A post by Luke Roelofs.

In last week’s post, Peter Langland-Hassan presented an argument for revising the widely-accepted view that emotional responses to fiction are driven by imagination. Rather, he argues, they are driven by beliefs - beliefs about the content of the fiction.

Although I disagree with Peter, I hope he may find this post indirectly supportive, since in trying to resist his argument, I find potentially revisionary implications in the opposite direction: rather than belief taking over what we thought was imagination’s domain, imagination might spill into territory we thought belonged to belief.

Peter’s argument, as I understand it, is that imagining the content of the fiction can’t be what explains our emotional reaction, since that reaction is conspicuously absent when we imagine, as we clearly can, various alternative completions of the story. If I’m sad at the end of Romeo and Juliet because I am imagining Juliet and Romeo dying, why is it so supremely unsatisfying to simply imagine, instead, that they survive? Of course someone might successfully change their feelings by elaborating, writing down, or even performing their alternative ending, but at that point they are creating a rival fiction and not just imagining. Peter consequently suggests that what drives our emotions is our beliefs about the fiction, not our imaginings in accord with it.

One response appeals to differences in ‘vividness’ or ‘engagement’: the imagining we do in accord with the play is ‘richer’ and more ‘powerful’ than our imagining of our concocted alternative. It’s easy to see why this would be the case with visual media like plays (which feature visible, audible, physical people and props acting in ways that match what is being imagined), and a bit less clear how it applies with written media where the imaginer has to construct the sights and sounds themselves. But it does seem right that there is a difference in how forcefully one is aware of something when seeing it written and when simply knowing or remembering it (consider someone who cannot bring themselves to look at their rejection letter, even though they know what it says: actually seeing the words makes it, for whatever reason, hurt more).

But I don’t think ‘vividness’ can be the whole story. I might be distraught over an event in the play that happens off-stage, mentioned only briefly, so that I have to supply images and sounds myself, just like I do with my own alternative ending. And I might still be upset even much later, when the perceptual details have faded from my memory and I can’t imagine the canonical story any more vividly than I can my alternative version.

Here’s a second factor to explain the difference in emotional force: with the canonical story, our imagining is governed by something outside us (the public social institution called ‘Romeo and Juliet’), while our private alternative version is, as Peter says, ‘up to us’. Something about the relative objectivity of Romeo and Juliet seems to matter.

This is pretty close to Peter’s view, that what drives our emotional responses is our beliefs about the fiction - beliefs which, unlike imaginings considered in isolation, are the right kind of thing to be judged true or false, because the fiction is an objective social entity. But an opponent of Peter’s belief-based account can agree that this is the right place to look for an explanation, while disagreeing about how to analyse it. They could say that something about the way that external facts about Romeo and Juliet dictate what we are to imagine gives the thus-dictated imaginings a sort of special force or solidity or importance.

Amy suggests this in a comment, suggesting that “non-fiction-prescribed imaginings result in considerably less affect than fiction-prescribed imaginings.” Peter replies that “We might then ask why some imaginings generate more affect than others.” It’s a fair question. I think the best answer would be something like: we are intensely social, imitative beings, and without even wanting to we tend to find ourselves looking for ways to align with others. Just as we spontaneously tend to synchronise our movements with others, and to catch the emotions of others, and to pick up the ideas, mannerisms, and value judgements of others, we also cannot help trying to align our imaginings with others. Fictions accomplish such alignment - either actually, when they are seen by many, or potentially, when they are seen only by us but have an objective form that others could view - and this gives them a greater hold over us than imaginings that remain ‘in our heads’.

I’m not sure how to decide between Peter’s position (that the attitudes that matter are beliefs with fictions in their content) and mine (that the attitudes that matter are imaginings which we spontaneously seek to align with others via fictions). But I think it’s worth having both on the table.

I also think it’s worth laying out this rival position because of what it tells us about ‘imagining under constraints’. Amy Kind and Peter Kung have distinguished two importantly different styles of imagining: imagining ‘constrained by reality’ (the ‘instructive’ use) and imagining without such constraints (the ‘transcendent’ use). Roughly, the idea is that although imagination is, in its own nature, ‘free’, it becomes suitable for epistemic service when we discipline it to match reality. 

But thinking about emotional responses to fiction underlines how simplistic it would be to take Amy and Peter as saying there are just two categories here: constrained (and a source of knowledge) and unconstrained (and epistemically impotent). For imagining in response to fiction is highly constrained, but not in a way that makes it a source of knowledge.

I find it useful to think in terms of a three-way division, between (i) instructive imagination, constrained by reality, (ii) ‘pure’ or ‘free’ imagination unconstrained by anything, and (iii) imagination constrained by rules like those of a game or fiction. Labels are never quite perfect, but I like the terms ‘simulation’, ‘fantasy’, and ‘pretence’ for these three categories.

What is particularly interesting about the rules of pretence - of a game, a fiction, an artwork - is that they are both compelling and arbitrary: they rest on nothing more sturdy than convention, intention, and expectation, yet we can experience them as fixed and involuntary. Even though there is no ultimate basis in reality for the fact that Shakespeare writing certain words and actors saying certain words ‘make it true’ that Romeo and Juliet are dead, it feels like a horribly unchangeable fact, that survives all our wishing it were otherwise.

I mentioned above that I thought defending this line might have revisionary implications for the relationship between belief and imagination. That’s because it suggests that a long-running, sophisticated pretence might be hard to distinguish from a belief, even by the subject themselves. While unconstrained fantasy is easy to distinguish from belief, and reality-constrained imaginative simulation at least seems to keep us closely-adjacent to belief (to simulate well, we need to draw on our prior beliefs, and if we succeed we can justify further beliefs), pretence can feel just as constrained as belief even if completely cut off from genuine beliefs.

For example, pretending that there is a benevolent deity is not believing it, but it might feel very much like believing to be engaged in a long-running ‘game’ in which I, and others around me, imagine things about God in a rule-governed way, in response to specific objects, places, or questions (e.g. Bibles, churches, and queries about religion).

Similarly, the analysis I’ve given suggests that people might read and write news and arguments for, e.g., conspiracy theories, might even take themselves to believe them, without having to actually believe anything. (An idea that Marianna Bergamaschi Ganapini wrote about in her post here last year, which for the record I should admit I pushed back against in the comments - views can change!)

In short: Peter suggests that the profile of emotional reactions to fiction is more belief-like than imagination-like; I think imaginations can have that kind of profile, but that in turn makes me wonder whether, in other cases, we might be less good than we think at telling our beliefs from our imaginings.

References

Bergamaschi Ganapini, Marianna. 2018. “Fake News and Imagined Narratives.” On The Junkyard: A scholarly blog devoted to the study of imagination. https://junkyardofthemind.com/blog/2018/10/29/fake-news-and-imagined-narratives

Kind, Amy, and Kung, Peter. 2016. “The Puzzle of Imaginative Use.” In Kind and Kung, Knowledge Through Imagination, Oxford University Press: 1-40.

Kind, Amy. 2016. “Imagining Under Constraints.” In Kind and Kung, Knowledge Through Imagination, Oxford University Press: 145-159.

Langland-Hassan, Peter. 2019. “Choosing your own adventure?” On The Junkyard: A scholarly blog devoted to the study of imagination. https://junkyardofthemind.com/blog/2019/10/18/choosing-your-own-adventure