Some philosophers, like Roger Scruton, famously deny that a photograph can be a work of art. On their views, whatever is truly photographic is sheerly mechanical: it is dependent on the objects of the world, not on the ideas, beliefs or intentions of the photographer. Photography cannot make art, because there is no way to intend something photographically. To help us grasp what is essentially photographic, Scruton suggests we consider what he calls an “ideal photograph,” which is (as he explains) a “logical fiction.” The “ideal photograph” is the product of photography stripped of all manipulation and reduced to what is specifically photographic. Some corresponding ideal could be posited for painting. The ideal painting, Scruton explains, stands in what he calls “a certain ‘intentional’ relation to a subject.” Among other things, that means that it stands in a certain relation to “a representational act, the artist’s act,” so that “in characterizing the relation between a painting and its subject we are also describing the artist’s intention.” “In characterizing the relation between the ideal photograph and its subject,” on the other hand, as Scruton explains, “one is characterizing not an intention but a causal process, and while there is, as a rule, an intentional act involved, this is not an essential part of the photographic relation.”1 Because of this, your interest in a photograph can always be reduced to an interest in what the photograph pictures, which we’ll call (following some philosophers) the “pro-filmic event,” or to some nonphotographic manipulation of the photographic image after it is formed. Your appreciation of the photograph can never be an appreciation of it as a photograph, because there is no way for a photograph to show something, precisely as it appears, exactly because the photographer means what the photograph shows. And that is because of the kind of action taking a photograph is, on Scruton’s account. You can take a photograph intentionally, but you cannot intend the photographic image itself. The image is an unwilled, merely mechanical copy of the pro-filmic event.

It might be worthwhile to pause for a moment to consider the nature of an action. In classical formulations of theories of agency, an action has to stand in a certain relation to an intention. So, to take an old example, let’s say a sound wakes me in the middle of the night, and I go downstairs to investigate. I turn on the light in my living room when I reach the bottom of the stairs, see nothing worth further investigation, turn the light back off and return to bed. We can say that turning the light on was an act and that I performed it.

If two burglars, who were hiding in bushes outside my door, waiting to break into my house, saw the light and ran away, can we say that I acted to frighten away the would-be burglars? No. Even though they were frightened away by the light I turned on, as an act and through my own agency, we cannot say that I performed an action in frightening them off. I did not know they were there, and so could not have intended to use the light to send them away. The right relation to an intention is missing.2

This, I take it is why Scruton will allow that taking a photograph is typically an action, but that no feature of the ideal photograph can be thought of as the product of the photographer’s agency. Making the photograph is like turning on the light downstairs; the particular qualities of the photograph are like the flight of the panicked burglars—outcomes that lie beyond the reach of my intentions.

Curiously, the very analysis of photographic agency—or nonagency—that makes Scruton want to rule out the possibility of photographic art makes photography particularly exciting to postmodernist artists and critics as an artistic medium. What they like so much about photography is the way it attenuates authorial intention. A recent special edition of the journal Critical Inquiry featured essays about artists since the 1960s exploiting photography’s “automaticity,” which they understand to be its diminished form of agency. “Chance,” “mechanicity,” “accident”—these are the terms that are valorized in this celebration of photography since the time of what the literary critic Roland Barthes in 1967 called “The Death of the Author.”

One of the artists featured in this special issue was Tacita Dean. She makes films and uses still photography. Her most famous project, and the topic of an essay by the art historian Margaret Iversen in this issue of Critical Inquiry, is a photo book titled Floh.3 That’s the German word for “flea,” and it is the title of Dean’s book because she bought all of the photographs in it at flea markets. The photographs are accompanied by no text or other meaningful intervention. The point is just to collect, arrange, juxtapose, in a matter-of-fact way, this assortment of photographs Dean found and bought.

Iversen explains that the way Dean finds the photographs reflects the artist’s commitment to chance.4 Like the surrealist poet André Breton, Dean values the chance encounter enabled by browsing the flea market for its ability to put one in touch with “external circumstances” (813). This openness to chance (815) is reflected in the photographs themselves in two ways: first, they “are,” as she puts it, “a regular inventory of technical errors” and, second, are damaged in ways that show their availability, as objects, to “accidents” (813). In the end, Iversen explains, “[i]t is as though the condition of the medium were being explored by illustrating everything that can go wrong” (813).

By saying that accidents in the making of the photographs and in their preservation reveal the “condition of the medium,” I take Iversen to mean that chance is essential to photography—specifically analogue photography. “Of course,” Iversen qualifies herself,

artists using analogue film exercise considerable agency selecting camera and film, in framing, focusing, and setting aperture size, time of exposure, and so on, as well as similar choices throughout the printing process. Yet, … all these forms of intervention do not compromise the analogue’s photo-chemical continuity with the world. The analogue is defined as a relatively continuous form of inscription involving physical contact (797).

Projects like Floh thematize this openness to the world. The haphazard quality of Floh’s photographs and the variable state of their preservation, Iversen argues, “associate analogue photography with a kind of attentive exposure to things in the world marked by chance, age, and accident” (799).

Strange as it seems, then, the photograph that best exemplifies the medium of photography turns out to be a bad photograph. It helps, of course, to find it by browsing and it doesn’t hurt if it’s a little worse for wear, but the as-it-were formal guarantee of photography’s openness to the world is an unambiguous “inventory of technical errors.” Errors show the withdrawal of the photographer’s intention from the photo-chemical contact between the film and the world.

Iversen explains that Dean courts this withdrawal in her filmmaking practice by setting up wide-angle static shots and holding the frame until the film runs out. Dean explains: “It is just allowing the space and time for whatever to happen, and that comes very much from the nature of film” (817).

In a certain sense, this is the opposite of what Iversen says about the flea-market photographs. If what Iversen likes about the flea-market photographs is that they’re full of errors, the great thing about the static shot as Dean does it, is precisely that nothing can go wrong. She’s out to film just exactly whatever happens. But in another way, Iversen’s being perfectly consistent. Cherishing technical errors in snapshots is just another way of avowing an interest in photography by embracing the absence of authorial agency in the photograph. And nothing can go wrong in the shooting of a film if all you really care about is showing that whatever ends up in the film is indifferently okay with you precisely because the project has nothing to do with what you wanted to show. The tautological embrace of whatever happens and the valorization of what the photographer does accidentally amount to the same thing—a positive revaluing of photography’s ability to keep authorial intention out the picture, so to speak.

Dean and Iversen essentially reproduce Scruton’s account of photography. The big difference is that, since they are postmodernists, they are not troubled by art without authorial intention or agency. Or, perhaps it would be better to say that, rather than no agency, the artist in Dean and Iversen’s mode has agency—even Scruton will allow that taking a photograph is an action—but the artist has a diminished or attenuated agency. One that extends to getting the photograph, but not to what the photograph looks like. Since, at least in our present paragon, Dean’s book Floh, this mode of getting is shopping, let’s call the mode of photographic agency they offer photography-as-buying. You pay your money; you get a photograph. What the photograph shows—its subject, its technical errors, its wear and age—are not the outcome of your action. The artist didn’t make them, and so the artist cannot mean them. In my view, as in Iversen’s, buying a photograph is a pretty neat metaphor for this view of photographic agency, and it is one she shares with Scruton and Dean.

I don’t share it with them. I don’t think it’s right. Dawn Wilson has recently offered a criticism of Scruton’s account of photography. Roughly, Wilson argues that, even without retouching (or some other nonphotographic manipulation), a photograph could indeed be made to show something that looked different from the pro-filmic event. She argues that the making of a photograph is a multi-stage process that includes key decisions about processing and materials and that it includes them not incidentally, as Scruton and Iversen treat them, but constitutively. Otherwise, holding an empty frame up to the space before one’s eyes could count as a photograph.5 Thus we have a multi-stage event—the photographic event—that includes a “merely causal relation” to the pro-filmic event and another part that may permit this “causal phenomenon” to “be mastered and creatively exploited by skilled artists” (“Photography and Causation,” 340). Thanks to this mastery, the photograph can, perhaps, come to stand in an “intentional relation to the subject” because the “causal relation places no constraints on what a photograph may depict.” I take this to mean that the photographic image’s causal relation to its subject does not necessarily determine what the photograph looks like; a skilled artist may be able to harness it to an intention and so make the photograph “transparent to human intentionality” (“Photography and Causation,” 340).

Wilson gives an example of a photographic event that produces a photograph that does not look like the pro-filmic event—a photograph of a moving train made with an exposure time too long to show the train clearly. It appears as a blurred streak, instead. Certainly, the resulting photograph is “transparent to human intentionality”—at least if we can understand Wilson to have intended to make a colorful photographic abstraction rather than a picture of a train. The guarantee of the transparency of Wilson’s photograph to her intention is its difference in appearance from the actual pro-filmic event. It proves it means—that the picture stands in an intentional relation to the photographer’s action—by failing to resemble a train.

I want to offer another example, one that supports Wilson’s thesis, but not because the photograph fails to resemble the pro-filmic event; rather, this photograph conveys a vivid impression of its subject. This is a very famous photograph by Walker Evans of Allie Mae Burroughs, one of the tenant farmers who hosted Evans and his collaborator, James Agee, in 1936 while they stayed in Hale County, Alabama, to research and document a story for Fortune magazine. That project, which resulted in the modernist masterpiece Let Us Now Praise Famous Men, was reviewed by Lionel Trilling shortly after its appearance in 1941. “Evans’ pictures,” he writes,

are photographic in the sense that people mean when they say “merely photographic,” they are very direct, they even appear to be literal, and how the moral quality gets into them I do not exactly know; I suppose it is because Evans wants it to be there.6

Not quite a page later, Trilling tells us that part of the picture’s greatness is the sitter’s refusal to “be an object of your ‘social consciousness’; she refuses,” Trilling continues, “to be an object at all—everything in the picture proclaims her to be all subject” (“Greatness,” 376-77).

Trilling shows how alive he is to the problems of photography. To speak of a picture being “merely photographic” or “literal” is Trilling’s way of alleging that it is simply a record, a mechanical record of the pro-filmic event. But he is careful to indicate that this is merely a common usage—just the way people speak about certain photographs. What he calls the moral quality gets in there, though, somehow. To put it in philosophical terms, Trilling supposes the photograph, literal as it is, to be transparent to Evans’ intentionality, at least where that moral quality is concerned.

But then, Trilling almost immediately reverts to something Scruton might say, giving all the credit for expression to the sitter. If she is not an object at all, she is a subject. That is, she is the one with the agency, not Evans, and our interest in the photograph turns out to be in the sitter, in the pro-filmic event, after all. Who expresses him- or herself in this photograph—Walker Evans or Allie Mae Burroughs? If Burroughs is “all subject” are we to understand this photograph the way Scruton understands an ideal photograph?

Let me propose another possibility. If Evans is to have agency in the photographic event, that agency need not consist in preventing the photograph from resembling the pro-filmic event. That was Wilson’s strategy in the photograph of the blurred train. Her strategy might make a case for photographic agency in a much wider field of photographic work if we apply it to straight photographs like Evans’. In other words, photographic agency might not consist only in making a photograph look different from the pro-filmic event. It might consist in securing its resemblance to the pro-filmic event in respect to some quality the photographer means to record, even quite literally.

When Trilling notes the similarity of the sitter’s mouth and eyebrows and fine wrinkles to the contours and texture of the boards behind her, he is noting features of the real objects in the picture, which a photographer of Evans’ masterful skill can make simultaneously visible, and even pictorially significant in their similarity, where a lesser photographer would very possibly fail. That Evans’ photograph revealed to Trilling the importance of wear and age, distress and wearing out, makes it specially suitable for embodying these themes, which are central to Agee’s text. Thus, Evans expresses a thought in even such a “merely photographic” image.

As I say, I have introduced this reading of a single photograph not against Wilson, but to propose that agency in photography can consist in making the photograph depict the pro-filmic event “merely” photographically, as in Scruton’s “ideal photograph.” And this is valid for the reason Wilson indicates: that, in the making of a photograph, there is so much room for agency that even the most literal depiction cannot be ruled out as the outcome of an expressive action. Even though a similarity exists between Burroughs’s face and the boards of her shack, a photographer who wants it to count pictorially will think hard about placing and lighting the sitter, about lens, aperture, and exposure time, about the choice of paper to print the photograph on, and other things, too.

Indeed, even to notice the similarities between the wear in Burroughs’ face and the wear of the boards of her shack would require a kind of artistic and specifically photographic insight. Her face will have been full of tans and reds and other hues. The boards of the shack will have been the gray of weathered and unsealed pine. She would surely have been animated by the mobility, intelligence, and emotional presence of a real person; the shack’s wall would have shone silently in the summer sun—a thing against which to see other things, not an interlocutor. For Evans to have constructed the metaphor he then captured by his skill in mise-en-scène and photographic technique—this is, taken altogether, the kind of expressive space Wilson imagines for the photographer’s agency within the photographic event. Even when the photographer’s aim is to produce a photograph that renders literally and makes pictorially effective certain features of the pro-filmic event.

Of course, it might have happened by chance. Evans might have set Burroughs against the wall without thinking of any kind of visual rhyme, so to speak, between her face and the wall. And, unlikely as it might seem, the light and his camera’s settings might have been just such that the likeness appeared by itself, unnoticed, perhaps, until Trilling spotted them. One might also say that, Wilson’s argument notwithstanding, a blurred train is surely the subject of countless photographic mishaps, too. Does the—admittedly small—chance that Evans’ photograph of Burroughs was a happy accident mean that Scruton was right, after all? Does it mean that a photograph can only ever be merely the equivalent of pointing at the pro-filmic event?

No. In fact, the possibility only underscores what I’ve been saying all along. If we want to argue that a photograph is to be a vehicle for meaning, we must put the photograph in some intentional relation to its author. Worrying that something might be meaningless because it might be nothing more than an accident is just a way of insisting on the centrality of intention to agency.

So what we need is an account of agency that is suitable for a meaningful photography. The account we have already seen—photography-as-buying—won’t do. Rather than emphasize the agency of the photographer, it likens the photographer’s place in photography to a customer’s place in a flea market, where one might buy a little packet of old photographs just to see what one gets in something like the way Dean sets up a camera to see what happens. What results may be interesting or evocative, but it won’t be an act of expression. If we want photography to be a vehicle for genuinely photographic acts of expression, we need a different account of photographic agency.

Really, what we need is to imagine a different relation between intentions and outcomes. The advantage, and the failure, of the postmodernist account—what we called photography-as-buying—is that the relation of the act—buying—and the outcome—the photograph—is one of indifference. Just like Iversen’s account of Dean’s filming whatever happens, where the point is to let the mechanical openness of photography to whatever happens deliver just that. The problem with buying as a model for photographic agency—or for artistic agency more generally—is the indifferent relation between buyer’s intention and the outcome. To put it slightly differently, buying cannot happen and fail. If I buy something, I may be sorry later, but that doesn’t mean I didn’t buy it. By contrast, if I make a work of art, and, looking at the finished work, decide it does not embody anything I can mean, I can say the making failed.

Maybe we need an economic model that captures the essence of artistic agency better than does buying. The Victorian intellectual Herbert Spencer wrote an account of the history, origins and nature of economies in his Principles of Sociology. Spencer attacks what he takes to be the typical conjecture about the origins of economic exchange—that it begins with “barter.”7 Spencer goes on to explain his theory that the root of economic exchange is “certain ceremonial actions originated by the desire to propitiate,” or to win favor with someone (388). The example he uses to make familiar to his reader this kind of ancient ceremony is what he calls “swopping,” which is to say, the practice of buying rounds among groups of men drinking in bars. “We have here, indeed,” he says, “a curious case, in which no material convenience is gained, but in which there is a reversion to a form of propitiation from which the idea of exchange is nominally, but not actually excluded” (391). Exchange is not “actually excluded,” of course, because once you’ve bought, or “stood,” your round, you clearly expect someone to buy the next, and so on, until everyone is even, as if he’d been buying his own drinks all along. So in effect, it’s just like buying your own drinks. That is why Spencer can say that no material convenience is obtained by swopping. Seen from a distance, so to speak, it’s just a way of buying yourself drinks. The difference, however, is that you may buy your round but not find yourself treated to a round in return. Your expectation of reciprocity is internal to the act of swopping, and yet, you may successfully stand your companions a round without that expectation being fulfilled. It is like buying, except that, even when the act is fully accomplished, you don’t necessarily get what you paid for.

The great Victorian photographer Peter Henry Emerson regarded Spencer as a kind of intellectual hero.8 And Emerson suffered a kind of crisis of faith in photography. The details are too much for this conclusion, but he was worried that there wasn’t enough room for the photographer’s agency in the making of a properly photographic picture. In fact, after fulminating about the kind of work that could count as properly photographic for several years, Emerson’s crisis led him to renounce photography as an art. He nevertheless continued for several years to publish photographic books. Actually, I don’t think he ever stopped wrestling with the question of photographic agency.9

So it strikes me as deeply important that, just as he was preparing to renounce photography as art, he published a photo book framed by a story about buying and selling. The book, Wild Life on a Tidal Water, recounts the adventures of Emerson, his friend, the painter T.F. “Dick” Goodall, and their factotum, “Joey,” who works for them on their houseboat on the broads near Great Yarmouth.

In the first chapter, Joey is frying some fish for Emerson and Goodall. Emerson is so pleased at the quality of the fish that he asks Joey how much he paid for them. Joey responds that they cost “a shillun’,” and Emerson tells him to keep them at that price. Joey goes back into the kitchen and then reappears, agitated, to say that he’d shopped widely to find a good bargain, which was why he was so long in town that morning. Emerson and Goodall think they understand why Joey was defensive about his long shopping trip—they knew he liked beer, so they assumed that he’d been drinking on his shopping trip.

The book ends with a coda to this opening anecdote. Joey has been dismissed, and Emerson has returned to London, where he receives a letter from Goodall, who thinks he can explain Joey’s good luck in finding fish. He had been in town with the new hired man, Ben, searching for fish to equal Joey’s. They are discussing the problem in the street when they are approached by a man who claims he can help them. The man leads them to a pub. Puzzled, Goodall and Ben watch the stranger, who is the landlord of the pub, pull a sole and a skate from behind the bar and give it to them. Goodall asks how much the landlord wants for the fish. Goodall relates the exchange that follows:

“He said, ‘Give me what you like; I don’t want them.’

“I declined, and pressed for a price, but with no result; so I stood him a drink, and Ben a drink, and myself a drink. This ceremony over, I said, ‘How much for the lot now?’

“’O, give the little girl something, and take ‘em.’

“At last, after much pressing, he named a ‘shillun’, and we dealt.” (WLTW, 121-22)

Goodall asks the landlord how he gets the fish. He answers that the fishermen leave them as gifts, and he ends up with more than he can use. Goodall immediately surmises that Joey got his fish the same way and drank his “shillun’” at the bar each week. The landlord positively identifies Joey from Goodall’s description, and Goodall reports his conclusion to Emerson, who accepts it and asks if we, his readers, do not agree that the mystery of Joey’s “festive manner on fish days” has been solved.

I want to propose that Emerson has constructed this story to show a relation between intentions and outcomes, between expectation and causation. Emerson and Goodall think they are buying fish—understood as receiving fish because they have paid money for them via an agent, Joey. In fact, however, they are standing Joey a round or two at the pub, and he is reciprocating by regifting them a windfall of fish. The theme of ceremonial gift-giving is repeated, ostentatiously, in Goodall’s account of his attempt to buy the fish from the barman. (He is permitted to pay for the fish in the end, but only after standing a round and being encouraged to give a little girl some change. The girl, by the way, receives no other mention in the story, and so is presumably to be thought of as some lesser figure in the pub—a bar-back, perhaps—and exists only to supply a target for unmotivated, and therefore ceremonial, gift-giving.)

This might distract us from photographic agency and tempt us to think about the nature of buying qua act. If I give another person a certain amount of money called a price for a commodity, and they give me what I ask in return, precisely on the condition that I give them the price of it, we will typically say I have bought something. Just like Tacita Dean in a flea market. Success is tautological. Swopping is different, as we’ve noticed—although an expectation of reciprocity is built into it, it can be successfully accomplished without that expectation’s fulfillment. The outcome stands in a certain relation to an intention—neither tautological nor indifferent. This is, I would claim, much more like art making as we have always understood it than buying is.

Think of painting. Now, Scruton, in discussing painting, said our understanding of such a work of art always led us back to the artist’s intention so that “in characterizing the relation between a painting and its subject we are also describing the artist’s intention.” Let me note, however, that he is far from saying that painters can make the paintings they want to make just by intending them. They bring expectations to the act of painting, and perform actions in making paintings, but their expectations may go unmet. Just as in swopping.

Emerson explains this in a series of propositions on art from the 1899 edition of Naturalistic Photography. This work was written in the depths of Emerson’s crisis. In his Proposition XII, a painter and a photographer set out to picture a stretch of shore:

The photographer at once sets up his machine, focuses and exposes; but in these very processes his ideal has gone. What results may be beautiful, but it is no more the representation of his ideal, the vision he first saw. It is something else, for the machine imposes certain conditions which were never in the photographer’s mind at all.10

In this, according to Emerson, the photographer is different from a painter. For the painter, “each touch helps to his desired or ideal end . . . everything is done unto one end, and all is certain from the first . . .” (I:185-86). That is because the painter’s work is guided by his or her intention at every step. That leads Emerson to say a painted work is “a perfect index of its creator’s mind.”

And yet, Emerson himself explains why this is not so. The key criterion of artistic success for Emerson is the full expression of what he calls a “sense of beauty” or the “ideal.” But this “sense of beauty” “will vary with individuals and in the individual from day to day, nay, from hour to hour” (I:182). So, painting isn’t so certain after all. As the painter proceeds, no matter how fine the original inspiration, his or her intention may vary, resulting in a work that does not correspond to the original intention or even to any of the shifting, partial intentions that informed and are indexed by the marks that make up the final work. This seems like what I have in mind if, looking at a student’s clumsy finished painting, I say (or more likely, think silently) he didn’t pull it off. He didn’t manage to perform all the little intentional acts that add up to the realization of an original, or at least integral, intention. The manugraphic character of painting isn’t enough to make it so transparent to an intention as to make it “certain.” Or even, to make it different from photography in this respect.11 You can’t make the picture you want to make by intending to make it. Not with a camera, not with a paintbrush.

If I understand Emerson correctly, his allegory of photography contends that, while photography make look like shopping at a flea market, it’s really more like standing a round at the bar. This is important because, as we have seen, if all is certain from the start, as in buying as we commonly understand it, the relation of intention to outcome is taken to be tautological. So the outcome cannot be understood as the successful embodiment of the artist’s meaning. If the act is left open to chance, so that the outcome is indifferently acceptable, it can’t be understood as the successful embodiment of the artist’s meaning, either. Only a model of photographic agency like the one Emerson proposes as it were allegorically can help us understand how we make photographs we mean.

I would like to thank audiences at the London Aesthetics Forum, the Los Angeles County Museum of Art, and elsewhere for questions and readers, especially Michael Schreyach and Elizabeth Mead for comments and questions.