The World and the Will: On the Problem of Photographic Agency
Walter Benn Michaels’s recent contribution to a nonsite symposium on photography and philosophy takes its title from G. E. M. Anscombe’s admittedly “paradoxical and obscure” sounding claim that in acting intentionally, “I do what happens”:
That is to say, when the description of what happens is the very thing which I should say I was doing, then there is no distinction between my doing and the thing’s happening.1
Michaels goes on to argue in his essay that the account of human agency that Anscombe endorses is sharply different from that of Donald Davidson, despite the common depiction of the two of them as allies in an “Anscombe–Davidson view” of intentional action, and that the difference in their respective positions can be brought out by considering the peculiar form of agency involved in photography. I believe that Michaels is right on both counts, but wrong in some of the details of his argument—partly because of what I will argue is a misreading of one of the texts on which that argument is premised.
Let’s begin with Anscombe’s “formula.” As Michaels notes,2 Anscombe writes in Intention that she “came out” with it in considering the following remarks in Wittgenstein’s Tractatus:
6.373 The world is independent of my will.
6.374 Even if everything we wished were to happen, this would only be, so to speak, a favour of fate, for there is no logical connexion between will and world, which would guarantee this, and the assumed physical connexion itself we could not again will.3
Here is how Anscombe glosses the position that she finds here—it is within the context of considering the strange idea that “I really ‘do’ in the intentional sense whatever I think I am doing”:
E.g. if I think I am moving my toe, but it is not actually moving, then I am “moving my toe” in a certain sense, and as for what happens, of course I haven’t any control over that except in an accidental sense. The essential thing is just what has gone on in me, and if what happens coincides with what I ‘do’ in the sphere of intentions, that is just a grace of fate. This I think was Wittgenstein’s thought in the Tractatus… (Intention, 52)
Having explained the Tractarian position as she understands it, Anscombe then calls it “nonsense,” and criticizes her teacher as follows:
if nothing guarantees that the window gets opened when I “open the window,” equally nothing guarantees that my toe moves when I “move my toe”; so the only thing that does happen is my intention; but where is that to be found? I mean: what is its vehicle? Is it formulated in words? And if so, what guarantees that I do form the words that I intend? for the formulation of the words is itself an intentional act. And if the intention has no vehicle that is guaranteed, then what is there left for it to be but a bombination in a vacuum? (ibid.)
I find it hard to be wholly persuaded by Anscombe’s anti-Wittgensteinian argument here, partly because I’ve been brought up in a philosophical tradition where we are comfortable thinking of the “vehicles” of our mental attitudes as the states of a person’s brain. But I’m nevertheless strongly attracted to the alternative position she recommends, with its refusal to construe intentions as mental items that exist alongside bodily movements and their downstream effects. For Anscombe, if we wish to explain what intentional actions are, it “will be a mistake to look for the fundamental description of what occurs—such as the movements of muscles or molecules—and then think of intention as something, perhaps very complicated, which qualifies this. The only events to consider are intentional actions themselves, and to call an action intentional is to say that it is intentional under some description that we give (or could give) of it” (Intention, 29).
This last claim is one that Dominic McIver Lopes, whose account of photographic agency Michaels means to be disagreeing with, professes his own allegiance to. Lopes suggests that the key to understanding intentional action is “a bit of action theory originally due to Anscombe, made standard by Donald Davidson”:
An action is an event, one that happens intentionally. However, attributions of agency and intentions serve different purposes. In attributing agency, we credit an agent or hold them responsible for what happens. In attributing intentions, we explain or justify what an agent does—we work out their reasons for acting. Moreover, one and the same event can be intentional under some descriptions but not others. The guard flips a switch, turns on a light, draws 5.4 Amps, illuminates the room, and alerts a prowler: these are one action described in four ways, and what the guard does is intentional under some but not all of these descriptions—she did not intend to draw 5.4 Amps or alert the prowler, for example. An event is an action as long as it is done intentionally under some description. We explain the action by attributing intentions relative to descriptions, but we credit the act under all descriptions. The guard did draw 5.4 Amps, though she did not intend to, because what happens is 5.4 Amps get drawn and that is the same event as turning on the light, which she intended to do.4
While this position does indeed coincide with Anscombe’s on many particular points, her fundamental disagreement with it can be brought out by looking at the third sentence quoted here. According to the Davidsonian position that McIver Lopes favors, we are to treat “what happens” in action as a particular event that can be singled out independently of its connection to a person’s activity, so that “attributing agency” is a matter of postulating some relation between the two: here, a relation of credit or responsibility. On this position, we find intentional activity whenever a person, or a certain group of her mental states, causes something to happen, where what happens corresponds appropriately to the agent’s internal representation of it. Anscombe, however, says that when we act, what happens is simply what we do. Yet this formula does sound pretty “paradoxical and obscure,” and this is surely a large part of why the dominant tendency among analytic philosophers has been to read Anscombe as having largely the same view as Davidson. Can we bring out the coherence of her position by reflecting on the nature of drawing as a form of image making?
Michaels seems to me to get right to the heart of these matters in his discussion of the work of Robert Morris in his essay “Blind Time (Drawing with Anscombe).” Morris’s Blind Time drawings were made blindfolded or with his eyes closed, the artist moving his hands across the canvas in a series of planned steps. In his essay, Michaels quotes Jean Michel Roy’s illuminating description of the philosophical vision behind this project:
Traditionally, [Roy] says, we might think that “drawing is a direct product of an intention” (136). We would, for example, describe ourselves as “drawing a horse” or “drawing a diagonal line” and we would think of the horse or the line as the product of our intentions. But the Blind Time drawings and the accompanying commentary show that that would be a mistake. The artist’s intentions are not to draw a horse but to move his hands on the paper in various ways (“upward” and then “outward”); the drawings are what happens when he does that. Thus, Roy says, they should be understood as “the by-products of the artist’s intentions and not its products” (137).5
Michaels then draws an instructive parallel between this position and that of Davidson in his paper “Agency.” For Davidson, the relation of what he calls a person’s “primitive actions” to any wider descriptions that hold true of what a person does is a relation of cause to effect:
When we infer that he stopped his car from the fact that by pressing a pedal a man caused his automobile to come to a stop, we do not transfer agency from one event to another, or infer that the man was agent not only of one action but of two. We may indeed extend responsibility or liability for an action to responsibility or liability for its consequences, but this we do, not by saddling the agent with a new action, but by pointing out that his original action had those results.
We must conclude, perhaps with a shock of surprise, that our primitive actions, the ones we do not by doing something else, mere movements of the body—these are all the actions there are. We never do more than move our bodies: the rest is up to nature.6
We should be struck by the similarity between Davidson’s position and the Tractarian one that Anscombe rejects. Both positions depict the core of agency (either “willing” or “primitive action”) as on one side of a divide and the world (the results, the wider descriptions) as on the other, with the connection between these things depicted as something further that a human agent must presume but is powerless to effect. And this is exactly the opposite of Anscombe’s own view. For her, “what happens” when we act, even under descriptions in terms of what goes on at a distance from the movements of our bodies, is itself the stuff that action consists in, and not a mere consequence or result of what we do. As Michaels puts it, Anscombe’s crucial idea is that “we should not think of what we’re doing [in acting] as moving our hands in such a way as to cause there to be a drawing or a word—we should think of ourselves as drawing or writing” (“Blind Time,” 3-4).
But I think that Michaels misreads, in a way that ultimately reverses the lesson we are meant to draw from it, a passage in Intention where Anscombe reflects on the role of perception in writing or drawing. That passage follows her introduction of the formula “I do what happens,” which she acknowledges was found “extremely paradoxical and obscure” by everyone who heard it. Anscombe continues:
And I think the reason is this: what happens must be given by observation; but I have argued that my knowledge of what I do is not by observation. A very clear and interesting case of this is that in which I shut my eyes and write something. I can say what I am writing. And what I say I am writing will almost always in fact appear on the paper. Now here it is clear that my capacity to say what is written is not derived from any observation. In practice of course what I write will very likely not go on being very legible if I don’t use my eyes; but isn’t the role of all our observation-knowledge in knowing what we are doing like the role of the eyes in producing successful writing? That is to say, once given that we have knowledge or opinion about the matter in which we perform intentional actions, our observation is merely an aid, as the eyes are an aid in writing. Someone without eyes may go on writing with a pen that has no more ink in it; or may not realise he is going over the edge of the paper on to the table or overwriting lines already written; here is where the eyes are useful; but the essential thing he does, namely to write such-and-such, is done without the eyes. So without the eyes he knows what he writes; but the eyes help to assure him that what he writes actually gets legibly written. In the face of this how can I say: I do what happens? If there are two ways of knowing there must be two different things known. (Intention, 53)
The second to last sentence quoted here should signal to the reader that Anscombe is not putting this argument forward in her own voice: it’s rather a natural line of reasoning that she thinks will lead a person to resist the formula “I do what happens,” since it gives expression to a picture of agency that makes that formula appear so paradoxical. On that picture, in writing a word a person may have to rely on her eyes to know what is written when she writes something, but the fact that she is writing a certain thing is something she can know even if her eyes are shut. And this reinforces the very division between “doing” and “happening” that Anscombe is out to challenge: it treats these as two different objects of knowledge, the first known from within by a person who may have no idea of the other, and the second known through perception in the same way as an outside observer might.
Yet Michaels makes the mistake, which I should emphasize is an easy one to fall into, of reading this passage as an expression of Anscombe’s own view. As he reads it, the proper conclusion of this passage is that in an activity like writing “‘the essential thing’…has nothing to do with observation—my eyes are useful in making sure that what I write is written legibly, but I’m writing what I’m writing (and I know what I’m doing) without them” (“I Do What Happens,” 2). And to say this is just to abandon Anscombe’s central thesis. If it were possible for me to be writing what I am writing even if the words weren’t appearing on the page (or screen), then the doing that is my action would stop short of the happening that is their actual appearance. And if it were possible for me to know that I am writing what I am writing even if I didn’t know whether the words were actually getting produced, then what I’d know in knowing this wouldn’t be “exactly the same thing” that can also be known “by observation of what takes place” (Intention, 51). We would be back to something like the “mad account” that Anscombe started off with, on which “what one knows as intentional action is only the intention, or possibly also the bodily movement; and that the rest is known by observation to be the result, which was also willed in the intention” (ibid., 52).
This is not the place for me to take on the very difficult question of how Anscombe wishes us to understand the role of perception in the knowledge of one’s own intentional actions.7 Rather, I have focused on the proper interpretation of the case in which Anscombe writes with her eyes shut because doing so is a way to bring out very clearly the commitment of Anscombe’s that Michaels wants us to endorse, even if his own understanding of the case ends up pushing us in what I think is the wrong direction. What makes Morris’s artistic practice so strange is precisely that it represents the artist’s drawings in the way that “successful writing” is treated by Anscombe’s imagined interlocutor: these things may result from a person’s activity so long as all goes well, but whether this actually happens is something that the agent must rely on observation to determine. By contrast, in an ordinary productive activity like drawing a picture, writing a word, constructing a building, and so on, the agent’s relation to her work is more intimate than that.
This brings us around to Michaels’s discussion of the philosophical problem of photography, i.e. of what he calls the “pressure” that photographic agency puts “on the relation between what I do and what happens.”8 As I understand his argument, the nature of the pressure is supposed to be as follows. In drawing or painting, what’s depicted in the image that an artist creates will be depicted only because the artist herself meant to depict it. Thus, for example, it simply makes no sense to suppose that it may have “just happened” that there’s a child in the background of Courbet’s The Grain Sifters who is peering into the machine in the corner. For such a thing to happen, Courbet must have painted the boy intentionally, and so Anscombe’s “certain sense of the question ‘Why?’”9 will have application to the boy’s appearance. Having seen the boy in the image, we are immediately invited to ask: What was Courbet doing, in putting him there? Why did he make this choice? What is the significance of showing this child, within this context of domestic labor, doing the thing that he is?
As I have argued, on the Anscombean position that Michaels and I share questions like these are not at all ones about the “mind” of the artist as distinct from what he or she actually does.10 The situation is just as with the people depicted in Courbet’s painting: when we look at the painting and describe what they are doing (sifting grain through a sieve, picking things off a plate, peering into a box) we thereby describe many of the intentions with which they are acting as they are. (For this point see Anscombe, Intention, 7-9.) Similarly, our questions about Courbet’s intentions in painting the boy aren’t questions about how it came to be that this bluish-green patch is off to the right of the center of the canvas. Rather, they are questions about why Courbet painted the boy—and so a discovery that this isn’t what Courbet intended to be doing after all (that he was painting with his eyes closed, perhaps, and this is just what materialized: surely an impossible thing to imagine in the case at issue!) would then lead us to abandon this as a description of what he did, and indeed of what is there on the canvas, after all. Not a boy, perhaps, but a splotch of paint that can easily be mistaken for one.
But things are quite different when it comes to photographs. When Cartier-Bresson took his photograph of the couple appealing to Cardinal Pacelli, he also took a photograph of everyone else in the surrounding throng, including the balding man with the unfortunate mustache who is shown in the upper left of the image. And our judgment that Cartier-Bresson photographed the mustached man will not be called into question by the discovery that he didn’t intend to do so—that, perhaps, he didn’t know the man was there, or knew he was there only after he observed his face in the corner when the photograph was developed (compare Anscombe, Intention, 11 and 14). While it’s true, as Lopes emphasizes, that in such a case photographing the man will nevertheless be something that falls within the scope of Cartier-Bresson’s agency, so that it is something we can attribute to him and judge him responsible for, this won’t be in a way that gives the Anscombean “Why?” question application to it. To the question, “Why is that mustached man shown in the corner?,” the answer “He just happened to be in the background when the picture was snapped” does not give a reason for photographing the man, but rather a cause of why the man appears in the photograph. It does not, however, impugn the judgment that he does appear there, in the way that the corresponding judgment about the boy would be, if we discovered that this splotch of paint was put there without intending it to depict what it evidently does.
I have tried to be very cautious in the way I’ve put this point. As I see it, the pressure that photography puts on the Anscombean understanding of how an artist’s productive activity relates to the work that she produces is by far the most acute when our concern is with what is represented or depicted in an artwork. That is, it is insofar as the bit of paint off to the right of the center of Courbet’s canvas is taken to depict a boy that we think it couldn’t be there unless he intended it to be so—and then, in the unbelievable case where we discovered that this isn’t what Courbet was out to depict, we would be forced to conclude, not that Courbet didn’t intentionally put the paint there (even in “in that shape,” perhaps), but that in so doing he didn’t paint a boy in that portion of the canvas. By contrast, the fact that a balding, mustached man is depicted in the top left corner of Cartier-Bresson’s photograph is entirely independent of the photographer’s own judgment: the man can be in that corner of the photograph without Cartier-Bresson having put his him there. In this respect, what happens when Cartier-Bresson takes his photograph is, to look back to that phrase of Wittgenstein’s, to a significant degree “a favour of fate” (Tractatus 6.734). The man may have happened to show up in the background of Cartier-Bresson’s photograph in a way that the boy can’t have happened to show up in the background of Courbet’s painting.
This is not just a point about the “descriptions under which” an act is intended. (For this phrase, see Intention, 10-11.) Pacelli was born on the 2nd of March, and so in photographing him Cartier-Bresson photographed a man who was born on that date: and what he did may have been unintentional under this description but intentional under the description “photographing Cardinal Pacelli.” But that point holds just as much in the case of painting as in that of photography, whereas the one I’m after is more specific. I might put my point by saying that, in painting or drawing, depiction is an essentially intentional act: the shape that one paints on a canvas, or draws on a page, may be only unintentionally a depiction of X, as one may not have known that “X” was a true description of whatever one depicted. It must, however, have been an intentional description of someone or something if it is to have been a depiction, intentional or not, of anyone or anything at all. (So the discovery that Courbet didn’t know the date of birth of the boy he painted wouldn’t bear on the question of whether a boy who was born on that date is depicted in his canvas in the same way as the discovery that, impossibly, Courbet didn’t know that he was painting a boy there at all.) By contrast, it is quite possible to take a photograph of someone or something while having no idea at all that you are doing this, since the status of a region of a photograph as depicting someone or something is secured just by its having come about through an appropriate mechanical process.
Yet, for all this, clearly it is possible for a person to photograph someone or something intentionally, and thus for Anscombe’s question “Why?” to have application to some aspect of what is depicted in a photograph, or to how this depicted thing appears. And this possibility opens back up the way of thinking about agency that the formula “I do what happens” was supposed to help us resist. When Cartier-Bresson pushes the button on his camera, a certain mechanical process begins to unfold, and this process results in an image that contains Pacelli, the couple in the foreground, the mustached man in the top left, and so on. And all of this could also have happened entirely by chance, or if Pacelli’s intention was to do something quite different from what he did. What, then, is Pacelli’s intention supposed to be, if it is not an inner state of mind, something “attaching to the action at the time it is done,”11 and which does not enter into the description of what happens when, thanks to Pacelli’s push of a button, the mechanism in his camera goes on to generate his photograph?
I think that Michaels’s response to this question is largely on the right track, and largely in line with Anscombe’s own. (This convergence is no accident: I am, like John Gibbons, the kind of person who’s inclined to think “that being incompatible with Anscombe is a little like being incompatible with the facts.”12 ) Toward the end of Intention, Anscombe encapsulates her position with another fairly obscure slogan, writing that “the term ‘intentional’ has reference to a form of description of events” (Intention, 84). The point of italicizing the word “form” in this context is to mark an opposition with the idea that in describing an action as intentional we posit a special element that exists alongside, and is somehow appropriately connected to, the agent’s bodily movements. Her idea is, rather, that the description of what these movements are, spelled out in a way that gives her question “Why?” application to them, is itself a description of a person’s intentional activity.
Here is how this position is supposed to apply to the case of taking a photograph. I have tried to bring out how, in photography, there is a distinctive kind of gap between a person’s immediate movements and the product that results from them—a gap that is filled by the operation of a mechanical process, with plenty of room for the “favour of fate” to intervene. Still, it is not as if a photographer’s contribution to this process is simply in the press of a button that sets it off: for there is also the process of setting up from an appropriate angle, focusing the lens, identifying successful images and then perhaps cropping and retouching them, and so on. None of this puts the photographer in quite the same position as the painter, since the way that her image comes to depict what she photographs means that the response “Oh—I didn’t see that this (he, she) was there; it’s just how things happened to be when I took the picture” can have a significance that isn’t available in relation to something that one has drawn or painted. But the status of photography as an intentional activity, and in some cases a form of artistic practice, is constituted by the wider context in which this activity takes place.
Of course, the wider context that’s supplied by what we might call traditional photographic practice, where the role of the photographer is to deliver a faithful representation of things as they anyway are, is only one among many of the contexts in which photographs can be taken. Lopes and Michaels identify several of these over the course of their exchange: among them we have Garry Winogrand’s street photographs and Robert Mapplethorpe’s nudes, which stand opposed to one another in the way they conceive of the photographer’s relationship to his or her subject matter. And I agree with Michaels that part of what makes this work philosophically interesting is precisely the way that its wider context includes an awareness of the photographer’s peculiar relation to the image that is the product of her work—so that the work comes to have the circumstances of its production as one of its own topics, and can’t really be understood independently of this reflexive concern.13