I continue with my critical review of Deleuze’s recently published seminar entitled On Painting. In my previous two posts I first sketched Deleuze’s account of the creative process of artistic painting, which (a) crucially involves the painter’s creation of an individualized and dated ‘diagram’ (an obscure and complex notion, as I have stressed), and the application of this diagram to the problems of painting; and (b) in the second post Deleuze’s explication of the concept of the diagram through what he calls its characteristics, its tendencies, and its positions. These two summaries only take us through the first half of the seminar, and accordingly in this post I’ll attempt to finish summarizing the text and begin the process of critical reflection upon Deleuze’s account of artistic painting. As in my previous posts, I’ll not attempt to capture the philosophical intensity or exuberantly experimental radicalism of Deleuze’s expressive speech; nor shall I so much as mention a number of interesting discussions that strike me as peripheral to the main line of thinking. The two guiding questions here are: What does Deleuze mean by the term ‘modulation’, the concept of which he introduces as constitutive of the activity of artistic painting? and What is Deleuze’s basic conception of painting?
Near the end of the fourth session Deleuze adds a further key feature to his characterization of the diagram: “the diagram is a modulator” (p. 136). What is modulation? I find this the most obscure of all the central discussions in the seminars, but let’s see how it goes: Deleuze says that it’s “hard to develop a concept for it”, but he initially opposes it to ‘articulation’, which he associates with codes consisting of units organized by binary oppositions (his example here phonemes in language, particularly as conceptualized by the linguist Roman Jakobson (p. 135). He also treats articulation as part of the ‘digital’, as “ “digital” is the binary choice determining the unit”, and opposes this to ‘analog’ (p. 122), and to what he calls ‘analogical language’ generally: “Analogical language is about modulation. Digital, or coded, language is about articulation.” (p. 136) So “painting is indeed an analogical language” and “when you say “to paint,” you are saying “to modulate.”” (ibid) Deleuze makes two stabs at positively characterizing modulation. First, “modulation refers to the values of a nonarticulated voice”. Second, and much more centrally for Deleuze, modulation occurs “on the basis of something else”, namely “a carrier or medium according to a signal” (ibid). With regard to painting, “[t]he signal is the model” or ‘the motif’, but also “the surface of the canvas”, and that what is modulated is light and/or color (a couple of dozen pages later Deleuze will seemingly deny this; see below). So Deleuze’s central characterization of modulation in the relevant sense is “that to paint is to modulate light or color—light and color—depending on the flat surface and (these aren’t exclusive) depending on the motif or the model that plays the role of signal.” (p. 137) The result of modulation in artistic painting is “[t]he figure on my [sic] canvas”, such as “Pollock’s line with no figure, or Kandinsky’s abstract figure, or Cézanne or Van Gogh’s figural figure” (ibid).
As he approaches his final characterization of artistic painting as constitutively involving modulation, Deleuze adds a further, and puzzlingly equivocal, feature. “To paint”, he says, “is to modulate to plan, on the plane, that is, on a surface, the canvas.” (p. 158) He elaborates this claim by recurring to a passage his most central contemporaneous philosophical source, the philosopher Gilbert Simondon. In his major (and quite difficult) philosophical treatise Individuation in Light of Notions of Form and Information, Simondon attacks the so-called hylomorphic model of explanation, that is (crudely put), the kind of explanation of phenomena initiated and developed by Aristotle wherein the explanandum is analyzed into (a) some matter that has some combined with (or embodied or been shaped by) (b) some form; on this account things generally, any thing whatsoever, is an instance of formed material as well as materialized form. Simondon argues that the schema of giving-form-to-some-matter fails to capture “the technical operation of form-taking”, and that any such operation can only be understood as further involving the energies or forces within an entire “energetic system” or “energetic regime”. (Simondon, p. 29) One paradigmatic example that seems well-suited to hylomorphic explanation is brick-making: clay (the material) is pushed into a mold (the form), and a brick is produced. Simondon counters that first of all the clay must be “in a state of complete internal resonance” wherein “what occurs at one point reverberates within all the others” in order that the clay has the requisite plasticity. Moreover the operation requires the worker who pushes the clay into the mold. Accordingly, the mold is not a form, but rather “a condition of enclosure, limit, halted expansion, and direction of mediation” that is part of “a system state that requires this realization of energetic conditions, topological conditions, and material conditions” (p. 29) Properly understood, the mold is not a paradigmatic instance of form in the operation of form-taking, but rather is a kind of modulation. More precisely, molding and modulating are extreme cases on either end of a dimension of form-taking. Molding is a kind of modulating wherein the form-giving dimension is fixed; modulating, as in some cases in electronics where the matter is electrons and the form is an electrical field, is a molding where at each moment there is also an unmolding (that is, of the immediately prior electrical state), and so the modulator is a ‘continuous temporal mold’. Between the conceptual extremes of molding and modelling there are intermediate kinds, such as a rolling mill that operates continuous (like a modulator) but produces elements with definitive profiles (like a mold); Simondon calls these intermediate types ‘models’. So the seemingly distinct conceptual types of ‘mold’ and ‘modulator’ are actually part of a continuum ‘mold-model-modulator’, with the place in the continuum determined by the relative fixity of the form-giving element and of result of the operation of form-taking (pp. 30-1).
Deleuze summarizes Simondon’s account and exclaims “Wonderful! That’s exactly what we needed!” (Deleuze (2025), p. 147), and then gives the definition of painting cited at the beginning of the previous paragraph. Yet puzzling and, as suggested previously, equivocally, he explicitly avoids deciding whether the term ‘modulation’ in painting refers only to the extreme end of the continuum, or whether the term covers the full spectrum of conceptual possibilities from molding to modulating (p. 159). To arrive at a definition of the concept of (artistic) painting, one final point is needed. Earlier (p. 137) Deleuze has said that to paint is to modulate light and/or color on the basis of some signal. What is the signal in artistic painting? Deleuze asserts: “The signal is space. A painter paints nothing but space—and maybe time as well, space-time. A painter paints nothing but space-time.” (p. 160) And so he gives his full definition of artistic painting: “to paint is to modulate light or color—or light and color—based on a signal-space. . . It produces. . . this resemblance to the thing that is more profound than the thing itself, this nonsimilar resemblance, produced through different means. The act of modulation is composed precisely of these different means. Modulating light or color based on a signal-space will give us the thing in its presence.” (p. 160; I have elided Deleuze’s references to what he calls ‘the Figure’. Earlier he has contrasted ‘the Figure’ with what is produced in Abstraction and Expressionism, but it seems to me that here and in other places ‘the Figure’ is synonymous with what he has called the ‘pictorial fact’ and/or ‘presence’—a full account of Deleuze’s thought on painting would need to take this into account, in particular his much more central characterization of it in his monograph on Francis Bacon).
All that remains for Deleuze are two historical problems: “What are the major signal-spaces of painting?” and “How does modulation work based on these signal-spaces?” Deleuze’s answers to these questions are heavily dependent upon and close to the accounts given in the early 20th century by the art historians Aloïs Riegl and Heinrich Wölfflin, and the phenomenologist Henri Maldiney. He considers in turn the arts of Ancient Egyptian, Classical Greece, the Byzantines, the fifteenth-, sixteenth-, and seventeenth-centuries, mostly in Italy, and finally the modern painting of Van Gogh, Gauguin, and especially Cézanne. As noted earlier, these accounts focus on the relative relations involved between hand and eye, as well as the relations among the planes of the surface support, pictorial background, and pictorial foreground. I find these accounts fascinating, although familiar and not especially distinctive of Deleuze’s astounding manner. So I’ll break off the summary here.
Returning to and elaborating the schema of the creative process of artistic painting that I proposed in the first post, we can now present Deleuze’s full schema as the following: attack upon clichés catastrophe creation and application of a diagram modulation of light and/or space creation of pictorial fact or presence = the artistic painting. In my final post, I’ll ask and reflect upon a few of the critical questions that immediately arise: Does this schema make sense? Is this schema afflicted by philosophical problems that, according to commentators such as Manfred Frank and Vincent Descombes, also afflict Deleuze’s general philosophy? How does Deleuze’s schema compare with what I called in my first post the common-sense understanding of the creative process?
References and Works Consulted:
Gilles Deleuze, Francis Bacon: The Logic of Sensation (2003)
--On Painting (2025)
Roman Jakobson, ‘Linguistics and Poetics’, in Language in Literature (1987)
Henri Maldiney, Regard Parole Espace (1973)
Alois Riegl, Historical Grammar of the Visual Arts (2004)
Gilbert Simondon, Individuation in Light of Notions of Form and Information (2020)
Heinrich Wölfflin, The Principles of Art History (1950)