be very large. This overproduction suggests that a process of selective attrition (Nelson et al. 1996) may be involved in shaping adult repertoires of
three to nine whistles. Calves’ acquisition of this large number of whistle
types may depend on auditory and social experience, although published
data are lacking to determine the importance of either factor in the initial
acquisition of whistles or the process of selecting which types to retain.
The contextual use of some whistles changes during development
(McCowan and Reiss 1995b). Calves call more when separated than in
socializing or other contexts until about 8 months of age. Their use of signature whistles in social contexts increases after this point. This pattern is
similar for signature and shared whistle types. The function of these whistle
types is uncertain, making it difficult to know whether developmental
changes arise from increasing accuracy of use or from changing social relationships and behavior as calves mature.
The ontogeny of whistle repertoires and production is more difficult to
study in free-living dolphins. Sayigh et al. (1990) repeatedly recorded 12
pairs of mothers and their one year or older calves during captures.
Dolphins were lightly restrained in shallow water and recorded with a
suction cup hydrophone attached to the melon. Within a single recording
session, renditions of individual calls varied in such acoustic parameters
as duration and absolute frequency even though whistle contour varied
little. The most commonly produced whistle of calves (presumably their
signature whistle) was stable, undergoing little systematic modification in
contour over the course of their 3-year study. Slight modifications to whistle
contour took place in a few individuals’ calls, but for most individuals, calls
did not vary more across years than within a recording session. Adult signature whistles were consistent over even longer time periods of up to 12
years. This study focused on stability of the signature whistle rather than
cataloging the entire repertoire. Signature whistles were recorded almost
exclusively—only 8% of calls recorded from female calves and 27% from
male calves were other whistle types. This finding contradicts the finding of
large repertoires (McCowan and Reiss 1995a, 1995b), perhaps because of
the different recording contexts. Although McCowan and Reiss (1995a,
1995b) worked in captivity, which can itself introduce artifacts, they
recorded in a variety of behavioral contexts, whereas Sayigh et al. (1990)
recorded only loosely restrained dolphins during captures.
Sayigh et al. (1990) found sex differences in the apparent acoustic model
used by males and females. Female calves produced signature whistles
that differed from their mothers’, whereas male calves were more likely to
produce whistles similar to their mothers’. This sex difference was less pronounced in a later study (Sayigh et al. 1995), where many male whistles differed from their mothers’. Sayigh et al. (1990) suggested that vocal learning
underlies this sex difference; however, their data cannot be used to determine directly whether learned acquisition or social modification occurred.
They studied calves older than one year and found little change in whistle
4. Comparative Vocal Learning
161
three to nine whistles. Calves’ acquisition of this large number of whistle
types may depend on auditory and social experience, although published
data are lacking to determine the importance of either factor in the initial
acquisition of whistles or the process of selecting which types to retain.
The contextual use of some whistles changes during development
(McCowan and Reiss 1995b). Calves call more when separated than in
socializing or other contexts until about 8 months of age. Their use of signature whistles in social contexts increases after this point. This pattern is
similar for signature and shared whistle types. The function of these whistle
types is uncertain, making it difficult to know whether developmental
changes arise from increasing accuracy of use or from changing social relationships and behavior as calves mature.
The ontogeny of whistle repertoires and production is more difficult to
study in free-living dolphins. Sayigh et al. (1990) repeatedly recorded 12
pairs of mothers and their one year or older calves during captures.
Dolphins were lightly restrained in shallow water and recorded with a
suction cup hydrophone attached to the melon. Within a single recording
session, renditions of individual calls varied in such acoustic parameters
as duration and absolute frequency even though whistle contour varied
little. The most commonly produced whistle of calves (presumably their
signature whistle) was stable, undergoing little systematic modification in
contour over the course of their 3-year study. Slight modifications to whistle
contour took place in a few individuals’ calls, but for most individuals, calls
did not vary more across years than within a recording session. Adult signature whistles were consistent over even longer time periods of up to 12
years. This study focused on stability of the signature whistle rather than
cataloging the entire repertoire. Signature whistles were recorded almost
exclusively—only 8% of calls recorded from female calves and 27% from
male calves were other whistle types. This finding contradicts the finding of
large repertoires (McCowan and Reiss 1995a, 1995b), perhaps because of
the different recording contexts. Although McCowan and Reiss (1995a,
1995b) worked in captivity, which can itself introduce artifacts, they
recorded in a variety of behavioral contexts, whereas Sayigh et al. (1990)
recorded only loosely restrained dolphins during captures.
Sayigh et al. (1990) found sex differences in the apparent acoustic model
used by males and females. Female calves produced signature whistles
that differed from their mothers’, whereas male calves were more likely to
produce whistles similar to their mothers’. This sex difference was less pronounced in a later study (Sayigh et al. 1995), where many male whistles differed from their mothers’. Sayigh et al. (1990) suggested that vocal learning
underlies this sex difference; however, their data cannot be used to determine directly whether learned acquisition or social modification occurred.
They studied calves older than one year and found little change in whistle
4. Comparative Vocal Learning
161