Primate cognition is believed to have adapted during evolution in resp
onse to complexity, stemming from either social or environmental chall
enges. Arguments favoring one or another domain as the predominant spu
r to increased intelligence have been proposed on the basis of the inh
erent complexity of problems facing primates and the sophistication of
their observed behavioral response. A review of recent findings shows
that cognitive differences between primates are well revealed by natu
rally posed problems. Current evidence shows a sensitivity to fine soc
ial distinctions in haplorhine primates, associated with complex socia
l manipulations and neocortical enlargement, compared to strepsirhines
. Several aspects of foraging behavior also suggests cognitive sophist
ication in various primates. Depth of understanding, however, is great
er in great apes than in monkeys; this applies to both spheres, social
(e.g., comprehending mental states) and nonsocial (e.g., comprehendin
g physical mechanism and hierarchical organization of behavior). The i
ncreased representational understanding in great apes is not associate
d either with more manipulative social actions or with neocortical enl
argement, compared to monkeys. This evidence supports theories of a so
cial origin of cognition for those aspects shared by monkeys and apes
but not for the unique qualitative differences of great apes. Task com
plexity is difficult, perhaps impossible, to measure or compare across
domains; any such comparison would certainly be premature at present
for primates. Behavioral skill, though in principle simpler to compare
, is at present difficult to assess when there is no common currency o
f theory-building. With the aim of encouraging comparisons of data fro
m laboratory and field and from social and technical problems, a preli
minary exploration is made with a notation derived from artificial int
elligence; this is shown to be capable of representing theories of com
plex behavior in both social and technical domains. (C) 1995 Wiley-Lis
s, Inc.