Jonathan M. Weinberg is Professor of Philosophy and Member of the Program in Cognitive Science at the University of Arizona. His research centers on the implications of empirical psychological work in aesthetics, epistemology, and philosophical methodology.
A post by Jonathan M. Weinberg
In the decades since Radford’s classic (1975) “How can we be moved by the fate of Anna Karenina?”, philosophers of the imagination have increasingly shifted away from confronting that titular question and its close relatives in terms of resolving a paradox, and more towards treating our affective responses to works of fiction as a complex set of phenomena to which a theory of the imagination must be adequate. My goal here is to raise such concerns of adequacy about one such theory’s treatment of the psychological particulars: Peter Langland-Hassan’s one-box theory, and its operator-theoretic approach to fiction.
Contra one of the dominant lines of thought in the imagination literature, Peter Langland-Hassan (PLH) has argued that we do not need to posit any sort of sui generis components of our cognitive architecture to explain the range of phenomena that generally get lumped together under the term “imagination”. Nothing like an “Imagination Box” (IB) or i-beliefs or any such cognitive machinery is needed, but instead a range of different and disparate cognitive elements can do the requisite work, such as our ability to elaborate the premises in conditional reasoning, or in forming the right sort of beliefs. As a long-time proponent of an IB view, I think PLH’s book is great, shaking up many elements of that orthodoxy and forcing us to think through what our evidence for it really is. (E.g., speaking just for myself, I had long thought there was a much quicker argument from basic separability phenomena – that one can imagine that p either with or without believing that p – than there turns out to be, on PLH’s compelling argument.) Having said that, I still think that orthodoxy prevails, and in this blog post I want to offer one argument for why. In short, it’s because we still need something like an IB to make sense of our emotional engagements with fictions.[i]
So: we are saddened when Othello kills Desdemona. How so? Everyone agrees it’s not because we believe that an actual real-life person Desdemona has been murdered by their husband Othello. On an IB account, we entertain in the imagination the proposition: Othello kills Desdemona. Carrying on a line that traces back to Neill (1993), PLH suggests that our emotions are engaged metafictionally, and not through an IB but rather through a belief in a proposition about the fiction itself: in the fiction Othello, Othello kills Desdemona. Because that “in the fiction Othello” is a sentential operator, such a sentence makes no reference to such entities as Othello and Desdemona, but just characterizes the content of a particular work of fiction.
There are some parallel questions, about which orthodoxy is less unified, as to the nature of our desires that play a crucial role in producing such reactions. If we simply don’t care about Desdemona and her plight, we would most likely not react as we do. Some (e.g., Kind 2011, Spaulding 2015) have argued that we have desires about Desdemona herself, albeit herself qua fictional character, which PLH calls the “Simple View”. Others have posited desire-like imaginings (e.g., Currie & Ravenscroft 2002, Doggett & Egan 2007). The former invites concerns about whether we can really rationally have desires concerning beings we know to be unreal; the latter makes an extravagant psychological posit on what seems to be thin grounds (Kind 2011, Weinberg 2014). One virtue of the operator approach is that prima facie it incurs no such costs – we are sad that in the fiction Othello, Othello kills Desdemona, because we desired that it not be the case that in the fiction Othello, Desdemona dies, and now we have come to believe that, indeed, in the fiction Othello, Desdemona dies. We desired not-p, and yet we come to see that p is true, and our sadness that p is thus completely explicable as an ordinary sort of desire coming to apparent frustration with an ordinary sort of belief.
Alas, it won’t do. And it won’t do because of a property of operators that they are affectively opaque: our psychological systems that generate emotional responses in us do not seem able to “see into” propositional representations that are sub-parts of larger propositional representations. As Aaron Meskin and I argue in our (2006):
“Fearful affective responses, for example, seem to require representations of the form S is in danger, where S is someone we care about. But it is not enough for the representation to be a subpart of another representation. For example, if you believe, not that the slime is threatening, but that a friend thinks the slime is threatening, your only fear will be for your friend’s sanity; similarly for a belief like it is metaphysically possible that a slime is threatening. So a belief like in the novel it is true that S is in danger is not of the right sort to generate correct affective responses.” (225) (We suspect that such operators are cognitively opaque more generally but will leave that to another day.)
So on an operator account, you can only get what might be called situation-directed affect, and not person-directed affect. This seems part of what Radford has in mind when he insists that “We cry real tears for Mercutio” (70; emphasis added; see also van Leeuwen 2021, p. 658). Our emotional responses to fiction are not limited to the monochrome palette being glad that in fiction F, p. or sad that not-in fiction F, p, plus perhaps some earlier stretches of hoping or fearing about the same. We are also often, and rightly, proud of characters, angry at them, disappointed in them. We can love them or hate them, far beyond any loving or hating any specific fictional event involving them. In short, more or less whole gamut of emotional attitudes we can take towards real persons, we can and do take towards fictional ones.
One can feel some of these emotions towards fictional works or events in them, but the palpable differences between such cases, and the cases more naturally described as taking such attitudes towards the characters, just underscores the inadequacy of trying to reduce all of the latter sorts of affect to the former. Having made the grave error of taking on the work for a 6th grade book report, upon reaching the concluding line “He loved Big Brother” I literally threw my copy of 1984 across the room in an act very much of anger directed at the work. This seems to me not remotely of a piece with my anger at Othello, say, for falling for Iago’s schemes and failing to trust his faithful wife. For Iago himself, I have not so much anger as contempt – which again feels not at all like the contempt I have for, say, Atlas Shrugged.
A further, if less vivid, bit of phenomena: we often seem to have desires about what happens to characters past the point of when we know the fiction to be fully closed. Even though the lingering concluding shot of The Graduate raises doubts about their future together, we nonetheless hope that Benjamin and Elaine will somehow find happiness and contentment with each other. I don’t see how to explain our uneasiness at that very ending, without considering that hope. But I don’t have any hopes for what will happen to them embedded behind an in the movie, The Graduate operator – I know full well that that was the concluding shot, and I’m not in any sense hoping for more movie to come. (Whatever one’s views of Rumor Has It may be, its events are not part of this film.) Similarly, if you don’t leave Casablanca deeply hoping that Rick and Louis’s friendship really turns out to be beautiful, and long, and highly contributive to the war effort – then you and I are made of different stuff.
Finally, a key component of the operator view is that we do indeed have desires about what happens in the fiction, or else it cannot even get off the ground, e.g., as described above, our desire that in the fiction Othello, Desdemona survives. But if we only have desires concerning the operator-bedecked propositions, it is left mysterious as to why we would have such a desire. It is clear, on the one hand, why we would desire that she not survive – her death is part of what makes it a great tragedy, pace Nahum Tate. But the most natural explanation for why we might have the metafictional desire is because we have the first-order desire about Desdemona herself. If there’s any sense in which I want her to survive in the fiction, Othello, it’s because I want her to survive, full stop, but I recognize further that, as a fictional character, she can only survive if she survives in her home fiction.
So, it looks like explaining our emotional responses to fiction in their full complexity will require something architecturally very like first-order, i.e., unoperationalized, beliefs. The operator account is, at best, inadequate for this job. But the IB account still provides a fine theory for doing so, since on that account the representations entertained in our imaginations are straightforwardly the likes of Othello kills Desdemona. It would, however, need to paired with that “Simple View” of desires and the imagination: that we do, literally, desire that Desdemona live (without hiding that content behind a fictionality operator). Although that has not been my view in the past (e.g., Weinberg & Meskin, 233-4), I have come to think that it must be be true.
But! The door may be opened here for a flavor of IB-rejecting view even more radical than PLH’s. For it may be worth considering a view like his, but in which we don’t form beliefs with an in the fiction F operator. Rather, what if we have beliefs in which fictional people and their going-ons are located by means of a locative prepositional phrase “in the fiction F” (or equivalent). That is, what if at the level of cognitive architecture, at the level of the structure of our mental representations, our minds are something like Lewisian or Meinongian after all?
This view haunts my dreams.
References
Currie, G. & Ravenscroft, I. (2002) Recreative Minds. Oxford: Clarendon Press.
Doggett, T. & Egan, A. (2007) Wanting things you don’t want: The case for an imaginative analogue of desire. Philosopher’s Imprint, 7(9), 1-17.
Kind, A. (2011) The puzzle of imaginative desire. Australasian Journal of Philosophy, 89(3), 421-39.
Langland-Hassan, P. (2020) Explaining Imagination. Oxford University Press.
Neill, A. (1993) Fiction and the emotions. American Philosophical Quarterly, 30, 1-13.
Radford, C. (1975) “How can we be moved by the fate of Anna Karenina?” Proceedings of the Aristotelian Society, Proceedings of the Aristotelian Society, Supplementary, 49, 67-80.
Spaulding, S. (2015) Imagination, desire, and rationality. Journal of Philosophy, 112(9), 457-76.
van Leeuwen, N. (2021) Imagining stories: attitudes and operators. Philosophical Studies, 178, 639-664.
Weinberg, J. (2014). All your desires in one box. In Aesthetics and the Sciences of the Mind, G. Currie, M. Kieran, A. Meskin, and J. Robson (eds.), Oxford University Press, 181-204.
Weinberg, J. & Meskin, A. (2006) “Imagine that!” in In Contemporary debates in aesthetics and the philosophy of art, M. Kieran (ed.), 222-235.