Michel Foucault's study of Velazquez's Las Meninas (1) was first published in the volume Les Mots et les choses in 1966 which was followed, in 1970, by the English translation titled The Order of Things. In "Las Meninas", which is the title of the opening chapter of The Order of Things, Foucault focused on the artwork itself as though it were before him, describing in extraordinary detail what he saw. His seemingly unobtrusive actions looking and describing elicited observations that, when positioned within the context of contemporary art-historical practice, were unprecedented. His examination of the painting is neither prescribed by, nor filtered through the various texts of art-historical investigation.

For example, the artist's biography is absent and there is no declaration of technical virtuosity and genius. Neither is there an acknowledgement of sources and influences, nor an exploration of questions of style and iconography. Nor is there interpretation, through the selection and interpretation of archival documents, of the relation between the painting, the artist's social context and his relationship with his patrons. In one instance, Foucault comments on the art-historical practice of identifying the subjects represented: "These proper names would form useful landmarks and avoid ambiguous designations; they would tell us in any case what the painter is looking at, and the majority of the characters in the picture along with him" (2002: 10). But the convenience of the proper name, in this particular context, is "merely an artifice: it gives us a finger to point with, in other words, to pass surreptitiously from the space where one speaks to the space where one looks; in other words to fold over the other as though they were equivalents" (p. 10). Foucault proposes a different relation of language to painting:

The relation of language to painting is an infinite relation. It is not that words are imperfect, or that, when confronted by the visible, they prove insuperably inadequate. Neither can they be reduced to the other's terms: it is in vain that we say what we see; what we see never resides in what we say. (Foucault 2002: 10)

Instead, Foucault proposes to "keep the relation of language to vision open", to "treat their incompatibility as a starting-point for speech instead of as an obstacle to be avoided" (2002: 10). Retaining a conception of the irreducible relationship between language and vision as a point of departure entails "eras[ing] proper names and preserv[ing] the infinity of the task" (p. 10).

Indeed it is through Foucault's language his meticulous, astute description of the visual world before him that the painting's self-reflexive acknowledgement of its artifice and crucially its status as representation emerges. His act of observing and describing draws from the pictorial surface a complex network of visual exchanges which simultaneously reinforces and dissolves assumptions about the relationship between painter, subject-model, world and viewer; between those who represent, those who are represented and those who look: From the eyes of the painter to what he is observing there runs a compelling line that we, the onlookers, have no power of evading: it runs through the real picture and emerges from its surface to join the place from which we see the painter observing us; this dotted line reaches to us ineluctably, and links us to the representation of the picture.

In appearance, this locus is a simple one; a matter of pure reciprocity: we are looking at a picture in which the painter is in turn looking out at us. A mere confrontation, eyes catching one another's glance, direct looks superimposing themselves upon one another as they cross. And yet this slender line of reciprocal visibility embraces a whole complex network of uncertainties, exchanges, and feints. The painter is turning his eyes towards us only in so far as we happen to occupy the same position as his subject. We, the spectators are an additional factor. Though greeted by that gaze, we are also dismissed by it, replaced by that which was always there before we were: the model itself.

But, inversely, the painter's gaze, addressed to the void confronting him outside the picture, accepts as many models as there are spectators; in this precise but neutral place, the observer and the observed take part in a ceaseless exchange. No gaze is stable, or rather, in the neutral furrow of the gaze piercing at a right angle through the canvas, subject and object, the spectator and the model, reverse their roles to infinity.

Foucault concludes: Perhaps there exists, in this painting by Velazquez, the representation as it were of Classical representation, and the definition of the space it opens up to us representation, freed finally from the relation that was impeding it, can offer itself as representation in its pure form.

Situated within the context of The Order of Things the major concern of which is Foucault's articulation of his archaeology of thought Velazquez's Las Meninas marks a threshold in the history of systems of thought. The painting's significance rests in its illumination of an epistemic shift what Foucault conceptualises as a discontinuity in the episteme of Western culture. (2) Its recognition of its status as representation is made possible by a reconfiguration of the structures that define the conditions, borderlines and possibilities of knowledge through time.

This essay does not situate Foucault's Las Meninas within the context of its publication in The Order of Things, Foucault's articulation of archaeological inquiry and his theoretical and methodological trajectory. Neither is it an attempt to engage with the painting itself. This essay aims rather to draw attention to the ways in which Foucault's Las Meninas has been situated within art history and to gauge its significance to the discipline.

It will focus specifically on the importance of Foucault's examination of Velazquez's painting to art historian Svetlana Alpers's (3) 1983 essay "Interpretation without Representation, or, The Viewing of Las Meninas" (4) and to Bryson's 1988 book of essays titled Calligram: Essays in New Art History from France within which Foucault's examination of Las Meninas appeared. (5) The volume was edited by art historian Norman Bryson. (6)

In their work as art historians, both Alpers and Bryson draw attention to the contribution of scholars writing about art, outside of the parameters of art history. Alpers introduces her essay thus: "Along with Vermeer's Art of Painting and Courbet's Studio, Velazquez's Las Meninas is surely one of the greatest representations of pictorial representation in all of Western painting" (1995: 285). She then poses the questions: "Why has this work eluded full and satisfactory discussion by art historians? Why should it be that the major study, the most serious and sustained piece of writing on this work in our time is by Michel Foucault?" (p. 258).

In a similar vein, drawing attention to the significance of work produced outside of art history, Bryson comments: When Michel Foucault, in The Order of Things, analyses Velazquez's Las Meninas, and Jacques Lacan, in The Four Fundamental Concepts, discusses Holbein's painting of The French Ambassadors, we find important theses being presented across what is to us an entirely unknown and unfamiliar idiom, a form of writing that is not art history as we in the English-speaking world know it (yet if it is not art history, what is it?).

Prior to Foucault's study, arguably the most well-known text on Velazquez's Las Meninas in the English-speaking world of the 1960s and early 1970s, was Kenneth Clark's essay published in the volume Looking at Pictures. The book, initially published in 1960, was reprinted in the early 1970s and is a compilation of essays that had appeared in the Sunday Times. While initially written for a newspaper and not for a strictly scholarly public, Clark was trained as an art historian. (7)

He was considered a conservative and controversial figure in part due to his perspectives on modern art. In 1966 he began writing and producing Civilisation for the BBC, a television series on the history of art that made him internationally famous when it was broadcast in 1969. The following is an extract taken from Clark's essay on Velazquez's painting: Each focal point involves us in a new set of relations; and to paint a complex group like the Meninas, the painter must carry in his head a single consistent scale of relations which he can apply throughout. He may use all kinds of devices to help him to do this perspective is one of them but ultimately the truth about a complete visual impression depends on one thing, truth of tone.

One cannot look for long at Las Meninas without wanting to find out how it is done. I remember that when it hung in Geneva in 1939 I used to go very early in the morning, before the gallery was open, and try to stalk it, as if it really were alive. I would start from as far away as I could, when the illusion was complete, and come gradually nearer, until suddenly what had been a hand, and a ribbon, and a piece of velvet, dissolved into a salad of beautiful brush strokes.

Prosaically minded people, from Palomino onwards, have asserted that Velazquez must have used exceptionally long brushes, but the brushes he holds in the Meninas are of normal length, and he also carries a mahlstick, which implies that he put on the last delicate touches from very close to. The fact is that, like all transformations in art, it was not achieved by a technical trick, which can be found out and described, but by a flash of imaginative perception. At the moment when Velazquez's brush turned appearances into paint, he was performing an act of faith which involved his whole being. (Clark 1960: 36-37)

On the network of exchanged glances or looks--so central to Foucault's description--Clark comments only briefly: There is, to begin with, the arrangement of the forms in space, that most revealing and personal expression of our sense of order; and then there is the interplay of their glances, which creates a different network of relationships. Finally there are the characters themselves. (Clark 1960: 38)

Clark's Las Meninas is a composite of his flamboyant and idiosyncratic voice (including a style of writing which in many instances reads like a work of fiction); anecdote; biography; connoisseurship; the reverence of the artist as genius; the art-historical practice of identifying influences and formal and stylistic analysis. It is significant that Foucault's method of observation and description, without the constraints of art-historical texts and methods of analysis, was able to derive from Velazquez's work a reading that, within the context of the discipline, was unprecedented. In fact, in his introduction to the critical anthology Art History and Its Methods art historian Eric Fernie draws attention to the most influential strands of art-historical practice from the mid-twentieth century to the early 1970s.

Fernie comments that the "decline of Hegelianism combined with the effects of modernism on art history gave renewed vigour to the study of the individual artist supported by the techniques of empiricism and connoisseurship (including quality, the canon, style, biography and sources)" (1995: 18). Fernie notes the significance of Erwin Panofsky's iconography; E.H. Gombrich's cultural history; the social history of art developed in the 1940s and 1950s by such Marxist art historians as Frederick Antal and Arnold Hauser whose work followed the "pioneering work of the American anthropological art historian Meyer Shapiro" (p. 18).

He notes "other aspects of Marxist analysis" which are "being applied in more detailed ways to questions related to the social function of art" (for example, the analyses of Walter Benjamin and Theodor Adorno) and to the "character and status of art" (the work of the critic Clement Greenberg) (p. 18). Fernie argues that from the early 1970s onwards, art history and its methods have come under scrutiny for a number of reasons: the narrowness of its range of subject matter and concentration on individual artists whom it classified as geniuses; for its restricted set of methods, consisting chiefly of connoisseurship, the analysis of style and iconography, quality, the canon, dating arguments and biography, for the uniformity of degree curricula offered by departments of the history of art, for its ignoring not only of the social context of art, artist and public, but also structures of power, especially those of relations between art historians and the owners of valuable works of art; and perhaps most important of all, for the lack of attention paid to the changes which had been taking place in the related disciplines of literature and history in the 1960s. (Fernie 1995: 18-19)

Fernie outlines the subsequent development of the "New Art Histories": The new art historians, as they have sometimes been called, shifted the centre of gravity away from objects and towards social context and ideology, that is to the structures of social power, and from there to politics, feminism, psychoanalysis and theory. (Fernie 1995: 19)

He comments on the ways in which theoretical developments in France impacted on art-historical practice and cites the examples of Roland Barthes,Jacques Derrida and Michel Foucault. Of Foucault's influence, he writes: [Foucault's discourse analysis describes] his view of the fractured and multifarious character of power relations in a society; in these terms a painting or a building can be seen as the nodal point of an infinite number of discourses, social, artistic, psychological and so on, and used as a means of identifying hidden agendas of power and control. Foucault's approach reminds us that the art of the past is the art of victors, and that the work of historians is itself conditioned by a web of discourses. (Fernie 1995: 20)

Yet while Fernie notes the significance of Barthes, Derrida and Foucault for the "new art histories", their work is not included in the anthology Art History and Its Methods and there is no mention, in his introduction, of Foucault's work on Las Meninas. However, included in the volume are excerpts from Svetlana Alpers's 1983 essay (8) in which she emphasises the importance of Foucault's reading of Las Meninas for art-historical methods.

In Calligram: Essays in New Art History from France, Foucault's essay features along with work by theorists such as Jan Mukarovsky, (9) Yves Bonnefoy, Julia Kristeva, Jean Baudrillard and Roland Barthes all of whom are not art historians. In his introduction to the volume, Bryson examines the significance of these writings for current debates about art-historical methods and interpretive practices. He suggests that "perhaps the most significant feature of such writing in France [is] the absence of the sense of threshold, of border police ready to pounce one feels the absence of the sense of apology with which the writer in England tends to marginalise his work in the visual arts" (Foucault 1988: xv). As an example, he cites Kenneth Clark's "grand refusal to allow the least whiff of the academy to compromise the pleasures of the cultivated amateur of the wonderful essays on art that in England crop up, yet always at the margins of the distinguished career elsewhere".

Bryson proceeds to express concern about art-historical methods within the English-speaking world on a number of levels: He argues that art history, in tending to emphasise the "context of the work's production" neglects its own "artistic and critical present" and further, that its persistent preoccupation with archival documents was restrictive (1988: xvi). Bryson, critical of what he conceives of as art history's insularity, its inability to reflect critically upon its methods and its disengagement from important scholarly debates, poses the questions: Why do we, in England and America, limit ourselves in this way? When literary criticism, for example, has by contrast become so broad in its horizons, so self-aware in methodology, so confident of its right to read from the present? One answer must be that for us the image is not yet particularly thought of in terms of signs, as something to be interpreted.