Deafened young horseshoe bats shift the frequency of their CF sonar
signals (Rübsamen 1987), indicating that auditory feedback does influence
vocal output, resulting in precise matching of cochlear tuning and vocalization frequency in normal individuals. However, vocal output has no
effect on cochlear tuning or responsiveness of the auditory system, so this
influence is unidirectional (Rübsamen 1987) and results in fine-tuning the
frequency of existing vocalizations rather than substantially reorganizing
signal structure.
Several researchers hypothesize that sonar signals develop from communication signals (Moss 1988). Isolation calls in infants who had no acoustic
input are normal, infants give these calls within a few hours of birth, and
species that cannot hear at birth produce isolation calls. Acoustic input is
clearly not essential for normal development (Ehret 1980). This appears to
be true also for sonar signals (Gould 1975). In none of these instances is
there direct evidence testing the role of vocal learning. Isolation calls do
show age-related changes (Jones et al. 1991; Scherrer and Wilkinson 1993).
Most changes are consistent with maturation of laryngeal and respiratory
function and of peripheral and central nervous control, allowing greater
control of vocal production (Could 1975). Functional considerations make a
reliance on acoustic input for call acquisition and normal development
unlikely. A bat with abnormal sonar signals would be severely handicapped
in foraging. An infant unable to produce normal isolation calls risks permanent separation from its mother, on whom it is completely dependent early
in life. Vocal learning, when it occurs, is probably restricted to social modification to increase individual distinctiveness (Masters et al. 1995) or increase
similarity to the infant’s mother (Jones and Ransome 1993; Esser 1994)
and to tune vocal output to the best frequency of the auditory system
(Rübsamen 1987). Thus, individual signatures and echolocation calls are
more likely to be socially modified than acquired through learning.
Very young infant bottlenose dolphins give whistles, suggesting that
acoustic input is unnecessary for normal call production; however, direct
tests have not been done. As with infant bats, a reliance on acoustic input
for normal development is risky, so social modification to adjust the degree
of similarity with social partners is likely to be the predominant form of
vocal learning. An interesting question is whether the preponderance of
social modification over learned acquisition in bats and cetaceans is due to
similar call function or because both echolocate.
5.2.5.2. The Process of Call Matching for Mammals:
Delayed Critical Period or Selective Attrition?
In both bats and dolphins, the critical period seems to be either nonexistent or labile to accommodate changing group composition and social relationships. However, the mechanism of call matching is unclear. In bats,
206
J.W. Boughman and C.F. Moss
signals (Rübsamen 1987), indicating that auditory feedback does influence
vocal output, resulting in precise matching of cochlear tuning and vocalization frequency in normal individuals. However, vocal output has no
effect on cochlear tuning or responsiveness of the auditory system, so this
influence is unidirectional (Rübsamen 1987) and results in fine-tuning the
frequency of existing vocalizations rather than substantially reorganizing
signal structure.
Several researchers hypothesize that sonar signals develop from communication signals (Moss 1988). Isolation calls in infants who had no acoustic
input are normal, infants give these calls within a few hours of birth, and
species that cannot hear at birth produce isolation calls. Acoustic input is
clearly not essential for normal development (Ehret 1980). This appears to
be true also for sonar signals (Gould 1975). In none of these instances is
there direct evidence testing the role of vocal learning. Isolation calls do
show age-related changes (Jones et al. 1991; Scherrer and Wilkinson 1993).
Most changes are consistent with maturation of laryngeal and respiratory
function and of peripheral and central nervous control, allowing greater
control of vocal production (Could 1975). Functional considerations make a
reliance on acoustic input for call acquisition and normal development
unlikely. A bat with abnormal sonar signals would be severely handicapped
in foraging. An infant unable to produce normal isolation calls risks permanent separation from its mother, on whom it is completely dependent early
in life. Vocal learning, when it occurs, is probably restricted to social modification to increase individual distinctiveness (Masters et al. 1995) or increase
similarity to the infant’s mother (Jones and Ransome 1993; Esser 1994)
and to tune vocal output to the best frequency of the auditory system
(Rübsamen 1987). Thus, individual signatures and echolocation calls are
more likely to be socially modified than acquired through learning.
Very young infant bottlenose dolphins give whistles, suggesting that
acoustic input is unnecessary for normal call production; however, direct
tests have not been done. As with infant bats, a reliance on acoustic input
for normal development is risky, so social modification to adjust the degree
of similarity with social partners is likely to be the predominant form of
vocal learning. An interesting question is whether the preponderance of
social modification over learned acquisition in bats and cetaceans is due to
similar call function or because both echolocate.
5.2.5.2. The Process of Call Matching for Mammals:
Delayed Critical Period or Selective Attrition?
In both bats and dolphins, the critical period seems to be either nonexistent or labile to accommodate changing group composition and social relationships. However, the mechanism of call matching is unclear. In bats,
206
J.W. Boughman and C.F. Moss