In the early editions of Principles of Art History, Wölfflin included the following passage in the preface.

Nothing marks so clearly the opposition between the art of the past and the art of today as the unity of visual forms then and the multiplicity of visual forms now. In a manner unprecedented in the history of art, the most contradictory [tendencies] seem to be compatible with each other. … But the loss of vitality compared to the one‐sided strength of earlier epochs is immeasurable.1

By the unity of visual forms, I take it that Wölfflin means, or at least includes, the coincidence of those general representational forms, or concepts as he more often called them, which he found typical of sixteenth‐century Renaissance “classical” art and of those he found typical of baroque seventeenth‐century art.2 By contrast, the art of his time (he was perhaps thinking of the eclecticism of Salon art at least as much as early modernism) combines in a “confusing jumble” the classical and baroque concepts.3

Wölfflin's remarks raise two questions that this article addresses. First, what links the concepts into classical and baroque groupings, beyond their historical coincidence? Second, what happens when classical and baroque combinations of concepts appear together—and in what sense, if any, are they contradictory? Wölfflin does not write of these at all, since from his point of view they obscure rather than illuminate the grand movement between classical and baroque, which he saw as central to art history. Reading Principles of Art History, one might have the impression that they do not exist; but they do, having a presence throughout post‐Renaissance European art. Moreover, we shall see that applying Wölfflin's concepts to these works illuminates them in significant and unexpected ways. This analysis will also give a new approach to understanding Wölfflin's concepts and provide the resources for understanding the classical and baroque groupings of the concepts.

Before starting, there are two points to note about the scope of my project. First, Wölfflin applied his concepts to architecture and sculpture as well as the pictorial arts—I will be concerned only with the pictorial arts. Second, Wölfflin describes five pairs of concepts. The first of each pair he associated with the classical and the second with the baroque: linear and painterly, plane and recession, closed and open, multiplicity and unity (of depicted parts, rather than representational forms), and absolute and relative clarity.4 Jason Gaiger's reconstruction, drawing on the work of Lambert Wiesing and others, eliminates the last two pairs, showing that they are implied by the first three.5 For reasons of space, I discuss only on the first two pairs: linear and painterly and plane and recession. A longer analysis would draw in closed and open.

Sections I and II examine linear and painterly and plane and recession, developing an account of them in terms of visual attention. In Section III this is used to make an analysis of pictures that combines classical and baroque concepts. This will depend on using visual attention to forge connections to a quality that Wölfflin held to be outside “pure” vision: expression. Section IV explains the “unity” of the classical and baroque concepts, and Section V identifies a limit to applications of my approach.

Linear style—as seen in sixteenth‐century painters such as Dürer, Raphael, and Holbein, to take some of Wölfflin's favorite examples—tends to delineate the outlines of bodies, whether using actual lines or through unambiguous discontinuities of color and tone between figure and ground. Painterly style, as seen in Rembrandt, Hals, and Rubens, treats outlines in a relatively indistinct way, instead focusing attention on the body or mass. Wölfflin founds this distinction, as he does his other concepts, in terms of vision.6 The linear and painterly styles register and communicate corresponding kinds of vision. As he puts it, “linear style sees in lines, painterly in masses.”7 He goes on:

Linear vision … means that the sense and beauty of things is first sought in the outline … while seeing in masses takes place where attention withdraws from the edges, where the outline has become more or less indifferent to the eye as the path of vision, and the primary element of the impression is things seen as patches.8

A concern about this approach is that in explaining variations of style in terms of variations in visual experience, one would be committed to an implausible account of vision that would require visual experience to vary in substantial ways between individuals and periods. My approach is one way of making it clear that a Wölfflinian account is not committed to this idea (and it is worth noting that it is not the only way).9

I propose that these picture‐makers working in linear and painterly styles bring different kinds of attention to their subject matter. That is, they attend to different features of their subject matter—either outline or mass, to use Wölfflin's way of putting it. These differences in attention are typically registered in the resultant pictures. (This is not necessarily the case of course, but bear in mind that a picture usually only depicts those features of the subject matter the picture‐maker attends to.) In turn, pictures shape viewers’ experiences. In attending carefully to what the picture‐maker has depicted, a viewer is bound to re‐create something of the visual attention of the painter.10

Wölfflin also describes how this attention moves. He observes that with the linear style the focus of attention moves along lines: “the eye is led along the boundaries and induced to feel along the edges.”11 That suggests the following condition for linear visual experience.12  Visual experience is linear if it involves visual attention directed along outlines of objects.

Movement also figures in the painterly style: “Then it is as if at all points everything was enlivened by a mysterious movement. … Whether the movement be leaping and vehement, or only a gentle quiver and flicker, it remains for the spectator inexhaustible.”13 Here there is no mention of attention moving, but I think it is a reasonable inference to make. For if the subject matter is not depicted as moving, what else can possibly be in motion, aside from the viewer's eyes and visual attention? Indeed it seems very natural to speak of one's vision “leaping” between spotlighted elements in a Rembrandt, and “quivering and flickering” across the textured surfaces of the faces, hands, and bodies, or even brushwork itself.14 This presents a number of elements one could draw on in defining painterliness. The definition I give here elucidates the idea of “seeing in masses” in terms of this second kind of movement. Rather than attention following outlines, it is guided over the bodies of forms, often following its contours via the grain of depicted textures or the actual textural marks of the paint. So, I give this account of painterliness: Visual experience is painterly if it involves visual attention directed over the bodies of forms, rather than along outlines.

Wölfflin's remarks also suggest this kind of attention could be diffused—“it is as if at all points everything was enlivened.” That is, that it is no longer focused but distributed over an area.15 That need not be the case on my account, which allows for focused attention tracing lines across forms; but so far as distributed attention involves a kind of palpitating movement over a surface, it could be consistent with my account.

A question remains about why Renaissance painters have one kind of attention and baroque painters another, but we shall see that it also arises in the next section, so I shall delaying addressing it until then.

Wölfflin introduces plane and recession in this way: “Classic art reduces the parts of a total form to a sequence of planes [whereas] the baroque emphasises depth. … [T]he eye relates objects essentially in the direction of forwards and backwards.”16

So, a planar picture will arrange its subject matter in a series of planes, parallel to the picture plane, so they appear to face the viewer and overlap another. A recessional picture would arrange the same subject matter so that the depicted forms appear angled away from the viewer, receding often into a deep pictorial space. One might think that this is the product of the discovery of methods of perspective: that is, that Renaissance art tends to depict depth less effectively than does the art of the seventeenth century. Wölfflin is careful to quash this thought, pointing out, rightly, that High Renaissance painters had full mastery of these methods, but chose to use them sparingly.17

For Wölfflin this distinction is also to be illuminated in terms of vision. Here too attention can be understood to play an important role. For Wölfflin, planar and recessional vision is more than a matter of the view one has of an object. In the case of recession, Wölfflin also holds that it involves movement. Viewing a recessional form does not alone suffice to achieve the quality Wölfflin intends: “Every picture has recession, but the recession has a very different effect according to whether the space organises itself into planes or is experienced as a homogenous recessional movement.”18

Again, since nothing else can possibly be moving when one is looking at a static picture, I take it that recessional movement is a matter of the movement of visual attention as it follows forms receding into deep space (or, alternatively, out of deep space toward the viewer). In a planar composition, visual attention presumably also moves, but it is typically more gently guided along these planes—Wölfflin speaks of the “repose” of planar compositions compared to the more dynamic movement into deep space of a recessional composition.19

So, guided by these points, I account for the distinction between planar and recessional as follows: Visual experience is planar if it involves visual attention guided along planes appearing to face the viewer. Visual experience is recessional if it involves visual attention guided away from or toward the viewer along forms appearing to recede from the viewer.20

Let me return to the question of why the vision of Renaissance painters should have certain attributes (linearity and planarity) and baroque painters should have others (painterliness and recession). I understand these different kinds of seeing not as products of an autonomous history of vision, as critics of Wölfflin sometimes understand him, but as patterns of interest, which are the product of cultural influence.21 There is nothing especially mysterious about this. For the most part, these patterns of interest are propagated through pictorial traditions. The attention painters give their subject matter is conditioned by the traditions they are part of, ensuring the relative consistencies in style that Wölfflin draws attention to. I will have more to say about why traditions may be drawn in one direction or another below. Change to these patterns of attention will also be explained through changes in culture. A complex story can be told around this, but it is enough to note that Wölfflin himself allowed that quite mundane forces might have a role here, acknowledging “the theory of the palling of interest and a consequent necessity of a stimulation of interest.”22

I now turn to pictures that combine classical and baroque concepts. Let me start with paintings that are both linear and recessional. Many pictures fit this description, using a linear style, but including vertiginous, recessional perspectives. Examples in modern painting include Munch's versions of The Scream (the lithograph version shows the linearity of his style well), de Chirico's Mystery and Melancholy of a Street, and, less well known, many paintings by U.S. surrealist Kay Sage. There are also examples in earlier painting, in particular mannerism. We see it in Parmigianino's Madonna of the Long Neck. To the left of the Madonna, a crowded group of angels appears pressed up almost to the picture plane. To the right, in what must be a conscious contrast, the composition dives away into a distant (and unfinished) perspectival view of receding columns.

In each of these a distinctive effect is achieved, quite foreign to the classical and baroque styles. One's attention is guided more or less slowly along the outlines of the subject matter in the foreground, before being plunged down into a vertiginous perspective. As one's gaze moves toward the vanishing point, it remains bound to the outlines of diminishing subject matter. Arising from this is an expressive effect. It has a somewhat different character in each of my examples, but it is, broadly speaking, a disturbing effect, which in the modern pictures at least can be described as expressive of anxiety.

Let me add two points about expression here. First, my inclusion of expression contrasts with Wölfflin, who endeavored to exclude expression from his account of the concepts. Gaiger discusses and criticizes Wölfflin's efforts to keep them separate, citing various critics of Wölfflin, from Erwin Panofsky to Michael Podro.23 As Gaiger puts it, “[u]nder closer examination, the idea of a completely neutral and ‘colourless’ organization collapses, for it can never be kept wholly distinct from expression.”24 My examples are an illustration of this. But I have another reason for including expression, because I am concerned with the value that the concepts can have in art. Without expression the concepts are only a matter of movement of visual attention—and that alone is not enough to sustain artistic interest. Understanding the expressive use to which the concepts are put, in the examples above and those to follow, is critical to understanding and accounting for their artistic value. Second, there is the question of how it is that the movement of visual attention through depicted space plays a role in generating expressive effects. Answering that question is beyond the scope of this article, and accordingly I put it aside. It will be enough for my arguments that these effects do arise and that the movement of visual attention does have a part in producing them.

Turning now to planar and painterly pictures, we can distinguish two kinds by the different expressive effects they achieve. The first is exemplified by intimist paintings such as Vuillard's The Yellow Curtain and Bonnard's The Table. In these paintings, visual attention is restricted to a series of plane‐like surfaces—a bed and yellow curtain and a tabletop set with plates and bowls are organized into planes facing the viewer in shallow represented spaces. But the edges of those depicted surfaces appear indistinct, and the viewer's attention, encouraged also by the surfaces’ textures and patterns, is guided across these surfaces. Other examples of twentieth‐century painting have something of this character, such as many of the works Matisse painted at Nice.

This type of painting merits the label intimism—not simply in its tendency to depict “intimate” interiors and domestic subjects, but in the way it depicts them. The depicted surfaces appear as close and tactilely available—a feature that depends on the planarity of their compositions, and the viewer's attention is encouraged to wander freely, in a relaxed way across the forms—a feature that depends on painterliness. These paintings also have a distinctive expressive effect: the expression of a mood appropriate to the subjects depicted, but beyond that difficult to describe. As André Gide put it, “Vuillard always speaks in a whisper,” and much the same may be said of the other works I have mentioned here.25 They produce a distinctively intimist effect that is dependent in part on a style that is both planar and painterly.

An objection could be raised that this effect depends only on the associations prompted by the subject matter, not the formal qualities. The best response to this would come in the form of a Wölfflinian double projection, showing an intimist painting on one side—say Bonnard's The Table—and on the other, a painting of similar subject matter in another style, such as a seventeenth‐century Dutch still life. However enticing or homely the subject matter of the Dutch painting is, it lacks the distinctive effect of an intimist still life, thus demonstrating the contribution of painterliness and planarity to expressive effect.

The second kind of planar and painterly picture has a quite different effect. Titian's late painting The Flaying of Marsyas is an outstanding example of this type. The composition is as stridently planar as that of a classical frieze. At the same time, the work is an exemplar of painterliness. The edges of the forms appear lost in murk, and the eye is guided across the depicted forms by the flickering brushmarks of Titian's late style. Despite sharing features of classical and baroque styles, the expressive effect it achieves is like neither of them. It is an example of Titian's terribilita. Its supernatural subject matter has a threatening and frightening quality that gains from its apparent claustrophobic proximity to the viewer (achieved in virtue of the composition's planarity) and its indistinct and obscure appearance (which depends on its painterliness).

So here too we have expressive properties—an expression of threatening and frightening power—in part dependent on the planar and painterly style. Again one might ask whether the painting in fact draws its impact solely from its subject matter, and again the answer is found in comparing a depiction of the same subject matter in a different style. Thus Ribera's Apollo and Marsyas, to take a rather harrowing baroque depiction of the same subject, does not inspire the dreadful awe of Titian's painting.26

I now come back to the question of why it is that linear and planar and painterliness and recession go together for Wölfflin. In what sense do they form a “unity of visual forms”? Wölfflin calls linear and planar and painterliness and recession “different roots of one plant,” and “one and the same thing, but seen from a different standpoint.”27 The significance of such comments is difficult to pin down, but a helpful way to understand him is to take his view as holding that these pairs of concepts are aspects of a single kind of vision, a “classical” vision and a “baroque” vision. The examples of the previous section make it abundantly clear that there is no necessary reason why linear must accompany planar and why painterliness must accompany recession either in style or in vision. Still, Wölfflin must have been aware of the kinds of examples I discussed in the previous section, even if he gave them little weight. So it is reasonable to understand him as claiming that it is vision instead that binds these concepts together. This view can be articulated as follows: (i) that pictures can combine linear and recessional and painterly and planar, but (ii) there is no kind of vision that corresponds with these combinations. That is to say, he understood human vision to exist on a continuum between classical and baroque, and these pictures do not occupy a place on that continuum. To be clear again about my own position, while I accept (i), I reject (ii). My rejection of the view I attribute to Wölfflin leaves me to find another account of the affinity of linear and planar and painterliness and recession. We have seen that these combinations are not the only ways of composing a successful painting, but they do seem to be effective ones. Why is that? Let me sketch a response to that question.

If there is such a connection between the concepts, it must have a basis in their phenomenology—since they are properly understood in terms of experience. On my account that will mean finding a connection between the kinds of movement of visual attention that I have described as characterizing the concepts. This can be done as follows:

In both linear and planar experience, attention is constrained in space—constrained to the edges of bodies, and constrained to planes that appear to face the viewer.

In both painterly and recessional experience, attention is (by comparison) unconstrained—it roves across bodies and moves back and forth, in and out of deep space.

So, linear and planar involve the constraint of vision while painterliness and recession do not. Let me make the now familiar move of introducing expression to my analysis. Although it would have especially dismayed Wölfflin in this case, it seems to me unavoidable. The terms in which I have made my analysis are not artistically significant: I can see no reason why painters or viewers would have cared whether visual attention is either consistently free or consistently constrained. As with the other combinations I have examined, it is in supporting expressive qualities that they can find significance. Constrained, visual attention proceeds in a slow, measured way. That kind of movement is apt to express some of the qualities we associate with classicism: calm, care, and meditativeness. Indeed, despite his desire to keep classical and baroque separated from expression in his analysis, Wölfflin habitually describes the concepts of the classical style in terms like this, speaking of planarity as “sobered down,” as having “repose,” and Raphael's linear style as having a “great noble gait.”28 Unconstrained, visual attention can still linger but tends to move over bodies and through space. That is apt to express some of the qualities we associate with the baroque: dynamism, energy, and vigor. In this way linear and planar, and painterliness and recession do have a natural affinity. Each on its own has an ability to express certain qualities, and together they reinforce each other's capacity to do this. It is out of this kind of consistency of vision and expression that the “unity of visual form” and “one‐sided strength” of the classical and baroque styles that Wölfflin admired comes forth.

It will be apparent from my account of the concepts that they are universally applicable. That is a feature of Gaiger's and Wiesing's accounts too. As Gaiger puts it, the concepts “allow the possibility of talking about style in general terms,” or to use Wiesing's phrase (which Gaiger quotes), they remain “universally valid.”29 This is also close to what Wölfflin thought, although he allowed that his formulation of the concepts would require “modification” when applied to different periods.30 This is a feature of Wölfflinian analysis that makes critics uneasy. Edgar Wind, for instance, called it “grandiose.”31 So I want to finish by adding a qualification to my account to defuse this worry.

My analysis here draws on Wollheim's concepts of thematization and deletion. Wollheim tells us that thematization happens when a painter “abstracts some unconsidered … aspect of what he is doing or what he is working on, and makes the thought of this feature guide his future activity.”32 Wölfflin's concepts can be regarded as thematized in this sense. The experiences of linearity, painterliness, plane, and recession will all have some presence, or at least exist as possibilities, in a naturalistic painting tradition. We have seen how certain painters and traditions fastened onto these experiences, making the thought of these features guide the painter's future activity. That is to say, they sought these concepts out, making them a feature of their paintings. Wollheim also says, “thematization is always for an end … the acquisition of content or meaning.”33 The immediate end of the thematization of Wölfflin's concepts, I have urged, is expression—since it is hard to understand how the concepts, as visual experiences, could have artistic value in themselves.

I introduce the idea of thematization because it allows us to understand how the concepts can be universally applicable to pictorial art, without being universally relevant to understanding it. In pictures where these concepts are not thematized in an expressive sense, they may be present only adventitiously—the fact that our attention tends to move in one way or another over a subject may not be meaningful. There are a great many pictures likely to fall into this category, including much painting outside the European tradition and early and pre‐Renaissance European painting (which, in fairness to Wölfflin, he excluded from his analysis), works by naive painters and outsider artists, and many photographs (including both art and nonart).

Wollheim describes deletion in this way: “When deletion operates … an agent thematizes some feature of the work and then goes on to ensure that this feature of the work does not show up on the surface or shows up in only an attenuated fashion.”34 Deletion too has its end, the avoidance of certain kinds of content. Thus, artists working in the baroque style delete the linear and planar, avoiding the expressive qualities of classicism. Deletion here necessarily accompanies thematization: for one cannot have the baroque concepts and their expressive effects without rejecting the classical concepts and expression.35

There is also at least one instance in the European tradition where painters have deleted a pair of concepts. The impressionists, for example, abandoned the systematic pursuit of recession or plane, rejecting neoclassical planarity as well as the “neobaroque” recessional compositions of a painter such as Delacroix. Instead, they took on a probably photographically influenced, “naturalistic” approach to composition. Their paintings of the 1860s and 1870s still contain planar and recessional elements—but they are not pursued in any systematic way. There is a sense that their subject matter is pointedly unarranged; planar and recessional elements mixed up, as one often finds in a photograph. And indeed it would be wrong to ascribe the kinds of expressive effects I have described to the great majority of impressionist compositions—they are quite foreign to the impressionist aesthetic.36 Thus plane and recession have both been deleted in this style.

So while I have argued that Wölfflin's concepts yield a significant understanding of art, there is also an important limit on their application. The universality of the concepts must be tempered with an awareness that their application will not always help us in understanding artworks. While they are universally applicable, they are not universally relevant.37

Quoted in Frederic J. Schwartz, Blind Spots: Critical Theory and the History of Art in Twentieth‐Century Germany (Yale University Press, 2005), p. 20. Schwartz quotes from Heinrich Wölfflin, Kunstgeschichtliche Grundbegriffe: Das Problem der Stilentwicklung in der neueren Kunst (Munich: F. Bruckmann, 1915), pp. ix–x.

This must be distinguished from a pair of Wölfflin's concepts: multiplicity and unity of parts. I follow Schwartz in this interpretation of Wölfflin.

Quoted in Schwartz, Blind Spots, p. 21.

Heinrich Wölfflin, Principles of Art History: The Problem of the Development of Style in Later Art 7th edition [1923], trans. M. D. Hottinger (New York: Dover, 1950), pp. 14–16.

Jason Gaiger, “The Analysis of Pictorial Style,” British Journal of Aesthetics 42 (2002): 20–36; Lambert Wiesing, Die Sichtbarkeit des Bildes: Geschichte und Perspektiven der formalen Ästhetik(Hamburg: Rowohlt, 1997).

Although Wölfflin refers to vision throughout Principles of Art History, in the preface to the sixth edition he observed: “It is preferable to speak of modes of imagination, rather than of modes of vision” (Wölfflin, Principles of Art History, p. vii, original italics). The account I develop is an account in terms of vision rather than imagination.

Wölfflin, Principles of Art History, p. 18.

Wölfflin, Principles of Art History, pp. 18–19.

See Gaiger, “The Analysis of Pictorial Style,” and Wiesing, Die Sichtbarkeit des Bildes. They present accounts that do not draw on differences in vision, instead basing them on differences in the features of the depicted subject matter or in the marks on the picture surface. So the distinction between linear and painterly, Gaiger proposes, “describes different types of transition between the marks on the picture surface. … [P]ainterly transitions are merging and fluid, such that it is impossible to determine exactly where one thing stops and another begins, [and] linear transitions are distinct, with each of the parts clearly isolated” (Gaiger, “The Analysis of Pictorial Style,” pp. 32–33). Such an account is potentially consistent with my own, since the kind of experiences I describe typically are occasioned by the kinds of marks Gaiger describes.

Broadly speaking, the kind of attention on which my account relies—object‐ and feature‐focused, aligns with sophisticated spotlight theories (for example, Gordon D. Logan, “The CODE Theory of Visual Attention: An Integration of Space‐Based and Object‐Based Attention,” Psychological Review 103 [1996]: 603–649). Note that movement of visual attention is different from, although related to, eye movement, so I avoid speaking of “the eye's movement” or “the eye being guided.” For an account of how vision is affected by changes in attention yet remains cognitively impenetrable in important ways, see Zenon Pylyshyn, “Is Vision Continuous With Cognition? The Case for Cognitive Impenetrability of Visual Perception,” Behavioral and Brain Sciences 22 (1999): 341–365.

Wölfflin, Principles of Art History, p. 18.

The accounts I give of the concepts take the form of sufficient conditions. I leave open the question of whether there are other ways of achieving the concepts (as some of Wölfflin's remarks may be taken to suggest).

Wölfflin, Principles of Art History, p. 19.

There is another candidate for movement: the movement of the brushstrokes, or, more precisely, the traces of their movement, legible on the picture surface. If one were to accept that our attention must in turn follow these movements to have the effects Wölfflin describes, it would bring this proposal close to mine, except that attention would then be on the paint rather than the subject matter. I think this kind of experience is an important part of our experience of pictures, but it is clearly not the experience Wölfflin is concerned with in formulating this concept, since it is an experience of medium, and he excludes that from his conception of form.

The distinction between focused and distributed attention comes from psychology and is discussed by Bence Nanay, Aesthetics as Philosophy of Perception (Oxford University Press, 2015), esp. chap. 2.

Wölfflin, Principles of Art History, p. 15.

Wölfflin, Principles of Art History, p. 73.

Wölfflin, Principles of Art History, p. 82.

Wölfflin, Principles of Art History, p. 73.

Another way of putting these would be to replace the relation of subject matter to the viewer with the relation of the subject matter to the picture plane (see Gaiger, “The Analysis of Pictorial Style,” p. 33). This would make planar and recessional experience picture‐specific.

For criticism of Wölfflin on this line, see Edgar Wind, “Warburg's Concept of Kulturwissenschaft and Its Meaning for Aesthetics,” in Edgar Wind, The Eloquence of Symbols: Studies in Humanist Art, revised edition, ed. Jaynie Anderson (Oxford: Clarendon Press, 1993), pp. 21–36.

Wölfflin, Principles of Art History, p. 230. Wölfflin alludes to Adolf Göller, “What Is the Cause of Perpetual Style Change in Architecture?” in Empathy, Form, and Space: Problems in German Aesthetics, 1873–1893, eds. and trans. Harry Francis Mallgrave and Eleftherios Ikonomou (Santa Monica, CA: Getty Center for the History of Art and the Humanities, 1994), pp. 193–225.

Gaiger, “The Analysis of Pictorial Style,” pp. 26–27.

Gaiger, “The Analysis of Pictorial Style,” p. 27.

Quoted by A. R., “Edouard Vuillard Retrospective Exhibition,” Calendar of the Art Institute of Chicago 66 (1972): 1–3, at p. 2. A. R. quotes from La Gazette des Beaux‐Arts, December 1, 1905. Gide was clear that the effect was a matter of expression: “[H]is brush never breaks free of the emotion which guides it; the outer world, for Vuillard, is always a pretext, an adjustable means of expression.”

Let me add a more surprising example of pictures in a painterly and planar style: some of the late work of abstract painter Rothko, especially the Seagram Murals. So far as one sees them as showing floating, proximate blurs of light indicating a supernatural presence, they suggest a similar analysis.

Wölfflin, Principles of Art History, p. 109.

Wölfflin, Principles of Art History, pp. 106, 73, and 12.

Gaiger, “The Analysis of Pictorial Style,” pp. 36 and 32. Gaiger quotes Wiesing,Die Sichtbarkeit des Bildes, p. 66.

Wölfflin, Principles of Art History, p. viii.

Wind, “Warburg's Concept of Kulturwissenschaft,” p. 23.

Richard Wollheim, Painting as an Art (London: Thames and Hudson, 1987), p. 20.

Wollheim, Painting as an Art, p. 22 (original italics).

Wollheim, Painting as an Art, p. 23.

Wollheim makes an analysis of baroque art along much these lines, drawing on Wölfflin in his characterization of the baroque and applying deletion (p. 26). It is also worth noting that Wollheim was effusive in his praise of Wölfflin (Wollheim, Painting as an Art, p. 359, n. 23).

This was not to last long—artists such as Cézanne and Seurat were again to thematize planarity in painting in the 1880s. There is an important exception in Manet (if he is counted as an impressionist), who retained a preference for planar composition through his career.

I am grateful to the members of the Anglo‐German Picture Workshop, especially Lambert Wiesing and Jason Gaiger, who gave generous and helpful advice on a forerunner of this article at the workshop's 2014 meeting, and also to Bence Nanay and an anonymous reader for this journal, whose comments have helped to improve and focus it.