‘‘outside,’’
etc., etc.

Where, according to Lacan, Foucault is mistaken, is in what he takes to be the stabilizing precondition
and genuine ‘‘subject’’ of Velázquez’s painting—the figures of King Phillip IV and his wife. In Foucault’s
understanding (a common if not predominant art-historical interpretation), the mirror at the back of the
depicted room, in which the King and Queen are reflected, tips the viewer of Las Meninas off to what’s
on the other side of the reversed canvas: the picture depicts Velázquez depicting the royal couple. Thus,
a veritable Foucualdian orgy of the pictorial, all of it turning around the stabilizing royal representation.
For Foucault, the substitution of a mirror image onto an invisible painted surface undergirds the
baroque system of the painting and, by extension, the epistemological system of the classical age. A
‘‘royal’’ subjectivity masters the epistemological landscape, dominating nature and the human.

But, according to Lacan, precisely in such substitution of mirror for painting, Foucault retreats from his
correct identification of the key issue for Velázquez himself, the turned canvas, qua representational
surface. What does that mean? First, and most simply, it means that Foucault has missed a vital moment
in the play of representations in Las Meninas—the compositional and perspectival framework that
actually forbids the obvious reading by which the hidden painting depicts the royal couple. Neither the
size, nor the position of the image in the mirror correctly represents the ‘‘viewpoint’’ structuring the
representational play of Las Meninas. Lacan argues that this and various details of the size and
positioning of the turned painting in the painting make it impossible that the king and queen form the
‘‘subject’’ of that mysterious canvas.

As to what, Lacan suggests, does occupy the occluded and hidden space of the turned canvas, his
seminar’s answer is at first unclear, but only because what eventually emerges is a Lacanian insistence
that that’s the wrong question to ask. In other words, Velázquez’s painting has two ‘‘aspects,’’ two faces
that it can offer a viewer. On the one hand, its labyrinthine play of representation reinforces the basic
historical phantasy of painting itself, the continuity between occupied visual space and the space of
representation. Inviting us into the ‘‘looking-glass world,’’ it tempts us to turn around the hidden canvas
and ‘‘see me,’’ as Lacan repeatedly puts it. On the other hand, though, the turned canvas in Las Meninas
also hits us over the head with another function of painting, of seeing and of knowledge. Lacan’s work in
Seminar XIII is to articulate this other painterly function, to work with Velázquez (and Rene ́ Magritte
and Balthus, as it will turn out) to distract us from the seduction of the representational labyrinth—all of
it in the name of a specifically psychoanalytic alternative. But to explain that other function demands
some background about the projects of both Foucault and Lacan as they conceived them in 1966.

1 The End of Man

The regal foundation for representation is not only key for Foucault’s reading of Las Meninas, it is also
what allows the painting to serve as synecdoche for the entire history recounted in The Order of Things
– the story of ‘‘the human sciences’’ and their emergence out of modern representation; for in the
painting a perfect exemplar of the classical episteme also explicitly points forward to the modernity that
will displace it. While it historically belongs within the world of classical subjectivity— for which
representation is still anchored to an overall ordering of nature (the taxonomy, the general table)—Las
Meninas anticipates the themes of what Foucault calls the properly ‘‘modern science’’ of the nineteenth
and twentieth centuries. He makes this explicit when, later in Les mots et les choses, in his introduction
to the modern episteme, he recalls his introductory reading of the painting. There he tells us that
Velázquez’s masterpiece is both the type for a classical knowledge, a knowing for the sovereign subject
who alone cannot appear within the table of things known4 and the promise of a final reflection of the
subject, one that will engulf the ‘‘knower’’ within a now-transformed science.5 After all, what else (for
Foucault) is indicated by the depiction of the royal couple at the viewpoint-situated mirror, a point
substitutive of ‘‘our own’’ presence in front of the painting? Phillip the 4th and his bride not only ‘‘rule’’
the scene of Las Meninas, they do so in a fashion suggestive of the future goal of the ‘‘human sciences.’’

If you can go that far with me, then you will be able to follow, also, Lacan’s implied suggestion that
hidden in the radicality of Foucault’s gesture in The Order of Things in relativizing ‘‘man,’’ is, in fact,
something that is hardly radical—an implicit end or telos guiding his entire account. Let’s take this one
step at a time: on a first reading, and certainly at the level of polemic, Foucault puts Las Meninas at the
beginning of The Order of Things in order to challenge historicism and particularly the smooth
historicism of Marxist humanism: it stands there as if to say that all efforts to discover ‘‘man’’ in
renaissance humanism or even Cartesian rationalism must fail—that, instead, knowledge has progressed
in a series of discontinuous ‘‘epistemes’’ of which only the most recent, emergent at the beginning of
the nineteenth-century, really contains the ‘‘man’’ who underlies contemporary (historicist)
understanding. Seen in light of this, Foucault’s defining project in Les mots et les choses, Las Meninas is
simply an introduction to the alienness of a classical knowledge and, as such, it can serve as entre é to a
sequence of chapters that examine the even stranger, more distant, world of renaissance ‘‘similitude.’’
In other words, the painting forces us to acknowledge the discontinuity of knowledge.

But the problem here (I’m not the first to point this out) is that Foucault’s interpretation of Las Meninas,
like the larger account of knowledge from the fifteenth–twentieth centuries that follows it, is too
good—too bound to a formal logic which it constructs. In fact, Foucault gives us a better story of the
continuity of knowledge (disrupted, yes, by radical epistemic shifts, but within a comprehensible and
comprehended framework), rather than a non-story about its discontinuity. It is in this way of seeing
things that Foucault’s synecdochal return to Las Meninas late in The Order of Things makes sense: we
are more struck with the familiarity of the paradigm announced in Las Meninas—its contemporaneity—
than with its strangeness.

Unacknowledged by Foucault, his tale (both in the painting and in his book) is the story of our increasing
liberation from views of ourselves positing us to be limited by any nature, ‘‘human’’ or otherwise. At
first, in the classical period, representation intervenes within the cosmological world of the renaissance,
insisting upon the mind’s (language’s) capacity to order beyond nature, to impose order on the natural.
Still, the limit of the classical sciences emerges (as Foucault brilliantly sees) in the very forms of such
knowledge, which inevitably seek, and find, a nature composing a whole: sign and nature separate, with
strict orderliness imposed by the mathesis universalis of classical science. As a result, for the practitioner
of classical ‘‘science,’’ we may rest assured that the ordering produced by language as an entire meaning
system reflects the orderliness of nature.

Given the hidden mainspring of Foucault’s argument, this residual naturalism relegates classical
knowledge to an intermediate status, demanding completion. And we have the inner necessity of the
‘‘modern’’ break—a new kind of science that, with life, labor and history robs from the previously
unmolested sovereign subject its excessiveness to representation and grants it to new, dynamic
sciences, above all, to the emergent ‘‘human sciences.’’ In other words, the chapters in The Order of
Things devoted to the transformation of knowledge in the nineteenth-century introduce an episteme
whose virtue is precisely its ability to absorb the subject, becoming, transience—in short all that
previously had resisted objectification—into a more subtle scientific apparatus. This logic of ingestion
fueling the explosion of modern sciences comes to its full fruition when it consumes the entire
‘‘subjective’’ universe (the ‘‘self’’ and its world) in the new ‘‘human’’ sciences—psychology, history,
anthropology, sociology, etc.

Thus, in a text whose avowed purpose is to challenge the reassurance offered by smooth historical
narrative—to disrupt historicism’s illusion of a single ‘‘subject’’ of history—The Order of Things gives us
a closely reasoned alternative narrative. It all leads to the ‘‘double murder’’—thrice announced—of God
and Man; it leads to the end of every religion (whether theist or humanist) insofar as the religious
assumes a privileged, a sacred and untouchable, subject. No doubt this story, leading inexorably to the
‘‘sacrifice’’ of that subject underlying it, is meant to be more disturbing than its humanistic forebear.
However, we cannot ignore the fact that this ‘‘death of man’’ foreseen at the end of The Order of Things
is there from the very beginning and thus already resurrects him in its very form, that it reintroduces a
subject for the history of knowledge precisely in sacrificing one.

1 Lacan’s discussion of Las Meninas occurs in the session of May 11th as well as the May 18th session.

2 For Foucault, they provide ‘‘the center around which the entire representation is ordered,’’ a center

which is sovereign.’’ Foucault (1967, p. 14).

4 Foucault (1967, p. 307).

5 Foucault (1967, p. 312).

2 Psychoanalysis in The Order of Things

Now, I don’t indicate this problem with The Order of Things simply as an essay in formal criticism; my
interest here is in the way that this formal problem leads us to a very specific question—namely, to the
status of psychoanalysis for Foucault. Of course, the discussion of Freudian and Lacanian analysis in
Foucault’s 1966 text is striking for its sympathy for and insight into analysis—a position that largely
disappears in Foucault’s later work. But, because it leads to the precise point where the ‘‘formal’’
problem of history is joined, the context of Foucault’s sensitive discussion demands exfoliation; and that
is, namely, the question of ‘‘finitude’’ and the ‘‘analytic of finitude.’’ If the power of human science
derives from its capture of the Cartesian subject’s heterogeneity to all representation in one or another
‘‘excessive’’ force—‘‘history,’’ ‘‘labor,’’ ‘‘life’’—then the ‘‘analytic of finitude’’ is something like the
reverse of this power. As Foucault puts it, it is the insistence that every advance in conscious science of
‘‘man’’ actually relies on an inexhaustible ‘‘unconscious’’ whose very excess over representation fuels
it.6 Moreover, he traces finitude itself as a modern concept to precisely this excessiveness to
representation now discovered by science: for example, if the historicity of history disallows its final
representation in a chart or table, then ‘‘finitude’’ names the temporal unconscious—the human
‘‘Event’’ underlying all histories—which explains this ‘‘fact.’’ Which means, of course, that the analytic of
finitude first of all ‘‘establishes a stable sojourn’’ for modern life—giving it the reassuring form that
Foucault, borrowing a phrase of Kant’s refers to as the ‘‘slumber of thought.’’7

On the other hand, though, ‘‘finitude’’ marks for Foucault the site of maximum instability within the age
of humanism and it is also precisely the site of Freudian and Lacanian psychoanalysis. One can already
hear this alternative evaluation of modernity in Foucault’s usage of the terms ‘‘conscious’’ and
‘‘unconscious’’ to discuss the development of modern science—in the necessity that the key concepts of
a dynamic nature and human nature should be retrieved from the dark space of the subject’s
heterogeneity to every representation. However, acknowledging the positive sense of ‘‘the analytic of
finitude’’ becomes most pressing in the few pages on ‘‘The Human Sciences’’ that Foucault devotes
explicitly to psychoanalytic practice: in the analytic situation, writes Foucault, the analysand is made to
confront her/his finitude in a manner precisely opposite to the prevailing practices of other modern
sciences: the point is not to mine finitude’s excess for new forms of consciousness but to come,
somehow, face to face with ‘‘death, desire and Law’’ themselves—with the constitutive nature of excess
for every self-understanding, every representation. In those pages Foucault gives us the full picture of
the de-stabilizing effects of psychoanalytic practice, practice which he describes as ‘‘unmak(ing) that
very man who is creating and re-creating his positivity in the human sciences.’’8
In other words, ‘‘finitude’’ both marks the site of a transcendental source for modern knowledge—one
which is to be overcome when ‘‘man’’ is ‘‘erased’’ ‘‘like a footprint upon the sand’’—and indicates the
point of contestation for ‘‘man,’’ against the ‘‘slumber of thought.’’ Foucault commits himself both to an
account of history which limits finitude to the moment of modern science, a moment beyond which we
must pass, and to an understanding of psychoanalysis that sees its contestation of modern humanism as
exemplary of precisely what Nietzsche means in suggesting ‘‘the death of man’’ and the birth of the
superman. Which is it to be?

6 Foucault (1967, pp. 371–372).

7 Foucault (1966, p. 385).

8 Foucault (1966, p. 379).

3 Seminar XIII and the object a

What’s interesting about Lacan’s 1965–1966 seminar—which is entitled ‘‘The Object of Psychoanalysis,’’
and which includes that discussion of Las Meninas at which Foucault is present as an invited guest—is
how apparently similar is the starting point of his thought there to Foucault’s. For Lacan, too, the project
of ‘‘Cartesianism,’’ of what Lacan calls a ‘‘suture’’ of the representational subject and its representation
(‘‘cogito’’ and ‘‘sum’’) is going ‘‘badly’’ today; indeed, it is bound to fail.’’9 For Lacan, as Mladen Dolar
paraphrases him, ‘‘in the place of the supposed certainty of the subject’s being there is just a void. It is
not the same subject that thinks and that is.’’10 All representation is for a subject that cannot be
included (even as excess to representation) ‘‘in the picture’’. Like Foucault, then, Lacan would have us
remain suspicious of modern science’s pretentions to transparency and its ‘‘universe,’’ founded as it is
upon the myth of such a transparent knowledge.

But with this like critical distance from modernist representation, the similarity between the positions of
Foucault and Lacan ends—an end that, perhaps, Lacan hopes to emphasize when, soon after the
resume ́ of his Descartes’-critique in the 14th session, he disparages Nietzsche’s ‘‘death of God.’’11
Whether or not he means to answer Foucault, Lacan’s turn in this session indicates the nature of his
divergence from any philosophical position—even from an ‘‘anti-philosophy’’ like that in The Order of
Things. But, unlike Foucault’s position, which sometimes collapses anti-modernism and anti-
Cartesianism, Lacan’s critique of Descartes combines an embrace of modern science together with its
critical attitude to the Cartesian project. The point, then, is to build upon modern science as somehow
proposing an alternative to such a mirroring relationship between truth and knowledge; and it is in this
project that the theme of Seminar XIII, the ‘‘object a’’ appears.
Our starting point this year was to render coherent what we have to affirm about the function of objet a
in the position of psychoanalysis, in so far as it originates from science and from science in its very
particular relationship to truth, science being understood as the modern science born in the 17th-
century.12

Leaving to others a full explication of just how Lacan’s response to Descartes both demands that he
(Lacan) embrace modern science and also pass beyond its self-understanding, let me turn in my remarks
to a kind of allegory for this subject in the consideration given in Seminar XIII to painting, perspective
and, in particular, Vela ź quez’s masterpiece, Las Meninas. Whatever the course of this allegory, though,
we must keep our eye on its end—the emergence of psychoanalysis itself as a (in a sense the) modern
science, because it is a science self-consciously split. That’s another way of saying, as will be seen, that
analysis demands the theorization of the object a. On account of the differential roles of analyst and
analysand, psychoanalysis marks a practice where, to oversimplify a bit, science can pursue knowledge
without self-contradiction, a practice not bound by the dialectical limitations of philosophical argument
as is, still, Les mots et les choses. One understanding of the object a is, thus, what appears only with this
double face, this double surface, for a newly modern knowledge/truth.

9 Lacan (SeminarXIII, Session of April 20th, p. 6).

10 Dolar, in Zizek (1998, p. 18).

11 Lacan (SeminarXIII, Session of April 20th, p. 12).

12 Lacan (SeminarXIII, Session of January 19, p. 22).

4 Window not mirror: Las Meninas

If we begin from the site of their agreement about Velázquez’s Las Meninas, if we start from a concord
on the fact that this is a painting about painting and one that takes up this theme in its representation of
the turned canvas in the canvas, the painting in the painting, then we are immediately struck by the
difference marked by Lacan’s polemic against mirrors. Or, to be more precise, Lacan’s objection to a
‘‘mirror’’ understanding of knowledge is itself related to an extended argument in Seminar XIII against
any equivalence between the ‘‘canvas’’ for painterly (or epistemic!) representation and the mirror
surface. For Lacan, as opposed to Foucault, the functions of mirror and painting must be strictly
contrasted. As Lacan puts it, ‘‘the relationship of the picture to the subject is fundamentally different to
that of the mirror.’’13

I will return below to the nature of this difference and Lacan’s argument for it in terms of the history of
painterly perspective, but it’s important here to indicate how Lacan’s understanding of this contrast
transforms his interpretation of Velázquez’s masterpiece: for Lacan, the mirror in the painting is
rhetorically polemical and non- substitutable for the painting’s ground. Its meaning is, ‘‘This is what I am
not.’’ Indeed, Lacan goes to considerable lengths to oppose, not only Foucault’s interpretation of Las
Meninas—according to which the turned canvas is a royal portrait—but also another popular art-
historical interpretation—in which the turned picture depicts the very scene shown in the painting but
itself as painted in a mirror. This second interpretation, too, depends upon a mirror, in this case a large
looking- glass placed at the ‘‘viewpoint’’ and stretching across the audience chamber from which we
(the viewers) seem to see the picture.14

Seeing it as an essay about painting and not about mirrors, how, then, should we interpret Las Meninas?
Lacan’s repeated pictorial explanation takes the form of an analogy with Magritte’s The Human
Condition, a painting, famously, of a painting of a window, with the depicted canvas placed in front of
the very window whose scene it correctly represents.15 The barely noticeable contours of the stretcher
disrupt our view into the represented space and through the window—both the Albertian window of
Magritte’s picture and the depicted window in that space. Our gaze is brought up short by this
‘‘representation of representation,’’ but in a fashion that doesn’t simply reinforce entrapment within a
representational labyrinth. The painting in the painting also displaces the subject from its tarrying in
imaginary depth and its symbolic elaboration to the very surface of representation.16 Or, to put this
more precisely, the effect of Magritte’s game in The Human Condition is both to intensify the
representational illusion (‘‘oh, if only that painting weren’t there we could see out the window!’’) and to
call it into question (‘‘There is no ‘outside’ of the window!’’)

Las Meninas painting by Diego Velazquez

Diego Velázquez, Las Meninas

Magritte, <em>The Human Condition</em>

Magritte, The Human Condition

Click on image to see larger.

It’s not a big jump to see the reversed stretcher in Las Meninas as playing the same role as the window-
picture did in The Human Condition.17 First of all, as Foucault announces, the combination of our
placement at the royal viewpoint and the turned canvas in the painting vivifies the representational
illusion.18 But, for Lacan, the task that Velázquez gives us is not to figure out ‘‘what’s going on?’’ in the
illusory represented space but rather both to investigate what this strong illusion hides and to examine
what produces the power of the representational illusion that we might enter the space of the artist’s
studio in the Alca ź ar palace, passing through the ‘‘open window’’ of the painting.19 In other words,
Lacan insists that we pay attention to the way Las Meninas, like The Human Condition, challenges mere
representational illusionism.

Velázquez uses the artist’s depiction of himself in a manner equivalent to the ‘‘slip’’ between the
depicted canvas and the window it covers in The Human Condition. Lacan notes the distance that the
figure of Velázquez in Las Meninas takes from his canvas: he (or his avatar) is not painting at the
moment of representation.20 This stepping back from representation becomes, indeed, the key for
Lacan’s interpretation of Las Meninas, whose primary device—the turned canvas—forces us to ‘‘look
again’’ at the painting, to see it, for ‘‘the first time’’ in the history of modern painting as Lacan boldly
claims, as something other than mere representation of some other reality.21 For Lacan, Velázquez
provides a token or ‘‘representative’’ of representation (Vorstellungsrepraesentanz, using Freud’s term)
that makes us aware that painting is not only a representation, not only something one looks through or
into but also something one must look at.22

It is in pushing this interpretation of Velázquez’s painting that Lacan introduces two related observations
that complete his ‘‘picture’’ of it: the first is that the key ‘‘space’’ in Las Meninas is the foreground area
between the depicted and the actual canvases or, if you like, between the reversed painting and the
actual ‘‘picture plane.’’ Just as was the case with the window in Magritte’s painting, in producing a space
or pause, Velázquez is inviting us to see that his painting—the one he painted, not just the one he
depicted—is an object: it is not the same as the space we look ‘‘into’’ but stands in front of that.
Secondly, by means of a fictional dialogue between the figures of the infanta and of the painter, Lacan
ties this empty liminal space in the painting to the question of painting itself: Dona Margerita’s gaze at
this space ‘‘behind’’ the depicted canvas seems to demand, ‘‘let me see’’ the representation itself23; to
which—so Lacan—the painting itself addresses her with the pronouncement, ‘‘you do not see me from
where I am looking at you.’’24

What is this injunction about? This is the central mystery of Lacan’s interpretation of Las Meninas, one
that, in the eighteenth session, he answers only with a double reference—first to phantasy and the
‘‘gaze’’ constitutive of that and then to Balthus’ painting, ‘‘The Street,’’ which was currently on
exhibition in Paris. In both cases, something about the painting qua painting, as ‘‘window,’’ seems to
confound our efforts (like those of the infanta in Velázquez) to see it as a mirror.

13 Lacan (SeminarXIII, Session of May 25th, p. 2).

14 See, Lacan SeminarXIII, Session of May 11th, pp. 25–36. Lacan argues both that the mirror-thesis is
incorrect because Velázquez does not represent his painting hand as reversed (right to left) and because
the royal couple could not be present if the mirror was.

15 In fact, Lacan claims that Magritte’s painting amounts to a rare reprise of the accomplishment of Las
Meninas: see, Lacan (SeminarXIIII, Session of May 11th, p. 26).
16 Lacan (SeminarXIIII, Session of May 11th, p. 26).

17 Lacan (SeminarXIII, Session of May 11th, p. 26).

18 Foucault (1967, p. 6).

19 Lacan first compares the effect of looking at the painting to participation in a game of cards, wherein
the turned canvas is equivalent to a ‘‘face down card.’’ Lacan (SeminarXIII, Session of May 11th, p. 21).
Immediately, though, he argues we should not be seduced by this invitation to ‘‘play,’’ that the role of
the ‘‘player’’ asking the question of the painter ‘‘what is he doing?’’ is ‘‘certainly the wrong position to
take up’’ (p. 24).

20 See, for example, Lacan (SeminarXIII, Session of May 11th, p. 21) and Lacan (SeminarXIII, Session of
May 18th, p. 48).

21 Lacan (SeminarXIIII, Session of May 11th, p. 26).

22 This reference to Freud’s ‘‘Instincts and their Vicissitudes’’ (Freud 1915) and Lacan’s persistent
interpretation of the term, Vorstellungsrepraesentanz as indicating a representative for the imaginary
representation providing the drive’s ‘‘object.’’ This comes to a head in relationship to painting in
sessions xvii and xviii, where it is allegorized by this possibility that the painting could provide a token for
that painterly function irreducible to representation. See, Lacan (SeminarXIII, Session of May 11th, p.
26).

23 Lacan (SeminarXIII, Session of May 11th, p. 28).

24 Lacan (SeminarXIII, Session of May 18th, p. 49).

5 The Street: Lacan reads Balthus reading Velázquez

Perhaps the best way understand what Lacan wants us to understand about Las Meninas is to look
briefly at Balthus’ La Rue. In what sense can Lacan legitimately claim of Balthus’ painting, as he does in
Seminar XIII, ‘‘that is Las Meninas’’?25

Las Meninas painting by Diego Velazquez

Diego Velázquez, Las Meninas

Balthus, La Rue

Balthus, La Rue
Click on image to see larger.

In some ways, that’s an easy question to answer; indeed, while, to my knowledge, no art historian has
followed out Lacan’s insight (there clearly are other sources for the painting, from Piero della Francesca
through John Tenniel of Alice in Wonderland fame), that observation nonetheless has a certain power to
bring La Rue suddenly into focus, to force our view of the painting into something approaching a Gestalt
shift. Beyond certain vague and inexact parallels (the space of the studio in the Alcázar Palace and the
Parisian Street depicted in La Rue, etc., etc.), two related details establish Balthus’ reference to Las
Meninas beyond question. In fact, the two features, collapsed together into a double group in the lower
left hand corner of La Rue, are responsible for much of the outcry surrounding this painting; they are the
figure reaching out to grab the crotch of the playing girl in red and the girl ‘‘her’’self (more on the scare
quotes below). In classic symptomatic fashion, the interaction of these two figures both calls attention
to their importance in the painting and distracts viewers from their real import—an importance that
Lacan alone insists upon; for the two figures are unmistakenly taken from Las Meninas and the
representations there of Velázquez himself and of the dwarf, Nicolas Pertusato, who occupies the far
right corner of Velázquez’s tableau. Both the facial features and gazes of these figures are unmistakable:
Balthus has ‘‘taken’’ them from the Meninas.

Still, this source for identification, while it suffices to explain Lacan’s ‘‘Eureka!’’ moment, leads, in the
final analysis, only to puzzlement; for, once one does shift perspective to see the whole of The Street as
a borrowing from the Meninas, the debt becomes increasingly urgent but also increasingly mysterious,
difficult to pin down with a rule for translation. Take, for example, the detail of the mirror itself, with its
image of Philip the Fourth and his Queen: Balthus has shifted the figures in the mirror to the right of his
painting, personing them with the figures of a manikin (King) being carried by a woman whose coiffed
hair replaces the queen’s face. Framed as they are against a blank storefront at the back of Balthus’
space and a door to right (and somewhat toward the foreground in the representation), these figures
seem both odd representatives of ‘‘the subject’’ from Velázquez and, more to point, oddly reversed.
They seem, in fact, to present a mirror-image of the mirror in Las Meninas. Indeed, once one notes this
mirror reversal, much in the painting comes into a different focus, as though La Rue were itself the view
of the studio in the Alca ź ar palace as it might appear in a mirror: the vanishing point for the space of La
Rue as a whole is to the left instead of the right as it is in Velázquez’s painting; there is a ‘‘second’’ figure
of the painter to the right of La Rue—identified by black attire and, in her hat, the ‘‘cross of Santiago’’
key to the earlier mise en scene—in precisely the reversed position of the Velázquez figure from the
classical painting; even the enigmatic figure located at the door leading from the studio in Las Meninas
corresponds in reverse to the flat figure of the cardboard chef now on the left. Still, this idea that La Rue
gives us a literalized ‘‘through the looking glass’’ vision of the Meninas only goes so far. It cannot, for
example provide an explanation for why Velázquez, the painter, has been split into two representations
(the molester and the turned woman on the right); nor can it do justice to, for example, the detail of the
girl/dwarf’s red ball, which occupies almost the right space to substitute for the red pitcher in the
infanta’s hand from the Meninas, or of the guard who, represented by the robotic striding figure, simply
moves toward the foreground in Balthus’ composition. And all of this simply leaves aside the really
enigmatic translations of Velázquez in La Rue, where the infanta seems to be substituted by a white-clad
carpenter figure holding a board which is all that remains of the depicted canvas from Velázquez’s
painting.

In other words, the identification between Balthus’ painting and Velázquez’s is both undeniable and
enigmatic, resistant to any simple rule of translation. We know it to be the case but can’t put our finger
on exactly how it happens. Of course, that works well for Lacan, whose entire interpretation of Las
Meninas is that it is meant also to call into question the possibility of reducing painting to a mirror of
truth. It’s ‘‘all there,’’ we might say—each of the elements of Las Meninas, but now mixed up in a way
that lacks any coherence, any sense of the totality of a ‘‘world.’’ This interpretation explains the main
fact in the history of La Rue’s early aesthetic reception, namely, its embrace by the surrealist group on
the grounds of the robotic disconnection between its figures.26

Here, finally, we can fill in at least one meaning of the contrast between painting as ‘‘mirror’’ and as
‘‘window.’’ The mirror lacks time, lacks memory: it shows only what is present to it as it is present to it.
It is thus incapable of the kind of mixing up that Balthus engages in with Velázquez’s figures. It cannot
take them out of a context and ‘‘paste’’ them (Lacan repeatedly plays with notions of phantasy and of
the painting here as ‘‘montage’’) into another. But the painter, like the surface of a film or a television
editor (Lacan references both) can retain the image; he can take the poses of figures from one time or
one place and reproduce them at another time and/or place. The window can ‘‘mount’’ (‘montage’)
different times and places into a new present, a new place.

Armed with this interpretation of La Rue as window, though, we can return to Lacan’s comments on Las
Meninas to see that, for him, Velázquez himself is already engaged in this kind of ‘‘windowed’’
treatment of the figures in his court painting; so understood, not only does the distance of the painter-
figure himself from his canvas gain new significance but we can also make sense of Lacan’s underscoring
of the way that the gazes of the figures in the scene of Las Meninas fail to interact.27 In other words,
the trick here, the virtuousity of the painter, is that, despite their having been correctly represented for
the expected tableau portrait— each in the proper perspective—these courtiers (and assorted fools,
animals, etc.) give all the indications of having been depicted from somewhere else and at a different
moment: Lacan speaks of the visible ‘‘furrow’’ of the painter’s ‘‘passage’’ through their midst as
recorded in those various visual foci.28

Furthermore, Lacan’s initial discussion of the Velázquez (session xvii) turns on the impossible necessity
imposed on us by the painter—Lacan refers to it as an ‘‘embarrassment’’—that we locate both a mirror
and the royal couple at the same spot ‘‘in front of’’ the picture plane. In other words, the painting itself
‘‘composes’’ its figures in a way that is meant to be seen as artificial at the same time that it seduces us
into imaginary identification.29
Vela ́ zquez introduces a temporal pause, a fundamental pastness, into his painting—one that, so Lacan,
forces us to reconsider the entire illusion that the Meninas gives us a matched present between the
scene in the painting and the invisible representation in the turned canvas. Instead, for Lacan, the
turned painting shows what could not be seen at the moment depicted in Las Meninas—a depiction
wherein the ‘‘figures’’ represented appear to us as they are after their depiction. And this reintroduction
of a non-presence, a lapse of time, into the painting ‘‘with a stroke’’ invalidates all the guessing games
filling in the content of the turned canvas: the painter could have painted anything at all there; or, more
radically, there need be no relationship between Las Meninas and the picture in the picture.

Isn’t this ‘‘disturbance’’ of our normal way of seeing things (whether a painting or our world) precisely
the virtue of painting as window, just as it is the most important potential of the experience of the
Lacanian analysand—forced to traverse the very fantasy holding reality together, to undergo a
bottomless anxiety, a suspicion whose ‘‘object’’ both does and does not exist? The first meaning of
Lacan’s analogy between painting and analysis is this experience of the analysand, of the one for whom
the arbitrariness of a world appears. To experience the window as window is to be subject to this
uncanny arbitrariness of experience. Elsewhere, Lacan famously proposes that anxiety is not
‘‘objectless’’ as both Kierkegaard and Heidegger take it to be: for Lacan, rather, anxiety has an object
that, in its sublime lack of focus, is non or even anti-objective. Anxiety before the window of painting
indicates the proximity of object a.30

25 SeminarXIII, Session of May 18th, p. 49, 51.

26 Both the painter Andre Lhote and Pierre Loeb, the gallery owner who mounted the 1933 exhibition
centered around ‘‘The Street,’’ wrote of the odd ‘‘automaton-like’’ quality of the figures in the painting.
They seemed, in Loeb’s words, like figures from ‘‘a strange dream in which everyone is a
somnambulist.’’ (Loeb, Voyages a travers la peinture, quoted in Weber, p. 156). On account of this
interpretation of the painting, the Surrealists featured it in the first Minotaur exhibition later in the
same year. Weber (1999, p. 191).

27 ‘‘If you look at things closely, you will see that except for the look of the maid of honour, Maria
Agostina Sariente, who is looking at Dona Magarita, no other look fixes on anything.’’ Lacan (SeminarXIII,
Session of May 11th, p. 34).

28 Lacan (SeminarXIII, Session of May 11th, p. 33).

29 Lacan (SeminarXIII, Session of May 11th, p. 33).

30 In fact, Lacan famously establishes this idea in a reading of Freud’s ‘‘Inhibitions, Symptoms and
Anxiety’’ Freud (1926/1953), where Freud himself argues for the ‘‘objectlessness’’ of anxiety as opposed
to fear. For Lacan, on the other hand, anxiety has an object, object a. (Lacan 2004). The concept is
further elaborated in Lacan 1973/1978.
6 The poetics of the window: truth of the analyst

Albrecht Durer, illustration from On Measurement

Albrecht Durer, illustration from On Measurement

It’s time to redeem my promise to fill out Lacan’s opposition of mirror and window with a deeper
understanding of the latter; for it is precisely in this, an investigation of Las Meninas as commentary on
the constitutive view of a painting as an ‘‘open window’’ first articulated in the Italian Renaissance, that
Lacan argues most extensively in his response to Foucault. Indeed, the introduction of Lacan’s
interpretation of Las Meninas comes in the midst of a discussion of renaissance perspective theory
(Session xvi)—and particularly of that method of constructing ‘‘correct’’ one-point perspective
developed by Leon Battista Alberti, the so-called ‘‘construzione legitima’’. Such pictorial perspective
seems to locate the absent subject (the ‘‘for which’’ of representation) in the vanishing point at infinity;
one could even say that the historical dialectic traced by Foucault assumes precisely such positioning of
subjectivity—so that its heterogeneity can be recouped as transcendence of or, later, excess to
representation. In other words, the entire ‘‘history’’ recounted in Les Mots et les choses depends upon
the implicit identification of the modern subject as the absent ‘‘pinnacle’’ of a representational pyramid,
an excluded point ‘‘for which’’ representation occurs.

Given such an assumption on Foucault’s part, it’s worth tarrying a bit with the Albertian
‘‘demonstration’’ upon which it is founded, an operation that we might reconstruct in several steps: first
comes an important theoretical insight, one implicit—but only implicit—in the painting of Masaccio and
other contemporary Florentines; it is the conception of a painting as, in Alberti’s words, an ‘‘open
window,’’ a planar cut in the ‘‘visual pyramid’’ running from the eye of the viewer (importantly, for
Alberti the viewer is always monocular, always a ‘‘cyclops’’) to a scene viewed.31 The conceptualization
of the painting as window is what allows Alberti a mathematically correct solution to the practical issues
facing the painter wishing to represent naturalistically, and also what produces the model of subjectivity
as point upon which Foucault depends in his account from Les Mots et les choses.

The Picture Window: Illustration from an 18th century treatise on perspective

The Picture Window: illustration from an 18th century treatise on perspective

Drawing the Floor: Perspective Grid for Peter de Hooch, A Woman Drinnking

Drawing the Floor: Perspective Grid for Peter de Hooch, A Woman Drinnking
Click on images to see larger.

By using this conceptualization Alberti arrives at the so-called ‘‘legitimate construction,’’ which responds
to an extremely common problem for the painter of Alberti’s day, namely, how to complete the correct
representation of a tiled floor in ‘‘one point’’ perspective. The idea was that, if a painter could do that,
then, using the floor tiles of known dimension as guide, on the basis of how many such squares they
occupied, he could also correctly represent the relative sizes of objects at various distances. And even
this problem was simpler than it sounded; for, though without theoretical justification, artists in the
preceding 30 years in Florence had figured out that lines of recession—lines running at 90 degrees to
the plane surface of the picture—all seemed, when the artist stood directly in front of that same picture,
to converge at a single point. The result of that insight is that those sides of the square floor tiles could
be correctly represented simply by having them move toward what we now call the ‘‘vanishing point.’’

Alberti’s ‘‘system’’ of perspective consists of his solution to the problem of correctly representing the
other dimension of the tile floor, the lines parallel to the picture plane. But the interesting thing about
this solution is its practical, not to say praxical, nature. In proposing his solution, Alberti brought his skill
from earlier work as a surveyor to bear. Effectively, figuring out the parallel lines correctly demands a
classical surveying ‘‘move,’’ where the one measuring angles projects his position to a point at the same
distance from the picture plane but at right angles to the ‘‘edge’’ or ‘‘frame’’ of the picture. From that
place, and imagining the canvas literally to be a window, the ‘‘surveyor’’ could physically notch it at the
points where the ‘‘rays’’ from his eye would intersect it on their way to the parallels marked on an
actual floor continuing on the ‘‘other side’’ of the glass.32

View two additional videos with alternative visualizations of Alberti’s legitimate construction.

Now, it is this displacement, this movement of the subject necessary to actually constitute perspective,
which sets up Lacan’s response to Foucault. Just as we already saw with regard to Las Meninas, indeed,
as the explanation for that effect, the construction of a ‘‘world’’ of representation demands not only a
dissymmetry between a representing subject and what it represents for itself as present to it, but also
the inclusion of a temporal disruption in the tableau of ‘‘presentness’’ between subject and world. The
world can only be present to the subject—the subject can only be present to itself—on the precondition
of a ‘‘move’’ not present to it. Indeed, Lacan goes to great lengths to demonstrate the irreducible
necessity of this second, unlocated and unlocatable ‘‘subject-point’’ for the construction of any concrete
painterly perspective.33

Or, to put this in another way, Lacan’s ‘‘counter-demonstration’’ to Alberti’s in the 16th and 17th
sessions of Seminar XIII shows how the ‘‘open window’’ of perspective actually only works when the
window is ‘‘closed’’—when the surface of the picture plane is materialized by a process of marking (both
of the receding and the ‘‘parallel’’ floor lines) which thickens the planar surface, making of it a kind of
transparent or ‘‘partial’’ object—the window’s ‘‘glass.’’ And this ‘‘thickening’’ is what Lacan himself plays
with constantly in the 17th and 18th sessions of Seminar XIII, using the variants of ‘‘montage,’’ and the
idea of painting as a ‘‘screen’’ for vision.34 The non-simultaneity of the subject is simply the reverse of
this constitution of a ‘‘partial object’’ in the surface of the picture plane, the space on which the figures
of representation then appear. What’s at stake here is something like the transcendental condition for
the possibility of the analysand’s experience of destitution, for the breakdown of ontological consistency
itself. But isn’t that condition, as it is discovered in the history of painting, precisely the constitution of
that space within which fragmented ‘‘pieces’’ of representation, regardless of value or context, emerge?
Only based upon this discovery can Velázquez play with the displacement of figures in his Meninas or
Balthus radicalize this play; and, as Lacan himself notes, only on the basis of this discovery does painting
increasingly become a matter of montage, of mounting things on a neutral surface, as it does in the
twentieth-century.35 In each of these cases, by forcing us to see the painting precisely as a surface for
the composition of anything, the artist makes the transparent or invisible part-object visible, allows us to
see what, as Lacan notes, we make invisible to look into the representation.36

In this sense, for Lacan, the practitioner of montage or collage is precisely like the psychoanalyst, who
makes manifest the ‘‘support’’ for the analysand’s fantasy, the thing that allows the world to appear
‘‘normally,’’ as a totality while itself withdrawing from presence. The analyst’s task is to represent the
‘‘truth’’ of that fantasy, to produce the non-object that now appears. As Lacan puts it,

… Something is produced in the construction of vision which is nothing other than what gives us the
basis and the support of the fantasy, namely, a loss which is none other than the one that I call the loss
of the objet a…37

If I wrote above of an experience of the analysand challenging any presentation as science because its
content challenges the very ontological closure assumed by every system of truth, then here we must
add the necessity that a certain truth always potentially accompany it. Vitally, however, this is not truth
for the same subject; rather that truth, which is quasi-transcendental understanding of the possibility of
the analysand’s anxious condition, can only appear for another, for one who accesses the ‘‘partial
object’’ in a fundamentally different garb than anxiety. To the analyst, the object a appears as part of a
regime of knowledge, but only of a knowledge once removed, a science of what ‘‘must be the case’’ in
order for an experience of ontological groundlessness to be possible for the analysand.

31 The notion of the visual Pyramid appears early in Book I of On Painting (Alberti 1991, p. 44) as does
the idea of the ‘‘open window’’ (1991, p. 48).

32 1991, pp. 54–59. For more information on Alberti’s insights underlying the development of linear
perspective, including his use of surveying to make the ‘‘legitimate construction’’ work, please see the
two additional videos.
33 Lacan (SeminarXIII, Session of May 4th, pp. 25–28).

34 For the idea of the ‘‘montage,’’ see, for example, the references to the ‘‘setting’’ (‘‘monture’’) of the
painting in the sixteenth and seventeenth sessions Lacan (SeminarXIII, Session of May 4th, p. 15) &
Lacan (SeminarXIII, Session of May 11th, p. 17). The idea of the ‘‘screen’’ becomes central at the end of
the May 11th session and the beginning of the May 18th one. Lacan (SeminarXIII, Session of May 11th p.
41) and Lacan (SeminarXIII, Session of May 18th, p. 4) It is a concept that Lacan had already developed
previously in Seminar XI. See, Lacan (1973/1978, pp. 105–107).

35 Speaking of the Vorstellungsraepresentanz, Lacan says ‘‘It is the representative (token) of the
representation in the mirror. It is not, in its essence as being, the representation. And modern art
illustrates this for you: a picture, a canvas, with a simple piece of shit on it, a real piece of shit, for after
all what else is a big splash of color?’’ Lacan (SeminarXIII, Session of May 25th, pp. 6–7).

36 Lacan (Seminar XIII, Session of May 4th, p. 16).

37 Lacan (Seminar XIII, Session of May 4th, p. 16).

7 Conclusion, References, Photo Credits

Recall, from my earlier discussion, the cul de sac in which the Foucault of The Order of Things found
himself; dedicated to challenging as deceptive the very idea of a single Western ‘‘history’’ unfolding over
time and space, his analysis nonetheless seemed to suggest just such a single story occurring for and to
a single subject. The term ‘‘finitude’’ marked the locus where this tension between project and the
hermeneutic with which it is pursued comes to a head. In a sense, my point here is simply that what
could only appear in Les Mots et les choses in a symptomatic manner—as the unresolved tension
between ‘‘finitude’’ as remainder of a historicism to be overcome and the same ‘‘finitude’’ as marker for
the very instability that defeats historicism in its rationalist presumption—emerges in Lacan’s Seminar as
defining psychoanalysis itself. Indeed, one could go further down this road, since Foucault himself, in
allowing the signifier ‘‘psychoanalysis’’ to mark the very place in his argument where the ambiguity of
‘‘finitude’’ emerges, seems to anticipate Lacan’s move. In Les Mots et les choses, we already find the
name for that practice highlighted as containing the positive possibility today for the ‘‘human sciences’’;
Lacan simply investigates what this ‘‘psychoanalysis’’ must be if it is to escape the imperative for a
‘‘death of man.’’

The answer to that demand comes in terms not directly of the ‘‘subject’’ of modernity but of an object
implied by but missed in earlier accounts of modernity. Lacan’s account of representation and its subject
differs from Foucault’s precisely because it turns on the enigmatic object, the object that equally either
appears to a subject as anxious ontological incompletion or as the transcendental condition for apparent
ontological completion. This is key; for neither position cancels out the necessity of the other; above all,
this means that the ‘‘science’’ granted to the analyst cannot be seen as substitute or replacement for
the destitute experience of the analysand. Lacan’s psychoanalytic theory, in demonstrating another
object—a ‘‘partial object’’ that shifts between the analyst’s (objective) understanding of it and the
analysand’s (subjective) experience (Lacan reminds us that the window looks at us)—also gives the lie to
Foucault’s paradoxically historicist impulse in Les Mots et les choses. It suggests that all along the
possibility for de-stabilization, for a challenge to modernist complacency, has stood within the
constitution of the modern subject. We should give up the post-modernist idyll to which the Foucault of
1966 sometimes succumbed—the dream of a world ‘‘beyond’’ the subject of representation. Not as
beyond or outside of representation, but in its innermost heart lies subversion of its limitations, of its
mania for domination.

References

Alberti, Leon Battista. 1991. On Painting, translated by Cecil Grayson, with an introduction and notes by
Martin Kemp. London: Penguin Books.

Dolar, M. 1998. Cogito as the subject of the unconscious. In Cogito and the

unconscious, ed. S. Žižek. Durham: Duke University Press.

Foucault, Michel 1966/1967. Les Mots et les choses. Paris: Gallimard. English Translation (1967), The
Order of Things. New York: Vintage Books. All references from the English translation.

Freud, Sigmund 1915/53. The standard edition of the complete psychological works of Sigmund Freud,
Vol. 14. Translated from the German under the general editorship of James Strachey. London: Hogarth
Press.

Freud, Sigmund 1926/53. The standard edition of the complete psychological works of Sigmund Freud,
Vol. 20. Translated from the German under the general editorship of James Strachey. London: Hogarth
Press.

Lacan, Jacques 1973/1978. Le Séminaire de Jacques Lacan, Livre XI, ‘‘Les quatre concepts fondamentaux
de la psychanalyse.’’ Texte e t́ abli par Jacques-Alain Miller), Paris: Seuil. English translation (1978) The
four fundamental concepts of psychoanalysis (edited by Jacques-Alain Miller), New York: Norton.

Lacan, Jacques. 2004. Le Séminaire de Jacques Lacan, Livre X, ‘‘L’angoisse’’. Texte e t́ abli par Jacques-
Alain Miller), Paris: Seuil.
Lacan, Jacques Unpublished. Transcript: Le Séminaire de Jacques Lacan, Livre XIII, 1965-1966, ‘‘L’objet
de la psychanalyse.’’ Available at http://nosubject.com/Seminar_XIII. Page Citations refer to the original
transcript. Unless otherwise notes English translations are from an unpublished text by Corman
Gallagher. Accessible at http://www.lacaninireland.com/web/.

Weber, Nicholas Fox. 1999. Balthus: A biography. New York: Alfred A. Knopf.

Photo Credits

Diego Velázquez, Las Meninas. (1656, Museo del Prado, Madrid) Image permission of Superstock, #4266.

Diego Velázquez, Las Meninas, detail. Image permission of Superstock, #1899-33145.

Rene Magritte, The Human Condition (1933, National Gallery of Art, Washington, D.C.), with permission
of the estate of Rene Magritte. Image permission of Superstock, #1158-2398.

Balthus, The Street (La Rue). (1933, Museum of Modern Art, New York), with permission of the estate of
Balthus (Harumi Klossowska de Rola). Image permission of Superstock, #260-1048.

Albrecht Du r̈ er, Illustration from On Measurement. Image permission of Superstock, #463-7149.

Perspective illustration after, Peter de Hooch, A Woman Drinking with Two Men. (1658, NationalGallery,
London). Image permission of Superstock, #1746-2988.

All other images are in the public domain.