Philosophical interest in the imagination has waxed and waned. It is there in the Ancients; indeed, one of them (Proclus) seems spot on as a contribution to current philosophy of fiction: “For stories which are entirely fictional are suitable only for those who live by the imagination alone and who in general have only passive intellect; the bright light of knowledge and the self-revealing quality of intellectual cognition are suitable for those who concentrate all their own activity in pure thoughts.” It is important in Kant, Hume, and Smith and, in the twentieth century, was the subject of an interesting book by Sartre called The Imaginary. In the past twenty-five years the imagination has been of considerable interest in two areas of philosophy in particular: in the philosophy of fiction and in the debate regarding how we understand each other (“simulation” versus “theory theory”). I work in the philosophy of fiction, and I will argue here for scepticism as to whether the dominance of the imagination is all to the good.
It might seem obvious that art and the imagination are closely linked, and that denying the close relation between the two is enough to give philosophy a bad name. Although I agree there is a link, I do not think matters are as straightforward as they appear. For a start, we should distinguish (to borrow some useful terminology from Gregory Currie) the “creative” from the “recreative” imagination. The “creative imagination” is used, amongst other things, in coming up with theories, inventing things, getting oneself out of trouble, and creating works of art. Whatever the creative imagination is, or whether it is just one thing, are not questions that concern me here. Rather, I am interested in the “recreative imagination”; do we need the imagination when we engage with works of art?
It is intuitively plausible that engaging with art, whether looking at pictures, watching a film, or reading a novel, involves the imagination. There have been some dissenting voices; Richard Wollheim, for example, doubted whether appealing to the imagination would throw light on the nature of pictorial representation. However, there is currently a near consensus that, with respect to fiction, this involvement is fundamental; that our engaging with fiction (whether with fiction films or with novels) is to be explained in such terms. I am not so sure; indeed, I am sceptical that it can be made to work. Another philosopher, Stacie Friend, has independently been coming to the same conclusion. I don’t know how many philosophers it would take to constitute a movement, but this might be a sign that the tide is turning against the imagination.
First, however, I want to enter a caveat. Nothing I say should be taken to imply that we are unable to use our imaginations when engaging with a work of art. Looking at Manet’s Bar in the Folie Bergere, I can try to imagine what the barmaid is thinking. I can visualise the characters in Pride and Prejudice (although arguably I need to stop reading in order to do so), imagine the scene from the cliff-top that Edgar conjures up for the blind Gloucester in King Lear, and so on and so forth. My point is only that these are not essential to engagement – although we might miss a lot if we do not exercise our imaginations in this manner. The imagination is not built into the fundamentals of our engagement with pictures, films, and (other) fictions.
The principal source for recent work on fiction and imagination is Kendall Walton’s book, Mimesis as Make-Believe. This is a long and complicated work, but at its core is a relatively simple idea. Walton first provides an account of games of make-believe. In a game of mud pies something happens in the actual world – Jane puts three globs of mud in a cardboard box – which makes it the case that in the game, Jane has put three pies in the oven. Walton’s claim is that many of our interactions with works of art are analogous to such games. This is a very powerful and appealing thought. In games of make-believe there are plenty of things which have one identity in the actual world and another identity in the “make-believe world”. There are globs of mud that are pies, and cardboard boxes that are ovens. Similarly, in the arts there are many things that have one identity in the actual world and another in the world of the arts. There are two-dimensional canvasses that are views of Delft; there are shaped pieces of stone that are Generals on horseback; there are melodies that are expressions of emotions; there are Penguin paperbacks that are eighteenth-century sailor’s journals and so on and so forth. Walton’s thought (which requires careful working out to fit different cases) is that we make-believe (or imagine – I will use the terms synonymously) of these things in the actual world that they something different in the world of the arts.
Let us take an example. What makes it the case that a rectangular piece of canvas hanging in The Hague is a depiction of a view of Delft? Walton’s account is not that we imagine of the picture that it is a view of Delft (it is difficult to work out what that would mean). Rather, in a similar way to that whereby I can imagine of an action of moving my hand in a certain way that it is the action of firing a gun, I imagine of my looking at the canvas that it is a looking at a view of Delft (and my so doing results in a rich and vivid game of make-believe). For my purposes here, we need not look at the details of the theory. All I want to discuss is the high-level claim that what it is for something to be a picture of something else is a matter of our using our imagination.
Walton’s view seems particularly at home when we consider cinema audiences. Consider the case of Josh, who is watching The Grand Budapest Hotel. What is the best way to describe Josh’s experience? In the actual world he is seated in a darkened room watching coloured lights flicker on the screen. That is not how he experiences matters; rather, he seems to be in perceptual contact with concierges and bell hops, grand dames and soldiers. So how does he experience matters?
Philosophers have come up with various different accounts. First, there is the “direct perception” model: Josh imagines that he can see the events depicted in the film. However, a host of problems now crowd in. If (in the game) Josh can see the concierge, why (in the game) can the concierge not see Josh? What happens if the film depicts a couple in an intimate moment? Surely, it is not true that (in the game) Josh is standing there watching them? If (in the game) Josh is in perceptual contact with a battle raging around him, why (in the game) is he not in danger? In place of this flawed model, some philosophers have suggested a “mediated perception” model: Josh imagines he is being shown a documentary of events that took place in and around The Grand Budapest Hotel. However, this is hardly much better. Is Josh supposed to imagine everyone was followed around by a camera crew – including that poor couple trying to have an intimate moment?
It is not so much that we have to balance one set of considerations against another, but rather than the whole debate just seems silly. So silly, in fact, it suggests we have gone wrong somewhere in the way we have set it up. Where I think it has gone wrong is in thinking that the imagination is required in order to see what we see in the cinema. What I am about to say is not decisive against Walton’s view, but it does have some puzzling features (this has been pointed out by Malcolm Budd, although Budd is generally quite sympathetic to Walton’s position). Here are two features of the experience of Vermeer’s picture. Anybody who looks at it, at whatever time of day, will have (roughly) the same experience. That experience will be highly stable and feature some detailed content. Furthermore, the nature of that experience is irresistible. It is not up to us, in any way, to determine the nature of our experience (except by closing our eyes or jiggling about with our eyeballs). Imagination is just not like that, which suggests that looking at the picture is not a matter of our using our imaginations. Ask 100 people to imagine something, and we would not expect them all to imagine exactly the same thing, especially if it requires imagining some detailed content. Furthermore, if asked to imagine something, we do not expect what we imagine to be irresistible. That is, what it is that we imagine generally feels to us to be under our control, rather than something imposed on us from without. The same applies to the cinema audiences. If they are exercising their imaginations, would it not be surprising that they are all imagining exactly the same thing to an arbitrarily complex level of detail? If they are exercising their imaginations, why does it seem to them that they have absolutely no choice as to what they imagine?
Let us drop the idea that pictorial representation is a matter using the imagination. I must admit, I have no account of the nature of pictorial representation – I am just going to assume that there is or will be some workable account somewhere or sometime. One advantage of this is that we do not have to start on the road that led to the silly debate we were engaged in earlier. When Josh sits in the cinema, he is not imagining that he is seeing something that, in the actual world, he is not seeing. Rather, he is really seeing what he appears to be seeing: pictures on the screen. In the jargon, his perception is veridical. In the actual world he is seeing pictures, and (apart from supplying an account of what it is to be a picture) that is all that need be said.
So much for cinema, what about novels? Surely there is a contrast between reading a newspaper and reading a novel. In the first case, we believe what we read and in the second case we make-believe (or imagine) it. This is the consensus view, which – as I said above – is currently coming under pressure from Stacie Friend and me. (The difference between Friend and me is roughly that Friend thinks there is a workable notion of the imagination that applies equally to fiction and non-fiction, while I do not think we have a workable notion of the imagination). Think about what we do when we imagine things. We can imagine unicorns (fictional creatures), but we can also imagine our child’s face, what it must be like to be attacked by a lion, what the world looks like from the centre spot at Twickenham, the death of Cleopatra, or pretty much anything that we are not experiencing right at this moment. In as much as we can be invited to imagine anything – fiction and non-fiction alike – what we are invited to imagine simply cannot do the work of dividing the fictional from the non-fictional.
We can run through this in a little more detail. The consensus view generally adopts the view of imagination as “simulation”. That is, we are able to simulate being in situations we are not currently in. If I am really being attacked by a lion, I have perceptual inputs (I see the lion bounding towards me), some internal workings (I decide to run as fast as I can), and some behavioural outputs (I turn and flee). If I simulate being attacked by a lion, I run my mental states “off line”. That is, I imagine the perceptual input (I visualise a lion bounding towards me or imagine that there is a lion bounding towards me), I go through the internal workings (although this time with make-beliefs rather than beliefs), and there is no behavioural output (it would simply be weird for me to turn and flee – as there is no lion, I would not be putting any distance between myself and danger). The view, then, is that non-fiction invites us to believe things, while fiction invites us to simulate things.
However, this is simply not going to work. While it might be true that all fictional scenarios require the effort of simulation, it is not true that all scenarios we are invited to simulate are fictional. There is no limit on the scenarios we can simulate except, perhaps, those we find ourselves in at the very moment we are attempting to simulate. If I read a novel that features our hero being attacked by a lion, I can simulate that scenario. However, I can equally simulate the scenario when described in a memoir of being attacked by a lion, when remembering being attacked by a lion, when observing someone else being attacked by a lion, when anticipating being attacked by a lion, and so on. There is simply no interesting link between simulation and fiction.
What then, can we put in place of the imagination? What is the right account of reading fiction as opposed to reading a newspaper? My view is that that is the wrong question to ask. The right question to ask is “what is the right account of reading a story (whether fiction or non-fiction)?” Once we are able to say what happens between the words on the page and our grasping the content of the story (that is, what the story is about) we have done most of the work. The difference between fiction and non-fiction does not lie in the mechanics of our engaging with the story (that is the same in both cases) but rather in the relation between the content that we grasp and our pre-existing structures of belief. Very roughly, if we are reading a non-fiction, we (generally) use the content to update our beliefs. If we are reading a fiction, we do not (generally) use the content to update our beliefs. At no stage in any of this need we mention the imagination.
What is the upshot of all this? First, it is good to get things right and, if recent philosophy of fiction has been mistaken in what it takes to be the relations between fiction and the imagination, it would be good for that to be corrected (although, curiously, Walton’s work arguably survives almost intact). However, it is not only philosophy that is affected. Work on fiction has drawn attention from cognate disciplines such as psychology and cognitive science. Much work has been done on finding the mechanisms that underpin our interaction with fiction and attempting to draw functional (“boxological”) maps of our psychology. If we have got the fundamentals of the philosophy wrong, then we will have set our interdisciplinary hares running in the wrong directions. Finally, it would be good to return to term “imagination” to the use it had before recent philosophy turned it into a technical term. We can be prompted by fiction into using our imagination (spinning yarns about the afterlife of the characters, perhaps) rather than it simply being part-and-parcel of what it is to engage with fiction.