Seeing things in pictures


A half century of theorizing about pictorial seeing has concentrated mainly on the central experience of seeing the subject in a picture. This consists variably of seeing an individual, identified or unidentified, real or fictive, in a picture, and seeing that individual as being of a given type or as having given properties. The latter, the depicted properties, are in general the most problematic. There is honorable dispute, as well as honorable uncertainty, about what properties determinate or indeterminate a subject is depicted as having in particular cases and in general. Can we see on-going movement as opposed to arrested poses in depicted scenes? Can we see states of mind, subjective visions as opposed to real or fictive realities? Can we see pastness or futurity in the subject or is that merely inferred from what we see? How far beyond the shown (on-stage) portion of a depicted scene can we see the subject as extending, and with what specificity? How far in front of the picture plane can we see the pictorial space extending? Can we see depicted persons addressing ourselves, the viewers? Can we see a pictorial subject as referring to its being a pictorial subject? The scale of such problems varies considerably depending on the kind of picture. They rise to near global proportions in cases of conspicuously textured or abstract modes of depiction, an ascent that begins with impressionism and post-impressionism and climaxes in cubism and other avant-garde manners, where the depiction is often bafflingly cryptic.

The questions about pictorial seeing do not stop with those about subjects and their properties. For as has long been recognized we can also see things in pictures that do not belong to the proper subject. Wollheim put it this way: “With any representational picture there is likely to be more than one thing that can be seen in it: there is more than one experience of seeing-in that it can cause.” (2001, 26) This raises the question to what extent, if at all, seeing other things in a picture than the subject or other properties than those depicted as belonging to a subject is relevant to the understanding and appreciation of the picture. I think the relevance of such seeing is substantial and somewhat under appreciated, as I will try to show.

At the outset a point of terminology needs to be made concerning the expression seeing something in a picture. As a piece of the vernacular this formula is often used for seeing a picture as depicting something. It does not offer a theory of pictorial seeing but merely expresses the fact to be explained. In this respect it stands in contrast to “see-in” formulas (often hyphenated) that are put forward to explain what pictorial seeing is, how much of it is “experiential” and how much interpretation. In the latter, more technical sense, the phrase allows for a conflict between what one sees-in a picture and what one sees to be its depictive subject. The vernacular usage is so deeply entrenched that it is extremely inconvenient to do without it in discussing pictorial seeing and I shall use it without apology where no confusion is likely to result. Thus I will speak of persons seeing something in a picture (vernacular sense) even when they also, knowingly and at the same time, see-in it something of a different type or possessed of different properties – without any error being implied on either side. Insofar as a seeing-in experience is taken to be representative of the subject, I can speak of it as subject-seeing. Pictorial seeing-in experiences which carry no presumption as to the subject of the picture or the properties it is depicted as having may be labeled phenomenal seeing-in whenever so doing serves the cause of clarity.


Robert Hopkins (1998) goes a bit further than Wollheim in regard to things seen-in a picture that are not taken to be its real subject, characterizing such seeing-in as a “separation” phenomenon. He cites cases where such seeing-in seems not just possible but inevitable or at least highly likely. One of his examples is a pencil sketch by Alfred Stevens in which the nude torso of a youth is rendered in a somewhat loose style.


Here is Hopkins 's account of what a viewer might see in the picture as she responds to the graphic peculiarities of the rendering.

a man of an outline shape as determinate as that she sees the drawing as having..Fn1. That man is semi-transparent, especially at the top of his leg and headpiece. His nose and mouth blur indistinctly into the rest of his face. Across his upper arm and midriff run parallel curving strands of something. The little finger of his left hand seems malformed. (139)

Perhaps not all of these descriptions will survive a scrupulous review. Certainly they fall short of precision. But they come close enough to make Hopkins 's general point. The imagined viewer's way of seeing is not without some foundation. The free-floating lines comprising the headpiece suggest neither hair or a turban nor anything else we can identify. The interior of some shaded parts of the arms can easily be penetrated by a probing eye. The left hand is seriously misshapen and joined to a dark V-shaped something or other, more solid than a shadow but without any other clear physical specificity. And these references are the tip of a substantial iceberg.

A competent viewer, says Hopkins, will know that many graphic features of such a drawing cannot represent the properties she knows the depictive subject to have, for her competence as a viewer incorporates practical knowledge of three broad kinds: (1) “knowledge of what sort of items the world contains,” e.g., “that things are very rarely limited in colour to differing shades of gray…that dogs are not green, and that people are not composed of very thin cylinders.” (2) “knowledge of the sort of items [that are] in general depicted,” namely    “items exhibiting something like the features real things enjoy.” (3) Knowledge that depiction-interpretations are conditioned by recognition of the limitations of verisimilitude achievable within a given medium and in rivalry with competing artistic aims. For instance, an artist may employ rapid, free-hand strokes that necessarily sacrifice accuracy or determinacy for the sake of verve. (138, 144)

This understanding of depiction practice, Hopkins says, will convince the viewer that the artist did not intend her to take an anomalous manlike being to be depicted, but rather a (roughly) normal man who is indeterminately depicted. Instead she will regard her (phenomenal) seeing-in experience as “eccentric.” Fn2. (139) Other accounts suggest variations of this general line. Wollheim's account, for instance, privileges the artist's intended form of seeing-in, providing that intention is fulfilled by the picture. (2001, 26-27). These responses serve to protect the depictive subject against misconstruction, but they invite attention to the equally vital question of what pictorial good, if any, is achieved by such wayward seeings-in and what the viewer's full response to them should be. Should we repress them, or let them slip into the background, or dwell upon them and cultivate our capacity to see them? Is our liability to entertain them a fault in our visual processing, or a virtue that can be “recruited” to worthy picture-viewing ends? Hopkins 's reference to limitations of media and to expressive non-depictive purposes acknowledges the subject, but leaves most of it unexplored.

Several preliminary points are useful to make in order to dispose of doubts that may be entertained regarding the idea that pictures like the Stevens drawing provide sufficient foundation for a viewer to have phenomenal seeing-in experiences as eccentric as the ones Hopkins claims. The first is based on the idea that separation-seeing, as I will call it, is incompatible with subject-seeing. But this is neither claimed nor is it justified, in my view. In the usual case separation seeing-in occurs when the viewer sees perfectly well what the picture depicts, including the properties the subject is represented as having. I do not suppose this is universally the case. The viewer's attention is variably distributed at different stages of viewing a picture. Also pictorial seeing can be more or less accurate, and more or less comprehensive. Where separation-seeing-in is taken as faithful to the depiction content, it is erroneously so construed. But where the viewer can only see a separation-“subject” in a picture, not the actual subject, her attitude is more likely to be puzzlement than conviction that she has got it right, for separation “as-if” subjects are typically anomalous. The content separation-seeing represents the picture as having is egregiously inapplicable to the real subject and sometimes to anything the viewer has ever encountered as a subject.

Another objection is based on the belief that claims for the possibility, prevalence or validity of separation-seeing embody a confusion of what is seen on the surface with what is seen in it. Here it is granted that certain design properties of the surface are non-standard for a depiction of a given subject. What is rejected is that there is any justification for our experience representing them in terms of an anomalous pictorial object or scene (or part thereof). It may be feared that accepting separation-seeing as valid or relevant would lead to chaos: every design property that falls short of full verismilitude would then be taken as an appropriate provocation for anomalous seeing-in. Only magic realism would escape and perhaps not even that.

The answer to this has several parts and leads inexorably to discussion of particular cases. Granted we can override some graphic features of drawings like the Stevens. Others we cannot, not without ignoring features that rightly claim our attention. Justified separation seeing-in, not just correct seeing of the subject, depends to some extent on the limitations of the medium. Compared with paintings, pencil drawings use a much diminished repertoire of marks. This affects what we can spontaneously see in them. As a result, wiry outlines enclosing part of a blank field do not generally support experiences as of seeing wiry strands enclosing empty space, as if depicting a wire drawing or sculpture by Calder. Recognizing as we do that we are addressing an outline drawing, we see solid three-dimensional forms depending on their contours' convexity, concavity and manner of intersection. In less spartan drawings modeling lines of various sorts add specificity, reinforcing or modifying the effect of bare contours. Where such devices are used, as in the Stevens drawing, absence of modeling, as in the upper leg, reinforced by the truncated contour, signifies incompleteness of depiction, not transparency. Fn3. This affects what we see in the drawing: we see incompletely drawn solidity. Thus the semi-transparency of which Hopkins speaks can be validly seen in the drawing only when it is supported by a conspiracy of cues relative to the medium, for instance where lines recognizably on the surface of a figure cross over smudges or turn under each other implying visible depth within the limb or other part. Thus we cannot properly see transparency in the upper leg. Fn4. Nor is it justified to see the torso as semi-transparent. Indeed that description applies to the youth's headpiece only with qualifications. For we cannot fail to see that whatever is there hides the scalp; to that extent the form isn't seen as semi-transparent. But the lines depicting whatever stands between us and the scalp curl and overlap in such a way as to indicate that its outer layer, if layer at all, is not opaque. In the end so far as I can see there is no credible interpretation of the particulars of that set of lines, even as strands of some unknown fiber. In short we have no sufficiently conceptualizable subject to fit the case. Yet I contend that they are things we can perfectly well see in the picture, ones that the picture invites us to see. Notwithstanding their resistance to any credible interpretation they are experienced as standing (lying, curling) in a fictive space, even by the skeptic. Or what is as good, they are so experienced when the drawing's details are attended to in a seeing-in mode. In fact the depth cues are strong enough to enable a competent viewer to find in the picture, including the headpiece, just as palpable a three-dimensional appearance as that afforded by the unproblematic parts of the figure.

Thus I am making three claims about valid but anomalous separation seeing-in. (a) It is the seeing-in of some three-dimensional object or scene differing in some respect from the depictive subject. (b) It is a seeing-in for which adequate support is provided by identifiable cues. (c) It is a seeing-in which has a role in any optimal pictorial seeing of the drawing: at some stage in the process the optimal viewer sees semi-transparency and the other anomalous properties in a drawing like the Stevens. Any viewing of such a picture that lacks this perception falls short of full and maximally perspicacious pictorial seeing, i.e., one that appropriately exercises our best aesthetic faculties.


Hopkins ' account leaves the impression that the separation phenomenon is relatively limited. It arises with free sketches, etchings, abstract representations such as in cubism, and the like. I think this is far too narrow an extension to do justice to the effect that design features have on what is constructively seen-in the works. I believe that a separation phenomenon is at least latent in all attentive picture-viewing where the picture possesses substantial pictorial or optical depth (and perhaps even more pervasively than that). All such pictures, viewed correctly, elicit experiences representing a figure or scene which in some way and to some degree is distinct from, because incompatible with, the picture's (propertied) subject. I also believe that it is not contrary to the knowledge or intention of the artist that we see in the picture these “eccentric” appearances. To the contrary it is something that an artist typically desires, requires, and often cherishes. It is, explicitly or implicitly, part of the depiction endeavor. Full and nuanced appreciation of the picture as a picture requires it, art criticism could hardly proceed without implicitly recognizing it, and in general the charm of pictures would not be half of what it is without such seeings-in being intended and taken up. I will call the first of these propositions the ubiquity claim and the second the aesthetic significance claim. At present it is the contribution of facture Fn5.-separation to ubiquity that claims our attention, without prejudice to further extensions of the basic separation theme.

To pursue the ubiquity thesis, let us examine a finely textured full-color painting such as Ghirlandaio's Portrait of Giovanna Tornabuoni Fn6. (1488). For all its fidelity of color and texture to the intended subject, which I take to be the sitter idealized, the flesh is presented as somewhat wooden and the hair as artificially, almost metallically stiff. Vasari complained of this in the paintings of this period. Fn7. The subject is not to be taken as wooden and stiff in these particulars. Nor is the lustrous gown to be taken as precisely like the real gown in all visible particulars. Even if we cannot precisely specify the intention, we know it cannot be exactly this. No gown's sleeve could present precisely that combination of clarity and confusion. Overall the painter subtly stylizes, combines a precision with a muted glow that confuses the detail of the sleeve. There is also the barely noticeable brushwork that textures the exposed skin in an unfleshlike way that the painter is not able to avoid, a fact we, his admirers, find ourselves far from regretting.

You may say, yes, but the viewer is supposed to overlook the discrepancies, to see Giovanna's soft, unstreaked flesh right through the brushwork, her silky hair through the two un-silky-stranded buns, and the tangible stitchwork through the complications on the sleeve. But I submit that any such “seeing through,” if possible at all, is either counter-artistic (naïve or negligent) or else impure. By impure I mean that the “eccentric” ostensible hardness never is expelled from the experience but persists like a veil or double-image, not just on the surface of the panel either, but there, in pictorial space. Perhaps you can visualize the subject full-fleshed, etc. Or you can imagine that your seeing the parts of the surface is seeing these properties of the subject. But there is a more perspicuous, sharply focused experience of the work that involves perceptual awareness of what is presented with more visual immediacy than the property intended for the subject. When all our best aesthetic faculties are appropriately deployed, the separation seeing-in is part of the ensemble. The depicted properties are what is ascribed to Giovanna, but where separation-seeing occurs, they cannot be seen there with such visual immediacy as the anomalous ones. The plain evidence of our senses prevents it..

With regard to the aesthetic significance claim, it seems equally certain that Ghirlandaio expected viewers to see the anomalously hard, clean properties that we see in the picture, though without intending to impute those properties to his subject. He certainly saw them, and there is good reason to think that what the artist sees the viewer should also see. Our appreciation of Ghirlandaio's distinctive style depends on not only seeing the surface facture as such but also seeing the impact of this in pictorial space. At the same time we find no difficulty in simultaneously seeing the true subject of the painting, Giovanna as she is meant to be. More of that in a moment.

The reason why separation-seeing is a part of pictorial seeing seems plain enough. The depiction-tradition governing pictures of this sort invites us to do a number of things: (a) to use our general capacity to see things in surfaces, bringing as many marks into the experience as fall within the normal (or advanced, superfine) repertoire; (b) to discriminate between ensembles of marks that are straightforwardly subject-descriptive and those that are equivocal or deviant or have other functions (say, symbolic or expressive or decorative) or only amount to visual noise marking the surface as a surface (e.g., the texture of paper, accidental marks or stains); (c) to navigate among these auxiliaries in appropriate ways, appreciating how they enrich (or alternatively, detract from) the total depiction-experience. Ghirlandaio's art belongs to a tradition that aspires to verisimilitude as perfect as the resources of painting make possible and which has developed those resources to the point where viewers are entitled to wring from the design as many visible properties, including lighting effects, as possible. The viewer therefore searches for the subtlest textural quality of the areas representing flesh and hair, and comes up with something too hard to be entirely convincing, whatever other virtues it may have.

As Vasari goes on to note, it was left to painters of the next period to achieve full mastery of the effect of flesh, as in Leonardo's Mona Lisa in whose throat the pulse seems to beat , and that of hair, as in Correggio's pictures where “it appears soft and downy, with each golden strand finely distinguished and coloured, so that the result is more beautiful than in real life.” Fn8. When we turn our searching attention to them, however, we see that achieving these effects has consequences, for now some part of the sfumato of shadows extends like a light haze over everything, the like of which we never encounter face-to-face. Better than life, perhaps, but in its own way subtly anomalous. Fn9.


In arguing for the (relative) ubiquity and significance of separation-seeing, I wish to stress that these claims carry no implication that there is anything illicit in suspending one's awareness of the slightly wooden look of Giovanna's flesh. To the contrary, I believe that at some point in one's negotiation with the painting it is entirely in order, in fact required, to give the lady the benefit of the doubt concerning her skin and imagine it to be as silky sleek as it may have been on the sitter's young neck, and even to see it so in the picture. How do we do it? By simply not focusing so unsparingly on the actual texture. This won't work if the painted texture is too harsh. We can't experience Giotto's female flesh as skin-soft. No manner of focusing or attending will yield that result. But many modes of depiction give us the option of subordinating the evidence offered by our senses when sharply and focally deployed and thereby enhancing the textural appearance. If, as I maintain, our address to pictures consists of many ocular fixes, slides, scans, and repositionings, then it suffices that among them there is one that yields the effect best representing the subject as the maker desires. But the full depictive character of the painting is revealed only by the ensemble of seeings-in it offers and the varying salience or faintness, persistence or transitoriness, of members of the ensemble. Exploration of a picture involves all of these modes of address appropriately staged, gathered, and reflected upon. That they cannot all be combined into a single moment of full and undivided comprehensive salience – a master synthesis – is neither here nor there.

What is (phenomenally) seen-in a design in separation-seeing has the appearance of a depiction subject, though by definition it is not one – or not the real one. It has only the appearance of being a subject (in the case of properties, of belonging to the subject); in other words it has only the potentiality of being a subject in another pictorial context. It is useful to refer to appearances nominally, as in apparent winners, apparent perpetrators and all the rest. If I am right about the relevance of separation seeing-in these appearances are important elements in our experience of pictures; and on the strength of this consideration I will take the liberty of referring to the objects of separation-seeing as apparent subjects, as Hopkins does at one point in referring to “the putative resembled object” (2003a, 160). In this spirit we may say that in seeing a picture compendiously we see not only the real subject but one or more apparent subjects. Adopting this terminology simplifies my task in what immediately follows as does referring to separation-seeing as visually more immediate than seeing the subject since the latter is mediated by a fuller, and authorized, set of subject-determining principles, such as those alluded to by Hopkins.

As an exercise in delving further into apparent subjects I propose a thought experiment. Imagine looking at drawing with highly conspicuous graphic facture and exploring the separation possibilities. Let us leave Stevens's drawing behind in favor of the much superior drawing by Rembrandt of Jan Cornelisz Sylvius, which also offers the advantage of prior discussion by Podro (1998, 16-7) and by Hopkins in the current volume. (Fig. .. , p. …). It is full of handles for separation-seeing. The project is first to assemble as rich an ensemble of separation-aspects as possible, picking up on the scribbles, ink strokes heavy and dry, and the other facture features, and then viewing the drawing from a favorable position bring the visually apparent subject to as high a degree of salience as possible.

Then imagine you can by flipping a switch replace this robust separation-seeing with a face-to-face experience of a three-dimensional counterpart that has solid forms and scenic features as closely corresponding to the elements in the apparent subject as possible. You find yourself confronting a wall with an oval opening in which there is a configuration of mark-like things which from your viewing station afford pretty much the same visual array. Since the latter experience is of an actual scene all the perceptual routines used for any actual scene may be deployed: moving a bit left and right, a little closer and farther, etc. Thus we are able to see how the heavy black vertical bar-like forms, the splotch-like and stained-wall-like surfaces and all the rest stand in relation to each other, and how they combine to yield, from our initial point of view, a fantastic life-size simulacrum of Sylvius' body in the setting. From the face-to-face seeing we can return to the separation-seeing – or if the two stand side by side, simply shift our gaze – and study the separation experience anew.

Granted, there are difficulties imagining real world counterparts that “realize” some of the graphic features. Since they stand at different depths without visible support they might have to float mysteriously or be marks on the surface of perfectly non-reflecting glass (non-reflective from the indicated point of view). There would not be any uniquely definitive counterpart. But the object of the experiment is only to bring home the relations between the experiences. And for this a good approximation should serve.

Envisaging such experiences has the benefit of forcing our attention upon the design peculiarities of the depiction and gauging their implications. For in imagining counterparts we must scrutinize the pictures far more intensively than we are likely to do otherwise. We ask more searchingly, what might this or that graphic feature be taken to depict? Of Jan Cornelisz Sylvius' outstretched left hand we ask, what could make it so dark that not even the fingertips catch the light and the hand itself virtually disappears in the splashy (six-pronged) ink-mark-like form? In rebuttal the skeptic may say that the features I attribute to that problematic apparent subject are merely features of the marks on the surface. The question asked about the hand is a question about the brush marks on the paper. My answer is that it is only when we see in that brush mark a three-dimensional form standing in pictorial space that the question can be asked: why don't those forms that look as if meant to be fingertips catch the light which is also seen-in the picture? The question concerns a seeing-in experience of a three-dimensional scene furnished with forms matching the marks. Given the right viewing conditions the experience can be coaxed into a quasi illusion not far from an after-image or ocular double-image. Anyone can verify this without difficulty.

The same applies to the global textural aspects of pictures of many sorts. In many of Dürer's engravings, for example, the engraving tool spreads over almost the entire surface a distinctive grain. Mostly it is seen as covering surfaces in pictorial space but sometimes as filling that space, but of course it does no such thing in the actual subject. A striking effect of this grain is to unify the scene, just as the grayscale does, in comparison with scenes in the world of the subject. Graphic or painterly textures and tones do this in general, thereby helping to give pictures their distinctively pictorial unity, a unity which extends to the whole pictorial space, not just to the picture surface. The salient point for my analysis is that the grain is depictively “recruited” but not directly for the real subject, but only for an as-if subject.


Is separation-seeing reducible to the real subject being represented as looking like an anomalous being of the sort in our thought experiment, or in the case of Ghirlandaio's portrait Giovanna's skin looking as if it were wooden. I think not. Aside from the fact that we do not normally think of these subjects in the X-looking-like-Y way, there is another, more crippling difficulty. There is no known way for the subject to look as it would have to look on the proposed solution without undermining the pictorial purpose. For a normal youth's headpiece to look like a snaky cordlike tangle overlying a featureless opaque covering is hard to develop into a coherent, visually realizable idea. The same is true of Sylvius looking like the fantastic being in the thought experiment. When the surface marks are as free as in the Stevens or Rembrandt examples, there is seems to be no good candidate for an X-looking-like-Y subject. We are stumped as to how a youth or an aged divine could look the way required, or be represented as looking that way. In the case of Giovanna the difficulty is different. The idea of Ghirlandaio's subject being Giovanna-with-wooden-looking-skin is deeply contrary to the reason for depicting her.

Another alternative to the extensive invocation of separation I am proposing might fall under the rubric of inflection, an idea that has figured prominently in recent discussions of pictorial seeing, e.g., in Lopes (2005/6, 123-4). The idea is prefigured in a general way in Wollheim (2001, 20) and somewhat more specifically in Podro (1991, 1998), where the key ideas are advanced of the surface markings being “recruited” to representational ends and of their interpenetrating or intermingling with properties of the pictorial subject in a way to which there is no analogue in face-to-face experience of a reality corresponding to the subject. The papers in this collection by Hopkins and Nanay moot this idea in some detail, giving to it a yet more specific meaning, as in this, by Hopkins : inflected seeing-in occurs when “what is seen in a surface includes properties a full characterization of which needs to make reference to that surface's design (conceived as such).” (…)

Hopkins applies the idea of inflected properties/seeing-in to Sylvius' outstretched hand, following Podro (1998), and offers an account of it that claims two things: (1) “the hand itself seems to be both body part and rising splash of ink” and (2) “what is seen in this picture is a hand composed of rising ink.” From these he concludes: “what is seen in needs characterizing in part by reference to properties of the picture's surface.” In support of this contention he observes that the “inky splash” can be seen as rising only if it is seen as the continuation of a movement beginning at Sylvius' left shoulder and strongly descending then rebounding outward and upward in the outstretched hand. All this must be seen in fictive space. If relegated to the picture surface the two motions would oppose each other rather than one being a continuation of the other.

While I fully endorse Hopkins 's openness to seeing anomalous things in pictorial space I find it hard to understand how it can be quite right to see-in the inky splotch a hand composed of ink . I suggest that what can be immediately seen-in it is a messy hand-like form, at the same time as one sees that the splotch depicts a normal human hand. For what could a hand composed of ink be, rising or still? (Actually the hand is thrusting forward rather than rising, but let that pass.) Certainly a hand cannot be a splash of ink and the idea of a hand being seen as that seems to me unintelligible. Perhaps a hand might be composed of congealed ink, but no such thing is seen-in the picture. Fn10. Further, I fail to see how the description of what is immediately seen need be taken as necessarily referring to the design. “Ink-splotch- like ” or even “composed of ink” certainly need not so refer. Nor is such reference demanded simply because the seeing-in is based on seeing the inky splotch on the surface. If that were so every mark depictive of a receding form would be inflected. Therefore I cannot see that Hopkins ' criterion of an inflected property (one the full characterization of which necessarily makes reference to the surface) is satisfied by the cited case. Also one other idea in the inflection story, viz. that some properties recruited to the subject cannot be found in (any conceivable) face-to-face experience seems refuted by the thought experiment. That shows that a splash of ink or a handlike something is quite possible to imagine oneself seeing face-to-face, even if what sort of thing there might be that would look that way is not by any means settled.

Yet I think that elements in the inflection story are real enough. There is reason to believe that pictorial seeing presents apparent subjects that cannot be seen face-to-face. And equally some kinds of transfer of surface properties to depicted subjects really occur.   Thus there is something real underlying the notion of inflected properties, which I will try to set forth immediately after making what I hope is a clarifying distinction between separation and other aspects of pictorial seeing that are easily confused with it. I believe that separation-phenomena form an aspect of pictorial seeing distinct from the inflection-phenomena typically cited in the literature.

In Michael Podro's seminal article “Depiction and the golden calf” (1991, 165-175) the author cites a number of interesting aspects of depictive subjects that are derived one way or other from the artist's activity, e.g. from pen or brush marks or intervals of dark and light. It is important to distinguish the ones that involve separation from those that do not, though Podro's metaphor of intermingled properties can easily be taken to apply to both. For instance, there is the effect of movement of the legs and arms of the dancers in Poussin's The Worship of the Golden Calf , which is an aspect of disegno traditionally discussed under the rubric of “invention” or arrangement of depicted elements in a scene. (168) The same is true of compositional movements (eye-tracks) such as the downward sweep proceeding from upper right to the seated figure in the foreground and thence upward to Aaron. (169) The artist arranges things so as to bind the composition together with connections such as this, exploiting the flatness of the picture surface, which gives more salience to the movement than face-to-face seeing of a comparable scene would. Do these effects involve separation? And can we experience the effects only in a picture? If the latter were so, then were we to confront a three-dimensional scene in which the figures, trees, rocks had the positions indicated in Poussin's painting we would be unable to see the sequence of forms corresponding to the one in the painting. But I submit that we can see a sequence very like it in such a real world counterpart, like enough to make it a felicitous compositional aspect of the view.   Even more plausibly we could see the effect of movement in a scene of real dancers frozen in the same limb and body positions. In the last case we might well need a period of adjustment in focusing our attention appropriately, since such a sight would certainly be shocking (real dancers frozen !), though in sculpture such impressions of movement-in-process occur easily enough. However that may be, in these cases we are not interpreting the marks on the surface in pictorial terms that conflict with those properly used of the subject, as we are in separation-seeing. We are not seeing transparency where opacity is ascribed, flatness where rounding into depth is ascribed, or a wiry something or other where an edge of a solid body is ascribed, or movement when stillness is ascribed.

Substantially the same points apply to Podro's seeing the shadowed ambience in Poussin's Eucharist as conveying both the historical remoteness of the scene and the viewer's task of making out what a picture depicts. (173, 175) This does not qualify as separation seeing-in; nor does his taking the off-palette brightness of the seated woman in the foreground of the Golden Calf to signal a choric commentator looking at the scene from without. (175) Whether or not these readings of the picture are correct, they exemplify not separation but rather a relation between two aspects (perhaps we should call them layers) of approved subject content, such as holds between pre-iconographic and iconographic content. Further, they do not sustain the idea of subject properties that are fully describable only via reference to surface properties.

I now return to the promise to set forth what I think comes closest to meeting the criteria of inflection. The signs of particular ways of image-making confer on the drawn or brushed marks dynamic and behavioral characteristics. Lines or brushstrokes are swift or slow, confident or hesitant, bold or timid, heavy or light, dry or wet. Calligraphy exploits these descriptively aesthetic properties within the two-dimensionality of the surface, with no implication of pictorial space. In pictorial contexts some of these characteristics are seen-in subjects. For instance, the behavioral-emotional property of boldness or timidity characterizig a brush stroke is often properly seen as characterizing the forms, whatever they be, that are seen-in pictorial space and thereby seen as possessed by the (real) subject. This is conspicuously so in Rembrandt's drawing. It is doubtfully possible that we could see Sylvius as exhibiting so forceful a personality, or possessed of so assertive a state of mind, without the brio of the brush marks. Here is a case where something like “interpenetration” occurs. When we see Sylvius amidst that turbulence of heavy stroke-like, three-dimensional forms we appropriately see the literal and the metaphoric intermingled, just as we do on the surface. This seems to me as close as we can come to the “symmetry” of subject and artistic process of which Podro speaks (seeing each in the other) or to Hopkins' idea of properties seen-in the subject that cannot be fully characterized except by reference to design properties of the surface conceived as such. Yet I am not convinced that even here the situation is quite what is claimed, for both base and descriptively aesthetic properties operate in parallel on the surface, in two- dimensional calligraphic space, and in three-dimensional pictorial space – not strictly across the border. They both figure in pictorial seeing by being co-present in the complex and extended process of seeing the picture. That co-presence may be all there is to inflection.


So far we have been considering only separation based on some part of the picture being less than maximally verisimilar of its kind. But there is another aspect of separation not mentioned by Hopkins but central to my account, which I will call “spatial separation.” This aspect would operate even if the picture offered no opening for separation within the constraints of what is possible given the medium (including the implements). It operates over all pictures because they are pictures, depending only on the projection system used. Consider Michelangelo's drawing of one of the nudes from the Sistine ceiling now in the Albertina Collection. Fn11. What position of the legs do we see in the picture?


From the correct viewing station the legs are seen extending out toward the viewer at a roughly 45º angle. But as we well know, seen from an oblique angle the effect is different. As we move to our left the legs swing closer to 90º and if we move to the right they swing closer to the picture plane. Manet's dead matador ( The Dead Toreador , probably 1864, National Gallery, Washington, D.C.) notoriously pivots similarly to right and left as the viewer moves laterally. What legs and bodies do, eyes do also. If they are ever directed toward the viewer they follow her as she moves laterally. More generally the space (phenomenally) seen-in the picture undergoes striking contortions, expansions, and compressions as the picture is viewed from different stations. Fn12. Yet the depicted subject does none of these things. These anomalies are a direct consequence of the discrepancy of dimensions (two vs. three), the projection system (perspective) and the fact that vision is irremediably perspectival. Even if the anomalies are not always noticed consciously, there is no question but that they are available, that they operate at a preconscious level, and that open-minded viewers can be easily taught to notice them. If this were not so viewers would experience a full illusion of an unchanging subject or else would be incapable of seeing a subject in a picture, i.e., would see the picture only as a surface. Fn13.

Reflection on this set of phenomena should lead us to inquire into the full menu of ocular and optical peculiarities of picture-viewing compared with those of ordinary perception. Accordingly, as naively reductive as it may seem to be, let us review and draw the consequences of the well-known facts that distinguish pictorial viewing from face-to-face perceptual viewing, a distinction Patrick Maynard christened “Gombrich's Gulf.” (Maynard 2005, 103). I limit the inquiry to perspectival works merely to fit within the allowable space.

            First (D1) our visual experience, binocular or monocular, of pictorial subjects lacks the phenomenon of selective clarity and blur that depends on the accommodation of the eye to distance.

            Second (D2) the experience lacks the binocular accommodation for distance and therefore any double-images of objects viewed out of focus.

            Third (D3) no parallax occurs within depicted scenes with change of the viewer's position: no seeing around objects or sliding of near objects in front of far ones as we move.

            Fourth (D4) as a result of this fixity when we move our position the visually ostensible subject deforms in response to the changes of the stimulus in the ways just remarked on in connection with the Michelangelo drawing, an effect which is more readily noticed in architectural pictures, especially big ones where the deformations are impossible for even a half-attentive observer to miss. Perceptual constancy is then no match for the power of the stimulus. These deformations contrast with the revelations of hidden surfaces produced by changed points of view in face-to-face perception. Far from random, the deformations obey strict principles of “invariance” (Gibson 1950, 152-4; 1979, 310-311).

            Fifth (D5) as a result of the foregoing ocular and optical constraints pictorial space is discontinuous with real space and is perceived as such whenever the viewer moves about in the normal area of viewing. Fixed depth of focus and fixed occlusion relations cannot join convincingly with their opposites, regardless of how forcefully the image may “intrude” into the viewer's space. Intrusions are seen as extensions of the pictorial space but not as ones that can merge with ours – when they are seen as pictorial, that is. Likewise our ability to establish a common metric for ambient and depicted space, so as to measure how far behind the picture plane a depicted object (fictively) stands, does not negate the visible break between the two spaces. Nor is the discontinuity dissolved by the fact that we can imagine ourselves stepping into pictorial space. The experienced structure of the two worlds remains incommensurable. Fn14.

            D6: Related but not identical to D1 is the absence of the reduced acuity of perception of things far in the distance, compared with nearer things.

            Finally, D7: Picture-viewing yields reduced illumination compared with face-to-face viewing of counterparts. All these features are properties of the mode of seeing-in as opposed to properties of the subject. Fn15.

Collectively D1-D6 tell much about how seeing depth in pictures differs from seeing it face-to-face. But it is necessary also to acknowledge that the total scenic depth effect (as opposed to more local ones) in pictures varies greatly. It varies with the properties of the picture, the circumstances of viewing, and the address of the viewer. Pictures with strong and consistent depth cues, including chiaroscuro and atmospheric gradients and well-managed color gradients are good depth-carriers, especially when the depicted depth is considerable. Surface texture that calls attention to the picture's surface weakens the depth effect. The size of the picture often matters. Small pictures are in general less potent than large ones, often because optically normal viewers cannot get a well-focused eyeful of the pictorial subject from close enough to the surface. Large ones helpfully fill the viewer's visual field, weakening the saliency of ambient space. Relevant viewing circumstances include how the picture is hung and illuminated. Pictures are often hung too high or too low for viewers to view them from the right height. The way they are illuminated can strengthen or diminish the depth effect. Depth flourishes when the picture surface is brightly but non-reflectively lighted and the ambient illumination in the room is kept low. Finally, as to manner of viewing, it is widely known that monocular viewing greatly intensifies the depth effect. But also a manner that can be called looking for depth is important. It consists in part of following out the sequence of planes, noticing near and far, keeping tabs of how things look in the periphery as one changes one's direct gaze, and in general ‘thinking depth.' Also relevant is noticing how the pictorial space squirms as one moves from canonical to noncanonical distances and lateral (or vertical) positions and how it resumes its proper relations when one reverts to the canonical point of view. Fn16.

I believe that anyone will obtain an impression of pictorial space being shallower than that in a comparable face-to-face scene. That is, it appears shallower than it should be judging by the projective indications of depth within the picture. Systematic vision research supports this claim. Fn17. I believe this deficiency results from D1-2 (virtual fixity of depth of focus and absence of binocular doubling). A related phenomenon is the tendency of farther planes to drift forward in picture space, a phenomenon found in face-to-face viewing only under exceptional conditions of deceptive alignment.   The result for separation is that we see a shallower depth in the picture than the perspectival indications allot to the real subject, even from the optimal point of view under optimal viewing conditions.

Gombrich Fn18. and others have spoken of pictures presenting viewers with relatively impoverished information, due to the fact that D1-2 and D4 arrest the flow of information available in face-to-face perception. In one respect this is true, since for one thing hidden aspects of the perceptual subject have no visible counterpart in pictorial seeing from any viewing station because the source contains none. But in another respect the remark misrepresents the situation. To every bit of the flow of information in face-to-face perception corresponds a part of the stream of information in seeing what I have called the apparent subject. As we change our position, our experience of it changes, only differently from what happens in face-to-face perception. A different flow is produced, obeying a different principle of invariance, one that is anomalous by real world standards but precisely determined by the perceptual situation. And notice, this is not the same stream as is provided by the surface of the picture, seen as such, i.e., by Wollheim's configurational aspect. That stream contains no three-dimensional compressions, expansions or squirm. In contrast, in seeing these deformations in (not of ) the design our experience represents them as having hidden aspects comparably deformed.

An interesting consequence of D1 concerns allegations concerning the visual field or outline or sensational shape that plays a central role in the theories of depiction of Budd, Hopkins and Peacocke. Each theory describes the shapes (let me lump them together under the rubric of stimulus shapes, with which each has a basic relation) as two-dimensional. Fn19. Each theory takes off from the reasonable idea that the stimulus shape is a common or at least resemblant aspect relating pictures to the three-dimensional views they represent. What D1 implies is that our experience of the two is importantly different. The stimulus shape abstracted from the ambient visual array in face-to-face perception may be two-dimensional, but it is not accessible in sharp focus at a single depth, or more precisely at the narrowly restricted range of depths that applies when we see the surface of a picture. Our focus has to leap out and back following the varying depth of the surfaces that generate the stimulus. This difference in ocular reception of the two is basic to the difference between the pictorial and the perceptual effect. Moreover, it implies that neither stimulus is as a whole two-dimensional but variously extended in depth, each according to the depth of the parts of the distal source. Whether this difference in stimuli of picture-viewing as opposed to face-to-face viewing contradicts the theories of depiction that use the concepts of visual field or outline shape is a question I prefer to dodge. To take up that question for the three theories would lead too far afield for present purposes. What is sure is that the difference of stimulus helps explain the difference between what the two sorts of visual experience are like, even apart from resemblance of shape. For note, no manner of perception of which eyes like ours are capable could eliminate this difference, given the difference of stimulus.

How then is it that by looking at the surface of a picture we can acquire (under optimal viewing conditions) a quasi-illusory experience in spite of the fact that there can be neither stereopsis nor any sliding of focus in or out relative to the visually apparent depth. An obvious possibility is that what greases the skids is the phenomenal resemblance between out-of-focus and peripheral blurring. When one's focus lights upon a part of a picture, say the knee of Michelangelo's ignudo, the thigh is somewhat blurry, aping the blurring that would occur if it were actually recessive, as in a sculptural equivalent. Given this resemblance it cannot be surprising that our recognitional capacity for the 3D aspect is activated, though its functioning in pictures is in a clear respect non-veridical. The result is a variably strong or weak misimpression of actual depth. At the same time our experience provides us with a vital contrary impression, an awareness of the first impression's non-veridicality, stemming from equally apparent indications that stir our flatness-recognitional capacities. As a result the object seen in the picture is seen as extended in a frail and insubstantial – unreal, merely virtual, teasing – depth.

Intimately implicated in this effect is our well-grounded confidence that moving our gaze to the thigh will create a comparable effect on the knee, resembling the experience of an ocular shift in perception of an actual knee. Given this confidence it is hardly surprising that the apparent subject is (i.e., seems) convincingly, but unstably, extended in depth, though none of this insubstantiality is (normally) ascribed to the real subject. Thought, or make-believe, or both intervene to confer true substantiality upon that.

Spatial separation seeing-in invites the following thought experiment, a counterpart to the one proposed regarding the separation-rich drawings. The experiment is to imagine visually experiencing the apparent subject seen in as fully naturalistic a picture as possible, one that furthermore has substantial pictorial depth (e.g. David's Napoleon in his Study , National Gallery, Washington), and doing this in as robust a way as possible; and then imagining that pictorial experience replaced by a face-to-face perceptual experience of as close a counterpart as possible. Imagine being able to adjust the face-to-face subject, tweak atmosphere and lighting and whatever else you choose, and switch back and forth between the two to check on the remaining disparity – all in the cause of making them look as much alike as possible. I predict that the ocular and optical features of pictorial seeing will cause you to experience a jolt as you switch back and forth even after “perfecting” the resemblance. The two presences would strike you as very different, even as you recognize that the picture is as naturalistic a depiction of that face-to-face subject as you can envisage.

The point of this is to bring out the reason why pictorial seeing-in must be different from face-to-face seeing. The ocular and optical peculiarities of the former are such as cannot, in principle, be found in the latter. This by itself suffices to insure one sort of incommensurability between pictorial seeing-in and seeing face-to-face, whether or not it is the same as the one Wollheim had in mind. Fn20. At the same time the peculiarities are such as dilute the resemblance between the spatial relations in the real subject and those we immediately see in the design. Equally, when we make believe our seeing of the surface is a seeing of the (real) subject, for us to include these properties would defeat the purpose.


Spatial separation seeing-in at the most basic level is afforded by even the most naturalistic picture. This confirms the ubiquity thesis, since what we see-in the design differs anomalously from face-to-face seeing of the subject both from given viewing stations and in moving from one station to another. Pictorial space is necessarily less robustly extended in depth, even from a single point of view, than the spatiality belonging to the subject. Though there is no way to overcome this relative debility, being a consequence of the only visual processes of which we are capable, it can be turned to good use. Even as it is, untweaked, it offers a more unified field. Artistic interventions often involve enhancement of the debility, that is, by further weakening the depth effect, accentuating shallowness, putting more stress on planarity, principally by appropriate use of the resources of facture. Fn21.  Matisse's The Egyptian Curtain discussed below is a paradigm case. In traditional modes the interventions more often aimed at compensating for the basic deficiency of depth, which gives rise to a different range of separation-seeing. Fn22.

As the apparent or ostensible subject (or if one prefers, the ensemble of properties seen-in the picture) passes through states of expansion, contraction, and contortion with the viewer's change of position, in almost all cases one state best accords with the real subject and therefore serves our interest in seeing what the pictures depicts. Call this the canonical state of what is immediately seen-in the picture. By this I do not mean that other states are to be avoided. Often they cannot be if we want to enjoy the picture at all, for instance when it is too small or placed too high. And much sensitive and insightful appreciation is possible from other viewing stations: ones too close or too far, or laterally or vertically eccentric. Fn23.  That this is so is sometimes mistakenly taken to show that our seeing-in is slavishly governed by “perceptual constancy.” Aside from that, however, it is striking how few viewers, even quite competent ones, take the trouble to view Rembrandt's drawing of Sylvius from below, to the left, and close up, as is clearly indicated by the rendering of the aperture framing the subject, or to see that aperture in its true circular form. Still, giving such credit as is due to non-canonical viewing should not blind us to the fact that something is lost when the canonical viewing station is neglected. Fn24.,Fn25 Non-canonical viewing prompts separation-seeing graduated in proportion to the eccentricity of the point of view.

As to artistic significance, two facts are worth noting. First the canonical state of what is seen in the picture is subject to the artist's choice of implied point of view, or in cases where parallel projection instead of perspective is used, of implied angle of view. Fn26. The variations available to the artist who uses perspective are lateral, vertical and in depth. Thus in some Dutch architectural painting the canonical point of view may be more to the side than is usual. Fn27. When such a picture is viewed head-on from the usual height and distance, the state of the apparent subject is decidedly non-canonical. But that implies no less canonicity, and also no separation beyond the basic. The situation is different for extreme anamorphs: in spite of the fact that the intended subject is better seen from the acute point of view than from any other in point of outline shape, the subject itself is not seen nearly so well as the same subject treated conventionally and viewed canonically. The disparity of focus between near and far parts of the anamorph is too great to give us optimal access to the subject. Thus what a viewer sees-in the anamorph exhibits spatial separation beyond the basic.

Canonical states may also differ for different parts of a picture. There may be more than one implied horizon (i.e., implied viewing level), not at all an infrequent occurrence. Then a given part of the subject is presented non-canonically from the point of view that is canonical for another part. Much good can come from these artistic choices, which imply not just eccentricity but inconsistency. Fn28.  When the picture is spatially consistent, full recognition and enjoyment of its artistry as a depiction depends on giving priority to seeing its subject canonically. In cases of inconsistency the story is more complicated: full appreciation of its artistry requires seeing each of the separate parts of the subject canonically at some point or other, as well as developing an awareness of the contribution they make to the whole. Fn29. That said, it is often difficult to determine the artistic benefit or detriment of inconsistent spatiality, as it also often is in the case of facture-separation.

To illustrate spatial inconsistency we need look no farther than Rembrandt's etching of Sylvius. Fn30.

Before detailing the inconsistencies, it is worth observing that the etching differs from the drawing in many ways. One concerns the mood conferred on Sylvius: he is shown as less energetic, more tentative than in the drawing. His extended hand, now the right one, is positioned well to the side of his face, not directly in line with it, and the gesture is concessive or pleading rather than assertive. The other elbow rests at his side with the hand on the surface of a desk or ledge instead of being cocked up and the hand more likely on his hip, as it is in the drawing. His open eyes, looking just barely to the left of us, are plaintive instead of being lowered or closed in concentration or resoluteness. His brow is furrowed and pinched instead of smooth, and the set of his mouth suggests regret or anxiety rather than the self-possession that comes across so well in the drawing. All of this fits the commemorative verse which appears below: “Exhausted by old age he has himself instructed old men./A lover of moderation he spurned pretense…/This was his principle: Jesus could be taught more properly through/ An improved [way of] life than by thundering speech.” Fn31. And of course the etching is far more determinate than the drawing. Many details of face, clothing and backdrop are far more specific, and none of the radical brush mark anomalies appear. In general what we can see in the picture is far more like the depiction subject than what can be seen in the drawing of its subject. The range of separation-seeing is narrower.

The inconsistency of point of view is immediately apparent to anyone practiced in judging projection schemes. The ledge or desk inside the opening is shown from above. This conflicts with the upward view of the opening itself, the flat edge of which drops out of sight at the bottom. What the latter does the ledge or desk should also do, or vice versa. Yet there it stands unoccluded, seen from above. This part of the scene is much confused by the loose brushwork of the drawing. The flurry of strokes near the bottom of the opening defines nothing in particular in the drawing's subject.

The consequence of the two points of view is that the space represented is not one we can encounter in face-to-face perception of normal subjects. It is true that with a skewed Ames-like mockup Fn32. the anomalous point of view appearances in the etching might be reproduced in face-to-face seeing from a selected point of view. And the prompt for trying to see that mockup in the picture might be the inconsistency I have cited. This is conceivable, but it is also so narrowly restrictive that it is hard to see how an artistic tradition could tolerate it. Any artistic tradition that imposed such a practice, requiring that point of view inconsistencies depict Ames-like constructions, would be cramped and impoverished compared with ours in respect of human relevance. Besides, in many cases substantial aesthetic advantages accrue to point of view inconsistencies, given the subjects that we wish to represent and the human capacity to overlook inconsistent points of view. What is relevant to my separation thesis is that this toleration of certain sorts of inconsistency reinforces the importance of spatial separation. Pictorial seeing as we know it depends importantly not just on basic spatial separation but, in due measure, more specifically on separation inconsistencies.

A second spatial inconsistency, this one concerning illumination, is also evident. The location of the light source of the shadow the head throws on the edge of the opening is represented as very far off, presumably the sun since only that would throw so focused a same-size shadow on a perpendicular surface. But this poses a problem: the source must be inside the room since Sylvius' head is not shown extending outside of the opening far enough for an external source to throw the shadow on the rim of the opening. Worse, the shadow of the hand is yet more anomalous, whatever light source one may postulate. No light shining across the opening on the inside could explain a shadow on the outside wall. A second source is required, but none can come close to explaining the form of the shadow of Sylvius's hand on the vertical face of the wall or partition. The shadows of the fingers are spread much too widely and the whole shadow would have to extend diagonally far down the vertical wall. Any such source would also completely wash out the first shadow.

Ideally I would now set forth the artistic advantage, if such there be, that Rembrandt gains by the inconsistencies I have just pointed out, citing authorities or concocting an account of my own. But the authorities are silent and the task of demonstrating the plausibility of any intuitive account is formidable. Ultimately it requires showing the result of the several ways in which the inconsistencies can be removed with minimal disturbance of the rest, thereby to gauge the effect on the whole. No such practice has been taken up by art historians, perhaps not surprisingly since the task is an exacting one requiring artistic skill and judgment. Also there is no assurance of how fruitful the practice would be in producing explanations favorable to the works. In the absence of explanations we are left with only a sense of the work's spatial waywardness that may (or may not) contribute to the value of the whole.   Perhaps there is a kind of allure in that, the allure of undisclosed depths. What is certain is that the spatial inconsistencies create uncertainty about what properties are ascribable to the subject (taking the subject to be the entire scene).


The question of artistic benefit or detriment arises across the board for all separation effects. Often there are reasonable explanations of benefit accruing from one or other instance of separation. Consider the questions generated by Matisse's The Egyptian Curtain. The curtain is seen as falling in front of the table, which is unusual, since normally curtains are next to the wall. But probably we should count that an oddity of arrangement, hence a property of the intended subject and not a case of separation. The leaf-shapes in the fabric design standing out against the untextured black expanse representing the curtain Fn33. are another matter. We know that they, being mere patterns on the fabric, follow the folds of the curtain. How could they not? But this implication of the subject does not constrain our immediate seeing of the leaves. Their contours twist and turn in space according to a subtle conspiracy of design indications. Thus the green leaf can be seen to curve along the rounding of the curtain, as it should, but also with equal force to extend out from the curtain like a scimitar. In so doing it links up with the pattern of   protrusive palm fronds, turning a broad, downward arch into a jauntily upturned, aggressively protrusive blade. Other anomalies are resident in the pattern, comparable to the scimitar, giving the picture a richly sportive suggestiveness. That same green leaf can also be seen (without forcing, without fault) as curling broadside to the viewer's gaze. Each of these gambits is immediately presented in depth, is seen depth-wise, as the eye plays with the pattern. This is abetted by the perceived flatness of the surface and painterly flattening of the curtain, without which that particular palimpsest of conflicting appearances would not be so salient. The conflicting readings replace one another as naturally as do the changing but entirely compatible visual impressions that occur as one explores an actual scene, and they too leave a residual presence behind. None can elbow the others off stage for long, and those upstaged merely retreat to the margins.

The Matisse example shows one way design complications can spawn a diversity of things seen in a picture to good effect. It is equally true that in many cases separation seeing counts against the efficacy of the depiction, as when colors that should stay in their proper depth refuse to do so, a complaint commonly levied in 17 th and 18 th century academies.


Pictures serve different purposes. A central use is that of visually conveying aspectual information about the subject, whether that subject be real or fictive. What information is conveyed will depend on the system of depiction, which Lopes and others plausibly take to consist of the aspectual commitments and non-commitments the pictures reflect regarding different sorts of properties. (Lopes 1996, 128) He points out that different systems select different ranges of content. They also vary in how detailed and multifarious is the content they convey. “Information” covers both real and make-believe or “groundless” information, the latter applying to fictive subjects. This aspect recognition theory, like its resemblance or seeing-in rivals, is dedicated to basic pictorial content, and in doing so it is faithful to the primordial function of pictures. That is, pictures are made in the first instance to convey information about actual things, real or prospective. Make me a picture of X, someone says. A draftsman complies. It is understood what sort of thing or scene is to be depicted. Methods of depiction arise in the first instance in that context. What counts and what does not depends on what works well enough for the purpose at hand. In this enterprise knowledge of the real world is an essential formative control or reference. Nothing else could possibly have as deep a connection to the pictorial or any other mode of communication. From such control and the selectivity it enforces, comes the normal pictorial subject.

But our capacity to see other things in pictures than the subject raises questions not answered by the normal systemic commitments. That the design properties of pictures can and should be seen in ways that disregard the normal aspectual commitments and non-commitments calls for an explanation. Viewers sensitized to depictive systems with a rich and subtle range of content, habituated to an artistic culture that retains the memory of many modes of representation, many styles, many expressions of artistic personality, and artistic self-consciousness, a culture that cultivates creative diversity, seek to find more in pictures than is an authorized part of the subject, and do so appropriately. What seems at first sight to be waywardness or idiosyncrasy turns out to be a consequence of advanced pictorial understanding. It arises primarily from the interest in the full and varying impact of graphic (or painterly) marks on our pictorial experience. It asks, what does this ensemble of marks suggest, of all the forms in my repertory of depictables? Thus design features are taken as a provocation to search out the various depictive suggestions created by the design, to test their strength, their consistency or inconsistency, using criteria based on elements of the very system that generated the depictive subject. Once one can see that subject one can see farther, “eccentrically” and yet with artistic-aesthetic relevance. This use of pictures seems no less valid, indeed just as compelling, for viewers and artists alike, than is adherence to the real subject. Strict limitation to the constraints of the depiction subject seems unacceptably restrictive when we find ourselves deeply absorbed in the visual experience of the picture itself.

The provocation depends greatly on the type of picture. For many pictures in which separation is particularly strong, this free-wheeling approach finds itself dealing with things uniquely accessible by way of separation-seeing. Not only that. The world of the actual subject becomes less dominant. It tends to hide behind or within the extravagant surface features. Or it recedes into bland indeterminacy and loses its allure. The restraints on pictorial subjects cited by Hopkins and listed above   (Section 1) seem inadequate to account for the interest we take in such pictures. More and more interest gravitates toward suggestions spawned by cryptic texture and problematic spatiality. The main point of the picture seems to be to present something much stranger (more lyrical, more surreal) than any normal, rule-compliant (even when fictional) world. At the same time these flights presuppose the normal information-transmitting pictorial practice as an underpinning, as poetry presupposes a prose paraphrase.

The case for this drift of pictorial seeing beyond the normal is easiest to make for modernist pictorial representations, for all varieties of cubism as well as less structured sorts of naturalism, Matisse and such. Consider again the latter's The Egyptian Curtain . Does the artist propose that we interpret the painting simply in terms of his hotel room (in Nice, perhaps) and the view outside, which is almost certainly the original setting, his starting point? Anyone who thinks so is invited to imagine sitting as the artist presumably did in that room looking through the large mullioned window at the crown of a palm tree in the blazing sun set off by a curtain close up on the right and a table with a bowl of pomegranates between window and curtain and about level with the bottom of the window. If one imagines this, working from the painting, and tries to see it as being what the painting means to engage our interest, one has done all one can to normalize the depiction subject in scenic terms. Does that do justice to the subject of this painting? I think not. When I try the experiment discrepancies leap out at me. No experience of a normal palm tree, normal table covering, normal pomegranates, normal curtain and wall surfaces can match up. The dissonance between that commonplace subject and the scene given to us by the painting is unresolvable. Can we take the given scene to disclose the artist's experience of the real world scene? But how could Matisse or anyone else experience the tabletop as having, or looking as if it has, that cushy texture, or the palm fronds as being, or looking as if they were, that stubby-leaved? The same I would say applies to Picasso's painting of Françoise Gilot in Femme Fleur (Lopes 1996, p. 95) and indeed every painting in that artist's mature oeuvre. Certainly the conventional remarks we find in the literature do not meet the need: for instance that Picasso's Vollard portrays the man from multiple viewpoints which are all related in the cubist manner (Lopes 1996, 126) or even the moderately gothic description of the same artist's Girl with a Violin as being “the shattered body of a woman” (Podro 1998, 171). These conventional descriptions hardly touch the fascinating strangeness of the images. Fn34.

The same result accrues from reflection on games of make-believe using the painting as a prop. Kendall Walton's admirably developed theory provides a plausible account of what I have called the real subject (which of course need not be realistic – fantasy subjects are just as real). His account suggests three ways to deal with the provocations that lead me to find non-normality dominant in the works under discussion. The first is “to defuse [the] paradoxes by disallowing the fictional truths responsible for them.” (1990, 179-180). That way leaves only the fictional truths unproblematically applicable to the subject. The second gambit is “to declare offending fictional truths deemphasized, rather than disallowing them.” (180) This way counsels us that such fictional truths as are applicable to the anomalous apparent subject are to be allowed but “not to be dwelt on or even noticed.” (182) The third gambit is “to accept and even emphasize fictional truths that clash with one another, but to mute the clash by disallowing the fictionality of their conjunction.” (182, 239) The anomalies are recognized and appreciated along with the real subject. Walton's pictorial examples include ones in which the subject is overtly paradoxical, as in Escher's Ascending and Descending (1960) or Hogarth's False Perspective and ones in which an entirely normal subject is presented in a manner that generates separation, e.g., Matisse's The Red Studio. The three ways of dealing with anomalies are recommended variously for different works and different aspects within a given work.

For separation cases the first gambit would take the hatching, impasto and scumbling, the two-way chair (in The Red Studio ), the gestural brush strokes and the like, to convey normal shadows, smooth surfaces, clear atmospheres, regulation chairs, all presented in a certain style. What is explicitly or implicitly depicted conforms to the reality and mutual belief principles. (144 ff.) The second gambit would marginalize separation-features, treating them as unimportant and not to be inquired into, on pain of one's inquiry being judged silly). The third is open to acknowledging the efficacy of the anomalies in conveying the expressiveness ascribed to the subject or to the artist's painting the picture that way rather than to any visible feature of the depicted scene. This might apply to the stubby strokes that render the palm fronds, the cushy pink texture of the table, the unruly behavior of the leaf pattern on the curtain, and much more. All three gambits what Matisse's picture authorizes us to make-believe systematically regularizes the scene (nothing special, strange). So far as the subject (the fiction) goes one is authorized to imagine oneself in Matisse's room taking in the sun-filled palm tree, the curtain, table and pomegranates.

I understand that from a certain point of view it is right and proper to disallow or marginalize the anomalies by sensitively responding to the information such a picture offers as to the normal subject. Doing so in this case respects the process of making the picture, since the artist certainly sat in his studio viewing the palm tree through the window, felt what he felt and painted accordingly wielding his brush as he did. Normal traffic in Nice went whirring by off in the distance, normal bodily intrusions of the household into the painter's space came and went. The picture plausibly fictionalizes some suitably generalized version of this background but keeps it in the background. To this may be added expressiveness: the artist plausibly produced the picture to celebrate the joy of the sun on the palm and the unruly energies that the pattern on the curtain stimulated in him. The real Nice, or some place similar, is celebrated in a particular way. So let the viewer make-believe her seeing the picture is seeing this selectively determinate-indeterminate real-world scene suffused by celebration (radiance, calm, delight).

I need not deny any of that. My thesis is merely that this is not all there is to this picture. This by itself does not do justice to its artistic presence and force. The viewer who cares about the particular textures, marks and design, the particular reticences and forthcomings, must find herself confronted by something else that arises more immediately from the design properties, something that sets aside the real world background, and that is dominated by the image in the artist's eyes as he contemplates what he has made, as he lives in that. That apparition is no part of the world of the normal subject. It may not need any readymade context of real or fantasy world clutter. It often seems to resist any set of determinations sufficient to constitute a possible world, however indeterminate and distantly backgrounded the non-focal surplus may be. And however much expressiveness may be conveyed, the means of such conveyance remain standing as visible features within that apparition. They insist upon characterizing something visible, however anomalous. This response is a game of make-believe, but a different one that the picture, as a work of art, invites us to put in primary position, as its special gift. Fn35.


Of course not all art-grade pictures with significant pictorial depth suggest such non-normal subjects. David's Napoleon in his Study Fn36. does not suggest any but the normal one, so far as I can make out. If any such can be discerned in the Ghirlandaio, whose deviation is mild, it is much closer to the normal one. The sort and degree of strangeness can be determined only by appreciative absorption of what the picture has to offer. I contend only that many, many pictures fall within the category of the non-normal seizing the limelight. Popular images are not excepted. Many comics are included (but not Prince Hal) as well as most animated cartoons and all anime. Early Renaissance paintings abound in separation that produces a problematic set of properties. Giotto's Arena Chapel Flight into Egypt gives us papier mâche mountains that do not match any real or fictive

topography. This is no fault. To the contrary their anomalousness has a positive effect on the work which any naturalistically rendered mountains (or hills) would ruin. Fn37.

Accepting the prominent role of non-normal as-if subjects in pictorial representations makes the works more fascinating and thus explains more persuasively their appeal. It enables the viewer to forge a tighter bond between surface and subject in cases where the normal subject is rendered banal by indeterminacy. The normal subject never entirely loses its grip, because only it can supply the default suppositions that make it possible to speak of a fictive world . The non-normal one generally has insufficient defaults to constitute a world. It remains too narrowly, too exclusively virtual or dreamlike to be a proper world. It is a matter of visions, thoughts and feelings that haunt our ambitious engagement with pictures.

Websites for the Poussin, the Matisse Portrait of Mme Matisse, and the Parmigianino.


Budd, Malcolm, 1992, “On Looking at a Picture,” in Hopkins and Savile, eds. Psychoanalysis, Mind and Art, 259-280. ( Oxford : Oxford University Press).

____________, 1993. “How Pictures Look,” in Knowles and Skorupski, eds. Virtue and Taste ( Oxford : Oxford University Press).

____________, 1995. Values of Art Pictures, Poetry and Music , ( London : Penguin).      

Dubery, Fred and Willats, John 1983 . Perspective and other drawing systems. (London : Herbert Press). 

Susan Feagin, 2001. “Presentation or Representation” in the Journal of Aesthetics and Art Criticism , 56: 234-240, reprinted in Richard Wollheim on the Art of Painting: art as representation and expression , ed. Rob van Gerwen, ( Cambridge : Cambridge University Press, 2001).

Gibson, J.J.1950. Perception of the Visual World . ( Westport Ct. : Greenwood Press).

_________ 1979. The Ecological Approach to Visual Perception. ( Boston : Haughton Mifflin).

Gombrich, Ernst, 1960/1988. Art and Illusion: A Study in the Psychology of Pictorial Representation . ( Oxford : Oxford University Press).

_____________ 1973, “Illusion and Art,” in Gregory and Gombrich 1973.

R.L. Gregory and E.H. Gombrich 1973. Illusion in Nature and Art . ( London : Gerald Duckworth.& Co., Ltd).

Hagen, Margaret A. 1978. (with Rochelle Glick and Barbara Morse) “Role of two-dimensional     Surface Characteristics in Pictorial Depth Perception” in Perceptual and Motor Skills 46, pp. 875-881.

________________, 1986. Varieties of Realism: Geometries of Pictorial Representation . ( Cambridge : Cambridge University Press).

Hockney, David 2001. Secret Knowledge: rediscovering the lost techniques of the old masters. ( New York : Viking Studio).

Hopkins, Robert 1997.   Review of Lopes 1996, British Journal of Aesthetics, 37, 3.

_____________ 1998. Picture, Image and Experience ( Cambridge : Cambridge University Press).

_____________2003a. “What makes representational painting truly visual?” Proceedings of the            Aristotelian Society Supplementary Volume 77 , 149-167.

_____________2003b. “Pictures, phenomenology and cognitive science,” Monist 86, 653-75.

Hyman, John 2006. The objective eye: color, form, and reality in the theory of art. ( Chicago and London : The University of Chicago Press).

Koenderink, Jan J. 1998. “Pictorial relief,” Philosophical Transactions of the Royal Society, London , Series A , 356, 1071-1086.

Kubovy, Michael 1986. The psychology of perspective and Renaissance art . Cambridge : Cambridge University Press.

Levinson, Jerrold 1998/2001. “Wollheim on Pictorial Representations,” Journal of Aesthetics and Art Criticism, 56: 226-233; reprinted in Rob van Gerwen, ed., Richard Wollheim on the Art of Painting: art as representation and expression. (Cambridge : Cambridge University Press).

Lopes, Dominic 1992. “Pictures, Styles, and Purposes,” British Journal of Aesthetics 32.

_____________ 1995. “Pictorial Realism,” Journal of Aesthetics and Art Criticism 53.

_____________ 1996. Understanding Pictures ( Oxford : Oxford University Press).

_____________2005. Sight and Sensibility   ( Oxford : The Clarendon Press)

Maynard, Patrick 1994. “Seeing Double,” Journal of Aesthetics and Art Criticism 52: 2.

______________ 1996. “Perspective's Places,” Journal of Aesthetics and Art Criticism 54: 1.

______________2005. Drawing Distinctions: the varieties of graphic expression. Ithaca , N.Y. : Cornell University Press.

Peacocke, Christopher 1983. Sense and Content ( Oxford : Oxford University Press).

__________________ 1987. “Depiction,” Philosophical Review xcvi, 3 (July 1987).

Penrose, Ronald 1973. “In Praise of Illusion,” in Gregory and Gombrich 1973.

Podro, Michael, 1991. “Depiction and the Golden Calf,” in Normal Byson et al, Visual theory: painting and interpretation . ( New York : IconEditions/HarperCollins Publisher). Originally published in Andrew Harrison, ed., Philosophy and the visual arts: Proceedings of the Royal Institute of Philosophy , ( Dordrecht , Holland : D. Reidel Publishing Co.), 1987.

_____________ 1998 . Depiction . ( New Haven and London : Yale University Press).

Shier, Flint , Deeper into Pictures , 1986. ( Cambridge : Cambridge University Press).

Van Gerwen, Rob 2001. Richard Wollheim on the Art of Painting: art as representation and expression , ed. Rob van Gerwen, ( Cambridge : Cambridge University Press, 2001).

Walton, Kendall, 1990. Mimesis as Make-Believe . ( Cambridge : Harvard University Press).

____________, 1987. 'Looking at Pictures and Looking at Things,' in Andrew Harrison (ed.),    Philosophy and the Visual Arts . ( The Hague : D. Reidel).

Strauss, W. and van der Meule( New York : Abaris Books).

Wollheim, Richard 1980. Art and Its Objects. ( Cambridge : Cambridge University Press.)

_______________ 1987. Painting as an Art (Princeton: Princeton University Press).

________________2001. “On Pictorial Representation.” See Van Gerwen 2001.

________________2003. “What Makes Representational Painting Truly Visual?” Proceedings of the Aristotelian Society , Supplementary Volume 77.

Wright, Lawrence 1883. Perspective in Perspective. ( London : Routledge & Kegan Paul).

White, Chrisopher 1999. Rembrandt as an etcher: a study of the artist at work. 2 nd edition. ( New Haven and London : Yale University Press).


1. Note how extreme this is, if we take “as determinate as she sees the drawing as having” strictly. It appears to imply that every surface mark seen by the viewer is seen in terms of a property of the anomalous “man” seen in the drawing. A more reasonable necessary condition for separation is that some marks that are not subject-indicating are seen as if they were subject-indicating. There are obviously degrees of separation. The Ghirlandaio example discussed below is clearly a mild case. Perhaps Hopkins would not consider it one at all, since another strand in his view is that separation occurs when the resemblance between the design and the outline shape ascribed to the subject is insufficient to support the viewer's seeing the subject in the design, as in a stick figure that depicts a human. In my view this resemblance condition does not obtain in the case of the Ghirlandaio.

2. From what Hopkins says about the “marriage” option (127-8) it seems possible that in some cases knowledge of the depiction practice may lead a person to reconsider her seeing-in experience and cease to see anything deviant in the design.

3. Other “process” cues signify second thought or correction , preliminary blocking out , etc. When these features have aesthetic properties they may confer them on the subject, as do works in Walton's “Sloppy Style.” (Walton 1990, 317-19)

4. The only transparency cue offered is the slightly cloudy background of the entire picture. If we see this as visible atmosphere then the upper thigh is fully (not semi-) transparent. But this is trumped by the evident incompleteness-cues. Still, I do not deny that one can force oneself to see emptiness even without invoking the cloudy background. Another complication is the sheer   incoherence of parts of the drawing, for instance the marking of the shadow in the left upper arm and the bristly-bushy treatment of the chest. These properties do not offer openings for us to see anything in the specific textures.

5. Any marking or other handling of the surface with depictive intent counts as “facture” as I am using the term, whatever implements and materials are used.. It contrasts with what was once called “invention,” namely the arranging of characters and décor in a depicted scene. Both are design features.

6. Domenico Ghirlandaio, Portrait of Giovanna Tornabuoni , 1488. tempera on panel, 77 x 49 cm. Museo Thyssen-Bornemisza , Madrid . The image is available at screen-filling scale on the following website:

7. Vasari 1965, “Preface to Part Three,” p. 251: “the hard, dry, harsh style that art had acquired through the excessive study.” Vasari's list of these artists includes Domenico Ghirlandaio.

8. Vasari   1965, 266, 253.

9. Correggio's Mystical marriage of St. Catherine in the Louvre serves admirably as an example for both skin and hair. A good image is available at: http :// . Click on the image to enlarge.

10. It is instructive to inquire whether it is in principle possible to depict, within any known depictive tradition, the apparent (phenomenal) subject of the Rembrandt drawing. Even when one allows for feasible extensions of the known traditions, the answer, I think, is negative.

11. A good image of this can be found by Googling “Bilddatenbank Albertina.” Once on the site select Michelangelo . The relevant image is in row 2 of the first page, farthest to the right. Enlarge as desired.

12. Anyone in doubt as to the perceptual impact of these distortions is invited to walk back and forth in front of the oversized photographs of Jeff Wall. The architectural features of the interior scenes are particularly subject to distortion, as are landscape features such as roadways and ditches depicted as orthogonal to the picture plane. Viewed up close the space seen in the photographs becomes radically compressed in depth, and so forth. Viewers who miss these deformations are not attending closely to what is there to be seen in the works.

13. I am aware that this flies in the face of the alleged perceptual constancy of the pictorial effect (Kubovy 1986 calls it robustness). But the effect of constancy obtains only when pictures are viewed with less than scrupulous attention. Also constancy depends on clear indications of the orientation of the surface, such as the rhomboidal projection of its format to the viewer, almost always reinforced by other environmental markers, or on viewing from so great a distance that the alterations in our visual field of the format and environment are slight. Not all features are as robustly inconstant as others. I find, for example that viewing pictures on a slanted surface more readily results in an apparent shortening of the image than does the impression of parallel sides of the format yields to an apparent non-rectangularity.

14 The examples Podro 1998, p. 79, discusses from the works of Rembrandt, e.g., The Syndics , fall into line with my point. The relation of the depicted characters to us, the viewers, crosses a boundary between two worlds. We are seeing ourselves as standing where the two worlds overlap without there being any actual relation. If we were in the pictorial space here is where we would be, at the mercy of the syndics' cold regard. But we are here, in our space, into which that stare cannot enter, beneficiaries of a secure but robust vicariousness.

15. Reduced illumination might be thought a characteristic of the subject, but in fact we do not ascribe it to the notional subject. Hence it must be counted an attribute of the mode of presentation. This does not prevent its being absorbed into the visually ostensible subject as an attribute of it.

16. Gombrich 1973 includes a number of the factors mentioned here in his discussion of the pictorial illusion.

17. See Koenderink 1998. As nearly as I can tell, it is also confirmed by the comparatively simple experiments reported by Hagen et al 1978. Magritte's famous painting The Human Condition of 1933, showing a painting against a landscape which it exactly depicts, suggests a simple experimental setup for testing the comparative shallowness of pictorial as opposed to real depth.

18. Gombrich 1960, 38ff.

19. Hopkins 2003 defines “outline shape” of an object in terms “of the directions of its various parts from a point in its environment” (corresponding to the point of view), hence objectively; Budd 1993 defines “visual field shape” as the viewer's visual world from a point of view with depth removed, hence incorporating whatever properties the viewer ascribes to the scene that do not imply depth; Peacocke 1987 takes visual field shape to be in “visual sensational space.” Each philosopher endorses the planarity of the shape. Though Hopkins requires relationship to a point of view, he specifically holds that “distance is irrelevant.” (2003a, 152). Thus the connection with the stimulus even if not identity, is close. I use “stimulus” rather than Gibson's “ambient visual array” only because it is shorter and more familiar.

20. Wollheim 1987, 47.

21. The reverse effect of continuous recessiveness into an indefinite distance, as described by Heinrich Wöllflin in his classic Principles of Art History is achieved by uninterrupted textural gradients and unstressed contours inter alia . See esp. Chapter II, Plane and recession, 73-106.

22. The well known practices of enlarging distant features or diminishing the convergence in near scenes are instances. See the discussion of Caravaggio's Supper at Emmaus in Dubery and Willats 1983, 80.

23. A remarkable example of an interpretation that gives equal significance to the canonical and one particular non-canonical point of view is proposed by Patrick Maynard of Dürer's much discussed engraving, St. Jerome in his study , 1514 . Maynard 2005, 176-183. This interpretation attributes two different scenes, not just two views, to the picture.

24. Podro 1998, pp 71, 75, gives testimony confirming this when he observes, regarding the depicted gazes that set up a communicative relation with the viewer: “it is only at a certain viewing distance and momentarily that we can imagine ourselves in … an exchange of looks. When we go too close we no longer experience their eyes as coordinating their gaze on us and they seem to look behind us.” Just so.

25. For pictures in parallel projection (whole or partial) the canonical station is a movable set of positions for different parts of the scene. Hopkins often uses “perspective” and “point of view” for what might better be termed a direction from which the object is seen.

26. John Hyman (2006, 78) holds that only the line of sight is relevant, but this is clearly wrong, since objects curving into depth reveal less of their margins from close-up than from a distance along the same line of sight. Occlusion patterns among object at different depths are also altered by distance.

27. A good example of this is Saenredam's The Interior of the Grote Kerk at Haarlem, National Gallery, London, discussed by Dubery and Willats c.1983, pp. 87-88. An image is available in that text and on the National Gallery website. Select “Collection,” type “Saendredam” in the Search box. Click on “Image only” for an enlargement. The text on the website erroneously says the artist altered the perspective. What he did was to choose an up-close viewpoint close to the left side. To minimize the considerable distortion resulting from viewing the picture straightaway at the normal distance, the artist set its lateral termination in the middle of the first and last of the three columns. The ameliorating effect of this is dramatically illustrated when we see an expanded version of the scene, such as is given here.

28. Examples can be found on my website: .   For starters, see “Getting deeper into pictures via digital transformation: a tale of two horizons” and “The role of monocular viewing in pictorial appreciation.” David Hockney 2001 displays many other examples.

29. I believe that all I say here about the importance of the canonical point of view is consistent with the valid core of Patrick Maynard's animadversions regarding the many extravagant, confusing and false claims about the canonical point of view in the literature. (Maynard 1996) But I have many reservations about the details in his complaint.

30.Jan Cornelisz Sylvius (Benesch 280 I), Etching and drypoint; 27.8 x 18.8 cm. British Museum . London . See Christopher White 1999 for discussion and illustrations. Reproduced also in Podro 1998, 19.

31.The Rembrandt Documents 1646/8.

32. Gregory and Gombrich 1973, p. 68. Or Google “ Ames room” and select .

33. An image of the painting, sufficient for the points I want to make, may be accessed at:,_Henri/image/Interior_with_Egyptian_Curtain_1948.jpg.html&img=36&tt= Or Google “Henri Matisse - Egyptian Curtain” and select among the many sites.

34. Much the same must be said of the treatment of pictorial illusion in Penrose 1973, whose account of cubism, for instance, is a fount of the generalities that are now part of the conventional wisdom without contributing any sharp-edge understanding of any particular work. The only thing we can take from this is the phrase: “this intricate game of illusion and counter-illusion.”   (255). We have to hold in abeyance the claims concerning cubists exercising precise control or conducting probing analysis and rigorous reconstruction of objects, until and unless they are cashed out. Yet along the way he says various apt things (“every style adopts forms of illusion appropriate to its attitude toward reality” 246) and cites suggestive statements by artists (Delacroix: “those things which are most real are the illusions I create in my painting” 245).

35. In the 2003 Aristotelian Society exchange with Hopkins concerning seeing-in, Wollheim makes two claims bearing on anomalies in pictures (Wollheim 2003, 142ff.); one conflicts with my understanding of the work and the other leaves me uncertain of the author's meaning. The first is that we correctly see Parmigianino's Madonna with the Long Neck only when we do not see-in it the Madonna as long-necked. That feature, he says, belongs only to what he calls the Presentational how, not the Representational what and not even to the Representational how, since the latter “corresponds to a property of the what of representation, possessed either permanently or transiently, whereas the Presentational how does not qualify the what at all” (143) That is, it should affect only how the picture is perceived, not how the normally necked Madonna is taken to be represented. This strikes me as an implausible and uncalled-for playing down of the artist's eccentricity. The second concerns Matisse's Portrait of Mme Matisse (1905), which features a stroke of green on the face. This is explicitly said not to be a property of the Representational what: “he was not representing a woman who had a green line down her face.” But it is never explicitly said whether it might be part of the Representational how, thus representing a transient state of illumination, say. More generally we are told nothing specific about the significance of these features of the Presentational how in the cases at hand. In consequence I cannot make out how my view of separation-seeing relates to the author's view. Since Wollheim says the Presentational how “may reflect a range of things from the expressive vision of the artist, through the artistic pressures of the day,..,” what I am contending for may conceivably fall under it. Much hangs on whether the picture's reflecting the expressive vision of the artist can be explained without recourse to separation-seeing.

36. Googling “David – Napoleon in his Study” will bring you to the National Gallery, Washington, D.C. site and a good image. Click to enlarge.

37. See “Experiencing the picture, experiencing the subject” on my website for text and illustrations that make a case for this proposition.