may vary is by being more or lessradical– being more radical if it relies
exclusively on simulation, and less radical to the extent that it also depends
upon the subject’s use of general knowledge about mental states. Since a
strongly radical form of simulationism is none too plausible, there is some
danger of simulationism and theory-theory coalescing with each other.
Indeed, since we do acknowledge that for one function our mind-reading
capacitymustrely upon a form of simulation, the position we advocate –
limited simulation as an enrichment of the operation of an innate theory –
is not too far away from a veryunradical form of simulationism.
The main issue between theory-theory and simulationism concerns the
sort of cognitive process involved in mind-reading – whether it isknowl-
edge-drivenorprocess-driven, to use Goldman’s nice way of making the
contrast. But we should remember that theory-theory not only furnishes
us with aphilosophicalaccount of what conceptions of mental state types
are: according to theory-theory, it is also the folk-psychological theory
which supplies the ordinary human mind-reader with those very concep-
tions. Simulationism cannot very well just borrow this functionalist ac-
count, according to which such states asbelief,desire,hope, andfearare
understood in terms of their general causal interactions with other men-
tal states, characteristic stimuli, intentions, and subsequent behaviour.
Simulationism has to give up on the functionalist account of how we
understand concepts in the vocabulary of propositional attitudes and
intentional states – because such an account eVectively involves implicit
grasp of a theory. This looks like a serious gap unless the simulationist
can come up with an equally plausible account of how we conceptualise
the propositional attitudes.
So, can simulationists provide any adequate alternative story as to what
our conceptions of mental states might consist in? This is clearly no easy
task, given that some form of functionalist or theory-theory account is
much the best that anyone has been able to come up with in the philosophy
of mind (see chapter 1). Simulationists who have grappled with this task
diVer in the extent to which they have relied uponintrospectionas a
possible source for knowledge of mental states. Our contention (see section
2.2 below, and also Carruthers, 1996a) is that, whether they invoke intro-
spection or not, simulationists cannot give an adequate account of self-
knowledge.
We noted above that if routine mind-reading runs on simulation, then
simulation cannot be taken as full-blown, conscious pretending that one is
in someone else’s shoes. Such full-blown pretending occurs all right, but
simply is not commonplace enough. So simulationists need a form of
implicit, sub-personal pretending to match the theory-theorists’ appeal to
an implicit body of theoretical knowledge. Such an account has been
The alternatives: theory-theory versus simulation 81