Arrangement and Timing: Photography, Causation and Anti-Empiricist Aesthetics

1. 

The causal theory of photography (CTP) claims that photography is a causal medium, in contrast with painting, drawing, etching and similar so-called ‘handmade’ modes of picture-making. According to CTP, a photograph of X causally depends upon X itself for its content, whereas this is not true of a handmade picture of X. By way of illustration: photos taken by delusive or hallucinating photographers depict what is in fact before them, whereas paintings or drawings by delusive or hallucinating painters depict the contents of such delusions or hallucinations (Currie 2004: 65–66).

Among CTP’s virtues are that it explains two differences between photographs and handmade pictures. First, it explains a difference in epistemic value. When Leland Stanford desired to know whether horses gallop with all hoofs aloft, he turned to the photographer Eadweard Muybridge rather than to a sketch-artist or painter. CTP explains why this decision was rational: photography is a method of picture-making that is causally sensitive to the world itself (the part before the camera, at least), whereas paintings and drawings are causally sensitive to their maker’s perceptual experiences of, and beliefs about, the world. The truth-value of the latter mental states concerning the proposition that horses may gallop with all hoofs aloft was precisely what Stanford sought to settle. Upon pain of circularity, those states could not themselves serve as (dis) confirmation of the proposition. This is how things would have stood, had Stanford relied upon a painter or sketch artist. What Stanford required was a method of picture-making directly responsive to the facts themselves. CTP explains how photography uniquely fits that bill.¹

Second, CTP explains why there is unexpected depiction in photography but not painting (Currie 2004: 66). Consider photobombs. These are photographs that are ruined, often in amusing ways, by persons or animals in the scene before the camera, and which the photographer was unaware of when taking their picture. It is striking that we have no correlative concept of a ‘paintingbomb’ or ‘drawingbomb.’ CTP explains why: photographs acquire their content from the world itself, rather than anyone’s beliefs about the world, hence why there can be divergence between what someone expects to be depicted in their photograph and what is in fact depicted. But because paintings acquire their content from their maker’s intentional mental states, the possibility of unexpected pictorial content does not arise, at least not in the same way.

Yet whatever CTP’s explanatory virtues, it suffers a simple, substantial problem: it says nothing positive about the role of photographers in producing photographs. CTP’s aim is to say what is distinctive about photography. It does so negatively: by reference to putative ways the photographer fails to be involved in making their pictures. Whatever aspects of photography this idea is apt to explain, it leaves us in the dark as to how the photographer is involved. So as a theory about the nature of photography, CTP remains incomplete.

When the problem is developed further, incompleteness can appear the least of CTP’s problems. Consider two representative statements of CTP. Currie says that photography is distinctive insofar as “events themselves are registered independently of the agent who otherwise might be the author of the process” (2004: 68). Walton similarly claims: “Objects cause their photographs and the visual experiences of viewers mechanically. By contrast, objects cause paintings not mechanically but in a more ‘human’ way, a way involving the artist” (1984: 261). As has been pointed out by CTP’s critics, defenders of CTP insinuate with such remarks that photographers are causally excluded from occupying any interesting causal role in the production of their pictures (Costello & Phillips 2009: 15). This is troubling. For in whatever sense it is true that a photograph of X is caused by X, analogous to how, say, a visual experience of X is caused by X—or, using Currie’s (2004: 66) analogy, a footprint is caused by a foot—a Moorean concern one might have is that we are more sure of the claim that photographers cause photographs. So, without some positive account of the role that photographers play in causing photographs, CTP will be at best incomplete. At worst, CTP will be false, if what its defenders insinuate is that photographers are causally excluded from occupying such a role.² Call this problem for CTP the Problem of the Missing Agent.

The Problem of the Missing Agent is particularly pressing when one considers the aesthetics of photography. Without some mention of the photographer’s role in producing photographs, CTP may explain photography’s epistemic value; what it will be at odds with is the idea that photographs are objects of aesthetic value. Photographs are artefacts. As such, they are to be admired, at least in part, as the products of a photographer’s agency, and not the mere agency of light or some other natural process.

My aim here is to revise CTP in order to help it overcome the Problem of the Missing Agent. The revision, which I outline in Sections 2–3, relies on Fred Dretske’s distinction between triggering and structuring causes: photographed objects are the structuring causes of photographs while photographers are triggering causes of photographs. Crucially, structuring and triggering causes are not in competition to causally explain their effects, but must work together to bring their effects about. So one can grant photographed objects a key causal role in the production of photographs, as CTP wishes to, while leaving room for the photographer to also occupy a central role. While one might finesse CTP with any number of tools from the philosophy of causation, the triggering/structuring distinction will prove a particularly valuable one. For in Sections 4–7, I will show that CTP, so finessed, can explain at least two aesthetic interests that we commonly take in photographs: an interest in how they have been structured and an interest in how they have been triggered. I shall explain how these interests figure in our appreciation of photographic portraiture and street photography, respectively. What is relatively unique about such aesthetic interests is that they do not consider photographs along empiricist lines, as sensory surfaces for the viewer’s perceptual pleasure. Rather, when taking an aesthetic interest in how a photograph has been structured or triggered we take a distinctly anti-empiricist interest that targets the achievement of the photographer in producing their pictures. A subsidiary aim of this paper will be to make a case for anti-empiricism about the aesthetic value of photography.

Two preliminaries. First, defenders of CTP often acknowledge that their view appears inconsistent with allowing the photographer to occupy a causal role in the production of their pictures. For instance, despite claiming that “objects cause their photographs” (1984: 261), Walton is quick to dismiss the implication that photographs are not also made by people. Likewise, despite claiming that in photography “events themselves are registered independently of the agent who otherwise might be the author of the process” (2004: 68), Currie is also quick to deny that this means photographs are created “independent of intention” (2004: 66). This is all reasonable. For CTP, as I said above, would be implausible, otherwise. Nevertheless, these remarks look purely stipulative. They do little to dispel the appearance that CTP entails photographers are causally excluded from the production of their pictures. In that respect, such remarks offer little that might form the basis of a reply to the Problem of the Missing Agent. This is a crucial point at which Dretske’s distinction can add value. Indeed, it might well be that something like Dretske’s distinction is what the above defenders of CTP have in mind with their remarks. Still, how the distinction works in the case of photography, particularly in terms of dissipating exclusion worries, remains to be made explicit.

Second, a caveat about the status of CTP in this paper. I will not engage with the question of CTP’s extensional adequacy. CTP has come under much fire in this respect.³ In particular, it has been claimed that CTP ignores non-causal, artistic uses of the medium and that many styles of photography, including photomontage, fail to fit its mould. Responding to this objection falls outside of the scope of this paper. Nevertheless, since this is another cause for anxiety about CTP, I will sketch two possible lines of reply. My aim, in doing so, is to make the modest case that the overall prospects for CTP are not hopeless.

A defensive reply to the question of extensional adequacy: the claim that non-causal uses of the photographic medium produce photographs, or that significant alterations to a photograph do not destroy its identity as a photograph, are not beyond doubt. After all, it is reasonable to think that some pictures (photomontages in particular) must occupy a middle-ground between photographs and handmade pictures in their being a hybrid artform (Levinson 1984; see also Anscomb 2019). And it is not obvious why the defender of CTP is precluded from placing such ‘problem’ cases here (though see Lopes 2016 for some dissent). Second, a concessive reply: suppose that, as some critics claim, CTP is not extensionally adequate because ‘photography’ is a family resemblance concept, with no defining essence, let alone a causal one (Costello & Phillips 2009: 3). Does this spell the end for CTP? Hardly. CTP might still aptly describe at least one branch of the photography family tree: traditional or documental or straight photography.

Yet no matter how defenders of CTP respond to the question of its extensional adequacy, the issue of what CTP can say positively about the role of the photographer and how CTP might explain (or at least be consistent with) aesthetic admiration of photographs, would remain unanswered. These are the matters that I am concerned with here.

2. 

In a discussion of the causal efficacy of mental content, Fred Dretske (1988; 1993) draws a distinction between triggering causes and structuring causes. The distinction is entirely general and independent of the particular use that Dretske puts it to: causally explaining behaviour. In order to illustrate the distinction, Dretske gives the following example:

A terrorist plants a bomb in the general’s car. The bomb sits there for days until the general gets in his car and turns the key to start the engine. The bomb is detonated (triggered by turning the key in the ignition) and the general is killed. … Although the general’s own action (turning on the engine) was the triggering cause, the terrorist’s action is the structuring cause. (1993: 122–23)

I shall be arguing that we can usefully understand photographed objects to be structuring causes of photographs and photographers to be triggering causes of photographs. First, it is necessary to tease out from Dretske’s example exactly what the structuring cause/triggering cause distinction amounts to. This section and the next will be fairly programmatic. The justification for applying Dretske’s distinction to photography consists in the explanatory gains outlined in later sections; in particular, its ability to elucidate at least two aesthetic interests we take in photography and which have so far gone overlooked by philosophers.

First, Dretske claims that triggering causes are related to their effects in a Humean way by being constantly conjoined to them. Given the circumstances, triggering events of a certain type, i.e., the turnings of keys in the ignitions of cars rigged to explode, are typically followed by events of a distinct type, i.e., explosions. As Dretske puts it, a triggering cause “tops up a pre-existing set of conditions”, a set that, with the triggering cause, is then sufficient for the relevant effect (1993: 123). This reflects the fact that explosions do not follow the turnings of keys in ignitions per se, but only in certain circumstances, e.g., when a bomb is wired into the ignition system.

By contrast, Dretske thinks of structuring causes as insufficient for their effects. The wiring of a bomb into a car’s ignition system is not enough to cause an explosion. One needs a further event to obtain, i.e., the turning of the key in the ignition, an event that will then trigger the explosion. Because of this, Dretske allows that we may prefer to call structuring causes mere background conditions (1993: 124). For instance, it is tempting to say that planting a bomb per se doesn’t cause an explosion. Only when bombs are detonated, i.e., causally triggered, are we liable to speak of causation. Consider how the bomb may sit in the car, unexploded, for days, weeks, months or indefinitely.

On the other hand, Dretske notes that we are sometimes willing to call structuring causes causes proper, not mere background conditions. For one, we say that the general was killed by the actions of the terrorist (or the fact of there being a bomb wired into the car), i.e., the explosion’s structuring cause. We do not say that the general, by turning the keys in the ignition thereby being the explosion’s triggering cause, caused their own death. Second, events counterfactually depend upon their structuring causes as much as their triggering causes: had the terrorist not planted a bomb, then the car would not have exploded and the general would not have been killed (1993: 125). While counterfactual dependence may not suffice for causation, there remains something cause-like about structuring causes in how they relate to their effects, though we may refrain from considering them full-blown causes. Crucially, both these points illustrate how we may be ambivalent about whether to call structuring causes causes proper or mere background conditions (a matter to which I return below in the case of photography). Either way, structuring causes are claimed by Dretske to be necessary, along with triggering causes, for their effects.

Here is a useful way to think about structuring causes: structuring causes are primed and ready to bring about their effects. But they require that extra ‘spark’, as it were, i.e., the triggering cause, in order to do so. No structuring cause? No effect. (No wiring of a bomb into the car? No explosion.) But likewise, no triggering cause means no effect either. (No turning of the key in the ignition? No explosion.) Structuring and triggering causes are thus not in competition to causally explain their effects—they are not at risk of causally excluding one another—rather, they must work together to bring about their effects.

I believe that it is natural to extend this thinking to photography in the following way: the photographer’s act of pressing the camera’s shutter release is the triggering cause of the resulting photograph while the objects before the camera are that photograph’s structuring cause. Turning the key in the ignition (the trigger) causes an explosion only conditional upon the obtaining of a pre-existing set of conditions (the structure), i.e., the presence of a bomb suitably wired into the car. Likewise, pressing the camera’s shutter release (the trigger) causes a photograph of a cathedral only conditional upon the obtaining of a pre-existing set of conditions (the structure), i.e., the presence of a cathedral suitably focussed on by camera. Of course, in both cases, many more conditions must also obtain as part of the structuring cause, e.g., there must be film in the camera and there must be a working battery and electrics in the car. We can prescind from such details.

In addition, I noted above how Dretske claims that we may be ambivalent about whether to call structuring causes real causes or mere background conditions. For although it is intuitive to say that the General was killed by the terrorist’s act, i.e., of wiring a bomb into the former’s car and thereby making the structuring cause a real cause, this can also seem a mistaken way of talking insofar as the bomb would have sat completely idle were it not for the General’s act, i.e., of turning the key in the ignition. This ambivalence carries over to photography. For while a photograph of a cathedral may seem to causally depend upon the photographed cathedral itself, the very problem for CTP under discussion, the Problem of the Missing Agent, is that we may be resistant to agreeing with the earlier claims of Currie and Walton that the cathedral was the cause of the photograph. For we then require some positive account of the role of the photographer, lest they be causally excluded, and the claim that objects (not photographers) are the causes of photographs seems a mistaken way of talking. Consider, in this regard, William Henry Fox Talbot’s remarks to the Royal Society as he proclaimed his house “to be the first that was ever yet known to have drawn its own picture.” Likewise, Louis-Jacques-Mandé Daguerre, another early inventor of the medium, wrote how photography “is a chemical and physical process which gives her the power to reproduce herself.” Compare also Rudolph Arnheim’s (1974: 155) remark that in photography “the physical objects themselves print their image.” There seems to be at least a grain of truth to such remarks, charitably interpreted, but of course they also seem to go too far, as the Problem of the Missing Agent attempts to make vivid. Dretske’s triggering/structuring distinction can explain why: the ambivalence is explained by photographed objects being structuring causes of photographs. For we have seen that just this kind of ambivalence over whether to consider X a cause proper or a mere background condition is what comes with the territory of identifying X as a structuring cause.

Finally, consider what I flagged above as a useful way to think about structuring causes: primed and ready to bring about their effects, but only given that extra ‘spark’, i.e., a triggering cause. This again seems to readily apply to photography. Consider the state of affairs of a cathedral’s being before a camera (one loaded with film, suitably focussed, etc.). Just as a car rigged with a bomb (with suitably working electrics, etc.) requires the triggering cause of the general turning the key in the ignition for there to be an explosion, the state of affairs of the cathedral’s being before a camera likewise requires the triggering cause of the photographer’s pressing the shutter release for there to be a photograph of the cathedral, along with many other background conditions.

This (just about) completes the tasks of applying Dretske’s distinction to photography. One delicate, but as we shall see, helpful, complication needs addressing: an ambiguity in the concept of a structuring cause concerning whether structuring causes are states or events.

Consider again the general/terrorist case. The structuring cause here could be identified with the state of the car. For instance, we might say that the general was killed because there was a bomb wired into the car at the time the general turned their key in the ignition. On the other hand, the structuring cause of the explosion might be identified with the event of the terrorist’s putting the car into that state. For instance, we might say that the general was killed because of the terrorist’s act of wiring a bomb into the car’s ignition (Drestke 1993: 124–25).

Call this distinction one between a stative conception of a structuring cause versus an evental conception of a structuring cause. How might this sub-distinction, not between triggering and structuring causes themselves, but between two different ways of considering structuring causes, apply to photography?

I suggest the following: just as the structuring cause of the car’s explosion could be identified with the state of the car at the time of the explosion’s triggering cause, the structuring cause of a photograph might be identified with the state of the world before the camera at the time of the photograph’s triggering cause. This is the stative conception of a structuring cause, applied to photography. It refers to the state of the world, before the camera, when photographed.

Similarly, just as the structuring cause of the car’s explosion might instead be identified with the terrorist’s action of putting the car into such-and-such a state, the structuring cause of a photograph might be identified with the photographer’s (or anyone else) action of putting the scene before the camera into such-and-such a state. Consider, for instance, how the photographer may arrange the scene before them a particular way. This is the evental conception of a structuring cause, applied to photography. It refers to the act of the photographer putting the world before the camera into some state prior to its being photographed.

There are some forms of photography where the evental conception of a structuring cause applies and others where only the stative conception applies. For instance, in street photography, photographers typically do not engage in rearranging the scene before them; they instead aim to capture life on-the-fly, as is it. Here, the evental conception of a structuring cause will not be relevant, while the triggering cause certainly may. But consider portraiture. In photographic portraiture, staging and intervention is an important factor in causally explaining why the resulting photograph is the way it is, why that particular item is in a certain part of the frame and another is over there, etc. and so the evental conception may be highly relevant. I will spend Sections 4–7 unpacking these ideas and explaining how one may take an aesthetic interest in the triggering and structuring of such photographs. In doing so, I will thereby show that CTP, revised with Dretske’s distinction, is not at odds with taking an aesthetic interest in photographs, but is well-positioned to positively explain at least two of them. For now, however, I return to the Problem of the Missing Agent.

3. 

Relating what I have said so far to my earlier discussion of CTP, when defenders of the view say that photography is a causal medium, they have had in mind the stative conception a structuring cause. They have been focussing, very narrowly, on the role played by objects situated before the camera in causally explaining relevant features of the resulting photograph; why what is depicted is, e.g., a cathedral, a snapdragon in bloom, or Marissa’s aunt. The result, I complained, is the Problem of the Missing Agent: defenders of CTP at best say nothing positive about the role of the photographer in bringing about their pictures. At worst, they insinuate that the role played by photographed objects leaves no room for the photographer to do any interesting causal work.

With the triggering/structuring distinction on the table, defenders of CTP are supplied with the resources to solve the Problem of the Missing Agent. Consistent with acknowledging that photographs owe their causal history to photographed objects is acknowledging that photographs also owe their causal history to photographers, and in such a way that, crucially, the two are not at risk of causally excluding one another. Without either the cathedral or the photographer’s pressing the shutter release, there would be no photograph of the cathedral. One is a structuring cause and the other is a triggering cause. They must work together to bring about their effects.

More precisely, on the revised way of understanding CTP, the two do not causally complete with one another since each explains something different about the resulting photograph. On revised CTP, photographs owe their depictive content to their structuring causes and yet owe their existence to photographers. What (proximally) explains why a photograph was taken on any given occasion is the action of the photographer in triggering the photograph, e.g., by pressing the camera’s shutter release. But what (proximally) explains why a photograph depicts this or that is the presence or absence of this or that object in the scene before the camera, e.g., the structuring cause, and this, crucially, may include any action of the photographer in arranging the scene this way or that. Structuring and triggering causes explain different aspects of their effects and so are not in competition, but must work in concert. Again, one might wonder if all this is extensionally adequate. My aim here is simply to revise CTP in such a way that helps it overcome the Problem of the Missing Agent.

For the sake of completeness, notice that the above claims are mirrored in the general/terrorist case. What explains why an explosion occurred was the triggering cause, i.e., the turning of the key in the ignition, otherwise the structuring cause would lie causally dormant. But what explains why that event was an explosion, and not the starting of a car, is something else; namely, the structuring cause, i.e., a bomb’s having been wired into the car. Without that structuring cause, what would have been produced by the triggering cause in this case would be very different. Likewise, without the presence of a cathedral before the camera, what would have been produced by pressing the camera’s shutter release would not be a photograph of a cathedral, but a photograph of something else. The triggering/structuring distinction thus vouchsafes a role for reality itself in the causal history of a photograph, while also vouchsafing a role for the photographer. (If one doubts that this latter role is very important, they should await the discussion of street photography in Section 7.)

It is important that the revision not only overcomes the Problem of the Missing Agent, but does so while remaining true to the original spirit of CTP.⁴ The aim here is not merely to solve a problem, but to solve it in such a way that is acceptable to those at whom the problem is directed. Does use of Dretske’s distinction allow for that?

Yes. First, we saw in Section 1 that defenders of CTP, like Currie and Walton, want the theory to at once capture the sense in which photographs depend upon the world itself, and not what anyone thinks the world is like, but without having to deny that photographs depend upon photographers. The proposed revision is certainly true to the spirit of CTP on that front. It says that photographed objects are structuring causes of photographs while photographers are triggering causes of photographs. Revised CTP is not guilty of injecting an undesirable degree of agency into our understanding photographs, by CTP’s lights, but just the right amount.

Second, defenders of CTP claim that photographs have a distinctive epistemic value (as the Stanford-Muybridge example shows) and that there can be unexpected depiction in photography (as photobombs show). Again, revised CTP affirms this too, since it says that photographs owe their depictive content, proximally at least, to their structuring cause (the objects depicted), not their triggering cause (the photographer). Hence one can learn something from a photograph about the photographed object, or scene, that no one previously knew, not even the photographer; hence, moreover, why what is depicted by a photograph can diverge from what the photographer expected to be depicted.

Third, and most importantly, defenders of CTP want there to be a clear difference between photographs and handmade pictures, both in the above two respects but also more generally. Dretske’s distinction supplies the resources for such a demarcation, in ways traditional CTP defenders should find acceptable. Revised CTP says that, in photography, structuring and triggering causes are distinct; one is the world itself, determining depictive content, and another is the photographer, determining the photograph’s existence. The difference with painting and drawing is that structuring and triggering causes are not distinct here, but fused; what causes a painting or drawing to have its content is the same thing that causes its existence, namely the actions of the picture-maker in intentionally marking the picture-surface.

Using Dretske’s distinction to demarcate photography from painting in the above manner thus captures two interrelated ideas central to CTP. First, that photographs are rooted in, while paintings and drawings exhibit a freedom from, concrete reality. Second, that paintings and drawings are saturated with intention and agency in ways that photographs are not. In explaining why a photograph depicts this or that, e.g., a cathedral, a snapdragon in bloom, or Marissa’s aunt, revised CTP says one must reference this or that object being in the world, before the camera, at the time the picture was taken; to explain why the photograph itself was taken, and caused to exist, one must talk about something else: the photographer. But in explaining why a painting depicts any one of those things, revised CTP can say that one must reference the picture-maker’s actions and intentions. These are the very same things one must reference to explain why the painting itself exists. In painting, all the explanatory action is given over to the painter, who both structures and triggers their picture with the action of painting. With photography, by contrast, the world itself plays a causal role. Yet both are made by people.

Having discussed how CTP, by availing itself of the triggering/structuring distinction, can solve the Problem of the Missing Agent, we are now in a position to address a further question: is this revised version of CTP able to say something of substance about the aesthetics of photography?

4. 

Defenders of CTP who want the theory to be consistent with, and maybe even explain aspects of, photography’s aesthetic value must fight a battle on two sides. First, as we have seen, they must handle the Problem of the Missing Agent, thereby responding to those who reject CTP through (i) being sceptical of its ability to vouchsafe any causal role for the photographer in the creation of photographs, and that therefore (ii) such a view must lead directly to scepticism about the aesthetic value of photography. Second, however, they must respond to internal challenges from defenders of CTP who embrace these conclusions. I have so far focussed on assisting CTP to fight the first of these battles. What of the second?

Notoriously, Roger Scruton is a defender of CTP who has developed a number of arguments against the possibility of taking an aesthetic interest in photographs, precisely on the grounds of CTP’s being true. Although considerable ink has been spilled in response to these arguments, some remain poorly understood and others have escaped the notice of commentators altogether. Below is one such argument. I call it The Manifest Argument. Consideration of this argument will help set the scene for showing that CTP, revised with Dretske’s triggering/structuring distinction, can illuminate at least two aesthetic interests in photography.

5. 

Aesthetic empiricism is the view that the aesthetic value of an artwork is a function of its experiential effect upon viewers. Aesthetic empiricism divides into pure and enlightened varieties. Consider the following remarks by aesthetic formalists Clive Bell and Alfred Lessing, expressing support for pure aesthetic empiricism:

To appreciate a work of art we need bring with us nothing but a sense of form and colour and a knowledge of three-dimensional space. (Bell 1914: 37)

Aesthetic experience is … wholly autonomous. It cannot take account of any entity or fact which is not perceivable in the work of art itself. (Lessing 1965: 470)

This form of empiricism is ‘pure’ in that it identifies a work’s aesthetic value with what is manifest in a perceptual encounter with the work that is free from the influence of any and all of the viewer’s beliefs, e.g., about who made the work, what genre or category of art it belongs to, when and where it was made, and so on.

Enlightened empiricists agree with the spirit of pure empiricism: aesthetic value is a matter of how a work looks, sounds, etc. But they deny that viewers’ beliefs about a work’s history cannot enter into their appreciation. Indeed, enlightened empiricists typically claim that the audience must take account of the work’s history if they are to experience the work correctly. Still, the relevance of such beliefs and the historical facts that they represent is only indirect. Enlightened empiricism says that facts about a work’s history only matter instrumentally, for grasping a work’s aesthetic value by correctly shaping how the work is seen. Consider the following remarks by Peter Lamarque and Kendall Walton:

[A]esthetic value does depend on what is directly perceived (or experienced) at least when the perception is suitably informed by beliefs about what kind of object is being perceived. (Lamarque 2010: 134)

[I]t may … require special training or a special kind of sensitivity. But [aesthetic] qualities must be discoverable simply by examining the works themselves if they are discoverable at all. [Artworks] are to be judged solely on what can be seen or heard in them—when they are perceived correctly. (Walton 1970: 336 and 367)

Contrast both forms of empiricism with aesthetic anti-empiricism. Following Davies (2004: 258), I understand anti-empiricism to be the view that “There are factors that enter into our assessment of aesthetic value [of artworks] that are not reducible to the experienced effects of such works” (see also Currie 1989; Sharpe 2000; and Roberts 2018). For instance, consider the position that one finds in Denis Dutton’s influential discussion of forgery:

[W]orks of art represent the ways in which artists solve problems, overcome obstacles … The ultimate product is designed for our contemplation.… But this ought not to blind us to a fact we may take for granted: that the work of art has a human origin, and must be understood as such. (1979: 305)

Dutton’s considered view appears to be that aesthetic evaluation is dual-aspected: it conjoins, though Dutton was not clear how, both perceptual pleasure in the final product and a distinct, cognitive, acknowledgement of the artist’s achievement in producing the work, where the latter involves awareness of the constraints within which it was produced. Such constraints include both those intrinsic to different media as well as ones that are purely conventional. The crucial matter is that Dutton’s view renders art-making as much a matter of problem-solving and achievement as it does the production of pleasurable experiences. As Dutton writes:

[I]n order to grasp what it is that is before us, we must have some notion of what the maker of the object in question has done, including some idea of the limitations, technical and conventional, within which he has worked. It may be perfectly true … that in a painting of the Madonna the pale pink of the Virgin’s robe contrasts pleasantly with the light blue-gray of her cloak. But it is far from irrelevant to know that the artist may be working within a canon … according to which the robe must be some shade of red, and the cloak must be blue. The demand (to juxtapose fundamentally warm and cool colors) poses difficulties for creating harmony between robe and cloak … To say that the resulting assemblage of colors is pleasant may, again, be true enough; a fuller appreciation and understanding, however, would involve recognizing how that pleasing harmony is a response to a problematic demand put upon the artist. (1979: 307)

Since it involves an empiricist aspect, Dutton’s view might be branded a hybrid of anti-empiricism and enlightened empiricism. But insofar as Dutton would undeniably agree with Davies’s claim that at least some factors entering into our assessment of a work’s aesthetic value are not reducible to its experienced effects, he counts as an anti-empiricist. Enlightened empiricists like Lamarque and Walton certainly would not accept the latter claim.

Anti-empiricism is by far a minority view, with enlightened empiricism being the most widely accepted. Why should anyone be anti-empiricist? Davies (2004: 260) explains: “In appreciating a work, we appreciate a particular … doing on the part of an agent.” Davies’s idea, though at odds with his remarks above on visually appreciating photography, is that aesthetic appreciation involves acknowledging parts of a work’s causal history, principally the actions of its maker, that fall outside of what can be gleaned from a perceptual encounter with the work, suitably informed or not.⁷

With the contrast between empiricist and anti-empiricist accounts of aesthetic value now acknowledged, we are able to develop a distinctive reply to Scruton’s Manifest Argument. The reply, in opposition to Davies’s, grants (4): photographers’ actions are not perceptually manifest in their works. However, we can now see that only if we assume empiricism (pure or enlightened) does this have any negative consequences for the aesthetics of photography. If one is an anti-empiricist, then no such consequences follow since one rejects the assumption that an artwork’s aesthetic value is exhausted by what is perceptually manifest in it. In short, the Manifest Argument, in claiming that (5) follows from (3) and (4), begs the question of anti-empiricism in its assuming an empiricist, experience-based account of aesthetic value. That Scruton presupposes empiricism is clear from his remarks, quoted above, about the impossibility of taking an aesthetic interest in the photographer’s action of arranging the scene before the camera, on the basis that such an act is not discoverable by looking at the resulting photograph.

Let’s give The Manifest Argument a minor tweak, turning it into the following The Anti-Empiricist Argument:

6. 

Scruton, we have seen, claims that it is not possible to take an aesthetic interest in how the photographer has arranged the scene, prior to photographing it. For such actions, and the intent behind them, are not manifest in the resulting photograph. This is to deny the possibility of taking an aesthetic interest in the structuring cause of a photograph, on its evental conception. But anti-empiricists will not agree. They may claim that photographs sometimes merit aesthetic admiration by virtue of being the product of the photographer’s expertise in arranging various items of the scene, particularly in knowledge of how those objects will appear in the resulting photograph.

Here is an example to focus discussion: Arnold Newman’s portrait of Igor Stravinsky. Stravinsky sits in the lower left corner of the frame, visible only from the chest up. Dominating the picture is a grand piano, with its lid propped open at a 45-degree angle. The picture collapses the distance between the piano and the background, seen through a triangular gap, the boundaries of which are formed by the piano, its lid and the lid’s prop. A visually striking triangular shape ‘pops out’ in the middle of the photograph, and the distinct tones of the background create a second, much smaller, triangle near the apex of the larger one. Another object of visual interest is how the angle of the piano lid, resting on its prop, mirrors the angle of Stravinsky’s arm, propping up his head.

The photograph may be claimed by empiricists (pure or enlightened) to be aesthetically arresting by virtue of the above formal properties being visually absorbing to behold. But anti-empiricists insist that it is aesthetically arresting for being a product of Newman’s photographic skill and intelligence. That is, the picture commands aesthetic interest by virtue of being the causal upshot of Newman’s intentionally and knowledgeably arranging the piano and Stravinsky in the way he did. A particular aesthetic problem, the problem of how to take Stravinsky’s portrait, is solved here in an interesting way. The photograph is a manifestation of Newman’s aesthetic expertise in general, and his photographic expertise in particular. On anti-empiricism, the photograph merits aesthetic admiration for these facts, and not because of the perceptually manifest, formal features it possesses. On anti-empiricism, the visual delight we take in the photograph is a reflection of our prior, cognitive grasp of what Newman has achieved in producing the photograph. The photograph’s disposition to produce visual delight is not per se what its aesthetic greatness consists in.

Compare a non-photographic example (see Davies 2004: 82): Turner’s Snowstorm is visually stunning, but one might think that its aesthetic greatness consists in a further, non-perceptual fact: how Turner substituted color for tone in order to achieve that composition, how he solved an aesthetic problem in order to complete the work. Snowstorm, like Newman’s portrait of Stravinsky, is a manifestation of an artist’s pictorial skill and intelligence. On anti-empiricism, the works merit aesthetic admiration for this fact, rather than their experiential effect on viewers, as pleasing as that may be.

Insofar as we ever admire photographic portraits for the skill that has gone into composing and staging the scene photographed, we thereby admire the actions of the photographer. Revised CTP identifies this action with the evental structuring cause of the photograph. The act is sufficiently far back in the photograph’s history that it is not plausible to say that it is perceptually manifest when seeing the photograph. Hence the act of scene-arranging is grasped cognitively and requires an anti-empiricist treatment.⁸ Revised CTP and anti-empiricism are thus natural allies, and anti-empiricists can flesh out the Anti-Empiricist Argument so that it does not merely lay the grounds for a reductio of the claim that photographs may be objects of aesthetic appreciation.

Both Scruton and critics of CTP may be unmoved. For it might be replied that to appreciate a photograph as a photograph is not merely to appreciate artefactual features, but requires appreciating artefactual features that are specifically photographic, whether intrinsic to the photograph or part of its causal history. Scruton and CTP’s critics may thus allow that an interest in the photographer’s action of arranging the photographed scene is an interest in a photograph’s causal history, construed along evental structuring lines, and may even allow that the interest is bona fide aesthetic in nature (though for Scruton some back-peddling on empiricism would be required). Still, both may deny that this is a distinctively photographic aesthetic interest, claiming that the source of that interest is not related to the nature of photography per se. After all, there seems no obvious connection between the act of arranging objects in the world, the target of the aesthetic interest in question, and the medium of photography. If that is right, then the Anti-Empiricist Argument remains to be fleshed out, and reflection of photographic portraiture fails to support an anti-empiricist aesthetics of photography, qua photography.⁹

The strength of this objection relies upon Scruton and CTP’s critics identifying specifically photographic features of such pictures, and then showing that arranging objects in the world fails to be among, or related to, such features.

First off consider what both Scruton and CTP identify as a specifically photographic feature: causal dependence upon concrete reality. Does arranging objects before the camera fail to be related to this feature? Hardly. Rather, Newman’s act of arranging such objects precisely serves to modify that dependence. For had the objects been placed differently, then the causal dependence would have differed. An aesthetic interest in Newman’s action of scene-arranging therefore does count, on CTP at least, as a photographic aesthetic interest. So by CTP’s own lights, the aesthetic interest in skilful arrangement characteristic of photographic portraiture ought to count as photographic.

This reply can be motivated in a second way, independent of CTP. Indeed, consider how one of Scruton’s critics, Paloma Atencia-Linares, understands the idea of specifically ‘photographic features’ or ‘photographic means,’ making manipulation of light the key factor:

Photographic means: any action or technique performed or taking place during the production of an image, including the stages of transduction and storing, that consists solely in the exploitation, manipulation, or control of the incidence of light onto, and its interaction with, a photosensitive material. (2012: 22)

Atencia-Linares’s primary use of this definition is to argue that some darkroom manipulations count as photographic and so do not turn photographs into handmade pictures, as some defenders of CTP, like Scruton, affirm. Such manipulations include over- and under-exposure, combining negatives, burning or blurring, and so on. These count as photographic means of manipulation, and the end result remains a photograph, on Atencia-Linares’s view, insofar as they all involve “the action of light on photosensitive material” (2012: 22).

Crucially, this understanding of ‘photographic means’ extends not only to certain darkroom techniques, but also to actions undertaken during the ordinary creation of a photograph, since the latter certainly involves “the recording of light” (2012: 22) on a light-sensitive surface.¹⁰ If so, then there is no difficulty in picturing an interest in Newman’s actions of scene-arranging as a photographic interest, since what those actions effected was precisely a certain pattern of light distribution being recorded when the photograph was triggered. The skilful arrangement of objects in a to-be-photographed scene just is a way of directing ‘the action of light on photosensitive material’, rendering that action photographic on this anti-CTP account and thereby rendering aesthetic interest in that action a photographic one as well.

Thus, by both CTP’s and Scruton’s own lights, as well as on independent anti-CTP grounds, an interest in how Newman arranged objects in the scene before his camera can be characterised as an interest in something photographic: skilful manipulation of the resulting photograph’s causal dependence on reality or skilful manipulation of light on a photosensitive material.

However, insofar as not all photographs involve a prior act of arrangement on the part of the photographer, this is not an aesthetic interest that applies to all photographs. Our next question should be what an aesthetics of triggering photographs would look like, and whether it too has an anti-empiricist flavour. In doing so, we will further flesh out the case for an anti-empiricist aesthetics of photography.

7. 

If there is an aesthetics of structuring photographs, might there be an aesthetics of triggering them too? I noted above that an aesthetic interest in how the photographer has arranged the scene will not apply to all photographs, since some are taken on-the-fly. Street photography is a case in point. In this section, I argue that revised CTP, armed with the triggering/structuring distinction, and combined with anti-empiricism, can explain our aesthetic interest in such photographs; specifically, as an interest in the actions of the photographer in triggering them. Such an interest is an interest in photographs being ‘well-timed.’ Insofar as the well-timedness of a photograph refers to facts about its triggering cause, and insofar as that is sufficiently far back in its causal history, it is not plausible to think it perceptually manifest in the resulting photograph. This feature of a photograph must be appreciated cognitively, in an anti-empiricist manner.

By way of illustration, return briefly to the formal elements of Newman’s photograph of Stravinsky; in particular, how the angle of Stravinsky’s arm mirrors the angle of the piano lid. One aspect of photographs that we often find absorbing is how these patterns or rhythms, sometimes called echoes or motifs, repeat throughout a photograph. Street photography often involves such repeating patterns, particularly as they relate to repeated bodily forms. On revised CTP and anti-empiricism, what we are taking an aesthetic interest in here is the action of the photographer in triggering the photograph at a particular time; that is, their action of pressing the shutter release to capture fleeting and repeating patterns or forms in the world, and not such patterns and forms on the picture-surface itself. Again, examples serve to make this idea vivid.

First example: the lead photograph in Henri Cartier-Bresson’s photo-essay China: The Great Leap Forwards. A sea of dancers is depicted in mid-leap, with their legs skipping, seemingly effortlessly off the ground and in near-perfect synchronisation. Formally, the photograph is visually compelling by virtue of these repeating patterns of bodily positions. But what merits aesthetic admiration, on anti-empiricism at least, is Cartier-Bresson’s having captured the scene at that particular moment, how he achieved a photographic record of repeating patterns that actually occurred in the world via pressing the shutter release at the precise moment he did. On revised CTP, this is what gives the photograph its aesthetic ‘wow-factor’.

The above idea is independently plausible. Cartier-Bresson’s pictures are frequently praised for being well-timed. This straightforwardly appears a term of aesthetic praise, one that concerns the photographer’s skill at capturing a fleeting scene in concrete reality at a certain point in time. In using the term ‘well-timed’ to describe this photograph of Cartier-Bresson’s, we refer to an achievement on his part, one that revised CTP identifies with factors relating to the triggering cause of the photograph.

Second example: another of Cartier-Bresson’s photographs, his picture of the artist Alberto Giacometti taken in 1961 at the Maeght Gallery in Paris.¹¹ The photograph depicts Giacometti in the gallery, arranging his sculptures of rough-textured, spindly, human forms. Giacometti is captured standing between two particularly angular examples. But the photograph was timed so as to catch Giacometti himself looking rough, spindly and angular. Just as the sculptures seem to lean precariously, so too does their creator, with the positioning of his body a near mirror to the positioning of the sculpted figures in the foreground.

Again, we can find the formal elements of this second photograph visually interesting. But the aesthetic impact of the photograph might instead be considered a matter of its being well-timed; that is, our aesthetic interest might be thought to concern not the formal, patterned elements of the photograph per se, but Cartier-Bresson’s skilful achievement of a visual record of the corresponding patterns in the world, patterns there for a passing moment and then gone.

It is instructive to consider at this point how the aesthetic interest in Cartier-Bresson’s photographs described above would vanish, were they revealed to be hyperrealistic sketches. For then, the term ‘well-timed’ would no longer apply. An important lesson follows, which is that revised CTP, combined with anti-empiricism, cannot be criticised for picturing photography as limited in its capacity for aesthetic achievement, relative to handmade modes of picture-making. The aesthetic value of being well-timed, and which only the combination of revised CTP and anti-empiricism illuminates, is one that can be instantiated only by a photographic picture. The aesthetic achievements possible via photography, on revised CTP and anti-empiricism, are thus pictured as different to those possible via handmade modes of picture-making; they are certainly not pictured as an attenuated selection of the latter.

An objection: for all we know, as viewers, a photograph’s depicting a scene in a dynamic and lively way may have been a matter of “dumb luck” rather than “enlightened union with the moment” (Kelsey 2015: 205). So how can one say that appreciation of photographs, in street photography or otherwise, ever concerns their being well-timed? ¹²

A reply: it is reasonable for anti-empiricists to both dig their heels in and bite the bullet on this point. Digging their heels in, they may point to the fact that the term ‘well-timed’ is one that many do apply to photographs. This is something that there is no getting away from. Biting the bullet, they may allow that one can never be completely sure whether a photograph is, in fact, well-timed or a product of luck. Still, one may nevertheless have a substantial degree of justification for believing that it is, insofar as all available evidence indicates that to be the case. This, the anti-empiricist, may reply, is sufficient.

The idea that one can never be sure whether a work is a product luck or skill falls out of the anti-empiricist’s insistence that its aesthetic-making features, the achievement of the artist, is not recoverable from a perceptual inspection of the work. To know such features, the anti-empiricist thinks that a degree of background knowledge, or reasoning, or inference, or imaginative reconstruction, is required. This may be guided by general knowledge of the medium in question, or by more specific knowledge of facts surrounding the history of the making of the particular work in question, the testimony of the artist themselves, or comparison with other works by the artist, and so on.

Barring implausibly severe scepticism, suitably informed viewers should be pictured, for the most part, as judging aesthetic achievement not only justifiably, but correctly. Still, even armed with the relevant facts, we may occasionally appreciate wrongly. All the facts may speak in favour of a photograph’s being well-timed, but we may simply be wrong; the skilled photographer got lucky that time and so the term ‘well-timed’ does not apply, despite putative evidence to the contrary. So be it. This is the price to be paid for picturing facts about making as that which determine aesthetic value. Sometimes such facts are difficult to grasp, though certainly not always, and sometimes they are not even within our ken. Even when they are, mistakes may still be made, as they are in almost all other areas of human inquiry.

8. 

Revised CTP and anti-empiricism are uniquely placed to illuminate two widely-ignored aesthetic interests characteristic of photographic portraiture and street photography: an interest in how the photographer has arranged the photographed scene and an interest in a photograph’s being well-timed, respectively. Revised CTP identifies these these interests with an interest in the evental structuring of a photograph and the triggering of a photograph, respectively. Together, revised CTP and anti-empiricism offer a unique reply to the Problem of the Missing Agent: consistent with CTP is that photographers are structuring and/or triggering causes of their photographs, and this is something that we can take an aesthetic interest in.

Although I have not attempted to argue directly in favour of anti-empiricism about the aesthetics of photography, my hope is that the anti-empiricist interests that I have identified, skilful arrangement and skilful timing, are recognisable by admirers of the medium as independently plausible ones. If so, then we ought to more seriously consider the case for anti-empiricism about the aesthetics of photography, if not art in general.

A final thought on approaching photography’s aesthetic value through the lens of anti-empiricism, one related to the above discussion of luck: photography has long suffered at the hands of its detractors, who have vigorously complained that the medium lacks any serious aesthetic credentials. But if aesthetic value were as the pure or enlightened empiricists think it is, and photographs are surfaces for our sensory pleasure (whether cognitively untainted or informed), then, Scruton aside, no one would have seriously doubted whether photographs could be objects of aesthetic appreciation. Yet the problem, many of its detractors claim, is not that photographs fail to be visually absorbing or fail to elicit perceptual pleasure. Photography’s problem was never that it is ugly to look at. Rather, the worry is that that making eye-catching photographs is no real achievement. This sentiment is reinforced from a number of directions: from the reactions had by many to the eye-catching filters one can now easily apply to one’s photographs with a simple click of a button, to anxieties about whether Vermeer relied upon a camera obscura, a proto-photographic technology, to compose his paintings. The lesson that such anxiety imparts, like anxiety about luck, is that the ‘how’ of depiction matters just as much, if not more, than the ‘what.’ And this is precisely what the anti-empiricist wishes us to recognise.

Nevertheless, the story doesn’t stop there. For if what I have said here is right, then revised CTP and anti-empiricism can provide therapy for this longstanding anxiety: photographs can be aesthetic achievements. For they can be the product of a skilful act of scene-arrangement on the part of the photographer or a skilful act of timing. There may be further aesthetic merits of photography that anti-empiricism can elucidate. But these two seem sufficient for us to take aesthetic anti-empiricism seriously, in the case of photography at least, and to reconsider whether CTP is without the resources to explain aesthetic interests in photography.