First,each one is produced in a context that can be clearly identified.
Thus,grunts are produced during affiliative interactions involving food
or a conspecific (Hauser and Marler 1993a,section 2).Shrill barks are
given exclusively in the context of alarm,and for rhesus monkeys on
Cayo Santiago,represent their only alarm vocalization (Hauser and
Marler 1993a;Bercovitch,Hauser,and Jones 1995).Copulation screams
are given only by adult males during copulation and in no other context
(Hauser 1993b).Second,quantitative acoustic analyses were already
available from published results (Hauser and Marler 1993a) and unpub-
lished data.Thus,before starting our experiments,we had a good under-
standing of the range of acoustic variation both at the population level
and in terms of specific features of the call.Third,in manipulating the
structure of a call away from its species-typical morphology,it is impor-
tant to avoid changing its structure into that of a different call from
within the repertoire.Thus,for example,adding a broad,frequency-
modulated component to the terminal portion of the end of a coo turns
the signal into a harmonic arch (see figure 6.2).For grunts,shrill barks,
and copulation screams,manipulating interpulse interval does not trans-
form them into different call types from within the repertoire.
Having manipulated one parameter of the call,we conducted playback
experiments using the design of our previous experiments (Hauser and
Andersson 1994).Specifically,calls were broadcast from a speaker
located 180 degrees behind the subject and head orientation was scored.
Figure 6.6 shows results from playbacks of each call type.For all three
types,playbacks of unmanipulated exemplars,and exemplars with inter-
pulse intervals reduced to the population minimum,subjects showed a
highly significant right ear bias.For grunts and shrill barks,eliminating
interpulse interval eliminated orienting bias,with some individuals
turning to the right,some to the left,and some not responding at all;for
copulation screams,however,right ear bias was preserved.When inter-
pulse interval was stretched to the maximum in the population,the ten-
dency was for subjects to orient with the left ear leading for both grunts
and shrill barks,but this pattern was not statistically significant;for cop-
ulation calls,right ear bias was preserved.Finally,when interpulse inter-
val was stretched to twice the maximum,subjects showed a statistically
significant left ear bias for the grunt and shrill bark,but a right ear bias
was preserved for the copulation scream.
For grunts and shrill barks,manipulating interpulse interval beyond
the species-typical range of variation (at least for this population) caused
a shift from right ear bias to either no bias (eliminating interpulse inter-
val) or to a significant left ear bias (two times the maximum interpulse
interval).This pattern of change was not observed in playbacks of cop-
ulation screams.93 Primate Vocalizations in Emotion and Thought