that may favor vocal learning. We follow with case studies of birds and
mammals. These in-depth case studies are mostly in the bird groups oscines,
psittacines, and trochilids and in the mammalian groups chiropterans,
cetaceans, and primates because that is where research has focused or
evidence suggests vocal learning may occur. We recognize that different
sources of selection may favor vocal learning at each level and so divide
the following sections into three functional levels: individual signatures,
group signatures, and dialects. However, in some taxa, data at multiple
levels indicate that vocal learning occurs. In these cases, it may be difficult
to determine where the primary function for vocal learning resides. For
instance, dialects may be a by-product of vocal learning that has evolved to
serve a function on a smaller scale.We include studies at the level we believe
is most important or where the system has been best studied, and we also
describe evidence at other levels.
2. Individual Signatures
Individually distinctive vocalizations have been described in many birds
and mammals. In fact, individual distinctiveness is ubiquitous and has been
found in almost every taxon in which it has been explored. Individual distinctiveness is enhanced when differences between individuals are large and
variation within individuals is minimal, so that each individual produces
a consistently unique vocalization. Distinctly different call types are the
most noticeable individually distinctive vocalizations (e.g., whistle signatures in bottlenose dolphins, Tursiops truncatus; Caldwell and Caldwell
1965). However, large differences are not essential; even subtle acoustic differences can generate sufficiently distinctive voices for individual identification in some species (e.g., Weary and Krebs 1992). This is especially true
when individuals differ on several acoustic dimensions, but subtle differences on a single acoustic dimension can be sufficient for individual
identification. Nor does variation within individuals necessarily obliterate
distinctive voice characteristics. Differences in several acoustic dimensions
are more likely in complex, frequency-modulated, or noisy calls, whereas
tonal calls may differ in a single acoustic dimension.
Individual distinctiveness is necessary but insufficient for vocalizations to
function as individual signatures. Conspecifics must be able to use vocalizations to discriminate between individuals. This is a more demanding criterion. Although many taxa have individually distinctive calls, far fewer
cases of individual signatures have been demonstrated, in part because
recognition is difficult to test. Yet, individual signatures are defined both by
their structure and their function. Individual signatures need not be learned
to serve their function. Individual distinctiveness can arise in two ways: by
designed individual signatures or by-product distinctiveness. Only designed
individual signatures are likely to rely on learning.
4. Comparative Vocal Learning
145
mammals. These in-depth case studies are mostly in the bird groups oscines,
psittacines, and trochilids and in the mammalian groups chiropterans,
cetaceans, and primates because that is where research has focused or
evidence suggests vocal learning may occur. We recognize that different
sources of selection may favor vocal learning at each level and so divide
the following sections into three functional levels: individual signatures,
group signatures, and dialects. However, in some taxa, data at multiple
levels indicate that vocal learning occurs. In these cases, it may be difficult
to determine where the primary function for vocal learning resides. For
instance, dialects may be a by-product of vocal learning that has evolved to
serve a function on a smaller scale.We include studies at the level we believe
is most important or where the system has been best studied, and we also
describe evidence at other levels.
2. Individual Signatures
Individually distinctive vocalizations have been described in many birds
and mammals. In fact, individual distinctiveness is ubiquitous and has been
found in almost every taxon in which it has been explored. Individual distinctiveness is enhanced when differences between individuals are large and
variation within individuals is minimal, so that each individual produces
a consistently unique vocalization. Distinctly different call types are the
most noticeable individually distinctive vocalizations (e.g., whistle signatures in bottlenose dolphins, Tursiops truncatus; Caldwell and Caldwell
1965). However, large differences are not essential; even subtle acoustic differences can generate sufficiently distinctive voices for individual identification in some species (e.g., Weary and Krebs 1992). This is especially true
when individuals differ on several acoustic dimensions, but subtle differences on a single acoustic dimension can be sufficient for individual
identification. Nor does variation within individuals necessarily obliterate
distinctive voice characteristics. Differences in several acoustic dimensions
are more likely in complex, frequency-modulated, or noisy calls, whereas
tonal calls may differ in a single acoustic dimension.
Individual distinctiveness is necessary but insufficient for vocalizations to
function as individual signatures. Conspecifics must be able to use vocalizations to discriminate between individuals. This is a more demanding criterion. Although many taxa have individually distinctive calls, far fewer
cases of individual signatures have been demonstrated, in part because
recognition is difficult to test. Yet, individual signatures are defined both by
their structure and their function. Individual signatures need not be learned
to serve their function. Individual distinctiveness can arise in two ways: by
designed individual signatures or by-product distinctiveness. Only designed
individual signatures are likely to rely on learning.
4. Comparative Vocal Learning
145