The study reported in this manuscript examined Fodor's (1992) argument
that standard false belief tasks used in developmental research serio
usly underestimate young children's understanding of false belief. The
problem of these tasks according to Fodor is that always a unique, ac
tual state of affairs (e.g., chocolate is now in cupboard B) is contra
sted with a believed state of affairs (e.g., chocolate is still in cup
board A). Fodor argued that this uniqueness feature may be critical be
cause young children with limited computational resources have to trad
e reliability of behavioral prediction for computational simplicity an
d, therefore, may rely on simple heuristics such as ''Predict that the
agent will act in a way that will satisfy his desires''. In the stand
ard false belief task such a heuristic will result in a unique, but in
correct (reality-based) prediction. Fodor's expectation is that when y
oung children are not misled into applying such heuristics by the poss
ibility of unique, reality-based prediction, then their true competenc
e for belief-based reasoning will become evident. The present study co
ntrasted for two different belief tasks a traditional unique version w
ith a non-unique version, but found no support for Fodor's expectation
as both 3- and 4-year-old children did not improve in the non-unique
version.