This week at The Junkyard, we’re hosting a symposium on Peter Langland-Hassan’s recent book Explaining Imagination (Oxford University Press, 2020). The book is available as an Open Access download. See here for an introduction from Peter. Commentaries and replies will follow Tuesday through Thursday.

* * *

Commentary from Nick Wiltsher

One way in which a philosophy book can be brilliant is by presenting claims, positions, and arguments with which you agree entirely, and yet deriving from them a view with which you feel you somehow disagree. By this criterion (and by many others), Peter Langland-Hassan’s book is brilliant. The pleasure and the puzzle lie in trying to work out just what it is that’s bugging you. I’m not sure that I’ve cracked the puzzle yet, so I’m going to indulge in some thinking out loud, and invite Langland-Hassan to help me straighten it out.

As he does (p. xii), let’s begin in the bakery, where there is a happy profusion of goods: cakes and croissants and buns and biscuits and scones and sourdough. If you want to know how to make one of these things, you ought to consult a recipe, which will tell you which ingredients to combine, in which proportion and by which method, such that the object of your appetite arises.

Is the recipe a reduction of the product? Sure, why not. It tells you what the constitutive parts are and how they’re put together. Is the recipe an explanation of the product? That really depends on what you want to explain. If you want to explain taste, flavour, colour or consistency, a list of ingredients will be very useful indeed. But if you want to know whether a brioche is bread, or whether croissants are necessarily crescents, or whether biscuits should be served with gravy or tea, and if you want to explain why any of these things are so or aren’t: well, finding out that everything concerned contains flour won’t be much help.

Langland-Hassan offers an explanation of imagination by way of reduction. He argues that complex attitudes are constituted by combinations of simple attitudes (ch. 1 §8), and that the job of “predicting, explaining, and rationalizing behaviour” (p. 31) ostensibly done by ascribing complex attitudes to people can equally well be done by ascribing to them the right combination of simple attitudes. All the behaviour explained by attributions of suspicion, for example, can be explained by attributions of probabilistic belief, and belief is a simpler state. That’s a reduction of suspicion (ch. 1, §§8, 9). Similarly, he argues, what people do when they imagine is explicable by ascriptions of combinations of beliefs, desires, and intentions. So imaginative mental states—those involved in understanding fiction, for example—are constituted by, and reduce to, such combinations. That’s a reduction of imagination. It doesn’t amount to elimination of imagination as a handy piece of descriptive apparatus, any more than providing an ingredients list eliminates crumpets from our conceptual repertoire. But it does tell us that, once the supermarket of the mind is fully stocked, we won’t find imagination in the same aisle as the basic, primitive cognitive attitudes.

I’m happy with calling this a reduction, perfectly happy with the claim that imagination isn’t thereby eliminated, and pluperfectly happy with the corollary that there is no distinctive cognitive attitude of imagination. And yet something about it all bugs me: perhaps something to do with the explananda, perhaps something to do with the explanatory method. I’ll try to find the bug by asserting some things about suspicion and imagination, and then comparing my assertions with Langland-Hassan’s arguments.

So: if an old friend I know drops by to say hello, I might then see suspicion in your eyes. Langland-Hassan’s view is that, really, what I am seeing—ascribing—is a probabilistic belief: you have some degree of credence short of certainty in the proposition that something is going on between my friend and me. This probabilistic belief is enough to explain all your subsequent snooping, and that’s all the ascription of suspicion was meant to do. So suspicion can be explained in terms of and thus reduced to belief.

However, I assert, suspicion is not just an ascription of such a belief, and isn’t just meant to explain behaviour. Ascribing suspicion also involves ascribing, inter alia, reactive attitudes directed to me and my friend, a particular sort of affect, propensities or dispositions to form further beliefs and interpret future behaviour in certain ways, a certain kind of epistemic relation to the grounds for suspicion, and so on. All this is what distinguishes suspicion from other ascriptions of probabilistic beliefs: hunches, guesses, intuitions. Ascriptions of suspicion are rich descriptions of what’s going on in your head.

Similarly, ascriptions of imagination are rich—although, curiously, I find it harder to specify how so. Saying that someone is imagining is perhaps in part to say that they believe, desire, and intend certain things, and to try thereby to explain their behaviour. But it is generally also to say a great deal more than this. Given a bit of context, it might inter alia be to say something about the content and affective character of what they’re doing, their purposes in so doing, the origins of their states, their orientation toward the objects of those states, their future dispositions, and so on. An ascription of imagination says something rich, complicated, and detailed about a person: it doesn’t just explain what they’re doing.

That’s what I reckon, and it seems to put me at odds with Langland-Hassan. But why so? One possibility is that I’m getting hung up on the word “behaviour”. Langland-Hassan characteristically uses such actions as savouring the scent of a mud pie to  exemplify what ascriptions of imagination explain (e.g. ch. 1 §10), but perhaps “behaviour” is supposed to encompass everything I have mentioned—affect, dispositions, purposes, the lot. So perhaps, if my point is just that there’s more to imagination than how one behaves, I need to be more generous about my understanding of “behaviour”.

But I don’t think that’s it. My point is really that there’s more to ascriptions of imagination—more of interest and substance that demands explanation—than could possibly be explained by combinations of beliefs, desires, and intentions. Just as lists of ingredients won’t explain why some pastries are Danish, lists of simple states won’t explain why, for example, imaginative activities are socially valorised.

So perhaps what I’m insisting on is this: Langland-Hassan’s strategy makes sense if imagination is aptly treated as a term in a folk theory of mind, and a folk theory of mind is aptly thought of as a theory aimed at explanation, prediction, and rationalization of behaviour (ch. 2 §§1, 2). I suppose if there’s such a thing as a folk theory of mind, that’s the kind of thing it is. But imagination is not a term in that theory: not a posit that’s supposed to explain behaviour, but rather a thick descriptor. And so imagination cannot be reduced to terms in that theory.

On the other hand, I agree that imagination can be informatively reduced to simpler states. So perhaps instead I’m insisting on modesty: reductions to simpler states can explain those things about imagination that such reductions are capable of explaining, but only those things. And perhaps Langland-Hassan would ignore the idea’s air of tautology and agree. After all, he approvingly cites Dretske’s maxim to the effect that, until you know how to make something, you haven’t explained it (p. xi). But that doesn’t entail that you have explained it once you do know how to make it. So perhaps I’m bugging myself by attributing to Langland-Hassan a more grandiose explanatory project than the one he actually takes on.

But even if so, an associated worry remains. I have unfairly concentrated on the first couple of chapters of the book, where the method is explained, and ignored most of it, where the method is applied. In those chapters, Langland-Hassan carefully explains how such things as the aims of an imagining serve to substantiate an ascription, and to distinguish (say) engaging in fiction from pretence. Rightly so, of course: a recipe tells you not only what ingredients to combine, but also how to combine them. However, if the method of combination is substantially different, you end up with a different product, even given similar ingredients. Yorkshire puddings aren’t pancakes, sponges aren’t scones, muffins aren’t manchets. The recipes that Langland-Hassan provides combine the same ingredients in very different ways to yield various different imaginative activities. But if the methods of combination are so different, the question of what unites the products as imaginative bulks large. So perhaps what’s bugging me is, in fact, the question of whether the reductions offered can be said to explain the unity of imagination: why, that is, these recipes all appear in the same book.

Don’t let the extended analogy fool you: the points I’ve raised aren’t especially cohesive, and may not even be consistent. They’re stabs at articulating just what it is that stops me from endorsing the view, even while I endorse so many of its parts. I look forward to hearing what I’ve misunderstood.

Thanks to Kevin Lande and Gerado Viera for comments on a draft.

Response from Peter Langland-Hassan

I’m grateful to Nick Wiltsher for his thoughtful meditation on what might be missing in my reductive account of imagination.  I see him as raising two general points:  first, there are things that need explaining about imagination that cannot be explained even by a correct and comprehensive cataloging of imagination’s more basic parts (see his point about whether brioche is bread); second, while imagination can perhaps be broken into more basic parts, he is not convinced that I provide a full catalog of those parts.  He thinks that my failure to do so may trace to my (admitted) tendency to treat imaginings as theoretical posits—elements in a folk theory of mind—whose raison d’être is to predict and explain behavior.    

I think these are reasonable worries and am glad to have a chance to say something about them.  To take the first point first, I agree that there are things to be explained about imagination that cannot be explained in my favorite way—by appeal to its parts.  (In my defense, I did not title the book Explaining EVERYTHING About Imagination.)  However, one such explanandum, noted by Nick, is particularly important.  “Perhaps what’s bugging me,” he writes, “is, in fact, the question of whether the reductions offered can be said to explain the unity of imagination: why, that is, these recipes all appear in the same book.”  I am committed to there not being a deep unity, of course, as I think the acts that get called ‘imagining’ are a heterogeneous lot.  But Nick’s question still has resonance:  on what basis can I say that pretense, hypothetical reasoning, fiction consumption, and creativity all involve imagination?  Without an answer, suspicions may be raised that I have let the very thing we wanted explained slip through my fingers.

My answer lies in the definition of attitude imagining as rich, elaborated, epistemically safe thought about the unreal, fantastical, and possible.  This understanding of imagination—superficial as it is—is important to unifying the phenomena under discussion, given that, at a deeper level, there is relatively little unity.  It provides a theoretically neutral means for pointing at, and distinguishing, a set of phenomena that we would like to better understand, while allowing for the possibility that we will best understand it in different ways in different instances.  There is a significant amount of philosophy at work in this initial characterization of imagination, to be sure.  Much is determined by these first steps down the path.  Across chapters 1, 2, 3 and 4, I offer reasons for setting out in that direction.  Nick may want to argue that my surface characterization of attitude imagining does not capture what unites the contexts where imagining occurs.  It will prove difficult, however, to offer a deeper account that remains flexible enough to cohere with the wide variety of roles normally assigned to imagination, as others have noted (Kind, 2013).

Now to Nick’s second concern, which is, I think, his primary one.  Many of my arguments for reducing imagination to other kinds of states work like this:  I point to a kind of behavior (e.g. pretense) or ability (e.g. conditional reasoning) that most people agree requires imagination.  Then I give an account of how that behavior or ability can be guided and/or motivated by beliefs, desires, or intentions of various kinds.  Voilà, the imagining is nothing over and above these states.  Nick worries that this method paints an artificially conservative picture of imagination and its parts.  “An ascription of imagination says something rich, complicated, and detailed about a person,” he writes, “it doesn’t just explain what they’re doing.”  (Note how close this is to the first concern, that my surface characterization of attitude imagining does not capture what we mean to say about someone who is imagining.)

The worry is not ill-founded.  I take the approach I do because it seems the least question-begging way to consider questions about imagination and its reducibility.  It is not a conceptual truth that imagination is reducible in the ways that I propose.  How else, then, might one show that such reductions are possible than to identify the work imagination does and show that it could be done by other kinds of states we already have at hand?  We are all in need of a method that is faithful to the full phenomenon of imagination while allowing for the possibility of surprises about its nature.  That is the core commitment that moves me.  The view that folk psychology is a quasi-theory coheres with it, but is ultimately dispensable.

To say that someone is imagining, Nick observes, “might inter alia be to say something about the content and affective character of what they’re doing, their purposes in so doing, the origins of their states, their orientation toward the objects of those states, their future dispositions, and so on.”  This sounds right to me.  It also sounds like most of these features—viz., a state’s content, affective consequences, relation to the agent’s purposes, causal origins, dispositions—can be well-tracked by the kinds of folk psychological states I posit as the proper parts of imaginings.  But my ears are open to counterexamples.  In the meantime, I’m heartened by Nick’s suggestion that what needs doing isn’t the positing of an irreducible kind of state—sui generis “imaginings”—but, rather, more careful work in untangling the complex cognitive, conative, and affective states that combine to constitute cases of imagining. 

Reference:

Kind, A. (2013). The Heterogeneity of the Imagination. Erkenntnis, 78(1), 141-159. doi:10.1007/s10670-011-9313-z