Keyboard Shortcuts?f

×
  • →Next step
  • ←Previous step
  • ↓Skip this slide
  • ↑Previous slide
  • mShow slide thumbnails
  • tShow transcript (+SHIFT = all transcript)
  • nShow notes (+SHIFT = all notes)

Please be cautious in using the transcripts.

They were created mechanically and have mostly not been checked or revised.

Here is how they were created:

  1. live lecture recorded;
  2. machine transcription of live recording;
  3. ask LLM to clean up transcript, and link to individual slides.

This is an error-prone process.

Click here and press the right key for the next slide.

(This may not work on mobile or ipad. You can try using chrome or firefox, but even that may fail. Sorry.)

also ...

Press the left key to go backwards (or swipe right)

Press n to toggle whether notes are shown (or add '?notes' to the url before the #)

Press m or double tap to slide thumbnails (menu)

Press ? at any time to show the keyboard shortcuts

 

Mindreading as Asynchronous Coordination

Silent Films (Devine & Hughes, 2013)

“Why do you think the driver locks Harold in the van?”

“Why do you think the driver locks Harold in the van?”
wrong answers include to kidnap.
fully correct answers require saying the driver did not know Harold was in the van.

-> predicts independently rated peer social competence

(Devine, Kovatchev, Grumley Traynor, Smith, & Lee, 2023)

Birmingham Advanced Mindreading Stories

interpret stories, measure alignment the authors

-> convergent validity, long-term stability

Lee et al. (2026)

The observation that some mindreaders are better than others was the starting point for my work with Ian Apperly and Rory Devine.
[NO: completely wrong. In categorical colour perception, there's no need for anything like social consensus through joint action; could just be the one colour dictator who sets the category boundaries for all the research shows (and the training experiment is actually like that)] In thinking about categorical colour perception as a model for mindreading we have arrived at the idea that mindreading is asynchronous coordination.
This is where I think the background assumptions in mindreading come from.

What constitutes successful practical mindreading?

constraints:

- limited evidence, bounded rationality

- we as researchers do not share knowledge of the truth

‘an individual is successful at mindreading if they ascribe states, predict behaviours, assign responsibility, and so on in just the way that would occur if they were involved in a joint mindreading activity with others around them’

(Apperly, Devine, & Butterfill, 2026)

The Mindreading as Coordination hypothesis (MAC) hypothesis includes some proposals about mechanisms but here I’m just focussing on the theoretical parts.
(^think am I right to be framing the question as where mindreading standards come from? Maybe it's more about what kind of thing they are?)
[In categorical colour perception, there's no need for anything like social consensus through joint action; could just be the one colour dictator who sets the category boundaries for all the research shows (and the training experiment is actually like that)]
This is where I think the background assumptions in mindreading come from.

Mindreading as asynchronous coordination

mindreading standards

are ‘the intuitions and principles with which groups of mindreaders will tend to agree, and to which individuals hold themselves in order to be considered rational and reasonable’

(Apperly et al., 2026)

This needs one qualification: I think it makes the background assumptions sound more like things people actually know. (Come to this later in norm-like patterns.)
So where are we with respect to explaining the thing to be explained?
Immediate task was to explain where background assumptions come from.
So where do those background assumptions come from? The Mindreading as Coordination hypothesis suggests that they arise from coordination with others on attributions, which aligns us on principles.
By analogy with categorical colour perception, you can see that there may also be a role for non-social—or even innate—constraints.

confident local agreement

+

variation between groups and experts

for

- agency attributions

- mental state attributions

How to explain both agreement and variation?

- need to ask where background assumptions come from

- Mindreading as coordination: local agreement results from coordination; variation from different interaction histories

The MAC hypothesis was not supposed to explain this pattern, but I think it does provide a partial explanation.
∞todo: state explanatory aims of MAC note that this does not include my explananadum (local agreement and variation)
UPDATE: collapse into one limit. How can the standards be a source of stable normative pressure if they are unknown?
The twin limits—unknown standards and normativity—fit belong together because they are difficult to fit into a single story.:
We need to recognize that there is criticism when attribution fail to meet a standard, and also that people generally cannot articulate the standard and might have false beliefs about what the standard is. (Not saying they do: just saying we do not want to require awareness of the standard.) It’s difficult to fit both unknown standards and normativity together.
 

limit: unknown standards

Preview: the proposal I'm aiming for is coordination in attribution explains local agreement and variation; norm-like patterns explain how the standards can be unknown but also stable.
Bring in the point about attributions being criticised later.
I think there are three limits: friction, normativity and unknown standards
The limits do not contradict MAC but justifies asking how the standards can be maintained.
(i) there is apparently a normative aspect to attributions: if I try to excuse a friend by denying knowledge or intention, you might reject my attribution and insist they did know; or you might demand that I justify my claim that their intentions were good.
This is easier to see by looking at groups that have quite different standards to ours. Schieffelin (2008) reports that among the Bosavi in Papua New Guinea, attributing mental states to others is generally taboo. ‘While children are encouraged to verbalize their own desires and intentions, they are explicitly socialized not to verbally guess at or express others’ unvoiced intentions and unclear meanings.’
To illustrate normativity missing, return to the catgeorical colour analogy
in that case there are standards but perhaps not norms in any interesting sense (although people do sometime get quite cross if you don't agree on the colours of things).
On the hypothesis that Mindreading is Asynchronous Coordination, every attribution contributes in a small way to setting mindreading standards. Suppose we disagree about whether Ayesha intended to harm the cat or whether it was merely an accident.
If we ask what kind of thing the socially shared mindreading standards are, we immediately notice that ordinary folk do not necessarily know much about them. They don't seem to be the kinds of thing that people can state.
(ii) mindreading principles are likely to be often things that people cannot articulate, and about which they probably have many false beliefs.
These limits raise some questions. How does the normative aspect of mindreading fit into the picture? And if mindreaders mostly do not know what the mindreading standards are, how are those standards maintained across large groups?