The Poem as Work on Paper: The Illustrated Esthétique du Mal

LIKE OTHER LONG POEMS by Wallace Stevens, the composition of his wartime poem “Esthétique du Mal” was partly shaped by the dimensions of the legal notepad sheets on which it was drafted. When critics address the genesis of Stevens’s poems, they reflexively—and rightly—recall the image of Stevens composing his poems on walks to and from work, his lines the product of a perambulatory rhythm and thinking. But equally pertinent to a poem like “Esthétique du Mal” is the image of Stevens as a kind of draftsman, rounding off his rhetorical flights according to the space of the notepad. Think, for example, of the retroactivity of Stevens’s phrase in an interview on the composition of the poem, where he explains that he “was writing on a pad of paper, and the contents of each sheet became a separate stanza” (qtd. in Jaeckle 236). This is not to say that each stanza or canto ends simply because Stevens ran out of space, but that each section of the poem fits the page in a loose but crucially determined way.

This is true of both the composition of the poem and the experience of reading it. From notepad sheet to page, the distinctive one-canto-per-page format of the poem was faithfully reproduced when it was first published in the Kenyon Review in 1944. The page joins other formal elements, like line and stanza, as one of the poem’s crucial organizing constraints. While the dividing principle of the page is dispensed with in anthologies such as the Collected Poetry and Prose, I want to make an argument here for its importance to our experience of a long poem like “Esthétique du Mal” as a formal, visual, and material element. In its original form, each canto is isolated in its own discrete moment of reading, and appears less a part of a cumulative argument than a separate instance in an ongoing series. And while some cantos recall and develop earlier cantos, any continuities are qualified by the borders of the pages. In general, unlike a break in line or stanza, the end of a page is an interruption from the outside, a material encroachment upon our reading that requires a manual response. This is all to say that the page-by-page distribution of Stevens’s poem suits its discursive mode of a nebulous process of thinking through a series of discontinuous and various postulations, situations, or what I am tempted to summarize with the word “tableaux”—a word whose double sense refers both to a figural scene and to a portable surface used for manual inscription.¹

The poem’s second publication in 1945—the subject of this article—is thus significant for the way in which it makes the poem’s originally paperbound seriality even more visible and aestheticized. The beautifully wrought limited edition published by the Cummington Press, with sixteen diagrammatic drawings by the painter Paul Wightman Williams, mounts the poem as a work on paper.² As an art-historical category, “work on paper” is as capacious as “painting.” The latter is defined broadly according to the material media of paint while the former is defined broadly according to the material support or surface upon which the medium is applied—whether it be ink, graphite, watercolors, or, indeed, print. With the Cummington Press, which was established after the acquisition of a handpress by the Cummington School of the Arts in Massachusetts in 1939, the emphasis was always on the artisanship and artistry of manual printing. Rather than a medium for mass reproduction, the book—as Stevens put it in a letter to Harry Duncan, the publisher and lead printer at the Press—“contains much more of the individual printer and individual artist” (L 523). Like all Cummington books, Esthétique du Mal was hand-set. Three hundred and forty copies were printed in total, forty of which were hand-colored by Wightman Williams and signed by Stevens. The edition is thus the site of a convergence: a long poem that takes the page as its functional unit and an artisanal approach to printing that constitutes the book as a work on paper, in the art-historical sense, both as a manually printed object and a collection of pen-and-ink drawings.

My focus in this article is therefore on developing the claim that the visual elements of the illustrated Esthétique du Mal may enrich our readings of the poem in underexamined ways. Illustrated literature of this kind requires, inevitably, an interdisciplinary lens, and I argue that literary-critical approaches to the page in the study of poetic form and art-historical approaches to paper may enter into a productive dialogue. In so doing, I am particularly reliant on the modern discourse on the art of drawing. Despite their auxiliary presence in the Cummington Press edition, Wightman Williams’s illustrations demonstrate a considered exploration of the possibilities of the drawn line in a way that is comparable with much more prominent figures in modern drawing. Additionally, as I will soon show, twentieth-century approaches to drawing—Walter Benjamin’s writing on the subject being a good example—are often preoccupied with the relation between the line and the material support in ways that are relevant to both the poem and the illustrations here.

I am aided by Glen MacLeod’s writing on the relation between Stevens’s long poems (albeit from a slightly later period) and the visual arts. Specifically, MacLeod finds a surprising but instructive analogy between the later long poems of The Auroras of Autumn and Jackson Pollock’s modification of surrealist automatism. As MacLeod is aware, one of Pollock’s innovations was to find a means of transposing from paper onto canvas the fundamentally graphic procedures of automatic writing and drawing. (Incidentally, it was another poet, Frank O’Hara, who was perhaps the first to champion Pollock’s painting as a kind of “draftsmanship”: “each change in the individual line is what every draftsman has always dreamed of: color” [26].) For MacLeod, Pollock’s poured paintings and their “mental process of free association and continuous automatism” mirror Stevens’s late, long poems and their “comparable form of speeded-up, continuous automatism,” a parallel exemplified by Stevens’s eponymous auroras as an image of lineal flux (181–82).³

Indeed, both “The Auroras of Autumn” (the poem) and “An Ordinary Evening in New Haven” are governed, in their processes of thinking, by metaphors of drift and motility (the northern lights, the walk). By contrast, the slightly earlier long poems from Transport to Summer, “Esthétique du Mal” and “Notes Toward a Supreme Fiction”—the two that were published by the Cummington Press—are too reliant on patient but fragmentary conceptualization and on constructing coherent theoretical stances to fit the Pollock comparison in this respect. What I want to propose instead is that, as Pollock’s action painting grew out of drawing practices, the long poems from Transport to Summer may be productively set alongside drawing as the art of the unexhausted if constraining surface. Of course, this analogy is already an internal dynamic of the illustrated Esthétique du Mal; it is, in fact, its leitmotif.

As I have suggested, the most immediate function of Wightman Williams’s drawings for Esthétique du Mal is to sensitize the reader’s attention to, and activate, the page on which they are drawn. As such, they fulfill what is, for Benjamin, the criterion of all “true” drawing, that is, they draw out the material or background on which they are traced. “The graphic line is defined by its contrast with area,” Benjamin writes in his 1917 essay “Painting, or Signs and Marks” (83), an essay in which he sets up an opposition to (in German) the Mal (mark) and the Malerei (painting). The latter category, in Benjamin’s view, completely fills the surface or coincides with it as medium, exactly in the way that Pollock’s overwhelmed canvases would celebrate a few decades later with what Clement Greenberg hailed as an unrivaled ability to turn the painting into a “single synoptic image” (217). The Baudelairean mal of Stevens’s title thus attains a surprising translinguistic resonance with its Benjaminian homonym. I make this point without being entirely frivolous. The aesthetic of Esthétique du Mal is often richly coloristic.⁴ Stevens aligns the incommunicable gradations and shades of color not just with the sensuous world but with other modes of incommunicability—especially those of experiences of pain and suffering. See, for example, the superlative rhetoric of canto VII’s opening, “How red the rose . . .” (CPP 281), or canto XIII’s “Versicolorings” (CPP 285), a word that refers to variegated or fluxional colors. More pertinent for me in this case, however, is the image in canto VI of the bird that “downwardly revolves,” “Evading the point of redness” and resisting the “country colors crowding against it” (CPP 281). Stevens is painting with words the distinction between disegno and colore, the line of form and the Mal of matter: the bird, a mobile agent, moves like the point of a pencil, leaving the trace of its movements among the ineffable patches of color of the external world. Canto VI, therefore, does not just recall the ending of “Sunday Morning” but the very first poem of Harmonium, “Earthy Anecdote,” whose graphic tableaux of “bucks” moving in alternating “swift, circular line[s]” so as to evade a “firecat” reads almost like a transcription, in advance, of a Wightman Williams illustration (CPP 3). One of the frictions between the poem “Esthétique du Mal” and its illustrations, then, results from the contrast between the occasional painterly register of the poem as a means of marking the immeasurable particulars of the sensible world, on the one hand, and the insistent grid-making of the linear operations of Wightman Williams’s drawings, on the other.

Contrast is certainly key in the illustrated Esthétique. In drawing out the page, Wightman Williams’s illustrations do not merely certify the visual and material dimensions of the text, but also challenge and harass it into forced enjambments. The pages of Esthétique du Mal correspond with W. J. T. Mitchell’s influential term “imagetexts” (9), in which the visual and the verbal are composited, on the page, as one field. Instead of the parity this term suggests, though, the dynamic between poem and drawing in Esthétique du Mal is one of dissymmetry, alterity, and mild antagonism. For his part, Harry Duncan likened the relationship between the poem and the drawings to “an old fashioned marriage . . . with one part quite dominant: the drawings hover around, comment on, almost haunt the poetry” (WST/1/43/2 43b).⁵ Although the poem dominates, its shape is altered by the presence of the drawings, whose appearance on the page recalls the marginal doodles of illuminated manuscripts, or the kind of idle graffiti we might find in misused library books. Perhaps the overriding thematic register of these drawings is their martial quality, not just in their fractiousness but also insofar as they tend to resemble military field diagrams, with strategic lines of assault and defense. Like a child’s drawing of a field of conflict and conquest, they suggest arbitrarily partitioned zones of sovereignty.⁶

This is, of course, entirely unsurprising given the wartime context and content of the poem. At the time of the publication of Esthétique du Mal, the Cummington School of Arts was closed owing to war conditions. Duncan, rejected from military service, continued working on the Press, and in the winter of 1944–1945 he and Paul Wightman Williams repurposed the school’s kitchen as an improvised pressroom (see Richmond). This is the backdrop for Duncan’s gravitation toward works, like Stevens’s, that underscored the importance of art and poetry as having a role to play, or something to say, as part of the war effort. When he drew the illustrations for Esthétique du Mal, Wightman Williams was fresh from drawing the title page vignette for William Carlos Williams’s The Wedge, which the Press published in 1944. Like Stevens’s poem, The Wedge marked a serious attempt by the poet to make a claim for the role of poetry in a time of global conflict. Indeed, Williams makes the striking claim in his introduction to The Wedge that “The arts generally are not, nor is this writing a diversion from [the war] for relief, a turning away. It is the war or part of it, merely a different sector of the field” (53). Stevens, meanwhile, as Rachel Galvin has shown, had by this point arrived at a more ambivalent attitude toward writing what Galvin calls the “civilian war poem.” Esthétique du Mal, she argues convincingly, “is marked by rhetorical gestures of inadequacy that vividly portray the civilian experience” (194). Between the two, comparing Wightman Williams’s cover illustrations reveals an insightful visual commentator on US wartime poetry, adapting his graphic register to suit his poetic subjects:

Fig 1.

Esthétique du Mal, title page.

Fig 2.

The Wedge, title page.

Note the contrast between the indelible, insignia-like triangle of the drawing for The Wedge and the malformed and chaotic diagrams of Esthétique du Mal, an apt distillation of the contrast between the two poets’ visions of aesthetic order in a time of war. Barrett Watten has written on Wightman Williams’s illustration for The Wedge, seeing it as a powerful reading of the “relation of form to discontinuity and destruction. . . . This schematic design—in its reduced referentiality and functionality—points somewhat ironically to the uncanny disjunctions and reinforcements of Williams’s poetic material, each positioned within a formally structured framework, much like the bomb sight” (199). What Watten identifies as the strengths of Wightman Williams’s illustrative art here—its interpretive and ironic capability, as well as its minimalistic and reflexive approach to the construction of form—applies equally to the drawings for Esthétique du Mal, but on a more comprehensive scale, since illustrating each canto of the poem gives Wightman Williams full rein to experiment with the syntax and seriality of recurring structures and linear formations. Taking the illustration seriously, Watten analyses Wightman Williams’s drawing in the context of a broader historicizing thesis. Watten’s claim, also relevant here, is for a moment in “mid-century modernism”—including T. S. Eliot’s Four Quartets, Williams’s Paterson, and Ezra Pound’s Pisan Cantos—that “advances from an anticipatory wish for destruction to an anguished positing of that which succeeds it: a historically contingent demand for universals” (198). Though omitted by Watten, Esthétique du Mal, surely, partakes of this same sensibility, from the opening canto—which derives its theoretical exigency from the anticipation of the symbolic eruption of Vesuvius—through to the capacious “we” of the final canto’s final line (CPP 287). From the precarity of the Neapolitan hotel to the poem’s closing sense of contingent but rooted dwelling, rather than being a “wish for destruction,” Stevens’s will in Esthétique du Mal is to find a means of persisting, poetically, while becoming reconciled to the permanent possibility of destruction. Ultimately, however, Stevens is too much of a Nietzschean in his disposition toward the withdrawal of transcendent categories—that is, “truths”—to find satisfaction in any universalizing thesis beyond the kind of perspectivism implied by the “fragmentary tragedy” of the poem’s thirteenth canto (CPP 285). This fragmentary perspectivism is visualized not just in the isolation of each canto on a separate page but also in the shifting forms of the accompanying drawings.

In the rest of this article, I will comment on a representative selection of the illustrations. My aim is twofold. First, I want to consider how the frictions between Wightman Williams’s and Stevens’s lines enrich the poem’s themes of relatedness, universality, and habitation. These three themes may be thought of in terms of three corresponding questions: how does the shared possibility of suffering relate us to the world around us—both the ontological world and the social world? Can this relatedness be bound and thought of in terms of universals? How should we inhabit a world that is, in essence, predicated on the inevitability of suffering? Secondarily, I want to begin to stake the place of Stevens’s bibliophilic edition within the development of graphic poetics in the United States. This is a broader project whose elaboration is outside of my scope here, but the Cummington Press edition of Esthétique du Mal constitutes a precursor of and rich comparator to collaborative and reflexive experimentations with illustrated poetry, such as those by John Ashbery (with Joe Brainard) on The Vermont Notebook, Frank O’Hara (with a number of artists, but notably the Italian painter Mario Schifano) on Words and Drawings, and Robert Creeley (with Archie Rand) on Drawn & Quartered.⁷ To a varying extent, what unites these works is that the images are less subordinate to, or visually explicatory of, the poems they accompany than a conventional notion of “illustration” would lead us to think. Instead, the image is part of the visual texture of the poem and vice versa.

First I want to return to the correspondence between Stevens and Harry Duncan. Having been sent, unsolicited, Wightman Williams’s drawings by Duncan, Stevens demurred from commenting on their illustrative felicity, on which he maintained, throughout the exchange, an attitude of scrupulous deferral (“Just how apposite they are I shall have to leave to [Wightman Williams’s] virtue”). He does, however, aver that he finds Wightman Williams’s drawings “extraordinary” and that, after rummaging in the attic, he “couldn’t find anywhere any line like his” (L 487–88). Note, here, the familiar trope, common to the discourses of both drawing and poetry, of the “line” as the singular stylistic signature and trace of the artist. Wightman Williams’s drawings are obsessed with the abstract and figurative possibilities of the rudimentary wandering line—what Paul Klee called “An active line on a walk” (16). Indeed, even though Stevens, a noted aficionado of Klee’s work, did not himself note any parallel, Wightman Williams’s minimal style of combining boldly traced, variously straight and sinuous lines into abstract and sometimes suggestively figurative formations is heavily redolent of Klee’s distinct brand of line-drawing, most famously exemplified by the 1922 painting Twittering Machine.⁸ It also recalls the surrealist trope of eliding, through the flow of the line, bodily gesture and mechanical schematism; in this respect, there are similarities with the drawings of artists like Francis Picabia.

It would be unfair to claim that one of Stevens’s motivations for proceeding with the illustrated publication was the prospect of participating in a trend—what in contemporary parlance might be called the Fear of Missing Out.⁹ Yet it is still notable that he expresses a consciousness of a broader “tendency to illustrate poetry”: “I notice that some one in London is doing lithographs for this and drawings for that, and that young Nicholas Moore is publishing a volume with drawings by Lucian Freud” (WST/1/38/2 38b). Notably still: Stevens’s solicitous transatlantic glance is contrasted immediately with a witty, bitter reflection on his lamentable subjection in Franklin D. Roosevelt’s America: “I have lived under the New Deal so long I can take Mr. Williams’ graphs, although I should probably not be able to stand up to Freudian analysis” (L 488).¹⁰ Stevens’s fitting family association between Lucian and his grandfather Sigmund, and therefore between illustrational and analytical interpretation, ends up reading like a deprecation of Wightman Williams’s artistic merits. His equally fitting use of the word “graphs” to describe Williams’s illustrations, in contrast to the conception of an “analytical” illustration (as it happens, Lucian Freud’s drawings for Moore’s The Glass Tower constitute a pretty harmless menagerie), reinforces the notion of a surficial illustration, plotting and mapping the rhythms of the poem’s thinking as well as the surface of the page itself.

In one of the few commentaries on Esthétique du Mal to focus on the Cummington edition, Jacqueline Vaught Brogan leans heavily on Wightman Williams’s drawings to argue that the illustrated Esthétique du Mal “shows a Wallace Stevens whose poetic and aesthetic responses to a world at war are more clearly in line with the latent feminist protest of Elizabeth Bishop’s poetry.” This argument is based on Brogan’s reading of the title page illustration as evoking “a Madonna-like figure in gesture and form with a feminized (emasculated) world that has been caught in her womb by the entrapping lines of war and words of the time” (125). Brogan’s broader political reading is fascinating and persuasive, but I am hesitant about this reading of the illustration for two reasons. First, to impute the drawings to Stevens as an “aesthetic response” is problematic since, as we have seen, Stevens had no say or creative input (beyond his enthusiastic approval) on Wightman Williams’s drawings; and second, Brogan’s interpretation rather forcefully imposes an elaborate figurative reading in a way that belies the liminal nature of the drawing. Figuration, in this instance, always seems to be a by-product of the necessary abstraction of the drawings’ basic linear elements. As I wish to show, the drawings teasingly hint at the recognizable appearance of both figures and faces. For this reason, I would suggest that we view the title page illustration as doing two things. On the one hand, it introduces, with its weblike network, the aesthetic of the graph or diagram, as well as the recurring image of densely webbed concentric circles (a world or a womb, as Brogan suggests, but equally an aperture or a void—it is both an image of origin and of negation). On the other, it suggests a ghostly, spectral figure, indicating the spectrality of figuration itself as a way of setting the terms of the illustrations that follow it. Duncan’s description of the drawings as “haunt[ing]” the poem is particularly apt here (WST/1/43/2 43b), but we could point, too, to the poem’s hauntological tendency, with its historical “shadows” and “phantoms” (CPP 281–82): the persistence of the past is represented elegantly by the trace of the pen.

The similarly encroaching illustrations to cantos I and II underscore the continuity of the subject of the “he” with the recurrence of a minimal face on the right side of the graph. In the first illustration, it is facing the triangular structure of Vesuvius; in canto II it looks at the doubled circular appearance of the moon. The diagonal lines of the first illustration are modified in the second to represent a pull toward linear perspective, a more realized construction of three-dimensional space. As with the direction of the poem itself, this sets up an early sense of continuity before a later withdrawal into “unpeopled” abstraction (CPP 284). If we consider Stevens’s famous reality-imagination schema, then throughout the edition, in drawings such as these which take a landscape format, the right side tends to reflect the originating locus of the imagination, whereas the left suggests an apprehension of the canto’s governing image for external “reality.”

Fig 3.

Esthétique du Mal, excerpt from canto I.

Fig 4.

Esthétique du Mal, excerpt from canto II.

In each case, the construction of the face—resonating with Gilles Deleuze’s description of the face as “a structured, spatial organization” (15)—is made of the same lines that mark and map out the inhuman totality of the diagram. In other words, there is a generally fluid relation in these drawings between figure and field. Thus, the “human” is essentially part of the indifferent field being marked out, as an elegant visual analogue of the “us” that constitutes the reflexive or affective part of the otherwise indifferent “total past”: “And yet, except for us, / The total past felt nothing when destroyed.” The moon is not just doubled but bisected, half excluded from the web, directly capturing the way in which the moon “evade[s],” inevitably, complete apprehension or assimilation (CPP 277). The lines that bar the appearance of the face serve to ironize the couched rumination, in the second canto, on “This freedom” (CPP 278).

In an article on five differing editions of the poem, Jeff Jaeckle rightly draws our attention to the ways in which Wightman Williams’s drawings bully and force the poem into breaking its blank-verse regularity for enjambments that it does not make in other editions. Rather than an internal poetic dynamic, enjambment becomes a material necessity, akin to the end of the page. About canto I, Jaeckle suggests that the first forced enjambment—“He was at Naples / writing letters home”—“creates a literal distance on the page that subtly emphasizes the degree of physical separation” between Naples and home (237). But as well as creating a literal distance, the dynamic here is also one of an even more literal proximity: the premature termination of the line serves to insert the illustration directly into syntagmatic relation with the poem, so that readerly attention scans into visual attention. This inaugural line break, isolating the act of writing from its subject, serves to introduce the way in which, throughout the poem, the two graphic acts—writing and drawing—are reunified here in uneasy relation. We identify the act of writing letters with the immediate presence of Wightman Williams’s graphic marks, so that the opening of the poem becomes more reflexive than it is in other editions, implying as it does a perspectivizing association between the act of writing letters—discrete missives to presumably numerous correspondents—and the page-by-page arrangement of the cantos. This is the first example of an illustration drawing out the formal specificity of the poem. The two forced enjambments of the second canto, meanwhile, quite dramatically alter the reading: decomposing what is a difficult sentence (with a pronoun that seemingly refers to the antecedent “pain,” but which arrogates to it various modes of agency) into more manageable phrases, but also riddling it with hesitation, separating subject from predicate and situating the isolated “never sees” within the sightlines of the face, so that, again, the doodle interpellates our reading.

Fig 5.

Esthétique du Mal, excerpt from canto VII.

The illustration to the poem’s famous canto VII provides the clearest example of the military field diagram effect. The pull toward linear perspective is here replaced by a cartographic view, with lines resembling ballistic trajectories and impacts, trenches, and battlefield lineaments—an impression of the brute geometry of conflict. The presence of “Concentric circles” adjacent to their mention in the third stanza invites us, again, to read the illustration synchronically with the poem. What is striking about this drawing is that it suppresses any semblance of figurative illustration, whereas the canto it accompanies cycles through a sequence of concrete, visual motifs—a refusal that thereby accents the unspeakable or unpresentable totality of death and suffering that Stevens gestures toward. Whereas it harasses the first three stanzas, the drawing does not bother the last two stanzas, adding a weight of emphasis and reverence to the poem’s turn toward the quiescence of “sleep,” and the symbolic gesture of the woman: “the soldier of time lies calm beneath that stroke” (CPP 281). In this sense, the drawing reinforces the redemptive weight of the canto’s sudden resolution—but a resolution that, here, seems all the more like an expedient and specious wrapping up necessitated by the end of the page.

Fig 6.

Esthétique du Mal, excerpt from canto VIII.

Fig 7.

Esthétique du Mal, excerpt from canto XIII.

Canto VIII deals with the death of Satan and, sure enough, a horned figure, replete with tail, may be discerned on the right side of the illustration. The withdrawal of a transcendent agent of evil is reflected by the way in which the vague figure of Satan is locked within the web of the drawing. The drawing, then, as with other drawings in the edition, becomes a representation of the organizing and orienting act of an impoverished and decentered imagination, forming shapes, figures, connections, and analogies. Wightman Williams’s repeated refusal of the encompassing outline, the stable contour (almost all of the drawings are loosely bound by irregular lines that overlap and protrude) highlights the contingency of the frame of reference, alluding to the incompletion of the drawing, and therefore complementing Stevens’s ethic of hesitation and preliminarity. Finally, the forced enjambment in canto VIII effects the most straightforward irony yet, that is, by breaking the “passion for yes / that had never been broken”—and therefore asserting the disruptive power of the imaginative faculty’s orienting acts in the drawing as well as its affirmative potential in the poem.

The forced enjambment of canto XIII, meanwhile, has the sobering effect of slowing down our reading so that the new, contingent line “life is a punishment” is isolated. In its landscape structure, with two zones of denser lineation on the left and right sections, this illustration marks a visual rhyme with the illustration to canto VIII, hence my pairing them here. It is an echo that yokes the respective theological considerations of Satan in the earlier canto and patrilineal punishment in the latter, as well as their syntaxes of sequence (the yes-no-yes of canto VIII, the father-son linearity of canto XIII). And though a circular shape has returned, what originated as a face in canto I, then morphed into a figure of Satan in canto VIII, has now been vacated for pure associative line-making and abstraction.

Fig 8.

Esthétique du Mal, excerpt from canto XV.

The illustration to the poem’s final canto has the last word, as it were—at the bottom of the page, acting as a visual coda to, and summation of, the poem. It is quite different from those that have come before it: with its relatively stable, impermeable outline, we are left with an image of totality and containment. Insofar as the individual sections and strata attain various patterns, from the repeated serpentine lines of its lower half to the crisscrosses of its upper half, it offers a semblance of order. It appears that Wightman Williams has salvaged, ultimately, an image of self-containment and persistence. With its sloped back and four limbs, we may also make out the appearance of an inchoate organism, nosing its way off the page. But, crucially, since it points in two directions at once—both away from and back through the poem—the drawing also suggests a double bind, an image of irresolution.

For Charles Altieri, the double bind of the poem’s closing section relates subtly to the paradox of any realism: “The quest for [a] lucid realism may force the poem to become an instance of the very problem of distance that it brings into being” (186). This is the paradox that operates in the final line, with its conjunctions “as” and “where.” Either we take this line as a quietist endorsement of humbly accepting one’s environment and situation as the grounds of a sensuously rich, free, and therefore valuable life, or as an understated assertion of the vital and enriching frameworks, perspectives, and stances afforded by language (the “gaiety of language,” as Stevens memorably puts it in canto XI [CPP 284])—particularly poetic language, insofar as it orients a creative sense of being in the world. What the drawing underscores in this regard is the tension that one route leads us back through the pages of the poem while the other leads us away from them.

Although my examination of the drawings has emphasized their interpretive and additive qualities—Stevens wrote in a letter to James Guthrie that the work of the Press “definitely adds to the text” (L 514)—let me work toward a conclusion by reflecting on the way in which they mark, at the same time, a shortfall. As I set out at the beginning of this essay, the primary function of the drawings, more than their occasionally literal illustrativeness, is to anchor each canto onto the page. Since a line drawing is a record of a gesture or act, drawing in general invites the viewer’s identification with the material processes of its creation and the presence of the artist in a specific moment in time and space. But in their irregular marginal positions and in their abstraction, the drawings carry with them a sense of superfluity and redundancy. The best way to characterize the illustrated Esthétique du Mal, then, is according to this oscillation between complementarity and superfluity. In exhibiting their superfluity, Wightman Williams’s drawings, as we have seen, demonstrate a free play upon the page which isn’t “automatic,” in the sense of the evacuation of attention desired by surrealist practitioners, but which is rather a redirection of attention into constructing images of relatedness—notably, the recurrent image of the concentric circle. The drawing is therefore determined as much, if not more, by the internal logic of its own line-making impulses as it is by the illustrational imperative of providing a representative image for each canto.

It is, as such, a distinctly modernist mode of illustration. As interrogative exercises in form, these drawings relate to Wassily Kandinsky’s modernist definition of “Good drawing” as “drawing that cannot be altered without destruction of [an] inner value, quite irrespective of its correctness as anatomy, botany, or any other science” (53). Illustration is not a science, of course, but it is predicated on the reproduction of “observed” effects. Here, the integrity and completeness of the work are derived from an immanent principle of formal coherence—with drawing as the art of form, par excellence—rather than constituting the work’s fulfillment of transcendent, mimetic fidelity, or any other exogenous aesthetic principle. Finally, there is something aptly considered about the sense of disconnection harbored by the drawings. Poems, after all, require illustrations as much as they require critical interpretations. But both activities are modes of participation that a poem like “Esthétique du Mal,” with its final “we,” invites.