Explain the emergence, in evolution or development, of mindreading.
So let me conclude.
The challenge we have been addressing was to understand the emergence of mindreading.
Initially this seemed straightforward: you learn this from social interaction using language as a tool (compare Gopnik's theory theory).
However, the discovery that abilities to track beilefs exist in infants from around 7 months or earlier initially suggested a different picture:
one on which mindreading was likely to involve core knowledge.
But, as always, things are not so straightforward. The evidence is apparently conflicting.
There are actually two conflicts, not one: developmentally (A- & B-tasks) and in adults, automaticity.
The existence of two puzzles gives us confidence that the conflict is not merely a methodological artefact.
The solution is to recognise that there are modules, but there was an obstacle to the hypothesis that mindreading could be modular (*cognitive efficiency)
In constructing minimal theory of mind we've earned the right to solve them by appeal to modularity. (NB: it's modularity rather than mTm that explains the discrepancy; mTm is important because (i) it explains efficiency; and (ii) it generates predictions via signature limits)
Now the idea that there are both modular and non-modular mindreading enables us to solve the two puzzles (developmentally (A- & B-tasks) and in adults, automaticity.).
However, this resolution of the puzzles doesn't answer our overall challenge about the origins of knowledge of other minds.
In fact it complicates the account of the origins of knowledge of other minds, makes the challenge seem harder rather than easier to meet.
We can't give a theory theory / learning account; and we also can't give a straightforward core knowledge account.
Instead we need something a bit more complex.
I haven't tried to offer an account of what that thing is yet, and that isn't the point of this lecture.
But let me close by describing how I would approach it.
A key issue is the relation between infant and adult competence ...