four years), independent of diVerences of intelligence and social input. Yet,
give or take the usual variation to be found in any genetically controlled
process of maturation, that is just what we doWnd.
Some people have pointed to the fact that wide cultural variations can
seemingly exist in folk-psychological beliefs, as evidence for an account of
theory-of-mind acquisition in terms of enculturation (Lillard, 1998). But
almost all of this evidence relates to cultural variations in theory-of-mind
vocabulary, and in variations in the beliefs about the mind which the folk
will articulate, consciously, in language. Whereas, if there is a theory-of-
mind module it is almost certain to be independent of language (see our
discussion in chapter 9 below, section 3.9). Moreover, the basic internal
operations of such a module may well be largely inaccessible to conscious-
ness. What we would predict, from the perspective of nativism/modular-
ism, is that there should be a core mind-reading capacity whose develop-
mental staging is the same across cultures, surrounded by a variable body
of cultural accretions and concepts. And insofar as there is evidence
bearing on this issue, this is just what weWnd (Avis and Harris, 1991; Naito
et al., 1995; Lillard, 1998).
Irrespective of their diVerences concerning the mode of theory-of-mind
acquisition, theory-theorists have shown considerable convergence in res-
pect of the diVerent developmental stages traversed on route to the mature
state, supported by an intriguing and diverse body of data (Wellman, 1990;
Perner, 1991; Baron-Cohen, 1995). There would appear to be three main
stages covered between birth and the age of four – stages which may
correspond to increasingly elaborate theories devised by a process of
theory-construction; or which may represent the maturational steps in the
growth of an innate module, perhaps recapitulating the course of its
evolution. TheWrst stage – which is probably attained in theWrst eighteen
months of life – is a simple form of goal-psychology. At this stage actions
can be interpreted as goal-directed, and predictions can be generated from
attributions of desire. (To see how this can work, notice that in general
what is salient to you in your environment will also be salient to others; so
if you attribute goals to others and work out what they will then do given
their surroundings – in their environment asyourepresent it, that is – this is
likely to prove reliable enough to be useful.)
The second stage – attained between the second and third years of life –
is a kind of desire–perception psychology. As before, children can at this
stage interpret and predict actions in the light of attributed goals; but they
can now make some allowance for the presence or absence of perceptual
contact with the relevant environmental facts. Yet there is no under-
standing of perception as asubjectivestate of the perceiver, which may
represent some aspects of an object but not others, or which may represent
The alternatives: theory-theory versus simulation 79