Skip to main content

John Klein
Professor, Department of Art History & Archaeology, Washington University in St. Louis

In fall 1905 critics dubbed Henri Matisse “King of the Fauves” for his leading role in a new tendency among young artists to paint with bright, arbitrary color and a seemingly inconsistent, hasty-looking style. Fauvism not only refuted the idea of painting as a record of the visual appearance of the world (as in Impressionism), but its apparent incoherence also challenged more subjective tendencies, such as Symbolism, in which an artist presented a personal “vision” of the world. Nature morte aux oranges (II) (Still Life with Oranges [II]) would seem at first to have all the characteristics of a Fauvist painting. But it is well established that Matisse made it around 1899, long before his Fauvist period. So how should we think about this vividly colored canvas, especially in light of its unfinished nature? Is it “proto-Fauve,” precociously anticipating Fauvist innovations and the increasingly abstract qualities of advanced painting? Or was it an anomaly, a one-off effort that the artist didn’t follow up on for another five or six years?

To begin with, as the “II” in the title suggests, this is the second of two nearly identical compositions, showing the same vessels and pieces of fruit, placed on a table in a similar way, situated against a wall and open window.1 But while the compositions are similar, Matisse painted the two canvases very differently. The first version is thoroughly and consistently, if roughly, painted over its whole surface. By contrast, Still Life with Oranges (II) has an uneven surface, with some areas receiving extensive treatment and some only sketchily painted, while in others the artist applied no color at all to the primed canvas. Moreover, even the completely painted areas look unresolved. As several writers have observed, this was possibly the first of numerous times when the artist painted two distinct versions of the same composition, a practice that he pursued periodically over the next fifteen years.2 This characteristic of Still Life with Oranges (II)—that it is a variation on a theme—leads to an understanding of its significance and to answers to some of the questions posed above. Matisse’s approach to his painting at this time had an experimental, probing quality that he valued as a respite from the demands on a young artist of the marketplace and the public exhibition system.3 In 1898 and 1899, working in Corsica and Toulouse, he made paintings in an enormous variety of styles. Many of these canvases are brilliantly colored and some are quite tactile, with thick, unctuous brushstrokes, emphasizing a painting’s materiality and physical presence and suppressing its function as a vehicle of representation, or a window on the world.4 Most of them are landscapes or still lifes. These genres burgeoned in the nineteenth century as sites of avant-garde innovation. They offered artists the privileges of isolation and lack of obligation to other people (as in portraiture) or to the authority of events or texts (as in paintings of historical, mythological, literary, or religious subjects). Moreover, landscape and still-life painting did not form part of the academic art-school curriculum in France. Paintings such as Still Life with Oranges were, typically, sidelines for artists determined to achieve success within the academic system. Matisse, well trained in the academic tradition, nevertheless pushed hard against the strictures of his formation while maintaining much of the greater ambition that elevates the practice of painting above the level of mere personal expression. This ambition led him to conceive paintings in relation to one another, developing them as reflexive investigations of the relationship between matter and meaning as well as depictions of things next to other things.

Of course, the materiality of things in a still life is not irrelevant, and Matisse found a way to integrate paintings such as Still Life with Oranges (II) into the framework of academic ambition. The still-life subject in general, consisting of objects that can be manipulated and arranged by the hand, acts as a surrogate for the artist’s control over the practice of painting itself, and the freedom to invent. As a student, Matisse had copied still-life paintings by Jean-Siméon Chardin (1699–1779) and Jan Davidsz. de Heem (1606–1683/84), pillars of the northern tradition. His first independent painting was a still life (Still Life with Books, 1890), and he would continue to paint such humble subjects by preference, even when he came to the point of making his first major exhibition painting, The Dinner Table, in 1897. Although The Dinner Table was conceived in the academic tradition of the chef d’oeuvre, the masterpiece that would demonstrate a young artist’s powers and mark the end of his apprenticeship, it was a large canvas showing a woman in contemporary dress setting an elaborate table for a sumptuous meal—like still life, a distinctly nonacademic subject. Here Matisse was trying to have it both ways—to buck the system but from within its patented structures.

Variations on the theme of The Dinner Table followed, all of them eliminating the woman but retaining the objects, arranged in various ways. Leaving out the figure leads to the suppression of narrative possibility and suspends in time the objects that remain. In Matisse’s work at this time, these are always objects of daily use and consumption that he would have had around him. The culture of domestic objects and routines is central to the still-life tradition. Even so, the arrangement of objects throughout the history of still-life painting is typically artificial and sometimes quite arbitrary, with knife blades hanging over the edges of tables, stacks of carrots balanced precariously on ledges, or gatherings of objects that seem to have no business together. In contrast to such artificiality of arrangement, traditional still-life painters rendered their motifs in generally realistic styles, sometimes with extreme clarity and a high degree of detail.

Matisse directly inverted this conventional relationship between subject and style. In Still Life with Oranges (II) and other still-life paintings at this time, the arrangements are natural looking, as if excerpted from daily life. These are examples of what Norman Bryson has called “rhopography”— the depiction of “the overlooked,” of “those things which lack importance, the unassuming material base of life.”5 Matisse’s arrangements would not look out of place in a middle-class home and would not seem capable of demanding attention in themselves. In other words, they do not seem to be “artfully” arranged. Rather, it is the painting style of Matisse’s still lifes that is artificial and that does demand our attention. In this experimental environment, the still-life subject acts as a constant, the style as a variable. The distance of Matisse’s move from his first version of a still life with oranges to the second one makes it clear that variations—experimental risks—are more meaningful when they are set against such a constant.

There is another variable as well. As in more traditional still-life paintings, light operates to signify the passage of time. Matisse’s treatment of light varies from one still life to another, marking them with a temporality outside narrative. Full daylight in one painting gives way, implicitly, to twilight effects in another. Matisse also positioned himself and the objects he painted in different relations to the source of natural light. Sometimes the light is ambient, and the way it illuminates the objects seems direct and uncomplicated. In other paintings the artist positioned the objects between himself and the inflow of sunlight into a room so that he saw his motif à contre jour (as is the case with Still Life with Oranges [I]). The relationships between the range of colors and different light effects multiply the possibilities for manipulating the means of representation. Their descriptive powers are stretched thin, and their properties become ends more than means.

Nowhere is this transposition of means and ends more evident than in Still Life with Oranges (II). Whereas in other of Matisse’s still lifes, including Still Life with Oranges (I), the spatial relations among objects retain at least a residual sense of order (signaled by tonal or coloristic modeling and logical, consistent lighting), this sense of spatial order, as well as the entire descriptive goal of painting, is what Still Life with Oranges (II) disrupts. As in the first painting, the main light source appears to be the open window at rear right, but here Matisse eliminated many cues of modeling and shadow that would normally contribute to a logical rendering of objects in this specific environment. The pieces of fruit that had previously been rendered with fullness are signified by nearly uniform orange disks, their modeling reduced to schemata. Its solidity dissolving in a patchwork of painted marks, the table surface is no longer a stable resting place for the objects on it. The walls and the view out the window (this last, paradoxically, the area of thickest paint) are reduced to indiscriminate fields of arbitrary color. Whole passages—cup and saucer, fruit dish, tabletop, and table front—are not fully realized, indeed are sometimes barely begun, and some painted passages float free of anything, signifying nothing, starkly calling attention to how they have slipped the anchor of representation.

The color in Still Life with Oranges (II) is no brighter or more arbitrary than that in its first version, but the striking thing about the color here is that Matisse mainly applied it in flat, uniform, thin areas, as if simply filling in outlines, like using crayons in a coloring book.6 This flat, decorative quality of the color, all hue and no tone, as much as its brightness, is the characteristic that has led to this painting being thought of as foreshadowing his later work and being dubbed “proto-Fauve.”7 Coupled with this decorative quality is the lack of a definitive light source or shadowed areas, resulting in a quality of light that seems immanent, as if it is generated by the color itself. Luminous color rather than illuminated color—this would become one of the hallmarks of Fauvism several years later.

But perhaps we should not take this analysis of Matisse’s achievement or intentions too seriously, since he clearly did not complete this painting. The whole character of the canvas is that of an ébauche, the rough outline for a painting, with the motif laid in but not elaborated. We do not know why Matisse did not finish Still Life with Oranges (II), but two main possibilities may be suggested. One is that he may not have needed to finish it—from the process he had already carried out, he may have learned what he needed to in this effort, so that his work was done even if the painting was not. His working method at this time, after all, was clearly exploratory, not necessarily to pose a problem that had a definite solution. The other is that he did not know how to finish it—that his work on this canvas took him down a path that did not lead to a place he could recognize at this time. The variable of the means in his experimental procedures may have strayed too far even for Matisse to consider it as an outcome that he could acknowledge as a painting.

Nevertheless, even though he clearly left the painting unfinished, the canvas is signed. Many artists, Matisse included, frequently did not bother to sign paintings until they delivered them to their dealers or otherwise released them into the public domain, whether by selling them, giving them to someone, or putting them into an exhibition. The painting’s early provenance is unknown, but we can speculate that Matisse later came to believe in this canvas as a just expression of his goals, especially as the brightly colored paintings of his Fauve period may have seemed to validate this earlier experiment.8 Whereas in 1899 he likely would not have risked its public exposure, in retrospect it must have seemed more coherent, and it may have seemed more conclusive. Looking back on Still Life with Oranges (II), Matisse may have “discovered” what he did not know, or could not know, he had been seeking. Whatever truth may lie in these speculations, there is no question that a transient period for Matisse was reaching an end. In May 1899, after nearly a year in the south, he returned to Paris with more than fifty small but incendiary canvases, failed to sell any of them, was discharged from art school because at age thirty he was too old, and applied at the Louvre to copy Italian Renaissance paintings with historical subjects. The experiment was over—for now.

  • 1 The first, Still Life with Oranges (I) (1899), is in the Cone Collection, Baltimore Museum of Art.
  • 2 For more on this and other issues of importance in the painting, see Jack Flam’s excellent analysis in Joseph D. Ketner et al., A Gallery of Modern Art at Washington University in St. Louis (St. Louis: Washington University Gallery of Art, 1994), 74.
  • 3 See Hilary Spurling, The Unknown Matisse: A Life of Matisse; The Early Years, 1869–1908 (New York: Knopf, 1998), 159.
  • 4 For a full display of this variety, see the extensive illustrations in Xavier Girard and Alain Mousseigne, Matisse, Ajaccio-Toulouse, 1898–1899: Une saison de peinture (Toulouse: Musée d’Art Moderne, 1986).
  • 5 Norman Bryson, Looking at the Overlooked: Four Essays on Still Life Painting (Cambridge, MA: Harvard University Press, 1990), 61.
  • 6 Matisse seems to have had in mind Paul Signac’s 1899 book From Eugène Delacroix to Neo-Impressionism, which he had read in serial form a few months earlier (see Spurling, Unknown Matisse). In this influential text, Signac cited the ringing declarations of Delacroix, one of Matisse’s gods of painting: “Gray is the enemy of all painting”; and “Banish all earth colors!” (Delacroix, quoted ibid., 176).
  • 7 On the varying thoughts regarding this idea of the “proto-Fauve,” see John Elderfield, “The Wild Beasts”: Fauvism and Its Affinities (New York: Museum of Modern Art, 1976), 18. For further discussion of this topic, see also Lawrence Gowing’s extended and sensitive description of the painting in his Matisse (London: Thames & Hudson, 1979), 24–27 (where it is called Le compotier et la cruche de verre [Fruit Dish and Glass Pitcher]).
  • 8 I thank Kimberly Broker of the Mildred Lane Kemper Art Museum for her assistance in researching provenance information on this painting.