Studies on children's social understanding tend to assume that understandin
g the minds and emotions of others are theoretically and methodologically i
ndistinct. This study, however, aimed to assess differential effects, with
particular reference to the quality of the attachment relationship and mate
rnal mental health. The participants were 45, 40-month-old children who wer
e assessed on tasks of social understanding (false belief, Bartsch & Wellma
n, 1989, and emotion understanding, Denham & Auerbach, 1995); attachment se
curity was assessed by a story completion task (Bretherton, Ridgeway, & Cas
sidy, 1990), and maternal mental health was assessed by Beck's Depression I
nventory (BDI) (Beck, Ward, Mendelson, Mock, & Erbaugh, 1961). The hypothes
es were (a) that insecure children would have mothers scoring higher on the
BDI than secure children and poorer performances on the social understandi
ng tasks; (b) that insecure children need not be any worse than secure chil
dren in mind appraisal, rather, it was anticipated that their difficulties
would be with emotion appraisal. It was found that, relative to secure chil
dren, insecure children had lower verbal mental ages, poorer performances o
n emotion understanding and had mothers scoring higher in depression. Inter
estingly, no significant effects were found for mind understanding. A regre
ssion analysis showed that verbal mental age and attachment were significan
t predictors of the child's emotion task performance whilst depression and
SES were not. In the light of these results, this paper considers the need
for a move towards an integration of theories and methods.