The Realistic Angel : Pictorial Realism as Hypothetical Verity

: My main objective in this paper is to formulate a view of pictorial realism I call ‘hypothetical verity’. It owes much to John Kulvicki but diverges from his view in an important respect: rather than thinking that realistic pictures are true to our conceptions of things, I hold that they are true to what things would be like if they existed. In addition, I agree with Dominic Lopes that diﬀerent realisms reﬂect diﬀerent aspects of reality, but restate the case without recourse to symbol systems. Together, the twin principles of hypothetical verity and aspectival absolutism constitute a theory of realism able to account for realistic ﬁctional entities, the problem of revelatory realism and images that teach new information.

that is less subjectively relativistic than most versions currently on the table.It also remains true, I believe, to a common sense notion of the term 'realism'.
When discussing art, the word 'realistic' is used to mean different things.In Lopes' 1995 essay 'Pictorial Realism,' he disambiguates the notion of realism by distinguishing four uses of the term: (1) real rather than ideal subject matter, (2) depictions of actual rather than fictional events or things, (3) illusionistic production of the 'effect of the real,' and (4) pictorial realism-that is, realism in mode of representation or system of depiction. 1As the reader will soon appreciate, these senses are not in practice easily kept distinct.In general, however, the conversation about realism that this paper addresses concerns the fourth sense of the word.What is it about a picture that makes it count as an instance of realism?While remaining separate from questions about exactly how pictures depict, this investigation asks what it is that allows them to depict realistically.In this paper I respond to theories put forth by Lopes and, especially, Kulvicki.I call my view 'hypothetical verity' because it advances upon the verity view offered by Kulvicki, which he ties to perceptual conception.
The pioneering work of Nelson Goodman casts a long shadow over these debates.Goodman believed that a picture is realistic if it follows certain conventions set up in a system of artistic representation. 2There is nothing intrinsically correct, he believed, about the correspondence between any artistic marks and the object they are meant to represent.It all depends on the system of meaning within which the marks make sense.If the interpreter is familiar with that system, then the marks will mean to her what they mean to other users of that system: they will make sense.There are multiple such systems and none has any better claim to truth or objectivity than any other.
A Goodmanesque approach to realism is formulated by Lopes as follows: '(i) realism is relative to a symbol system, and (ii) pictures in standard systems are realistic'. 3In Goodman's case, the symbol systems are conventional.While agreeing with (i), Lopes rejects (ii). 4He rejects the second criterion because it fails to account for the fact that pictures from an unfamiliar system of representation can strike viewers as much more realistic than what they are used to seeing.If it were the case that just being in line with the accepted standard makes something look realistic, then works from unfamiliar systems would always look relatively unrealistic.Yet this is not so.This is known as the problem of 'revelatory realism.'It will be important later.
Lopes ends up saying that realism is achieved through a high level of informativeness in a system providing context for a particular picture.I believe that Kulvicki's theory is more useful, since he ties the verity of realism to perceptual conception.This will be discussed more fully in section three.The next section, section two, will clear the ground by explaining aspectival absolutism, my view that insists on the objective, universal status of realism despite its apparent cross-cultural variation.
First, I shall say a few words about depiction.Holding the view that images depict as parts of a symbol system may not commit one to believing that realism is merely a matter of convention.While the onus is most urgently on the proponent of the symbol-system view to advance a plausible theory of realism, the question about realism is in principle distinct from the perennially troublesome question about how pictures depict, whether under the auspices of a symbol system or by a more direct relation of resemblance.Kulvicki calls these the 'structural' versus the 'perceptual' theories of the image, respectively.He supports a structural theory, stressing the role of symbol systems, which he takes to be an improvement over that of Goodman in that it takes into account strong intuitions about the perceptual nature of images. 5In his book Understanding Pictures, Lopes advances a hybrid theory, aiming to incorporate the strengths of both approaches. 6While I am partial to a perceptual account, my twin theses on pictorial realism-that realistic images can convey the truth about various aspects of reality and that they do this by showing what the object would be like if it were to exist-ought to be compatible with both perceptual and (many) structural positions.If my claims seem hard to reconcile with structural accounts, then this will also be true for any other theory of realism that eschews appeals to convention.
To speak more generally: the answers to the questions I am asking (what do we depict when we depict realistically, how should we define 'realism'), while intertwined with questions about the nature of depiction, do not depend on their resolution.As Catharine Abell points out, it is acceptable to 'take a picture's depictive content for granted' when formulating a theory of what makes that content realistic. 7What I am after is not a theory of depiction, but of pictorial realism.

II
It is no doubt true that convention plays a role in determining both the meaning and accuracy of pictures.Various symbol systems make possible iconographic standards in different cultures and different historical moments.That being said, too much is made of the differences between systems and the supposed incommensurability of pictures in different systems.In fact, it would be better to drop the word 'system' altogether: it suggests a totality and impenetrability that seriously overstates the relation of different sets of pictures.A better way of looking at it, and one that lends to realism a degree of universality, is to think of the different types of pictures (photorealist, cubist, even caricature, not to mention the products of different cultures) as each aiming at capturing some particular aspect of reality.
This thought is in many ways similar to Lopes' aspectival realism except that his conception of realism as informativeness is too narrow and he concedes too much to Goodman's relativism.Lopes retains the idea that conventional systems determine what counts as realistic because he thinks that the opposite view necessarily implies a violation of cross-cultural respect.He writes, 'the absolutist will simply reply that the Haida, the Japanese, or the seventeenth-century Dutch are wrong.However, I find this an unattractive position since it relinquishes any hope of explaining diverse judgments of realism as anything more than bald errors'. 8e certainly want to avoid a chauvinist insistence on the exclusive realism of one culturally limited type of depiction, but it is perfectly possible to think that both a) different cultural judgments are correct and b) judgments of realism are universal, that is, their truth is not restricted to a particular system or culture.I will call this position 'aspectival absolutism'.Different cultures expect art to represent particular aspects of reality, both because they find varying features of reality to be important and because art functions in different ways.This explains why people from different times and places find particular pictures realistic.They are asking the pictures to speak to different realms of experience.This is not to say that every picture is as real looking as every other, but that genres or styles that are taken by people to be unrealistic because of their commitment to one paradigm of realism often represent some distinct feature just as realistically.When we accuse some unfamiliar style of being unrealistic, we often simply fail to see what it is realistically depicting because we expect it to do what our art does.It is perhaps tempting, for example, to add an implicit 'from the point of view of the human subject' to the question of whether an object is depicted realistically.There are other aspects of reality we can represent, however, beyond or besides the view from a fictional human observer in receding linear perspective.That is just one way for a picture to be realistic; it is one aspect of reality it can capture.
It will be clear that capturing the geometric shapes and their size relations as they appear to the situated human viewer will count as a familiar version of pictorial realism.Representation of a cross-section of a constructed visual pyramid is typical of this kind of realism as Alberti and other renaissance artists advocated it.Their method aims at providing the illusion of threedimensional lived space on a flat panel, wall or canvass.There should be no doubt that the term 'realism' applies here.
This should not be taken to mean, however, that all versions of pictorial realism must capture this one particular aspect of visual experience.Some art might instead deliver a very realistic portrayal of certain colours or textures, and again some others might portray with exactitude some phenomenon of subjective (psychological, emotional) experience.As long as the artwork succeeds in whatever task it can properly be expected to be attempting to accomplish in terms of the object of depiction, and as long as it meets the other main criterion of realism, namely that the object is depicted in its relevant aspect in the way it would appear in reality under certain presupposed or stipulated conditions, then the art is realistic.Aspectival absolutism will account for why different cultures will call different works realistic: they are looking for a depiction of different aspects.
I do not want to overstate my difference from Lopes's view on this point.I fully agree with his claim that 'by taking different points of view, pictures make different commitments and embody different aspects'. 9I do, however, wish to make clear that there is in fact a difference, in that I don't think Lopes requires the added apparatus of systems of representation inherited from Goodman and Gombrich to establish this claim. 10It is enough to say that different artworks represent particular aspects, and that the position of those artworks within structures of meaning is a secondary concern that is made possible in the first place by the fact that they successfully represent reality as they do.The question of how exactly a picture represents or corresponds to something in the world has no single answer because it will differ depending on the kind of thing or aspect of it that it attempts to portray.At its most direct level, it can certainly be as simple as a one-to-one match between colour or spatial form in the visual field (or, in somewhat more complex cases, structural matches in the differential relations of colours, etc.).This fact does not rule out the possibility of more sophisticated forms of realism.None of them, in my view, will depend on symbol systems.

III (A)
Realistic pictures need not always correspond to our 'perceptual conceptions' of an object (a restriction which would, incidentally, introduce an unwarranted level of relativism to the question of what counts as realistic: the question of whether an object matches a conception is always asked in reference to the conceptions and perception of an individual observer) but rather to what a thing would really be like (insofar as it were available to perception) under certain conditions.Realism is conformity to the laws of nature.Pictorial realism is achieved when the image depicts a thing in accordance with those laws.If it is the case that in order to exhibit the physiology we associate with dragons, a dragon's blood would in the real world have to be cooled by vents in its dorsal armour, the picture that depicts those vents (assuming that they would be visible in the depiction if they were there) is, in that respect, more realistic.This is so whether or not the spectator knows anything about the counterfactual biology of flying fantasy reptiles.
A word must be said on what the 'laws of nature' might be.My view should be compatible with a fairly wide range of views on the status of such laws.There has long been a lively debate on this topic, with many philosophers advocating views rooting laws in either deductive systems or in the instantiation of universals; others hold more or less radical antirealist positions. 11The viability of the view under consideration in this paper does not hinge on the resolution of this debate.The only truly conflicting positions would be the most relativistic of the antirealist views, because they would not allow for any laws universal or permanent enough to hang our aesthetic hat on.Any reader sympathetic to the objective status of realism in art would probably be inclined to dismiss radical nomological antirealism anyway.All that is needed is the belief that the world is such that there are some general rules or regularities in nature that determine the distinction between necessity and accident and allow for truth and falsity with respect to counterfactual entities and events.The ground of these laws is not at issue.
We may safely return to the main course of the argument.Hypothetical verity's difference from Kulvicki's position has as a significant and desirable consequence that a picture can be realistic with reference to a given spectator even if that spectator fails to adequately interpret it as realistic.There can be failed interpretations in this respect.

III (B)
Kulvicki's version of verity goes as follows: Within a given system, a picture realistically depicts something as being q for some observer if (1) it depicts its object as being q and (2) some other quality the picture ascribes to that thing is included in the observer's perceptual conception of things that are q. 12It will be advantageous to rescue from this formulation its main benefit, which is that it allows the term 'realistic' to be applied to fictional entities.Additionally, it gives a more concrete and objective account than those relying on participation in conventional symbol systems.The image, for Kulvicki, actually has to match a conception; it has some standard of measure outside the symbol system.I find this a necessary characteristic that an adequate theory of realism should be expected to include.Unfortunately, however, the definition leaves realism still too relativistic: for different people with different perceptual conceptions of a sparrow, a single picture of a sparrow might be realistic for one person and not the other.The about sparrows does not enter into it at all.Is there a way to save realism for fictional entities yet still demand that we honour the truth about sparrows?I believe there is.

My modified version of verity goes like this:
A picture realistically depicts something as being q if (1) it depicts its object as being q and (2) some other quality the picture ascribes to that thing is as it would be in the real world if that thing were q.
Of course, even if the depiction of the quality in question is perfectly realistic, this says little about the total level of realism of the picture.A picture will typically ascribe many qualities to a thing, and some of them will be depicted more realistically than others.This definition, like Kulvicki's, is about the realistic depiction of some quality q, not about the picture as such.If a picture depicts a kitten as being furry and flat, it is realistic in respect to the furriness, but unrealistic in respect to the flatness.A judgement about the overall level of realism of a picture will balance the combined effect of the depiction of the many qualities.As experience suggests, no formula for making such a judgment exists.
Note that depictions of actual things must meet the same requirement as depictions of fictional entities.The definition works for both existing and non-existing objects.Note also that any degree of detail and specificity might be built into q, including even changes in the laws of nature themselves.There will, for example, be more or less realistic ways of depicting objects in universes operating with different physics than the kind we know.
What are the further advantages of hypothetical verity?Kulvicki rightly identifies two points at which his theory might be assailed: its consistency with what we often believe about a) revelatory realism and b) the apparently realistic nature of images that teach us things about the world (e.g.photos of a cell that instruct scientific investigation).In the former case, he comes up with an explanation that does, admittedly, work, but is not as satisfying as that possible with hypothetical verity.In the latter, he asks us to accept the intuitively awkward position that informative images that do not match up to our preconceived notions are accurate, but not realistic.This sacrifice is made in the name of rescuing verity to conception for the sake of its main advantage: it is possible in his theory to call depictions of fictional entities realistic.This advantage is shared, as we have seen, by hypothetical verity, which offers better answers to questions about revelatory realism and informative images.Hypothetical verity should therefore be accepted as a modest improvement over Kulvicki's original view of verity.
Kulvicki correctly believes that the two objections he identifies reveal his argument's weakest points.Despite the merits of his account, I do not think that he is completely successful in overcoming the objections.First, he needs to establish that his theory can account for revelatory realism.If a realistic picture simply corresponds to our premade conception of what a thing is, how could the level of realism achieved in a picture ever surprise anyone?Wouldn't people know already what the thing is supposed to look like?Kulvicki answers that 'when we manage to depict objects veritably in respects that we had not been able to depict them before, the result is surprising'. 13In other words, it is merely a question of improved technique.He adds to this that pictures that at one time seemed realistic may later seem unrealistic if a new interpretive scheme-that is, a new symbol system-has come into play. 14In order to explain why closer approximations to a preconceived perceptive conception might lead to qualitative leaps in what seems realistic, Kulvicki must here resort to the Goodmanian deus ex machina of conflicting schemes.
Hypothetical verity does a better job of addressing both a) why some pictures are surprisingly realistic and b) why they might no longer look realistic in the future.Remember that realism for Kulvicki is always truth to a conception.This conception limits and determines the image that corresponds to it, which always falls short in its approximation (at least if there is to be room for a more realistic picture of the same thing to come along later).For hypothetical realism, contrastingly, realism is measured by the world itself.It is typically the case that our conceptions do not correspond fully with reality.In that case, even if an image were perfectly true to our conception of something, it would fall short of perfect realism by the same distance that the conception falls short of its object (or, in the case of fictional entities, how that object would actually be if it existed in our world and was subject to all the laws of nature).A picture that manages to depict something in a way that surpasses in truth the conception the spectator has of the thing depicted would have a real reason to surprise her.Art can, in this way, show us things about the world that we have failed to notice in the past.This is an important artistic function and discloses the full significance of revelatory realism.
So how is it that a painting can look hyper-realistic at one point in time and less realistic later?This may happen when a picture is deemed realistic but replaced in viewers' estimation by other, more realistic pictures.It might then be wondered how the old picture seemed so realistic before.One reason is that people gain a greater knowledge of how the world is and a greater (or at least different) capacity for perceptual acuity.It takes a cultivated eye even to pick out forms and textures as they really are.And, of course, people with different occupations and interests will perceive things in different ways and concern themselves with different aspects of reality.Different psychological concerns will demand realisms pertaining to different aspects of the world.
I do not deny that pictures sometimes do not fully correspond to conceptions.This certainly might make for some lack of realism, if the conception is closer to reality than the resulting picture.Two points must be noted in this regard, however.First, such failure of execution is less frequent and less severe than it might at first seem.Developing techniques of art creation is difficult, but it happens fairly rapidly when people set themselves diligently to the problem.They usually do so whenever their pictures clearly fall short of their conceptions.It is a safe bet that the images people make and the conceptions they work with are fairly close.Second, it is not by the distance between the picture and the conception that the picture falls short of perfect realism in any case.It is the distance between the picture and its being an accurate portrayal of a thing (real or hypothetical) that determines how far it is from full realism (merely a regulative ideal, to be sure).The distance short of the conception may merely add to this total interval.
Kulvicki's theory of verity does not explain revelatory realism as well as hypothetical verity does, even if it does give a workable account in this respect.The second potential objection to conceptual verity is, to my mind, more decisive: it forces Kulvicki to say that pictorial verity is different from the kind of verity that can teach us things about the world.If scientists look at a photograph of a cell and learn from it how the molecules are arranged, does that photograph count as realistic?No, Kulvicki must say: it does not live up to the conception we had of molecules in a cell prior to viewing the photograph.This is Kulvicki's example.The scenario is even stranger if we think about historical photographs.A picture of a battle that shows us things about the battle that we did not already know would fall short of realism-and for that very reason.Its pedagogical usefulness would stand in inverse proportion to its realism.This is a strange and, it seems to me, counterintuitive conclusion.Hypothetical verity reaches the more sensible conclusion that pictures that tell us things about the world are realistic.This is because realism is measured by reality, not by our conceptions of it.
In conclusion, we can say that a picture is realistic if it depicts its object in accordance with the laws of nature.This goes for both real and fictional entities.Much of the variation in works that are called 'realist' is due not in fact to differences in the underlying conceptions of things or even differences in the symbol systems used to represent them, but rather to the particular aspects of reality its creators are attempting to render.Together, these two theses make up a robust theory of realism that allows for objectivity and universality in pictorial representation, while also recognizing the legitimacy of cultural variation.ccbuckman@gmail.comNOTES