A report by Luke Roelofs

Interdisciplinary conferences are often marked by a sort of fertile confusion. Nobody’s quite sure whether they’re talking about the same things, making the same assumptions, or pursuing the same interests, but the process of trying to figure that out can be very illuminating. The philosophy of imagination, as a sub-field, often involves a similar fertile confusion: philosophical raccoons clambering through the junkyard of the mind, trying to find, clean off, and exchange discarded fragments of ideas about fiction, modal epistemology, fantasy, thought-experiments, creativity, mindreading, and anything else that someone might describe in terms of ‘imagination’.

So interdisciplinary study of the imagination makes a lot of sense, and this was well in evidence in Bristol over September 15th-17th. Max Jones, Deb Marber, and Sam Wilkinson (of Junkyard fame themselves) assembled 15 talks and 2 panels, about half online and half in person, into The Science and Philosophy of Imagination.

What most struck me about the talks was how consistently certain themes emerged. Despite their different theoretical backgrounds, I think almost every speaker was concerned with one of five topics:

1. One theme, to nobody’s surprise, was visual imagery. Crawford Winlove and Margherita Arcangeli (also of Junkyard fame) presented masterful surveys of the current state of empirical research on visual imagery - Winlove summarising what neuroimaging tells us about the involvement of visual perception areas in visual imagery (tldr; it’s still not completely clear), and Arcangeli discussing the evolving literature on various forms of aphantasia. More provocatively, “Against Imagination” by Bence Nanay (yet more Junkyard fame - about half the speakers were regulars here) argued that we should consider dropping the concept ‘imagination’ altogether, in favour of the more scientifically-tractable concept ‘mental imagery’. He didn’t say that we should leave the junkyard of the mind on the trash heap of history, but he might as well have!

2. Nanay’s talk also connected with a second theme: defining imagination. By this point it’s a familiar refrain that nobody can quite agree what this term covers (the only thing philosophers of imagination know for sure about the imagination is that we’re writing about it…). Moreover, much of whatever the term does cover might potentially be covered by some other notion, like that of mental imagery - or that of ‘internal simulation’, the central notion in roboticist Alan Winfield’s presentation on ‘Robots with an Artificial Imagination’. If a robot builds a model of its world and itself, uses it to run simulations of its possible actions prior to doing them, and then acts based on the outcomes of those simulations, should we say it has used imagination? (Either way, it was nice to know that Winfield’s robots had independently evolved the awkward I’m-trying-to-get-around-you dance that humans sometimes do.)

My own talk was also squarely in this territory, examining Peter Langland-Hassan’s (Junkyard fame, once again) engagement with accounts of reasoning developed by Ruth Byrne and Philip Johnson-Laird. The question was how closely the thing we call ‘imagination’ is connected to the thing we call ‘reasoning’, and whether, if the connection is very close (as Byrne and Johnson-Laird argue), we should conclude that imagination is really just reasoning, or that reasoning is really just imagination. Ruth Byrne herself had been scheduled to deliver one of the keynote talks, but to everyone’s sadness was unfortunately unable to attend. 

3. A third thing that, like imagery and simulation, is sometimes identified and sometimes contrasted with imagination is creativity, the third theme. Paul O’Dowd demonstrated his robot that draws, even while carefully clarifying the limitations of its ‘creativity’, and two different talks directly addressed the role of creativity/imagination in scientific practice. Mike Stuart (of Junkyard fame) explored the epistemological frameworks scientists use to think about and evaluate their use of imagination to assist in research, while Alexander Bird sounded a somewhat more cautionary note: in light of the replication crisis, we might need to worry about too much imagination, if it leads to the framing, testing and often, just by the law of averages, apparent confirmation, of too many false hypotheses.

4. As well as all these talks about science, brains, and robots, there was a rich vein of philosophical engagement with the phenomenology of everyday life. One major theme here was affect and motivation in imagination. Three talks examined different areas where the imagination seems able to engage our feelings and change our desires:

• Edward Armitage explained how imagining the things we worry about can make our worries either more, or less, proportionate to their objects.

• Alex Fisher explored the controversy over whether performing actions in video games influences our feelings about those actions in real life.

• Andrea Rivadulla Duro reviewed experimental evidence of common behavioural impacts from imagining and experiencing, in cases like exposure therapy for phobias, or imagining eating food and feeling appetite for that food.

In each case, there was interest both in how far, empirically, imagination seems to be able to influence affect and motivation, and also in how far it should - how far such influence was appropriate, rational, or beneficial.  

5. Closely connected to these discussions of affect was the final theme, and perhaps the most frequent: imagination and embodiment. A remarkable number of speakers argued either that imagination ‘is embodied’ (always or often) - including Tom Schoonen (of Junkyard fame), Adriana and Maria Jimena Clavel Vázquez (also of Junkyard fame), and I-Jan Wang and Ying-Tung Lin (yet again of Junkyard fame).

Others argued simply that imagination can be usefully illuminated by drawing on theories which, in different ways, centre bodily experience in their account of the mind - Tom McClelland drawing on ecological psychology to explore the idea of ‘imaginative affordances’, and a whole panel on imagination within the framework of predictive processing.

A key idea was that more embodied imagining, in virtue of its richness, might be in some ways epistemically more powerful and in some ways epistemically more limited. If our imagining typically draws on multimodal, motor and somatosensory systems, and if these are not only integrated with each other but with affective and motivational systems, all reflecting an agent’s particular social history, then it becomes easier to see how it might give us reliable knowledge about our possibilities for action and perception, or even enhance our athletic skills (as argued by Lin and Wang). But it also becomes potentially harder to see how it could give us reliable knowledge about remote metaphysical possibilities (a point argued particularly by Schoonen), or about the experiences of differently-embodied others (a point argued particularly by the sisters Clavel Vázquez).

The theme of embodiment was particularly appropriate for this conference, since about half the attendees were embodied in the normal human manner, while the other half were embodied only through audio feeds and zoom screens, and one day of the conference was entirely online. Several attendees remarked that it was their first in-person conference since the beginning of the pandemic, and that it brought with it a certain sort of intangible joy in the tangibility of things. Coffee breaks, post-conference drinks, and walking to the restaurant together are not especially glamorous experiences, but after a year stuck at home they can come to seem oddly precious. Imagine that!

***

(Apologies must go to Adam Toon, whose talk I didn’t manage to fit into any of the thematic clusters: he gave a fascinating account of how theories of pretence along the lines of Kendall Walton’s could be used to better develop the radical idea that all our talk about minds and mental states is really just a useful fiction - rather like the religious fictionalism discussed in last week’s post.)