Vocalizations can be specifically designed by learning or by natural selection to enhance individual distinctiveness and facilitate individual recognition. These designed individual signatures appear to be relatively rare. They
include the cases where individuals produce distinct call types. The more
common mechanism is for individual distinctiveness to arise as a by-product
of individual variation in vocal tract morphology and body size. Such variation has a purely biophysical basis and appears to be sufficient in many
cases to generate a large number of unique signatures or distinct voices.
Such by-product distinctiveness can serve the signature function quite well,
although it is unlikely to involve learned acquisition. Individual distinctiveness is a simple consequence of morphological variation that may have
other functions when by-product distinctiveness is involved. In contrast,
selection must directly favor individual distinctiveness, or learning must
occur specifically to enhance it to produce designed signatures. Designed
individual signatures are likely to be found primarily when by-product distinctiveness is insufficient or unreliable. Even taxa that have repertoires of
call types may rely, in part, on by-product distinctiveness. For example, call
types usually fall into classes so that some individuals sound roughly similar.
Individuals that share call types may be differentiated on the basis of
by-product distinctiveness (e.g., Weary and Krebs 1992).
Understanding how selection has shaped vocalizations to serve as individual signatures requires knowledge of the potential benefits to conveying
individual identity offset against the costs that might arise. From the
signalers’ point of view, indicating identity could be beneficial to facilitate
interactions between social partners. Good examples are isolation calls that
infants give to facilitate reunions with caregivers in many species of bats
(e.g., Scherrer and Wilkinson 1993). However, indicating identity might not
always be favored. A signaler might also benefit from concealing identity
or from imitating another individual. Vocal learning can facilitate both
of these kinds of cheating, making it costly from the receivers’ point of
view. Vocal learning will only be favored if its benefits outweigh its costs.
In developing and testing hypotheses for the evolution of vocal learning
these potential costs should not be overlooked. Costs can be a powerful
force shaping signal design and development (see Fitch and Hauser,
Chapter 3).
Vocal learning is one mechanism to enhance differences between individuals (e.g., Janik 1999), and this mechanism operates on the time scale of
individual lifetimes. However, individual signatures need not be learned,
and in fact several demonstrated cases have a strong heritable component
(e.g., Scherrer and Wilkinson 1993). Heritable individual signatures reflect
the action of selection over evolutionary time to enhance differences
between individuals. As an example, learning plays almost no role in structuring squirrel monkey calls, but individuals vary in voice (Winter et al.
1973; Symmes et al. 1979) and are likely to be recognized by their calls
(Snowdon et al. 1985; Boinski and Mitchell 1997), so the calls function as
individual signatures. Studies of vocalizations in several taxa have shown
146
J.W. Boughman and C.F. Moss
include the cases where individuals produce distinct call types. The more
common mechanism is for individual distinctiveness to arise as a by-product
of individual variation in vocal tract morphology and body size. Such variation has a purely biophysical basis and appears to be sufficient in many
cases to generate a large number of unique signatures or distinct voices.
Such by-product distinctiveness can serve the signature function quite well,
although it is unlikely to involve learned acquisition. Individual distinctiveness is a simple consequence of morphological variation that may have
other functions when by-product distinctiveness is involved. In contrast,
selection must directly favor individual distinctiveness, or learning must
occur specifically to enhance it to produce designed signatures. Designed
individual signatures are likely to be found primarily when by-product distinctiveness is insufficient or unreliable. Even taxa that have repertoires of
call types may rely, in part, on by-product distinctiveness. For example, call
types usually fall into classes so that some individuals sound roughly similar.
Individuals that share call types may be differentiated on the basis of
by-product distinctiveness (e.g., Weary and Krebs 1992).
Understanding how selection has shaped vocalizations to serve as individual signatures requires knowledge of the potential benefits to conveying
individual identity offset against the costs that might arise. From the
signalers’ point of view, indicating identity could be beneficial to facilitate
interactions between social partners. Good examples are isolation calls that
infants give to facilitate reunions with caregivers in many species of bats
(e.g., Scherrer and Wilkinson 1993). However, indicating identity might not
always be favored. A signaler might also benefit from concealing identity
or from imitating another individual. Vocal learning can facilitate both
of these kinds of cheating, making it costly from the receivers’ point of
view. Vocal learning will only be favored if its benefits outweigh its costs.
In developing and testing hypotheses for the evolution of vocal learning
these potential costs should not be overlooked. Costs can be a powerful
force shaping signal design and development (see Fitch and Hauser,
Chapter 3).
Vocal learning is one mechanism to enhance differences between individuals (e.g., Janik 1999), and this mechanism operates on the time scale of
individual lifetimes. However, individual signatures need not be learned,
and in fact several demonstrated cases have a strong heritable component
(e.g., Scherrer and Wilkinson 1993). Heritable individual signatures reflect
the action of selection over evolutionary time to enhance differences
between individuals. As an example, learning plays almost no role in structuring squirrel monkey calls, but individuals vary in voice (Winter et al.
1973; Symmes et al. 1979) and are likely to be recognized by their calls
(Snowdon et al. 1985; Boinski and Mitchell 1997), so the calls function as
individual signatures. Studies of vocalizations in several taxa have shown
146
J.W. Boughman and C.F. Moss