Peter Langland-Hassan is Associate Professor of Philosophy at the University of Cincinnati. He has written a number of articles on imagination and is currently at work on a book he calls ‘Explaining Imagination’ (for Oxford University Press). The project of that book is to show how imagining can be reduced to (and explained in terms of) the use of more basic folk psychological states, such as beliefs and desires.
A Post by Peter Langland-Hassan
Here is a popular view: when we take in a fiction, we do so by imagining the propositions it contains (perhaps in addition to others it suggests or implies). These imaginings—and not any beliefs—are then partly responsible for the emotions we experience in response to the fiction. In this post I want to explore some tensions in this view as it appears in the work of some influential philosophers (e.g. Nichols (2004a, 2006); Weinberg & Meskin (2006); Schroeder & Matheson (2006), Carruthers (2006), Kind (2011), Spaulding (2015), Van Leeuwen (2016)).[i]
Let’s start with the “single code” hypothesis of Shaun Nichols (2004a, 2006). On Nichols’ view, if you imagine that p, your emotional reaction to that imagining will be much the same as if you had just come to believe that p. This is because, “affective systems can receive input from the imagination, and affective systems process input from the imagination as they would process isomorphic beliefs” (2006, p. 463). Tabling the concern that our responses to fictions are often milder than they would be if we believed the events to be actual—to which Nichols (2006) responds at length—I want to note an apparent idiosyncrasy of Nichols’s picture. For Nichols, the “affect generating system” takes input from our imaginings and beliefs, but is blind to our desires. (See Figure A, from Nichols (2004), where there is no line from Desire Box to Affect Systems; compare Figure B, from Meskin & Weinberg’s (2003) closely related proposal, where there is such a line).
Figure A: From Nichols (2004a) “Imagining and Believing: The Promise of a Single Code” Journal of Aesthetics and Art Criticism. The “Affective mechanisms” do not take input from the Desire Box.
This view seems problematic on its face. Very often when we acquire a new belief, our emotional response to the belief seems to depend crucially upon our desires. Whether we’re pleased to discover that asparagus is on the menu seems entirely dependent on whether we desire asparagus; whether we were thrilled, or indifferent, that the Cubs won the World Series depended on whether we desired them to win. But, for Nichols, there is no line indicating that the Desire Box provides input to the Affect Generating System—the latter can only “look at” one’s beliefs and imaginings. How, then, does it know what to do?
There are possible ways around this issue—some hinted at by Nichols (2004b) himself. Perhaps a belief that asparagus would be great right now, combined with a judgment that asparagus are on today’s menu, could suffice to generate positive affect. Let’s ignore the possible counterexamples—cases where you have both beliefs but still don’t get the affect. If it really turned out that we could explain affect without appealing to desire, it seems we could equally well explain action without appealing to desire. Why did I jump up and down, cheering for the Cubs? Because I believed that the Cubs are highly virtuous and deserving of victory, and I had just judged that they won. Why did I eat the asparagus? Because they were there for the taking, and I knew they would hit the spot. Are these not reasons (and causes) enough?
Whether or not we in fact find these desire-free explanations of action plausible, they ought to be plausible if the corresponding explanations of affect are. [Readers, do you disagree?] So, if we follow Nichols in denying that desire is an essential input to our affect generating systems, we seem headed somewhere quite radical: a reduction of desire to belief. Why ever attribute a desire in an explanation of behavior, if a value-laden (i.e., “evaluative”) belief can do the same work? I sort of like that direction, but I don’t think it’s where Nichols aims to be heading.[ii]
Yet there are problems in the other direction as well—problems Nichols was perhaps eager to avoid. Suppose we are convinced that the Affect Mechanisms need advice from both our beliefs and desires when generating affect (this is the architecture mapped out in Meskin & Weinberg (2003) (see Fig. B). If our imaginings are also to generate affect, some desire—or desire-like[iii]—state will need to pair with them as well. (One idea proposed in this connection, but which I won’t consider here, is the notorious “i-desire”—see end note iii).
Figure B: From Meskin & Weinberg (2003) British Journal of Aesthetics. Here the Affect systems do take input from the Desire Box.
But problems lurk when we are forced to specify the content of this desire. Suppose that I am watching a film where terrorists storm Buckingham palace. The film’s hero hatches a plan to drive the terrorists out. I feel anxiety as I follow along, imagining that the terrorists haven’t yet been driven out of the palace. Which desire pairs with this imagining to generate anxiety? It cannot be a straightforward (de dicto[iv]) desire that terrorists are driven out of Buckingham palace.[v] For while I do have a general desire for Buckingham palace to be free of terrorists, that desire is currently satisfied (for I know there are not, in fact, any terrorists in Buckingham palace). The unsatisfied desire—the one responsible for my anxiety—must be something other than the desire that Buckingham palace is free of terrorists. One possibility is that it’s the desire is that in the fiction, the terrorists are driven out of Buckingham Palace; a second is that it’s the desire that the fictional terrorists are driven out of the fictional palace (if this is indeed different the first).
But if either of these is the correct characterization of the desire (and I think one must be), its content is orthogonal to that of my imagining, which concerns terrorists and Buckingham Palace, not any fiction or fictional characters. (To my knowledge, no one proposes to embed an “in the fiction” operator, or reference to the fictionality of a character, within the de dicto content of our imaginings—tell me if I’m wrong!). As the imagining concerns terrorists and Buckingham palace, and my desire either concerns the course of a certain fiction, or certain fictional characters and settings, there’s no reason they should combine to generate affect. Cognitively, they are two ships passing in the night.
So, here’s the dilemma: either we hold (with Nichols) that emotions are blind to our desires and that they respond “directly” to beliefs and imaginings; or we allow desires an essential role in generating emotions. If we go the first route, we seem headed toward a reduction of desire to belief. (I see no reductio there—just a surprising result). If we go the latter route, we face a problem characterizing the content of the desire in a way that would allow it to pair with an imagining to account for our emotional responses to fictions.
The best response to all this, I think, is to leave imagination out of our explanation of emotional responses to fiction. (At least insofar as imagining is thought to involve a kind of mental state distinct from one’s beliefs and desires). Instead we can just note that we have beliefs about fictional characters, and desires about fictional characters, and at times these pair to generate affect.[vi] This is a good result to me, as I’m working on a more general project of reducing imagination to beliefs and desires. But I wonder what others make of this situation: can imagination indeed have a role in generating affective responses to fiction? One possibility—noted above, and not recommended by me—is to invoke corresponding i-desires. Are there any others?
[i] A notable exception to this trend is Matravers (2014), who does not find it useful to appeal to imagination in the explaining responses to fiction.
[ii] Similar questions arise for Van Leeuwen’s (2011, 2016) view, on which mental imagery can cause emotion and motivate action without the influence of belief and desire. Van Leeuwen’s view is difficult to assess, however, until we know the psychological force or attitude of the mental imagery that is invoked—is it, e.g., belief-like imagery, desire-like imagery, neither, or both?—and the limits that pertain to imagistic modes of representation.
[iii] This point is made by both Currie & Ravenscroft (2002) and Doggett & Egan (2012) in their arguments in favor of an imaginative analog to desire—“i-desires”—that can play this role. While I don’t recommend i-desires, I won’t say anything here to challenge the notion. My focus in this post is on problems that arise for those who wish to avoid i-desires, while still appealing to a sui generis mental state of “imagining” in the explanation of affective responses to fiction.
[iv] I assume that de dicto ascriptions of beliefs, desires, and imaginings—which create opaque contexts—ascribe the states relevant to explaining action and emotion (as opposed to de re ascriptions).
[v] Here I take myself to be disagreeing with Kind (2011) and Spaulding (2015) when they propose that the desires at work in generating affective responses to fiction need not make reference to fictions themselves. As Kind puts it, when describing her concern for the characters of Peter Pan, “My desire isn’t about the fiction, it’s about Tinkerbell” (2011).
[vi] Doggett & Egan (2012) argue that this commits us to irrational desires; but I think those arguments are well parried by Spaulding (2015) and Kind (2011).
References
Carruthers, P. (2006). Why Pretend? In S. Nichols (Ed.), The Architecture of Imagination. Oxford: Oxford University Press.
Doggett, T., & Egan, A. (2012). How we feel about terrible, non-existent Mafiosi. Philosophy and Phenomenological Research, 84(2), 277-306.
Kind, A. (2011). The Puzzle of Imaginative Desire. Australasian Journal of Philosophy, 89(3), 421-439.
Meskin, A., & Weinberg, J. M. (2003). Emotions, Fiction, and Cognitive Architecture. The British Journal of Aesthetics, 43(1), 18-34. doi:10.1093/bjaesthetics/43.1.18
Nichols, S. (2004a). Imagining and Believing: The Promise of a Single Code. Journal of Aesthetics and Art Criticism, 62, 129-139.
Nichols, S. (2004b). Review: Recreative Minds. Mind, 113(450), 329-334. doi:10.1093/mind/113.450.329
Nichols, S. (2006). Just the Imagination: Why Imagining Doesn't Behave Like Believing. Mind and Language, 21, 459-474.
Schroeder, T., & Matheson, C. (2006). Imagination and Emotion. In S. Nichols (Ed.), The Architecture of the Imagination (pp. 19-40). Oxford: Oxford University Press.
Spaulding, S. (2015). Imagination, Desire, and Rationality. Journal of Philosophy, 112(9), 457-476.
Van Leeuwen, N. (2011). Imagination is Where the Action Is. Journal of Philosophy, 108(2), 55-77.
Van Leeuwen, N. (2016). The Imaginative Agent. In A. Kind & P. Kung (Eds.), Knowledge Through Imagination (pp. 85-109). Oxford: Oxford University Press.
Weinberg, J., & Meskin, A. (2006). Puzzling Over the Imagination: Philosophical Problems, Architectural Solutions. In S. Nichols (Ed.), The Architecture of Imagination (pp. 175-204). Oxford: Oxford University Press.