This chapter focuses on problems of mechanism.I briefly discuss some
traditional views of animal vocal communication,and then show by
examples that many of these ideas must either be modified or rejected.
The empirical work is divided into two sections.The first explores how
the acoustic space of a nonhuman primate’s vocal repertoire can be cap-
tured by quantifying both affective and referential components of the
signal.The second draws on our understanding of call meaning and func-
tion to assess whether brain asymmetries underlie acoustic perception
and vocal production.All of the empirical work centers on one species,
the rhesus macaque (Macaca mulatta).I will,however,refer to other
species where relevant.Cries of the Heart and Mind
By the late 1970s,studies of trained apes and dolphins revealed that the
conceptual tools required to produce a referential system of communi-
cation were present in these animals,and could be expressed by means
of an artificial language (Premack 1986;Gardner,Gardner,and Van
Canfort 1989;Savage-Rumbaugh et al.1993;Herman,Pack,and Palmer
1993).In contrast,there was no evidence that natural vocalizations
produced by these animals were referential.The general consensus,
dating back to Aristotle,Descartes,Darwin,and other luminaries,thus
remained:animal vocalizations reflect changes in the signaler’s affective
state,emotions,and motivations.In 1980,however,a crucial experiment
(Seyfarth,Cheney,and Marler 1980) forced this view to undergo a sig-
nificant facelift.The first insight emerged from Struhsaker’s (1967) obser-
vation that vervet monkeys produce acoustically distinctive alarm calls
in response to three predatory classes:big cats (leopards,cheetah),birds
of prey (martial and crowned eagles),and snakes (pythons,mambas).
On hearing such calls or seeing the predator,individuals reacted with
equally distinctive escape responses.Tight pairing between call type and
response suggested that such calls might function as labels for a preda-
tory type.Using taperecorded alarm calls,playback experiments were
conducted.If the calls provide sufficient information about the predator
encountered,playbacks should be sufficient to elicit behaviorally appro-
priate responses.They were.In essence,when vervet monkeys hear an
alarm call,they are not only struck by a salient emotional event (i.e.,they
experience fear),but they are provided with information that enables
them to make a highly adaptive response.If a leopard is about,the best
place to be is high up on the thin branches of an acacia tree.If an eagle
is near,the best place is under a bush;eagles can scoop vervets out of
trees.Finally,if a snake is in the vicinity,the response is to stand bipedally
and scan the ground nearby.78 Marc D.Hauser