The Patron Saint of Superheroes

This is not a calligram. But it’s close.

A calligram, according to James Harkness, is a “poem whose words are arranged in such fashion as to form a picture of its ‘topic'” (60), such as Guillaume Apollinaire’s 1918 “Salut monde”:

Arranged to evoke the Eiffel Tower, the words translate: “Hello world, of which I am the eloquent tongue which your mouth, O Paris, will forever stick out at the Germans.”

Apollinaire, according to Harkness, “was, in fact, one of Magritte’s favorite writers” (60), and so the painter would presumably have read the poet’s collection Calligrammes: Poems of Peace and War 1913-1916 before painting his 1929 The Treachery of Images:

The painting is also known as Ceci n’est pas une pipe or This Is Not a Pipe, the title of a 1968 essay by Michel Foucault, which James Harkness translated in 1983:

I have no particular interest in Foucault and only came across the essay during some tangential Google search (probably relating to my recent obsession with appropriation art, which I’ll try very hard not to talk about here).

Foucault is not a comics scholar. He did, however, provide one of the earliest analyses of the comics gutter:

“On the page of an illustrated book, we seldom pay attention to the small space running above the words and below the drawings, forever serving them as a common frontier. It is there, on these few millimeters of white, the calm sand of the page, that are established all the relations of designation, nomination, description, classification. […] The slender, colorless, neutral strip, which in Magritte’s drawing separates the text and the figure, must be seen as a crevasse — an uncertain, foggy region now dividing the pipe floating in its imagistic heaven from the mundane tramp of words marching in their successive line. Still it is too much to claim that there is a blank or lacuna: instead, it is an absence of space, an effacement of the “common place” between the signs of writing and the lines of the image.” (28-9)

To Foucault, the background color of the image is somehow both “colorless” and “sandy,” while, if “foggy” is visually connotative, also gray. The color (which Magritte painted on a canvas but may evoke the beige of a book page) is also conceptually “neutral” — what Foucault further calls “a neutral, limitless, unspecified space” (15), a space “possessing the neutrality, openness, and inert blankness of paper” (20) — recalling the racial paradox of racial Whiteness, which while literally “sandy” is also sometimes treated as conceptually “colorless” and so outside of race as Color. The “frontier” could be literal but is also as metaphorically suggestive as the “crevasse,” while “an absence of space” is pure abstraction and further paradox.

Foucault is not describing a comics gutter directly but a more general phenomenon that includes comics gutters as a subset. Magritte’s painting, he writes, “is as simple as a page borrowed from a botanical manual: a figure and the text that names it” (19). Foucault is describing any page that combines image and text. This more general approach is useful to comics theory, in part because comics theory has a tendency to describe and analyze gutters as if they were unique to comics, while also sometimes mistaking their merely conventional features for their definingly formal ones.

Foucault also recognizes the shared quality of the rendering hand, which further unifies the juxtaposed elements: “the words … underneath the drawing are themselves drawn-images of words the painter has set apart from the pipe, but within the general (yet still undefinable) perimeter of the picture,” and “conversely, the represented pipe is drawn by the same hand and with the same pen as the letters of the text: it extends the writing more than it illustrates it ….” (23). The same is true of many many single-author works in the comics medium.

In comics terms, Magritte’s painting consists of an unframed panel and an unframed caption box. Adding conventional frames defines those perimeters but without altering the juxtapositional relationships:

Foucault does indirectly reference comics conventions when coining the term “word-bearers” to analyze another Magritte 1928-9 painting, Personnage marchant vers l ‘horizon:

Though the shapes of the “bearers” might suggest speech balloons or thought bubbles, their label-like content (rifle, armchair, horse, cloud, and horizon) might better suit caption boxes or, given the paradoxical shadows they cast, the metafictional play of sound effects or splash-page titles. The fact that Foucault does not acknowledge any of these well-established conventions for “bearing” words within images reflects the unbridgeable distance between comics studies and visual arts analysis in 1968.

Yet Foucault does theorize image-text relationships, a consistently core quality of works in the comics medium. He notes “the inevitability of connecting the text to the drawing” (20) in order to address how Magritte undermines “the traditional function of the legend” (23). For comics, Scott McCloud coined “duo-specific” to name what Magritte describes as “the simple correspondence of the image with its legend” (23). For correspondence that favors either image or text, McCloud coined “word-specific” and “picture-specific.” Foucault describes the same relationships over the past five hundred years of painting:

“Two principles, I believe, ruled Western painting from the fifteenth to the twentieth century. The first asserts the separation between plastic representation (which implies resemblance) and linguistic reference (which excludes it). […] The two systems can neither merge nor intersect. In one way or another, subordination is required. Either the text is ruled by the image (as in those paintings where a book, an inscription, a letter, or the name of a person are represented); or else the image is ruled by the text (as in books where a drawing completes, as if it were merely taking a short cut, the message that words are charged to represent). […] What happens to the text of the book is that it becomes merely a commentary on the image, and the linear channel, through words, of its simultaneous forms; and what happens to the picture is that it is dominated by a text, all of whose significations it figuratively illustrates.” (33)

Foucault rejects the possibility of duo-specificity: “verbal signs and visual representations are never given at once. An order always hierarchizes them, running from the figure to discourse or from discourse to the figure” (33). Magritte breaks that hierarchical dichotomy because his paintings are an “art more committed than any other to the careful and cruel separation of graphic and plastic elements. If they happen to be superimposed within the painting like a legend and its image, it is on condition that the statement contest the obvious identity of the figure, and the name we are prepared to give it” (35).

Foucault describes “the nonrelation” of such image-texts, what McCloud terms a “non-sequitor” word-picture relationship, conceding that some relation may be inevitable, albeit a “very complex and problematic” one (37). Rather than corresponding, image and text may diverge: “Word and object do not tend to constitute a single figure; on the contrary they are deployed in two different dimensions” (42). Where “Kandinsky dismissed the old equivalence between resemblance and affirmation, freeing painting from both, Magritte proceeds by dissociating the two: disrupting their bonds, establishing their inequality, bringing one into play without the other…” (43).

Foucault calls the relationship “nonaffirming,” where “nonaffirmative painting” employs “nonaffirmative verbal statements” with images that break the other ruling principle of resemblance-based representation (53). Though in 1968, the approach may have been “an unmapped space” (54), it has been mapped multiple times since.

In Creating Comics (2021), I summarize four kinds of interactions:

• Duplicate: the two sets primarily overlap each other, neither contributing uniquely to the whole.
• Complement: the two sets primarily correspond, one or both providing additional but congruent qualities to the whole.
• Contrast: the two sets primarily contradict, each providing incongruent qualities to the whole.
• Diverge: the two sets appear primarily unrelated, neither contributing to a whole.

In The Comics Form (2022), I show how the four harmonize a range of approaches, including McCloud’s, Barthes’s, Kloepfer’s, Schwarcz’s, Nikolajeva and Scott’s, and Bateman’s. The earliest is Eco’s, whose 1965 comics analysis accounted for the first three interactions but not the fourth — which three years later is the core of Foucault’s analysis of Magritte. “This Is Not a Pipe” should follow Eco’s “A Reading of Steve Canyon” in a historical survey of comics theory.

If I had read it before publishing The Comics Form, I would have used Foucualt’s categorical insight to subdivide interactions into two clarifying subsets:

The essay also suggests a theory of comics representation and layout. Foucault describes Magritte’s 1962 painting Representation as a “representation of a portion of a ball game, seen from a kind of terrace fenced by a low wall. On the left, the wall is topped by a balustrade, and in the juncture thus formed can be seen exactly the same scene, but on a smaller scale” (44).

The repetition is central to Foucault’s understanding of similitude, the quality that Magritte uses to replace traditional resemblance. Foucault contrasts the two:

“Resemblance has a ‘model,’ an original element that orders and hierarchizes the increasingly less faithful copies that can be struck from it. Resemblance presupposes a primary reference that prescribes and classes. The similar develops in series that have neither beginning nor end, that can be followed in one direction as easily as in another, that obey no hierarchy, but propagate themselves from small differences among small differences. Resemblance serves representation, which rules over it; similitude serves repetition, which ranges across it. Resemblance predicates itself upon a model it must return to and reveal; similitude circulates the simulacrum as an indefinite and reversible relation of the similar to the similar.” (44)

Magritte’s Representation illustrates similitude through juxtaposed repetition. “In the same painting,” writes Foucault, “two images bound thus laterally by a relation of similitude are enough for exterior reference to a model — through resemblance — to be disturbed, rendered floating and uncertain” (44). Roy Cook, after initially proposing a “panel transparency principle” for comics (“Characters, events, and locations within a fictional world described by a comic appear, within the fictional world, as they are depicted in typical panels within that comic” [2012: 134]), instead adopted a position that aligns with Foucault’s analysis of Magritte: “our access to the physical appearance of drawn characters in general is indirect, partial, inferential, and imperfect” (2015: 25). A comics image disturbs the impression of a stable and accessible model. The possibility of transparency collapses under the strain of small (and sometimes large) differences between multiple representations of the same subjects common in works in the comics medium — a rarity in fine art painting but overt in Magritte’s Representation where “Similitude multiplies different affirmations, which dance together, tilting and tumbling over one another” (46).

Foucault’s similitude (Harkness acknowledges that the word is imperfect and “might best be translated as ‘likeness,’ ‘similarity,’ or perhaps ‘a likeness'” [59]) includes the juxtaposition of image-texts too, since Foucault also calls The Treachery of Images “an open network of similitude” (47). I’m tempted to apply “the network of the similar” to comics layout generally, especially since the smaller image in Representation is framed by dietetic elements — a technique common in comics, where series of images and texts representing the same subjects are routinely arranged within a single, subdivided page/canvas.

Foucault centers his discussion of linguistic nonaffirmation and pictorial similitude on his most idiosyncratic assertion, that the core of Magritte’s work is “The Unraveled Calligram”: “Whence comes this strange game, if not from the calligram?” (24); “The operation is a calligram that Magritte has secretly constructed, then carefully unraveled” (20); “Magritte redistributed the text and the image in space. Each regains its place, but not without keeping some of the evasiveness proper to the calligram” (25); “Magritte reopened the trap the calligram had sprung on the thing it described” (28). The last assertion immediately proceeds Foucault’s description of a gutter I quoted first — and so comics generally might fit his calligram-based theory of nonaffirming image-texts.

My own illustrations (one begins and the other concludes this post) are more similar to calligrams than Magritte’s paintings or most image-texts found in or out of the comics medium. Each is composed of two styles of lettering: white letters defined by black negative space and black letters defined by white negative space. A third element emerges in the representational outline produced where the two kinds of negative space meet.

I have no idea what Foucault would think.

[And if you’re really interested in Foucault and comics theory, I explored related points in a 2018 post “Foucault Comics.”]

Tags: Calligram, Guillaume Apollinaire, James Harkness, Magritte, Michel Foucault, The Treachery of images, This is not a pipe