Figure vs. Gestalt
The slippage from figure to body requires a system of representation. The two figures of the snakes appear as bodies because they represent real snakes; figures in a book are visual references to the real (or sometimes ideal) things they depict. Going beyond simple resemblance, Gestalt theory combines the sense of completeness and unity of a shape that is more than the sum of its parts with the notion of representation, in so far as ‘Gestalt’ represents the order or the belonging-together inherent in the shape of a thing. The status of the represented in Gestalt theory, however, is not clear. In the Working Notes to The Visible and the Invisible, Maurice Merleau-Ponty asks himself: “What is a Gestalt?” and answers that it “is the Etwas of which the fragmentary phenomena will be the manifestation”. Although the Gestalt “is not a spatio-temporal individual … it has a certain weight that doubtless fixes it not in an objective site and in a point of objective time, but in a region, a domain, which it dominates, where it reigns, where it is everywhere present without one ever being able to say: it is here. It is transcendence.” By looking at the experience of a Gestalt, Merleau-Ponty concludes that “my body is a Gestalt and it is co-present in every Gestalt; it also, and eminently, is a heavy signification, it is flesh”. 
Both as representation and as Gestalt, the ‘figure’ is tied to a body, but also to a visibility, in which an image in one way or another resembles a body. Or, in other words, if the notion of ‘figure’ is understood as representation or Gestalt, seeing is conflated with knowing, and knowledge with the (visual) presence of a body.
If we apply such understanding to a drawing such as Place Roger Prijou-Valjean, which presents us with a figure, the drawing should represent something that we can know about the site. If I do not focus on the lines but rather on the pattern of corner points, despite its bend the figure is regular enough to reveal a constructive principle, which we could call, for example, ‘stacked squares’. With this particular drawing, I may be giving a geometric analysis of the square. However, in a different drawing, such as 43–49 Rue de la Glacière, the conclusion is not so simple because no constructive principle is apparent, while the same mapping process has taken place. The only thing I ‘know’ in this case is negative: I know that there is no constructive principle apparent in the figure. However, despite this, the figure is a particular visual presentation quite distinct from the others. We may also say, going back to the question of Gestalt raised by Merleau-Ponty, that in 43–49 Rue de la Glacière less of my body is ‘co-present’, making it less of a Gestalt than Place Roger Prijou-Valjean – or a Gestalt in trouble. Now, this may be the case, but in terms of ‘figure’, I would not make such a distinction; that is, both drawings are comparable with regard to their status as figure.
The difference that is made between these two drawings is not due to the act of drawing; it is, in fact, already invested in the material – i.e. the sites on which the drawings were made: Place Roger Prijou-Valjean follows a classical design, while 43–49 Rue de la Glacière is modernist in nature. Thus, it is the material that makes us see a Gestalt in a figure; it is not the figure itself that is always already a Gestalt. Moreover, it also becomes apparent that in a classical pattern, such as Place Roger Prijou-Valjean, perhaps because of its emphasis on Euclidean geometry, the body is already anticipated and made ‘co-present’, while a modernist site refuses this conclusion. It may even be said that the produced appearance of a garden as ‘wild’ forest, as is the case with 43–49 Rue de la Glacière, is akin to the notion of the monochrome, insofar as both forest and monochrome threaten to make the Gestalt disappear into a ‘ground’, which amounts to a negative ‘co-presence’.
Just to complete the picture, a postmodern design such as Les Jardins de l’Arche, for example, perhaps not surprisingly operates less on the notion of the Gestalt and more on that of the figure. It is as if the gardens anticipated not situated bodies, but potential drawings evoked through a play with an often fragmented geometry. Having said this, just to repeat, it is not the sites that make a difference to the figure as evoked by the drawings; the sites may confuse them, or rather us, with a Gestalt that they offer as a more easily obtainable refuge.
Going back to the question of visibility and knowledge, then, the notion of the ‘figure’ that I am proposing is constructed independent of the register of the body. If knowledge qua representation, however, is dependent on the body, the meaning of knowledge and, in fact, that of seeing, has to be re-negotiated, in particular in the context of a practice such as mine that calls itself ‘artistic research’.
Jean-François Lytoard conceptualizes three different types of ‘figures’ : (1) The “image-figure” as figurative representation, such as when opposed to a ground, the most conventional use of the notion of the figure in art theory; (2) the “form-figure” as the constitutive principle, such as the constellation or the Gestalt of a image-figure; it “is present in the perceptible, it may even be visible, but is in general not seen”; (3) the “matrix-figure”, which is invisible since it is the differential principle of disruption of the binary relation of the visible and invisible and indeed any binary relation. The matrix-figure’s “formal condition”, as Rosalind Krauss notes, is a “rhythm or pulse”.  Writing about the artists of the “optical unconscious”, such as Max Ernst or Marcel Duchamp, she states that: “the pulse they employ is not understood to be structurally distinct from vision but to be at work from deep inside it”.  As a consequence, according to David Carroll, “each of [the three aspects of the figure] is a complication of the visual nature of the figure”.  Although all types of figures essentially belong together in what Lyotard calls “the figural”, the “form-figure” and more especially the “matrix-figure” disrupt simple visibility within representation. In other words, inscribed in visibility is disruptive visuality, which is not of the same order.