of other groups (likely to involve learned acquisition), or can share more
subtle acoustic features of a single call type (likely to involve social modification). As with individual signatures, to demonstrate that calls are group
signatures requires demonstrating convergent call structure and recognition of group mates by call.
Using a repertoire that is partially shared with other groups is an inefficient way to convey group identity. If repertoires are shared, to unequivocally identify the group affiliation of a caller requires hearing much of the
repertoire or just the few calls unique to that group. In this situation, the
function of shared call types appears to be something other than group
identity. That function is currently unknown.
As with individual signatures, potential benefits of indicating group identity need to be considered in conjunction with potential costs and the
possibility of cheating. Imitating a group’s signal can allow an outsider to
gain access to group resources. Therefore, mechanisms for cheater prevention—a means of detecting imitators and preventing their access to
resources—should evolve in concert with group signatures.
Group signatures need not be learned to serve their function. In species
where social groups are formed of relatives, heritable signals can effectively
indicate group affiliation and may also be used to determine the level of
relatedness between individuals. Heritability of group signatures has not
been studied directly, but studies described earlier in several taxa have
shown a strong genetic basis for individual signatures (Jones and Ransome
1993; Scherrer and Wilkinson 1993). When social groups are organized
along kinship lines, acoustic similarity among group mates can result from
shared genetics. Thus, a pattern of convergence within groups or differences
between them is insufficient to demonstrate that vocal learning occurs if
group mates are close relatives.
When call similarity is favored, we expect vocal learning in two cases:
when groups are composed of unrelated individuals who cannot rely on
genetics to produce similarity in calls, or when group composition changes
slowly over time and group members use calls to indicate group affiliation.
In the latter case, the immigrants may change their vocalizations to match
the new group, or the entire group may accommodate the new group composition by changing call characteristics. In both of these instances, the most
relevant type of vocal learning is social modification. Learned acquisition
can occur, but is unnecessary unless new call types are added to an individual’s repertoire. Certainly, vocal learning can occur in groups of relatives
and may increase similarity or increase distinctiveness depending on the
primary function of calls. Separating the effects of heredity from those of
vocal learning in such taxa is likely to be a difficult but fruitful task.
In the examples that follow, we begin by describing the species’ social
biology and the benefits of group living, and consider when learned group
signatures are expected. Then, we describe evidence that calls are groupdistinctive and present data on call function. We conclude each example by
discussing the evidence for vocal learning and discuss alternative hypo168
J.W. Boughman and C.F. Moss
subtle acoustic features of a single call type (likely to involve social modification). As with individual signatures, to demonstrate that calls are group
signatures requires demonstrating convergent call structure and recognition of group mates by call.
Using a repertoire that is partially shared with other groups is an inefficient way to convey group identity. If repertoires are shared, to unequivocally identify the group affiliation of a caller requires hearing much of the
repertoire or just the few calls unique to that group. In this situation, the
function of shared call types appears to be something other than group
identity. That function is currently unknown.
As with individual signatures, potential benefits of indicating group identity need to be considered in conjunction with potential costs and the
possibility of cheating. Imitating a group’s signal can allow an outsider to
gain access to group resources. Therefore, mechanisms for cheater prevention—a means of detecting imitators and preventing their access to
resources—should evolve in concert with group signatures.
Group signatures need not be learned to serve their function. In species
where social groups are formed of relatives, heritable signals can effectively
indicate group affiliation and may also be used to determine the level of
relatedness between individuals. Heritability of group signatures has not
been studied directly, but studies described earlier in several taxa have
shown a strong genetic basis for individual signatures (Jones and Ransome
1993; Scherrer and Wilkinson 1993). When social groups are organized
along kinship lines, acoustic similarity among group mates can result from
shared genetics. Thus, a pattern of convergence within groups or differences
between them is insufficient to demonstrate that vocal learning occurs if
group mates are close relatives.
When call similarity is favored, we expect vocal learning in two cases:
when groups are composed of unrelated individuals who cannot rely on
genetics to produce similarity in calls, or when group composition changes
slowly over time and group members use calls to indicate group affiliation.
In the latter case, the immigrants may change their vocalizations to match
the new group, or the entire group may accommodate the new group composition by changing call characteristics. In both of these instances, the most
relevant type of vocal learning is social modification. Learned acquisition
can occur, but is unnecessary unless new call types are added to an individual’s repertoire. Certainly, vocal learning can occur in groups of relatives
and may increase similarity or increase distinctiveness depending on the
primary function of calls. Separating the effects of heredity from those of
vocal learning in such taxa is likely to be a difficult but fruitful task.
In the examples that follow, we begin by describing the species’ social
biology and the benefits of group living, and consider when learned group
signatures are expected. Then, we describe evidence that calls are groupdistinctive and present data on call function. We conclude each example by
discussing the evidence for vocal learning and discuss alternative hypo168
J.W. Boughman and C.F. Moss