Hausberger 1997). If our premise here is correct—that call function reveals
where selection operates—then social interactions should have a very
strong part to play in call learning because calls often function to mediate
complex social interactions. When the degree of similarity affects call function, social partners are likely to serve as call tutors, and the absence of
social interaction is likely to have a profound influence on call development, especially when learned acquisition occurs.
Primates may not show vocal learning because other cues are used for
identification and social facilitation, with vocalizations playing only a supportive role. Alternatively, individuals’ distinctive calls are a by-product of
morphological and body-size differences, which do not require vocal learning. Long lives and close association allow troop members to recognize
group mates’ individually distinctive calls. Under both hypotheses, vocal
learning simply is not necessary.
5.2.4. Relevance to Call Learning in Birds
We would next like to apply these ideas to the few cases where data on the
process of call learning are available for birds and mammals. This allows us
to evaluate the extent to which auditory template or sensorimotor models
describe call learning in mammals and birds. We also describe the process
of call matching to evaluate the universality of critical periods and the
extent of social influences on call development. Ideal studies for evaluating
these hypotheses include replicated data from deafened, totally isolated,
acoustically isolated, socially isolated, and normally reared individuals.
Unfortunately, we have these comprehensive data for very few taxa; therefore, our conclusions are tentative. Currently, insufficient data are available
for most taxa to test the selective attrition hypothesis for call learning.
5.2.4.1. Sensorimotor Model in Birds
Evidence that sensory input is essential for normal call learning comes from
studies on contact calls in budgerigars (Dooling et al. 1987; Brittan-Powell
et al. 1997; Heaton and Brauth 1999; Heaton et al. 1999) and chickadees
(Ficken et al. 1985; Hughes et al. 1998) and for song learning in Anna’s hummingbirds (Baptista and Schuchmann 1990). Early vocalizations are often
structurally simple and highly variable, and they become more complex and
stereotyped with age (Clemmons and Howitz 1990; Brittan-Powell et al.
1997). Deafened budgerigars developed abnormal contact calls that bore
little resemblance to the species’ typical call (Dooling et al. 1987; Heaton
and Brauth 1999). Calls of adults without auditory feedback deteriorated
(Heaton et al. 1999). Juvenile chickadees who did not hear adult calls developed abnormal variable see, gargle (Ficken et al. 1985), and chickadee calls
(Hughes et al. 1998), indicating that appropriate acoustic input is essential.
Isolated juvenile Anna’s hummingbirds sang songs that differed from
normal song in several acoustic features (Baptista and Schuchmann 1990),
4. Comparative Vocal Learning
203
where selection operates—then social interactions should have a very
strong part to play in call learning because calls often function to mediate
complex social interactions. When the degree of similarity affects call function, social partners are likely to serve as call tutors, and the absence of
social interaction is likely to have a profound influence on call development, especially when learned acquisition occurs.
Primates may not show vocal learning because other cues are used for
identification and social facilitation, with vocalizations playing only a supportive role. Alternatively, individuals’ distinctive calls are a by-product of
morphological and body-size differences, which do not require vocal learning. Long lives and close association allow troop members to recognize
group mates’ individually distinctive calls. Under both hypotheses, vocal
learning simply is not necessary.
5.2.4. Relevance to Call Learning in Birds
We would next like to apply these ideas to the few cases where data on the
process of call learning are available for birds and mammals. This allows us
to evaluate the extent to which auditory template or sensorimotor models
describe call learning in mammals and birds. We also describe the process
of call matching to evaluate the universality of critical periods and the
extent of social influences on call development. Ideal studies for evaluating
these hypotheses include replicated data from deafened, totally isolated,
acoustically isolated, socially isolated, and normally reared individuals.
Unfortunately, we have these comprehensive data for very few taxa; therefore, our conclusions are tentative. Currently, insufficient data are available
for most taxa to test the selective attrition hypothesis for call learning.
5.2.4.1. Sensorimotor Model in Birds
Evidence that sensory input is essential for normal call learning comes from
studies on contact calls in budgerigars (Dooling et al. 1987; Brittan-Powell
et al. 1997; Heaton and Brauth 1999; Heaton et al. 1999) and chickadees
(Ficken et al. 1985; Hughes et al. 1998) and for song learning in Anna’s hummingbirds (Baptista and Schuchmann 1990). Early vocalizations are often
structurally simple and highly variable, and they become more complex and
stereotyped with age (Clemmons and Howitz 1990; Brittan-Powell et al.
1997). Deafened budgerigars developed abnormal contact calls that bore
little resemblance to the species’ typical call (Dooling et al. 1987; Heaton
and Brauth 1999). Calls of adults without auditory feedback deteriorated
(Heaton et al. 1999). Juvenile chickadees who did not hear adult calls developed abnormal variable see, gargle (Ficken et al. 1985), and chickadee calls
(Hughes et al. 1998), indicating that appropriate acoustic input is essential.
Isolated juvenile Anna’s hummingbirds sang songs that differed from
normal song in several acoustic features (Baptista and Schuchmann 1990),
4. Comparative Vocal Learning
203