ÉDOUARD VUILLARD’S EXPERIMENTAL ARABESQUES
Among Édouard Vuillard’s earliest exhibited works was a painting titled Woman Sleeping (Femme couchée, 1892, fig. 43). Now catalogued under the less gender-specific title Sleep (Le sommeil), it is one of a number of works that Vuillard made in the early 1890s depicting a single individual presumed to be a woman sleeping. Sleeping women were hardly unusual in nineteenth-century French art. Ordinarily licensed by a mythological narrative and more often than not depicted nude, they repeatedly offered, as in a number of paintings by Gustave Courbet, opportunities for realist or naturalist voyeurism.1 Maurice Denis’s Décor, the centerpiece of his début at the Salon des Indépendants of 1891, seems to fit this category of sleeping or semiconscious female nudes. In the large study for the painting, Denis had carefully worked through various possibilities for the figure’s state of somnolence (fig. 20). In the previous chapter, I considered Denis’s Décor at length, as well as the gentle pastel parodies that Vuillard made of Denis’s ambitious work, in conjunction with late nineteenth-century debates over symbolist practices, and in particular the hyperbolic critiques of deformation. Vuillard’s Woman Sleeping was a very different proposition, but one nevertheless in dialogue with symbolist theories and determined by the assumptions underlying the experimental methods of the new French psychology.
| 43 | Édouard Vuillard, Woman Sleeping (otherwise known as Sleep), 1892. Oil on canvas, 33 × 64.5 cm. Musée d’Orsay, Paris. |
With Woman Sleeping, Vuillard refrained from using the image of a sleeping woman as a pretext for representing an implicitly available female nude. In fact, with the figure tucked up in bed, he gave viewers no discernable body at all, and, the original title notwithstanding, one could hardly be sure that the work depicted a woman at all. That this constituted a kind of violence that Vuillard had enacted upon the figure was visualized in a contemporary caricature that pulled back the sheets to reveal a “Sleeping Woman in Pieces” (“Femme couchée en morceaux”) (fig. 44). A grisly play on contemporary headlines,2 the caricature highlights one particularly unusual formal element of the painting that constitutes the central problematic of this chapter. In the wake of cloisonnism, synthetism, and Nabi theories extolling the virtues of the arabesque line, Vuillard might have been expected to render the sinuous contours and folds of the bed linens and other elements in Woman Sleeping as distinct curving lines emphasizing the planarity of the support, as indeed he did in a related watercolor and charcoal drawing (fig. 45). But upon closer inspection, many of the painting’s ostensible lines, especially the curves describing the folds just below the figure’s head, are better described as negative linear spaces, areas between larger zones of paint where a thin ground, and even in places the bare canvas, shows through. These formal oddities, where painted areas seem to shrink from one other, opening up gaps in the material surface of the work, appear in many of Vuillard’s early paintings. But they are especially visible in Woman Sleeping and even more manifest in a portrait of Vuillard’s and Denis’s friend, the actor and soon-to-be theater director Aurélien Lugné-Poë (fig. 4).
| 44 | Caricature from Mirliton, “Troisième exposition des peintres impressionistes et symbolistes,” Le journal, no. 9, literary supplement (26 November 1892), 1. |
| 45 | Édouard Vuillard, Woman in Bed, 1891. Watercolor and charcoal on paper, 14.7 × 22.8 cm. National Gallery of Art, Washington, D.C., Ailsa Mellon Bruce Collection, 1970.17.179. |
In his early journals (also known as carnets), which combine text and sketches, Vuillard not only grappled with naturalist, symbolist, and Nabi theories but also interrogated his own psyche with a view to establishing an artistic method.3 Through this interrogation, he developed—or, more precisely, told himself that he was developing—a method that relied upon his instincts, his habits, and his will: in other words, a procedure that combined both unconscious and conscious mental functioning. This method began with quickly drawn arabesques signifying the direct and immediate observation of nature, “arabesques of observation,” to borrow a term from Roger Benjamin.4 And it proceeded by transforming those arabesques into the curving linear gaps that we see in works such as Woman Sleeping and the portrait of Lugné-Poë. Still reading as lines and often taking the place of contour, these gaps in fact resulted from deliberate and often repetitive acts of painting, the very antithesis of the automatic arabesques with which Vuillard began. In this chapter, I argue that this device is best understood as an arabesque of experiment, an element licensed and determined by the late nineteenth century’s fetishism of experimentalism and, in particular, experimental psychology’s reliance on a pathological method. After a lengthy period of questioning-cum-theorizing revealed in the carnets, Vuillard effectively dematerialized the arabesque line, producing a formal element that Denis might have been tempted to argue was pure idea, hypothesis rigorously proved true. But Denis could have no truck with the place that Vuillard had assigned to instinct and the unconscious in his symbolist visual practice. While Denis’s experimentalism forged a path toward formalist justifications of modernist deformation, Vuillard’s experimentalism reveals the epistemological foundations of the avant-garde’s explorations of the unconscious.
Instinct, Habit, and Will
In 1898, well after the heyday of Nabi group activity, Denis received a letter from Vuillard, which Denis later called a “veritable psychological confession,”5 that disclosed the important roles Vuillard ascribed to instinct and habit over and against the will in his artistic practice. In private, as if he were intimately familiar with Vuillard’s carnets, Denis had previously critiqued Vuillard’s habit of rapidly recording his impressions of the world around him.6 Denis had come to believe that Vuillard had taken impressionism to a sensationalist extreme, reducing painting to “a kind of journal . . . a kind of stenography of everyday sensations,” which had then become a license for any and all symbolist flights of fancy.7 This “sensualist” obsession with subjectivity, Denis was convinced, derived from an outmoded theory of the role of nature in art, but its “primordial error” was its failure to cultivate a slow, willful, and intellectual creative process and consequently to produce significant works of art.8 This failure, Denis claimed, was reflected in what he saw as Vuillard’s inability to finish anything. By 1898, Denis was increasingly judging art like Vuillard’s the unfortunate consequence of the theories that he himself had put forth in 1890. Hoping that his friend had already begun to see the error of his ways, Denis had invited Vuillard to join him in Italy that year to partake in the “lessons in strength” that Denis was then deriving from the Tuscan masters. “What makes a work of art important,” Denis insisted to Vuillard, was the “artist’s effort, the power of his will.”9
Vuillard’s response to Denis’s invitation, his “veritable psychological confession,” questioned whether it was really possible to untangle the complex psychological and physiological functioning of the will. In the mode of working that he knew best, Vuillard wrote to Denis, faith in the value of his ideas tended to abandon him, and he often experienced profound doubts and intense lassitude. At those times, he tended to rely more on his working habits, proceeding “in fits and starts” in a process that stripped “formulated ideas” down to “elementary” ones. Often, Vuillard confessed, he hardly knew the will at all and was guided rather by “instinct, pleasure, or . . . a . . . sense of satisfaction.” Vuillard surmised that their differences came down to divergent understandings and experiences of the will that he ascribed not only to their distinct personalities and circumstances but also to their different educational backgrounds. Agreeing to disagree in a subsequent letter, Denis conceded that “the problem of the will cannot be posed in the same way for you as for me.”10
“The problem of the will,” as Vuillard intimated when he referred to differences in their education, was a centerpiece of French philosophical training. Although the philosophy curriculum had incorporated positivist methods and the new experimental psychology by the late 1880s, it nevertheless maintained its emphasis, derived from eclecticism, on a knowable self, constructed through the introspective methods still fostered in the classroom.11 While Sigmund Freud would later give equal if not greater weight to the unconscious in the construction of the self, the lycée philosophy course, as Jan Goldstein has argued, continued to privilege consciousness in subject formation. The will, in opposition to instinct and habit, took its place in the curriculum as a characteristic, even defining, function of the conscious mind. At the same time that the curriculum foregrounded a healthy functioning of the will as constitutive of the self, however, it also included an outsize place for nature’s experiments and for discussions of abnormal functioning, including automatism, somnambulism, and other supposed diseases of the will.12
Vuillard and Denis had in fact received very similar educations at the Lycée Condorcet in Paris, with one glaring exception. While Denis, along with a number of the Nabis, had taken the philosophy baccalaureate, Vuillard had terminated his formal philosophy training in order to concentrate on mathematics in his final year.13 So, while Denis was learning to construct a robust self through an introspective method that privileged knowledge about the conscious functioning of the will, Vuillard was conscious of constructing a self at one remove from this particularly influential French institution. In 1898, with the benefit of at least a little hindsight, Vuillard had confided to Denis not so much that his will was weak but that he had little experience of the will as Denis had been trained to understand it, and had thus embraced instinctual and habitual work as well as willful creation. Vuillard thereby established himself as deeply invested in, but at a slight remove from, the philosophical foundations of Nabi-symbolist theory as proclaimed by Denis,14 a remove that encouraged Vuillard to establish an important place in his working method for the unconscious encounter with nature. Vuillard’s recognition of his unconscious as a source of inspiration parallels the turn-of-the-century work of experimental psychologists such as Théodore Flournoy and Théodule Ribot, who were soon to identify the unconscious as the seat of creativity.15
Vuillard would synthesize unconscious and conscious functions in what remained a highly idiosyncratic naturalism. For despite some art historians’ claims about Vuillard’s “return” to naturalism after 1900, Vuillard only ever rejected a narrow, almost caricatured view of academic naturalism as the depiction of the externally visible world.16 Vuillard’s Nabi practice sought rather to reorient naturalism toward the self in order to focus on his own perceptions and sensations of nature—not simply to represent them but above all to know and to present them as a form of truth. While, visually, the difference with impressionism was stark, there was but a fine methodological distinction, as Vuillard himself recognized as early as 1895, when he referred to his practice as “interior impressionism.”17
On 21 November 1888, Vuillard noted in his journal that he had woken up with a “gueule de bois,” a French idiom for a hangover (fig. 46).18 A coincidence, perhaps, but the day before, Denis had written the final exam for his philosophy baccalaureate, and it is tempting to imagine the two friends celebrating Denis’s achievement with a night out. Earlier that fall, Vuillard had written to Denis, teasing him about his studying and looking forward to the day when his friend could put philosophy aside and they could go “rambling” outdoors together. In the same intimate letter, Vuillard had complained of his own laziness, nervousness, troubles externalizing his “impressions,” and reluctance to exhibit his work. But Vuillard had also confided to Denis that he finally felt that he was becoming a painter, even though he had not yet “given birth” to anything substantial.19
| 46 | Édouard Vuillard, journal, 22 × 17 cm. Institut de France, ms 5396(1): Carnet I.1, fol. 10r [21 November 1888]. |
Denis’s characterization of Vuillard’s method as journal-like and “stenographic” well describes the early carnets, which contain rapidly done sketches, summary notations, course notes, brief anecdotes, and much longer passages that all reveal a prolonged search for a method to guide his production.20Although Vuillard would reject the mantle of public theoretician, his journals shed light on three especially intensive periods of questioning, all beginning before Vuillard exhibited his first paintings in the fall of 1891. During the first period, from late 1888 into 1889, Vuillard interrogated the assumptions and methods of naturalism but also revealed himself to be closely tied to nature as a model. Although he was not among the group of Académie Julian students to whom Sérusier supposedly “revealed” The Talisman (fig. 37) in October 1888, at some point later that year or, more probably, during the following year, he had begun thinking about the implications of Gauguin’s lesson, which led to his official entry into the Nabi brotherhood, in all likelihood sometime in 1889.21 During this period, he tried working from memory as an alternative to working directly from nature, but quickly found himself disillusioned with this approach.
The second period coincides with the phase of Nabi theorizing leading up to and immediately following Denis’s publication of the “Définition du néo-traditionnisme” in late August 1890. At this point, in the “year of Sérusier,” as Vuillard called it, he pledged near-total allegiance to the Nabi-symbolist rejection of academic naturalism and the consequent redefinition of painting as both surface and objective truth. At the same time, however, anxious about originality and his desire to create a body of work (an oeuvre), Vuillard expressed a pressing need for a method that was entirely his own.
The beginning of the third period corresponds with the public exhibition of the first recognizably Nabi works by Denis and Pierre Bonnard at the 1891 Salon des Indépendants, which opened in March of that year. Lengthy entries continued through the summer of 1891, ending just before Vuillard started exhibiting his own paintings that fall.22 During this period, Vuillard confessed his ongoing dependence on nature and began to synthesize a method that would involve an instinctual, automatic approach to nature, habitual work, and a series of deliberate choices about the formal and material elements of his paintings. Woman Sleeping and the portrait of Lugné-Poë, with their experimental arabesques, encapsulate, even allegorize, the method at which he had arrived. When Vuillard began exhibiting his paintings, revealing himself to the public, the intensity and frequency of his confessional journal entries seem to have diminished. He had come to some semblance of a painterly practice that he judged, at the very least, adequate to both Nabi-symbolist theory and his own artistic needs.
Naturalism as Delusion and “Dirty Little Bits of Memory”
From its earliest pages, dating to late 1888, Vuillard’s first journal contains small sketches—rapid renderings of outdoor scenes, people, and works of art that attest not only to Vuillard’s wanderings through the streets, parks, and museums of Paris but also to his attempts to “fix,” as he would say, his “impressions” on paper. Short notes accompanying many of the sketches elaborate on specific scenes: “evening, immense sidewalk, wall on the right, reddish-brown trees, dusty, large woman all black, evening, night effect.”23 These “stenographic” notations often refer to different weather and lighting, suggesting that, like the impressionists, Vuillard took a particular interest in how varying atmospheric conditions affected his impressions of the world around him.
This method of notation also extended to Vuillard’s observations of himself. In some of the first entries, he made careful note of instances when his perception had been altered by either psychological or physiological means. On one occasion, he noted working while one pupil was dilated.24 On another, he tried to describe a certain dusty ocular effect.25 Elsewhere, he recorded a “bad state of mind” and “considerable irritation.”26 Such notes continued well beyond these early years. In 1894, he noted, “Disorder in my brain,” and in 1912, “interior collapse. . . . Nervous.”27 After waking up with a hangover the day after Denis’s baccalaureate, Vuillard expressed surprise at the relatively “normal” functioning of his eyes.28 In the upper left-hand corner of the page, marked off by a curving line suggesting a distinction between the night before and the morning after, he sketched a man slumped against a wall, possibly a self-portrait remembered from the night before (fig. 46).29 But while Vuillard was observing and recording the world around him, at times relying on his memory to sketch scenes from a previous day, already in late 1888 he had begun to question the validity of both of these approaches. In trying to fix his impressions on the page, Vuillard also recorded a great deal of frustration: “while working lose the impression . . . find it while sketching slowly, didn’t fix one single object.”30
In a lengthier entry from late November 1888, often tantalizingly excerpted, Vuillard expressed serious doubts about the naturalist project under scrutiny. He assumed, in line with Hippolyte Taine’s lectures on art, that a subject perceives nature as a set of relationships between forms, colors, sounds, and so forth.31 It followed, Vuillard argued, that in order to “reproduce” nature, one had to “fix” not only the sensations provoked by individual elements but also one’s perceptions of all of the relationships between elements. In recording one set of relations, Vuillard fretted, one inevitably lost the perception of the others, thereby compromising the veracity of the general impression. The problem was inherent in the temporal nature of creating art. In painting, by definition an act of some duration, one could not easily fix the instantaneity of perception unless, Vuillard wrote somewhat facetiously, one could transform the brain into an automatic recording instrument: “If the mental apparatus is not in a state to capture these relationships,” he mused, “to hold on to them for a moment and then to record them onto paper or canvas like a somnambulist, it’s a waste of time.” Far from advocating somnambulism as a creative strategy at this stage, Vuillard’s statement was on the whole a lament about the challenges of naturalism. “Doing nature,” he had concluded, was therefore only “to upset [and] to delude yourself.”32
Vuillard’s critique resonated with one frequently made by symbolist critics, who accused impressionists of trying to be simple recording instruments, taking naturalism to an absurd extreme of instantaneous observation. Paul Cézanne’s famous remark about Claude Monet—“He is only an eye, but good God what an eye!”—underlined the late nineteenth century’s perceived limitations of the impressionist project.33 But while Vuillard sowed the seeds of doubt about naturalism, thereby laying the groundwork for a turn to symbolist theories, he did not reject wholesale all attempts to faithfully reproduce nature; nor, at this point, did he question the desire to paint nature in some form. Rather, Vuillard was already hinting at an ambivalence that would eventually become the root of his continuing adherence to many aspects of naturalism and impressionism, those sensualist approaches to nature that Denis later criticized as outmoded.
Vuillard had already combined working from nature with working from memory—for instance, in his hangover sketch—but henceforth he made a more self-conscious effort to use memory as an antidote to the “delusional” methods of naturalism. For many artists and theorists of art, memory afforded a means of synthesizing sense perceptions before attempting to record them as both an ensemble and an ensemble of relationships. But Vuillard’s concerns about naturalism were soon mirrored by an even greater ambivalence about working from memory. In late 1888 or early 1889, Vuillard noted that he was attempting to paint his sister Marie from memory.34 The accompanying sketches, however, give little indication of this memory work, although a simplified and stylized face, probably a self-portrait,35 might suggest an attempt to sketch what he remembered of his own countenance (fig. 47). Evidently, Vuillard endeavored to work from his memories of things that were most familiar to him, composite perceptions of the faces he knew best.
| 47 | Édouard Vuillard, journal, 22 × 17 cm. Institut de France, ms 5396(1): Carnet I.1, fol. 32v [December 1888]. |
Since the middle of the nineteenth century, working from memory had gained an increasingly large foothold in French artistic training, but it had only recently emerged as a key strategy for artists who sought to liberate themselves from the schema of naturalism. Gauguin recommended working from memory as a way to inscribe the self in the work of art, and many in his circle followed his lead in the hope of liberating themselves from the model and orienting their own awareness and the viewer’s toward inner states and experiences.36 But vanguard artists, including those who admired Gauguin, did not universally embrace mnemonic strategies. While working with Gauguin in Arles in the fall of 1888, Vincent van Gogh had tried to apply Gauguin’s method, but, holding fast to his belief in the crucial importance of the model, at least for his own work, he struggled to put Gauguin’s recommendations into practice.37
Just as soon as Vuillard noted that he had begun working from memory, he recorded his disappointment that there had been little or no change in his work.38 And indeed, the facture, coloration, and overall atmosphere of the many portraits and self-portraits from late 1888 and 1889, including a portrait of his sister (fig. 48), recall the competent and relatively conventional naturalist portraits and still-life paintings that Vuillard had produced throughout much of the previous year. In autobiographical notes written in 1908, he marked 1889 as the year he “started to work from memory,” but a few years earlier, probably in 1905, he had recalled that moment in more negative terms, writing that 1889 was the year of “dirty little bits [petites salissures] of memory.”39
| 48 | Édouard Vuillard, Portrait of Marie Vuillard, 1888. Oil on canvas, 32.5 × 24.4 cm. Sold by Christie’s Paris, 20 May 2009. |
Regardless of whether he was aware of Gauguin’s ideas in late 1888 or early 1889, Vuillard’s interest in and appeal to memory remained firmly in the service of naturalism rather than of symbolism or self-expression. But he found working from memory an inadequate solution to the problem of faithfully representing nature. Despite his qualms about naturalism, Vuillard discovered that, like Van Gogh, he had no real desire to give up working from nature, and he soon admitted that “the idea of working from nature . . . pursues me.”40 Sketches in the journals attest to his ongoing desire to record the world around him, as he instinctively maintained his habits of naturalist notation. By early 1889, still an art student with little or no knowledge of symbolist theory, Vuillard had independently come to a symbolist critique of naturalism, but he had also reaffirmed the central role of nature in his art practice.
Simply from the number of times Sérusier’s name appears in the journals in 1890, it is clear that Vuillard was dealing that year with the full import of Sérusier’s ideas, Gauguin’s teachings, and symbolist theory more generally, via Denis and in conjunction with his more active participation in the social and artistic activities of the Nabi brotherhood.41 Vuillard was more implicated in the publishing of Denis’s “Définition” than is generally acknowledged. Possibly at the behest of Lugné-Poë, he acted as middleman between Jean Jullien, editor of the journal in which Denis’s article appeared, and Denis himself, relaying proofs back and forth between the two men.42 But more important, as evidenced by the carnets, he involved himself in the manifesto’s intellectual elaboration prior to its publication in the late summer of 1890.
Vuillard’s second concentrated period of reflection in his journals began in the spring or early summer of 1890, with anxieties over originality that were probably sparked by his increasing involvement with the Nabi group and his subsequent concerns over the visual similarity of some of their earliest works.43 He told himself “not to worry about Sérusier and the gang,” because, although they were all working from the same theories and toward similar goals, their works would be different simply because of their individual personalities, and in particular because of the unique ways in which they drew lines: “the same arabesque drawn by 2 different individuals [caractères],” he reassured himself, “will provide 2 different expressions.”44
In July 1890, Vuillard set down an antinaturalist definition of art entirely in keeping with what Denis would soon publish in the “Définition,” albeit much less elegantly formulated. His declared desire to break from naturalism now necessitated, Vuillard argued, a new definition of art as something other than a visual representation of the natural, external world: “the work of art starting from a motivating sensation . . . must be cherished by the artist not as a renewal of that sensation,” he wrote, but “as decorating a piece of paper [or] canvas.”45 A work of art, in other words, above everything else, is a compendium of its formal and material properties rather than a representation of an object in the world or even an attempt to represent or provoke in the viewer a sensation experienced by the artist before an object. In late August 1890, with Denis’s “Définition” newly published, Vuillard refined this statement, providing a much clearer directive, one more explicit in its rejection of naturalism and even more in line with the “Définition”: “Let us understand a painting as an arrangement of harmonies, distancing ourselves once and for all from the naturalist idea.”46
Both Denis’s and Vuillard’s definitions share much with one offered earlier by Taine: “A painting is a colored surface, in which the various tones and the varying degrees of light and shade are distributed according to choice.”47 While Taine emphasized color in this short passage, in the longer passage from which it is taken, the key formal element is line. As Vuillard would do, Taine declared line to be fundamentally expressive: “lines . . . have a value in and of themselves and they have different effects on us depending on whether they are straight, curved, broken, or irregular . . . just as an arabesque of lines [sic] that do not imitate any natural object can be plentiful or meager, elegant or heavy. Our impression varies with their arrangement; their arrangement therefore has an expression.”48 Throughout this section of his Philosophy of Art, Taine asserted the seemingly infinite artistic possibilities related to the use of the painting elements of line, form, and color. An artist chooses, Taine insisted, to highlight or to downplay this or that element in order to bring out certain characteristics and expressive capabilities already inherent in the depicted object and its characteristic form. While Denis’s statement at the outset of the “Définition” (“a painting . . . is essentially a flat surface covered in colors arranged in a certain order”) seemed to minimize the agency of the artist, Taine insisted upon choice as a central element of creation, emphasizing that the painter proceeded by making innumerable decisions about how each element was to be arranged on the surface of a canvas.
At the beginning of September 1890, “still under Denis’s influence,” he wrote, Vuillard expanded on the ideas of the now published “Définition” to include a similarly hyperconscious awareness of choice, a truly extraordinary degree of volition, in his account of the artistic process. Despite how difficult he was finding it to set aside “the mental habits of naturalist ideas,” Vuillard argued that one could overcome the “obstacle” of naturalism by “working with care. . . . Prepare your colors, your palette, and ensure that the forms . . . are precise.” “Only what is on the surface of the canvas,” he stressed, “affects our eyes . . . but this must be completely conceived . . . (old idea of deliberate [voulu]); the application, the execution, therefore, becomes a matter of great patience and care and that is all it is, the work of art.” This is the closest that Vuillard would ever come to Nabi-symbolist theory in writing.49 But, as if to mitigate that dependence, he also insisted emphatically that he had been thinking and working in this way for a while: “Already put into practice 6 months ago (portrait of Mimi in Yellow).”50
The most striking aspect of the September 1890 statement, given Vuillard’s later skepticism about the will and his claims about the importance of his instincts, is the overwhelming emphasis on conscious activity, on the slow, deliberate, and very careful arrangement of forms and colors on a surface. The so-called Schematized Portrait of Marie Vuillard (fig. 49) comes closer than any other extant work of 1890 or earlier to Vuillard’s attempts to create something that is at once a portrait of his sister (whom he called Mimi) and a “colored surface.”51 The restricted tonality, deliberate flatness, and reduction of elements to a minimum, however, combine to achieve a stiltedness that was neither particularly interesting nor original, recalling as it did the work of the neoimpressionists. The painting might have put into practice the Nabi doctrine of a “colored surface” and the Tainean corollary of art making as a series of deliberate choices, but Vuillard seemed just as unconvinced of this direction as he had earlier been about fixing his impressions on paper and working from memory. While he expressed his commitment to Nabi theories and the resulting definition of art as an arrangement of formal elements and materials, like many artists seeking to translate symbolist theory into a visual practice that could be recognized and justified as such, he was still struggling with form. While Denis and Bonnard débuted their Nabi works at the 1891 Indépendants, Vuillard held back, waiting until a full year after the publication of the “Définition” to make a tentative début, and nearly six more months passed before he showed a more significant group of paintings.
| 49 | Édouard Vuillard, Schematized Portrait of Marie Vuillard, 1890. Oil on canvas, 92 × 73.5 cm. Private collection. |
“I Need a Method of Which I Am Convinced”
His journals suggest that Vuillard underwent a crisis of method in 1890–91, probably stemming from multiple sources: his disillusionment with academic training (he left the École des beaux-arts at the end of 1889 after nearly three years of study); his personal critiques of naturalism and memory; the Nabi critique of naturalism; the very real challenges he was encountering in translating Nabi-symbolist theory into visual practice; and his ongoing desire to rapidly record the visible world. Overcoming naturalist habits, Vuillard recognized, was difficult, and adhering to a theory of the artwork as primarily a “colored surface” arranged through a series of choices had so far resulted in derivative works of art that he scarcely found convincing. His ongoing concerns over originality—“a petty idea to which I am nevertheless attached, the idea of the pélichtim [Nabi code word for the bourgeoisie]!”—fed his need to develop a method derived from, rather than beholden to, Nabi doctrine.52
In late October 1890, Vuillard recorded his concerns about his lack of a workable theory and method, terms that he tended to conflate so often that the awkward conjunction “theory-method” is justified, in no small part as a reminder of the unsystematic, even at times contradictory nature of Vuillard’s journal reflections.53 “It is necessary,” he told himself, “to have a method for producing where one cannot know the outcome in advance . . . I must imagine, see, the lines, the colors that I place, and do nothing by chance. . . . I must consider all my combinations. But in order for this work to be possible, I need a method of which I am convinced . . . a firmly held theory that would relieve me of the vulgar concern for originality. Only a well-established theory can put my mind at ease and allow me to develop.”54 Evidently, Vuillard anticipated a theory-method as a liberating force for his art, a way to obviate his concerns over originality and final results to a point where he could simply get down to work. Oddly, he situated this theory-method as prior to deliberate practice, almost as a precondition both for working and for his artistic development. He envisioned applying it consistently in every “humble” ouvrage such that he could achieve an oeuvre: “The oeuvre has only one method, the sum of all actions. Therefore all ouvrages will have the same method.” But if this theory-method was to result in original works, Vuillard argued, it must be utterly individual and consequently derive from the ensemble of the artist’s mental faculties: “This method depends on the mind [esprit], on the harmonious development of the faculties.” The rather opaque passage that follows goes on to imply that in order for an artwork to be individual, rational thought (“the mind”) must give way to less conscious functioning and become “a slave to the senses.”55
Evidence suggests that in 1890–91, as he would do periodically throughout his life, Vuillard revisited earlier pages of his journals.56 In doing so, he came to recognize his continuing dependence on the observation of nature and began to embrace his instinctive desire to record and to note the world around him stenographically. His recognition of the role of nature in his art came, characteristically, in the form of a confession: “It is the idea of working from nature that pursues me,” he admitted to himself.57 In order to justify this instinct, Vuillard turned to the terms of experimental psychology, arguing that because observation was a fundamental psychic fact of existence, an automatic rather than a willed act, it held an inevitable place in any artist’s practice. “One observes, one does not will oneself to observe. . . . Pure and simple observation is a simple act of life,” he wrote, “and it is this simple and primitive observation that is necessary, that is.” Therefore, Vuillard reasoned, the emotions or sensations provoked by observation must constitute “the first condition of a work of art prior to the methodical mind and practical intelligence born from the same soul that is under the influence of that emotion.”58 If observation constituted a primordial, automatic human function, if it derived directly from an individual psyche or “soul,” then it could function as a guarantee of originality. Observation, Vuillard concluded, therefore had a fundamental role to play in his emerging method: “Simply pay attention or go walking and faire ses pieds—There’s a good system.”59
Not long before Vuillard began to interrogate his own psyche in his journals, Léon Bélugou, the future philosophy correspondent for both La revue blanche and the Mercure de France, began attending Théodule Ribot’s courses in experimental psychology. Faithful student and ardent defender of the pathological method, as I argued in my introduction, Bélugou took copious notes from Ribot’s lectures, which support and amplify Vuillard’s concerns as expressed both in the journals and in his 1898 letter to Denis.60 They provide a useful counterpoint to the lycée philosophy curriculum, especially in recording Ribot’s views of consciousness and the unconscious as a continuum rather than two separate states, and in Ribot’s alliance of inspiration with less conscious states of mind. In the course offered in 1887–88, under the heading “Consciousness and Unconscious States,” Ribot assigned the act of seeing to “organic” rather than “conscious” memory, clarifying that vision existed at the threshold of consciousness in normal adults. Observation, as Vuillard would argue a few years later, was in essence for Ribot an act that the body “automatically” “remembered” how to perform.61
By 1894, Vuillard was giving a slightly different account of the role of nature in his practice. By this time, he had firmly established observation and the automatic recording of his impressions as the initial steps in his method, but he now clarified the ongoing role of nature by using the rhetoric of experiment and experimental idea(l)s: “I am only confident in ideas and reasoning when they are controlled by direct impressions of nature. I must nourish myself with new observations. . . . I tried this and I was astonished at the prospect for discovery that I discovered. Because an observation that applies to a particular case is applicable to all sorts of facts.”62 Discovery, observation, control, case, facts: this is the late nineteenth-century terminology of experiment. Patently, Vuillard understood the direct observation of nature not only as the starting point of his creative process but also as the ongoing source of necessary controls against which to check his results. Nature, in Vuillard’s view, thus supported the imagination by checking its tendency to generalize.63 Deriding the naturalist “caricatures” of his artistic training, Vuillard echoed Denis’s belief that academic naturalism had constrained the positivist search for truth by limiting the observation of nature to external forms. And, like Denis, Vuillard seemed keen to reclaim positivism by way of incorruptible experimentalism in search of a more truthful form of naturalism that was nevertheless in the service of Nabi-symbolist theory. The observation of nature, and of one’s own sensations and emotions provoked by nature, were to be the building blocks of Vuillard’s “interior impressionism.”
As Vuillard would do in his 1898 letter to Denis, Ribot refused to make a clear-cut distinction between conscious and unconscious states, arguing that just as there was only a quantitative difference between the normal and the pathological, there was only a “difference of degree” between consciousness and unconsciousness. In everyday experience, Ribot insisted, it was virtually impossible to distinguish between the two states. Individuals experienced a perpetual imbrication of the conscious and the unconscious that they nonetheless more often than not identified as something like a doubling or a splitting. But in reality, Ribot argued, the notion of two (or more) states was only a scientific abstraction, a structural convenience that gave the lie to what was in fact a continuum. In his 1887–88 course, Ribot previewed the arguments later put forth in his Essay on the Creative Imagination (1900), assigning a particularly generative role to the unconscious in inspiration, be it literary, artistic, or even scientific. To prove his point, he quoted from the writings of Goethe and Mozart.64 He would never have relied on the writings of a young, unknown artist, and certainly had no access to them, but Ribot might as well have cited from Vuillard’s journals.
From the Automatic to the Experimental Arabesque
At the same time that Vuillard attempted to reason his way to a viable working method, he also sought a way to translate his own emotional and sensorial experiences provoked by the observation of nature, his “impressions,” into form. In late 1890, Vuillard reconsidered his somewhat facetious conclusion of two years earlier. Could he in fact turn himself into an automatic recording device, not of the external, visible world, but of the internal experience derived from the observation of nature? If Vuillard were to step through the door he had opened in late 1888, were he to attempt to record his psychological experiences by drawing like a somnambulist, the journals might reveal an accompanying change of form. And indeed, among the final entries in the first journal, those written immediately after the publication of the “Définition,” there appear drawings unlike anything Vuillard had produced on earlier pages, evidence of a formal breakthrough, tentative at first, that would be carried over into the second journal (fig. 50).
| 50 | Édouard Vuillard, journal, 22 × 17 cm. Institut de France, ms 5396(1): Carnet I.1, fol. 57r [September 1890]. |
Opposite Vuillard’s statement from early September 1890, the one advocating a very deliberate, careful approach to the arrangement of forms and colors on a flat plane, appear thoroughly new approaches to drawing, linear signifiers of nondeliberation very different from the tight, naturalistic, and often modeled sketches of earlier pages. They anticipate many of the drawings in the second journal, as well as Vuillard’s theater drawings, which suggests that they are intimately bound up with Vuillard’s early, radical Nabi style. A strange humanoid figure dominates one page (fig. 50), accompanied by a skeletal arm reminiscent of earlier journal sketches that doubtless relate to the anatomy courses Vuillard attended as a student,65 and, in the lower left corner, an indistinct shape evoking a hunched, crouching figure. Vuillard initially drew the arm in pencil, like many of the earlier carnet sketches, but then retraced it with the pen and ink used for the other two drawings on the page. The shape in the lower left corner is especially free-form, its seemingly accidental inkblots having left a mirror trace on the opposite page.
The large figure, however, proclaims the most significant formal departure. With the most economical of means, with long, thin, rapidly executed strokes of the pen, Vuillard conjured up a highly stylized man-beast. The bare, emaciated torso perched atop thick legs resembles the body of a satyr. A thick, sinewy neck supports a head with small, round eyes, a large outlined moustache or downturned mouth, and the hint of a pointed chin or possibly a beard. The outlines of long ears are visible on either side of a head sprouting a few lines suggesting long, wiry hair. The small, round eye sockets with no eyeballs, the facial hair, the simplified, elongated ear shapes, and the slight upper body all echo Vuillard’s early self-portraiture, especially the decidedly schematized Self-Portrait with Walking-Stick and Boater (ca. 1891–92) (fig. 51).
| 51 | Édouard Vuillard, Self-Portrait with Walking-Stick and Boater, ca. 1891–92. Oil on cardboard mounted on canvas, 36 × 28.5 cm. Private collection, United States. |
In the painted self-portrait, Vuillard played with the viewer’s expectations regarding how the material surface of the work might represent its subject naturalistically. The eyes are unexpectedly empty areas of cardboard surrounded by paint, a literal reference to the Nabi claim that a painting is primarily a surface to which paint has been applied, but also a negation of vision, giving material form to Vuillard’s rejection of a narrow version of naturalism. The ink sketch declares its separation from the “normal” natural world even more emphatically. The sagging man-beast negates academic naturalism, especially by its placement adjacent to the skeletal arm, a reminder of Vuillard’s training in “correctly” representing human anatomy. Like Denis’s study for Décor on the reverse of a prizewinning drawing of a classical cast, it performs Vuillard’s act of freeing himself from old strictures, carrying out a farewell to constraining mental habits. It seems as if Vuillard has figuratively flayed himself, stripped himself of his academic education in the hope of revealing a new interiority.
Were this sketch the single instance of this kind of linear abstraction in the journals, it would be interesting but not especially significant. But because it inaugurates a new, apparently much less deliberated approach to form that dominates the second journal, it signals some of the structures and epistemological conditions of Vuillard’s new formal language. It announces that Vuillard will henceforth actively court a somnambulistic approach to nature, not to record it automatically as in a photograph, but to give form to his automatic impressions of nature. In late 1890, he associated this recording of his immediate sensations with the first inklings of an idea, confirming that he conceived of the rapidly drawn trace as the visual signifier of interior observation: “Let’s express what I feel (it is a simple designation, a fact pure and simple to designate the thing I have in my head): an expression of tender feelings caused by such and such an object. . . . Then I trace on the paper all at once [d’un coup] (in the same way that one speaks poorly when an idea presents itself) the line or the mark imagined and desired.”66 Because it was instinctual, automatic, and therefore seated in the unconscious mind, this mode of expression held out the promise of being utterly individual and could therefore function as the basis of a highly personal method for transforming theory into visual practice.
From here on, Vuillard would figure the starting point of his method as the arabesque. When he began to seriously employ the arabesque in the final pages of the first journal, he often chose to use the more fluid medium of ink rather than pencil, charcoal, or pastel, which had prevailed in the journal up to that point. The second journal begins with pages in which Vuillard explored contour and arabesque using pencil (fig. 52). But he soon returned to ink, now applied with a thick brush. The energy, density, and rapidity with which Vuillard set down these arabesques suggest his enthusiasm for the exercise. Lines intersect with one another; figures emerge out of a tangle of lines, as if the artist turned them into people or animals only after unconsciously moving the brush across the paper. In one of the first arabesques in the second journal, thick ink brushstrokes and shapes resolve into a figure in a long cloak wandering through an indeterminate landscape of curving black forms (fig. 53). Motifs straight out of Japanese woodcuts (fig. 54) suggest that Vuillard may have attended the exhibition of Japanese art held in April 1890 at the École des beaux-arts, but they also connect the arabesque to primitivist, exotic, and decorative forms of mark making.67 “In the beginning,” Denis had written in the “Définition,” “pure arabesque, as little trompe-l’oeil as possible; a wall was empty: fill it with symmetrical marks of forms, harmonious colors.”68
| 52 | Édouard Vuillard, journal, 22 × 17 cm. Institut de France, ms 5396(2): Carnet I.2, fol. 3v, ca. 1890. |
| 53 | Édouard Vuillard, journal, 22 × 17 cm. Institut de France, ms 5396(2): Carnet I.2, fol. 5v, ca. 1890. |
| 54 | Édouard Vuillard, journal, 22 × 17 cm. Institut de France, ms 5396(2): Carnet I.2, fol. 12v, ca. 1890. |
Like the line itself, the history of the arabesque as a visual device is a sinuous, twisted story. Borrowed from Italian, the French adjective arabesque, meaning “of the Arabs” or “in an Arab style,” occurs as early as the sixteenth century but was used more often in the eighteenth to describe elements of rococo decoration. The Encyclopédie, still recognizing only the adjective form, denoted a European borrowing from Arabic painting and sculpture in which curvilinear vegetal forms proliferated owing to the presumed injunction against depicting humans and animals. The encyclopedia entry emphasized the contemporary use of arabesque forms in decorative works, including royal tapestries, but also pejoratively identified their use as a capricious, less exalted form of decoration, suited only to small, private spaces.69
By the early nineteenth century, the term arabesque was more frequently being used as a noun, and was associated with the arrival of Islamic architecture in Europe during the medieval period, thus marrying the exoticism of the arabesque with the by then more familiar exoticism of medievalism.70 German romantic writers and artists such as Friedrich Schlegel and Philipp Otto Runge linked arabesque forms to a quasi-mystical search for the very origins of creativity. The arabesque was, in the words of Schlegel, “the most ancient and primitive form of the human imagination . . . the true mother, the embryo of modern painting.”71 In its supposed echo of the forms and rhythms of nature, it apparently signified a world beyond the visible.72 Romantics cherished the arabesque above all for its suggestiveness, as a form that could give free rein to the human imagination, and they increasingly ascribed to it the value of musicality. Musical and ballet arabesques are likewise romantic inventions, as the term continued to accrue rich meaning.73
Symbolism’s widespread promotion of the arabesque line as a multivalent signifier of artificiality, planarity, musicality, primitivism, and the imagination confirms symbolism as, at least in part, a renewal of romantic ideals. Symbolist artists were particularly drawn to what they saw as the ambiguity of the form, the potential, to use Dario Gamboni’s term, of a sinuous curving line to enable different, even conflicting, readings as it wound its way across a surface.74 Artists and critics often spoke of the arabesque as an animate being, as a moving, otherworldly creature with its own mercurial soul. And, as Gurminder Kaur Bhogal has recently demonstrated in great detail, using the Nabis as a key example, the ornamental, curving line enabled links to be forged with the other arts that had courted the allure of the arabesque—poetry, dance, and especially music.75 Not surprisingly, it held a privileged place among artists such as Gauguin and many of the Nabis who sought to ally painting with musicality under the banner of Wagnerianism.76
As for many symbolist and art nouveau artists of the 1890s, the arabesque performed multiple functions for Vuillard. It signified the rejection of modeling and thereby repudiated academic and naturalist iconicity, and it squared with Denis’s promotion of almost atavistic form, which held up the arabesque as the formal outcome of a process of devolution that would occur if one could return to the primitive origins of art. It courted a putative non-European exoticism but also signified, as Vuillard made clear, that the primitive, automatic side of himself had been mobilized and captured in visual form. It is the automatic arabesque that has smoothed Vuillard’s way into the modernist canon, winding itself like a red thread, from Vuillard’s ink arabesques, to his theater drawings, to fauvist arabesques, to surrealist automatism, and finally to the high modernist webs of Jackson Pollock.77
Initially, Vuillard developed the ink arabesques of his journals in works on paper that reflected his fascination and involvement with avant-garde theater, showcasing the first steps of his emerging method. These ink drawings, sometimes enhanced with watercolor, captured his actor friends at work, often fixed in awkward poses and contorted facial expressions (figs. 55 and 56). Vuillard’s theater arabesques, which repeatedly approach caricature, endeavored to represent the spontaneity and ambiguity of the radical forms of theater being explored at Paul Fort’s Théâtre Libre and at Lugné-Poë’s Théâtre de l’Oeuvre, the latter of which Vuillard helped found in 1893.78 But as much as there was important cross-fertilization between Vuillard’s theater work and his painting, especially in terms of subject matter, atmosphere, or mood, painting retained its elevated and critical status as a separate realm of ambition for Vuillard, as it did for many late nineteenth-century vanguard artists, including Denis and Munch.79 When Vuillard wrote of needing a method, he wrote of his need for a method of painting. Once he had instituted the arabesque as a signifier of instantaneous observation, how, he wondered (using photography as an analogy), would he move from the “instantaneous” to the painted “Image”?80
| 55 | Édouard Vuillard, Grisélidis, 1891. Brush, pen and ink, watercolor on paper, traces of graphite, 26.5 × 18.5 cm. Private collection. |
| 56 | Édouard Vuillard, Lugné-Poë, ca. 1891. India ink on paper, 19 × 9 cm. Private collection. |
Immediately after he likened the automatic arabesque to the rapid setting down of an idea, Vuillard clarified in his journal that this was only the first step in a creative process that would subsequently demand careful thought and deliberate composition. Painting, in his view, still necessitated making the kinds of slow, deliberate choices dictated by Nabi theory as inflected by Taine. “I develop it, if necessary,” Vuillard wrote, “and I compose, a fact quite simple in and of itself to compose, and all my patience (Oh that’s easy) is absorbed by a concern to do well and then the possibility of really very prolonged manual work.”81 Painting’s ultimate purpose, Vuillard reminded himself, was the production of objective, universal truths and an objective beauty guaranteed by belief in the existence of a god. “If we were simple beings, unspoiled by prejudice and habits, our art would easily be beautiful, because there would be no struggle to distinguish within it the truth from the lie. . . . Would the beautiful then be only subjective? . . . Not if we believe in objective existence, if we believe in God.”82 The automatic arabesque of observation could only ever be a starting point for truth, the basis on which truth could be laboriously founded. Nabi-symbolist theories and Vuillard’s own views of painting as also involving the will necessitated the transmutation of the automatic arabesque into something signifying the will. Enter the arabesque of experiment, a highly deliberated lack of paint, which transformed the automatic arabesque into a signifier of willed but increasingly repetitive acts of painting.
Vuillard’s Woman Sleeping (see fig. 43) and his portrait of Lugné-Poë (see fig. 4) depict two extremes of unconscious and conscious activity while at the same time instantiating radically different gendered subjectivities. The sleeping woman, unconscious, hardly visible, and unidentifiable either as an individual or even as a woman, is the very antithesis of the individualized, albeit caricatured, portrait of Lugné-Poë, which shows the actor hunched over his writing in a state of intense mental concentration. These are far from the only instances of works by Vuillard that include the odd feature that I am calling the experimental arabesque. But not only is the device particularly evident in these two paintings; together they allegorize the two poles of Vuillard’s method. In doing so, they mark out a continuum rather than a clear-cut distinction between conscious and unconscious states. Among the earliest works exhibited in the artist’s career, they provide the means by which Vuillard’s nascent method might be parsed.83
Woman Sleeping and its cut-up caricature draw special attention to the painting’s form, which is not the same as saying that the subject of sleep is mere pretext for form.84 But neither is the subject matter a pretext for looking at a female nude, which in and of itself calls attention to the phenomenon of sleep rather than the sleeping body as its ostensible subject matter. Sleep was a classic problem for philosophy and, increasingly in the nineteenth century, for experimental psychology. Was sleep a state of mental inactivity and therefore akin to death, or did the mind, as dreams seemed to suggest, remain active during somnolence?85 If the latter, what accounted for that activity? By the second half of the nineteenth century, the analogy with death was all but relegated to literary metaphor. And sleep’s pathological other, somnambulism, had evidenced not only the unconscious but, for psychologists like Ribot, the coexistence of consciousness and unconsciousness in all aspects of psychological life.86 Ribot had even located personality and some fundamental notion of the self in this coexistence, even though individuals experienced this coexistence more often as alternation and at times, more perniciously, as a doubling of personality.87
The automatic arabesque would therefore seem to be an ideal form with which to represent the subject of sleep. Indeed, Vuillard composed the related drawing Woman in Bed (fig. 45) around seemingly rapidly drawn charcoal arabesques that defined the contours of the figure, the bed linens, and its folds, providing a schematic layout “filled in” with watercolor. These curving lines are maintained in the painting In Bed (fig. 57) and to a lesser extent in Sleeping Woman (fig. 58). But they have virtually disappeared in Woman Sleeping, significantly, I think, the sole painting of this subject that Vuillard chose to exhibit in the 1890s.88 In Woman Sleeping, Vuillard scaled up the automatic arabesque, widened it, and transformed it into negative space by carefully, although not precisely, avoiding painting adjacent areas of paint too close together. At times, Vuillard appears to have filled in these spaces with thin pigment, but he continued to make it clear that there had first been a void, and that the “fill in” was the result of an equally painstaking act of essentially coloring the lines in the lines. The effect is not unlike that produced in slit weaving, where a gap between two warp threads opens up when the weaver introduces a new color.89 Once apprehended, the oddness, originality, and artificiality of these gaps arrest the viewer, but so does the sheer effort involved in producing them.
| 57 | Édouard Vuillard, In Bed, 1891. Oil on canvas, 73 × 92.5 cm. Musée d’Orsay, Paris, RF1977–374. |
| 58 | Édouard Vuillard, Sleeping Woman, 1892. Oil on board mounted on cradled panel, 26.8 × 26.8 cm. Private collection, Washington, D.C. |
This effort is especially evident in the portrait of Lugné-Poë, which depicts a state of intense mental concentration by way of Lugné-Poë’s hunched shoulders and his tense, misshapen hands pressing against the top and sides of a horizontal surface. In their tensed state, the shoulders and hands signify the concentrated labor of both mental and manual work. Nearly all the linear elements in the portrait, the outlines and folds of Lugné-Poë’s smock and the contours of his hands and fingers, are negative spaces through which the cardboard support and at times a thin ground show through. Despite its overt focus on a willful, conscious creativity, the painting retains references to an opposing mental state. Lugné-Poë was known for his stylized, trancelike stage movements, which in 1895 earned him the nickname “the somnambulist clergyman.”90 And some semblance of perspective on the left side of the painting, along with a shape resembling a pillow and the dawn or twilight lighting, makes it likely that Lugné-Poë, pictured in the cramped studio space that he shared in 1890–91 with Bonnard, Vuillard, and sometimes Denis, was leaning not over a desk but over the foot of a bed.
The negative arabesques are narrower, more even, and more sophisticated than in Woman Sleeping, comprising nearly all the contours of the work, in a way similar to the still life painting Kettle and Pot from around the same time (fig. 59). Their use is virtuosic, an intentional play on the artist’s lively theater arabesques, which include numerous depictions of Lugné-Poë (for example, fig. 56). Vuillard has taken the rapid ink arabesque and slowed it to a crawl, figuring the intense concentration with which Lugné-Poë is shown conducting his own creative work. The most intriguing passage of painting occurs just to the right of the bed’s footboard. An attenuated isosceles triangle of light blue, similar in hue to both the background and Lugné-Poë’s smock, seems to hover on the surface of the work, confounding the viewer’s efforts to distinguish it as part of either the figure or the ground. Further inviting concentrated looking, Vuillard has painted the triangle with parallel horizontal brushstrokes, the material traces of a series of repetitive, initially willful, but eventually habitual acts of painting.
| 59 | Édouard Vuillard, Kettle and Pot, ca. 1892. Oil on cardboard, 24 × 32.5 cm. Formerly Museum of Modern Art, New York. |
In both Woman Sleeping and the portrait of Lugné-Poë, Vuillard offered the viewer an experience of visual play. At a distance, the negative arabesques resolve into contours—positive, curving, seemingly rapidly drawn outlines describing the shapes on the canvas. But especially in the case of the tiny portrait of Lugné-Poë, the viewer must get very close for the “lines” to reveal themselves as labored gaps in the material surface of the work. This viewing experience sets in motion a dynamic oscillation between an imagined positive line and a negative linear space, a perceptual uncertainty that begs questions about both the artist’s practice and the visual experience of the work. But the questions provoked are nevertheless resolved into what, for Vuillard, seems to have been a universal truth. Beginning with an automatic, unconscious act, the spark of an idea figured as an arabesque line, Vuillard synthesized that idea and tested it in a willed process that could also be repetitive almost to the point of becoming mechanical. He courted a creative process that relied on both conscious and unconscious functioning, which Vuillard intuited and which Ribot argued could not, in practice, be readily distinguished from each other. This, for Vuillard, was a truth about human experience discovered through a complex and deeply paradoxical process that necessitated the “willed willessness” of the experimenter.91
Vuillard’s Doubt
His painting was paradoxical: he was pursuing reality without giving up the sensuous surface, with no other guide than the immediate impression of nature, without following the contours, with no outline to enclose the color, with no perspectival or pictorial arrangement.
—MAURICE MERLEAU-PONTY, 1945, TRANSLATED BY HUBERT L. DREYFUS AND PATRICIA ALLEN DREYFUS, 1964
Maurice Merleau-Ponty’s description of Cézanne’s “paradoxical” practice might equally apply to Vuillard in the early 1890s. But while Merleau-Ponty identified doubt as an inevitable accompaniment to Cézanne’s artistic project, Vuillard’s doubt (and I would wager Cézanne’s as well) might also be read as intrinsic to the artist’s experimentalism. It might be sufficient to argue that the creative process that Vuillard developed in the early 1890s courted a formal analogue of the pathological method of experimental psychology insofar as its starting point was an othering of form figured by the automatic arabesque, and that it proceeded by using the observation of nature as a control for the imagination in a series of deliberate and often repetitive acts ultimately aimed at producing objective truths. But still another aspect of Vuillard’s method, or, more precisely, his attitude, found a rationale in late nineteenth-century experimentalism. When Vuillard asserted that making art involved a series of carefully considered choices about the arrangement of elements on a surface, he also embarked upon an interrogation of how, in practice, one came to make those choices. Throughout his life, as Gamboni has noted, Vuillard expressed frequent and almost paralyzing doubts about his ability to carry through with his work,92 but his concerns over his own will to work were especially acute in 1890–91.
As early as the October 1888 letter to Denis, Vuillard had identified the career of a painter as particularly suited to a “hardened idler” like himself. He frequently made statements such as “no strength in the head,” and would often use a single word, such as flemme or paresse, to sum up his state of mind and body.93 Both flemme and paresse are usually translated as laziness, but the latter also connotes indolence and apathy. Later, he would give more elaborate accounts of his ongoing difficulties in summoning up the strength to think and work: “Annoyed can’t concentrate on a point, an idea. . . . Slave to my sensations and a pride condemning [me] to powerlessness. At the studio laziness and pride.”94 Vuillard deprecated himself, often to an extreme degree, over his severe lack of will.
In his journal, Vuillard would come to couch his concern with the mechanisms of his own laziness in terms of the new psychology’s understanding of the will as both psychological and physiological, and as one of the most complex and highly evolved human functions. In a series of questions to himself, Vuillard wondered whether his mind “resisted” working because of “physical and moral fatigue.” He asked what brought about a change in volition, and he considered whether the amount of will needed to overcome laziness was equal to the amount of laziness itself. “Can one . . . increase this power just as one exercises the biceps?” he asked. He suggested that only life, an “unknown force,” could bring about the necessary effort to overcome the overwhelming instinct to resist action. A string of question marks at the bottom of one page (fig. 60) conveys genuine frustration as he probed his own mind: “what about the consciousness of that effort. What is underneath these words efforts, cravings, desire! Confusion of the words and the ideas that they represent!”95
| 60 | Édouard Vuillard, journal, 22 × 17 cm. Institut de France, ms 5396(1): Carnet I.1, fol. 73r [31 August 1890]. |
These pages fairly bristle with anxiety as Vuillard voiced his torment. But he also implied that he felt it should be otherwise, that he should not only have better control over his own will but should understand his own will better, should have a tighter grasp of the meaning of words such as effort, craving, and desire. Written on 31 August 1890, immediately after Denis’s “Définition” appeared, the “should” is of course relative. “Should,” he meant, in comparison to Denis and to “Sérusier and the gang,” many of whom Vuillard seemed to imagine would be much less confused than he about the functioning of the will. But Vuillard underestimated himself. His skepticism toward an easy comprehension of volition, and his intuition that volition was both physiological and psychological, were more up to date than Denis’s professed certainties. For Vuillard, his purported mental laziness was something akin to physical fatigue, an insufficient amount of “moral energy” that undermined his ability to act consciously at any given moment. He understood that a healthy will was continuous with an unhealthy will, each a function of greater or lesser amounts of “moral energy.” In The Diseases of the Will (1883), Ribot had argued that “in every voluntary act there are two very distinct elements, the ‘I want,’ which states a situation, which in itself has no effectiveness; and a very complex psychophysiological mechanism, the only place in which the power to act or to resist resides.” Like all of Ribot’s monographs, his text on the will deployed a pathological method. Among the pathologies of volition that Ribot examined was the “madness of doubt,” which he defined as “a constant state of hesitation, for the most vain reasons, with the inability to arrive at a definitive result.”96 In his journals from around 1890–91, Vuillard effectively diagnosed himself as suffering from such a disease.
By 1894, however, Vuillard had come to see his doubt as essential to, even constitutive of, his artistic methods. His lack of will, he had begun to insist to himself, enabled him to develop and sustain the “habit of machine-like work,” adding, alongside instinct, another automatic function to his method. This completed a triad of conscious and unconscious mental functions; as Ribot had argued in his 1891–92 Collège de France course, “The Will from a Psychological Point of View,” the will was everything besides instinct and habit.97 The solution to the problem of a lack of will, which made Vuillard’s laziness in and of itself useful, was simply to work habitually, to circumscribe the problem of volition by cultivating a different kind of automatic behavior. “In thinking about work,” Vuillard wrote in his journal, “I will only be able to put my mind at ease by being delivered . . . by the habit of machine-like work, which my long periods of idleness have not yet been able to destroy. Most of the time, tired from my mind’s laziness, my desire for work is only a hope born out of my laziness that renders me easily content. Perhaps habit has its benefits, but the habit of an activity.”98 Vuillard’s solution, in 1890–91, to the problem of developing a method at once Nabi-symbolist and wholly individual was a process involving his instincts, habit, and will, and when he found his will to work especially vulnerable to laziness, he exhorted himself to “try to return to the work of 90–91,” instead of becoming, “in the face of sensations or new impressions,” a “slave to certain dimensions, to certain materials (cardboard, oil).”99 Vuillard had come to identify thinking coherently, being conscious of his work, as at times detrimental to the new ideas emerging, and he considered habitual work his touchstone. The hard-won, deeply personal method at which Vuillard arrived early in his career remained his lodestar in the 1890s, a source of comfort when he felt doubt and laziness getting the better of him. A lack of will enabled Vuillard to renew his work, to return to the method of 1890–91, and always to see the simple act of working as something quite marvelous.
In his Introduction to the Study of Experimental Medicine, Claude Bernard had assigned to doubt a central role: “the unique and fundamental rule of scientific investigation,” Bernard wrote, “can be reduced to doubt.”100 Doubt and a professed nescience about final results enabled the experimenter to retain his objectivity, to constantly question the results obtained during an experiment such that rigorous testing could continue and a final truth emerge. The late nineteenth-century image of the experimenter, which Émile Zola had co-opted, was that of both lone genius and humble worker. His flash of inspiration, his hypothesis, was subject to extensive tests and eventually proved true through a repetitive, almost machinelike process during which he willed himself to “willessness.” Vuillard cultivated for himself a very similar methodological stance that aped experimentalism, while at the same time situating his practice closer to the unconscious end of the spectrum than any of his contemporaries did. Vuillard seems to have delighted in subverting viewers’ expectations about the painted surface in conjunction with subject matter. Where the viewer might anticipate line, she gets negative space or nearly imperceptible boundaries between areas of pattern. Where she expects at least a daub of paint, signifying, for instance, an eye, she gets bare cardboard. Walter Sickert called attention to Vuillard’s studied childishness, praising his “purposed negligence” as the crux of his practice.101 Ultimately, Vuillard’s practice, like Cézanne’s, was one of paradoxes, but paradoxes central to, among other things, nineteenth-century experimentalism.
In the letter that Denis took as a guideline for his “Définition,” Sérusier offered two methods for arriving at truthful symbolist form, or, in his parlance, the “immutable principles.” Sérusier argued that an artist could proceed either by a kind of self-deduction, by searching for the immutable principles within the self, or by induction, by observing how other artists had successfully externalized the immutable principles in their works.102 Denis, more often than not, chose to work from the latter method, canonizing a host of primitifs from Fra Angelico to Cézanne. Vuillard, by contrast, strove to deduce a method from an intensive and concentrated exploration of his own interior mental processes. Sérusier had argued that only the artist “unspoiled” by academic training, a modern primitive like Gauguin, could deduce universal truths within the self without the need to check his results against those of past masters. Vuillard’s automatic arabesques constituted his attempt to establish the raw data for his artistic practice as he experienced it in the guise of the “unspoiled.” While the willful arrangement of colors and lines on a flat surface remained a central feature of his practice, Vuillard, cognizant of having received a philosophical education different from that of his peers, retained a critical place in his method both for his instinctual approach to nature and for habitual work—in other words, for those automatic, unconscious mental functions that late nineteenth-century psychologists opposed to the will.
Vuillard came to fetishize his method as the unique way in which he could work in order to attain truth. In a passage that could have been written by any number of the artists discussed in this study, he made abundantly clear the connection between his practice, his self, and the truth:
My only desire now is the life of my soul. Work is at once the sign and the product of that life. . . . Truth is life, life is truth, the truth is one or it wouldn’t be the truth. Dogmas are only methods for arriving at the truth; any method contemplated by the Spirit offers up a fascinating spectacle[;] it’s a harmonious ensemble of more or less pure ideas that makes up the more or less great and vast Beauty of dogma. Dogma is not an end, dogma is a work of art, at least that is how I understand it, it is the result of expériences towards a final aspiration that is and can be nothing other than God.103
Although he held his own against Denis in their 1898 exchange, Vuillard eventually lost faith in certain aspects of his method, and in particular in automatism and the arabesque of observation as starting points. One might argue that he loved the natural world and everything in it too much not to return to an increased visual correspondence between his painting and the visible world. One might argue that successful decorative and portrait commissions put too much pressure on a method that, at least in part, was originally designed to translate symbolist theories and his own interior experiences into visual form. When Nabi-symbolist theories became less imperative, the method had to change. And it did so in ways that ultimately threatened Vuillard’s desire to create a body of work that might without hesitation be called an oeuvre.
The level of intentionality that I have ascribed to Vuillard and to his development of a creative method might be criticized by those who choose to take the artist too much at his word regarding his lack of method. But the work of the 1890s, along with the ruminations of the journals, attests to the careful consideration that Vuillard brought to developing a viable working practice. Examining the method and its underlying assumptions neither negates nor diminishes the success of his results. As theorists of experiment have argued, in the face of discovery, of increased knowledge of truth, method and process are destined to fade into the background—indeed, by design should become entirely transparent to the results.104 Like Sérusier’s Talisman, the central subject of such paintings as Woman Sleeping and the portrait of Lugné-Poë is the creative process itself, vexing, paradoxical, neither wholly conscious nor wholly unconscious but existing somewhere along the wide spectrum of human psychological functioning.
A final comment about sleeping women and thinking men is in order. An undercurrent of Vuillard criticism has often held that he was not quite man enough. Shy, we are told, harboring secret passions for unattainable women, possibly including his mother, whom he called his “muse,” and his sister, Vuillard is nevertheless suspected of having done violence to female bodies by deforming them, a charge that gains him entrée into the modernist canon.105 My analysis implies that Vuillard’s stance on the female body might be part and parcel of a creative method structured by a continuum between automatic and willed states nevertheless perceived as an alternation. As these states were, and often still are, gendered feminine and masculine, as they are in Woman Sleeping and the portrait of Lugné-Poë, such an alternation might engender, or be perceived to engender, a process of “androgynization.” The uncertain gender of the figure in Woman Sleeping thus emerges as a sign of the gender instability inherent in the creative method that Vuillard developed and cultivated for himself, which the caricature of the cut-up Woman Sleeping attempted to stabilize by ascribing to the artist masculine violence. Such “critical cross-dressing” was, of course, possible only when performed by a white, bourgeois male member of an avant-garde “brotherhood,” but one nevertheless who had less purchase on the robust French republican male self than his colleagues.106 This is not to say, however, that the performance has exacted no price. In Vuillard’s case, one cost has been the stubborn persistence among art historians to address either form or content in analyses of his works, a tendency that renders his symbolist visual practice inscrutable.
