Ben Alderson-Day, Jamie Moffatt, Marco Bernini, Kaja Mitrenga, Bo Yao, and Charles Fernyhough, ‘Processing Speech and Thoughts during Silent Reading: Direct Reference Effects for Speech by Fictional Characters in Voice-Selective Auditory Cortex and a Theory-of-Mind Network’, Journal of Cognitive Neuroscience, 32 (2020), 1637–1653: https://doi.org/10.1162/jocn_a_01571
This is a huge achievement, and I am proud to say that when my friend Charles Fernyhough described it to me, I felt no pang of envy, even though it achieves something I have dreamt of myself. Here we see an idea about how literature works tested effectively in brain-scanning experiments. I’ve written about the challenges faced in my own small adventures in this sort of direction in a few posts (this one was the first). Alderson-Day et al. managed both to retain the key characteristics of the texts convincingly, and to generate results that seem both solid and suggestive.
This paper arises from the excellent the Hearing the Voice project, which examines the phenomenon of the voice in the head, and voices (plural) in the head, in a wide range of contexts and disciplines. Even if Charles weren’t my friend, I would still think this is a flagship for work spanning literature and psychology. I’ve touched on the project and Charles’s work in a couple of posts before, such as here and here. They have been working on the ways that a voice in the head might be caused and experienced and understood, and the key question in this paper goes to the heart of some important issues: when we encounter the written representation of voices, do we experience them differently from other written evidence?
They found that there was more elevated activation in the auditory cortex when subjects read direct speech, than when they read indirect speech. The indication of a fictional present out-loud voice, in quotation marks or whatever, resulted in brain activity suggesting it was processed in the same way as speech. The variation in effect was not only observed in the auditory centres, but also in brain areas ‘previously associated with inferring intentions from communication’. Previous studies had suggested the first effect, but the further extension was new, implicating ‘higher-order processes associated with gauging character intention and meaning, for now at least’.
They connect their finding works to the approach taken by , who is interested in how inner recreation of voices and thoughts in fiction work, and how they might be different. I suppose it is important to presume that while indirect speech and represented thoughts may not be activating brain mechanisms for interpreting and appraising speech, they are doing something else instead. ‘Speech in fiction may be special’, said the mysterious figure, ‘but we should probably think of this as a difference in kind, rather than in degree.’
Writers of course have a kind of deep knowledge about the difference made by representing speech directly rather than reporting it. They find ways of making that distinction less clear, by having a third-person indirect representation convey more or less manifestly the words and demeanour of the voice itself. Direct speech may also be evasive, giving very limited clues about how the stylings of the speech event beyond the words. Such manoeuvres surely add nuances to questions about what may or may not follow from the emergence of a character’s voice within the prose. I think it’s in those grey areas, the not-quite-direct speech, that some extremely interesting things might be happening.