This is true even though the initial response modulation can still be observed
for other conspecific songs (Chew et al. 1995, 1996; Stripling et al. 1997).
These stimulus-specific changes in the response properties of NCM neurons
have led to the hypothesis that NCM may play an important role in individual vocal recognition (Chew et al. 1996). Consistent with this idea is the fact
that many neurons in NCM (and cHV) show a rapid up-regulation of the
immediate early gene (IEG) zenk in response to the presentation of conspecific songs (Mello et al. 1992) that is tuned to the acoustics of particular
conspecific song syllables (Ribiero et al. 1998). Interestingly, the genomic
response also habituates to the repeated presentation of the same conspecific song (Mello et al. 1995) and is elevated during specific components of
the vocal-recognition task described above in starlings (Gentner et al. 1999).
The mammalian homolog to zenk is required for expression of late LTP and
long-term memories in mice (Jones et al. 2001). This suggests that zenk
expression in NCM and cHV may be related to learning about conspecific
songs and implicates these structures in concomitant processes.
3.4. The Nonhuman Primate Model
3.4.1. Vocal-Recognition Behavior
Various forms of vocal recognition are also prominent among many species
of primates. In humans, Homo sapiens, the ability to recognize the sex and
the identity of a talker is anecdotally obvious. Acoustic cues to a talker’s
sex are present in both the fundamental frequency (reflecting larynx size)
and vocal tract length (reflecting body size). Recent data on speech perception suggest that acoustic cues to individual recognition within sexes are
due to supralaryngeal vocal tract filtering caused by anatomical variation
between talkers (Bachorowski and Owren 1999). Individual vocal recognition (and kin recognition) has also been demonstrated in rhesus macaques,
Macaca mulatta (Rendall et al. 1996), and acoustic cues related to vocal tract
filtering have been suggested as the basis for this behavior (Owren et al.
1997; Rendall et al. 1998). Vocal recognition is also apparent in the referential call systems of many primates (e.g., Seyfarth et al. 1980a, 1980b;
Hauser 1998; Rendall et al. 1999; Fitch and Hauser, Chapter 3). In these
cases, rather than having signals that are associated with specific individuals or groups of individuals, particular calls are used to refer, for example,
to different classes of predators, food types, or various other behavioral
events. This aspect of primate vocal communication has been reviewed elsewhere (Cheney and Seyfarth 1990; Ghazanfar and Hauser 1999), and we do
not address it further except to point out that such behaviors provide
another set of natural categories for vocalizations that is amenable to future
neuroethological study.
A third class of vocal-recognition behavior has been studied extensively
in Japanese macaques, following the observation that several subtypes of
the “coo” vocalization in this species can be defined on the basis of acoustic
364
T.Q. Gentner and D. Margoliash
for other conspecific songs (Chew et al. 1995, 1996; Stripling et al. 1997).
These stimulus-specific changes in the response properties of NCM neurons
have led to the hypothesis that NCM may play an important role in individual vocal recognition (Chew et al. 1996). Consistent with this idea is the fact
that many neurons in NCM (and cHV) show a rapid up-regulation of the
immediate early gene (IEG) zenk in response to the presentation of conspecific songs (Mello et al. 1992) that is tuned to the acoustics of particular
conspecific song syllables (Ribiero et al. 1998). Interestingly, the genomic
response also habituates to the repeated presentation of the same conspecific song (Mello et al. 1995) and is elevated during specific components of
the vocal-recognition task described above in starlings (Gentner et al. 1999).
The mammalian homolog to zenk is required for expression of late LTP and
long-term memories in mice (Jones et al. 2001). This suggests that zenk
expression in NCM and cHV may be related to learning about conspecific
songs and implicates these structures in concomitant processes.
3.4. The Nonhuman Primate Model
3.4.1. Vocal-Recognition Behavior
Various forms of vocal recognition are also prominent among many species
of primates. In humans, Homo sapiens, the ability to recognize the sex and
the identity of a talker is anecdotally obvious. Acoustic cues to a talker’s
sex are present in both the fundamental frequency (reflecting larynx size)
and vocal tract length (reflecting body size). Recent data on speech perception suggest that acoustic cues to individual recognition within sexes are
due to supralaryngeal vocal tract filtering caused by anatomical variation
between talkers (Bachorowski and Owren 1999). Individual vocal recognition (and kin recognition) has also been demonstrated in rhesus macaques,
Macaca mulatta (Rendall et al. 1996), and acoustic cues related to vocal tract
filtering have been suggested as the basis for this behavior (Owren et al.
1997; Rendall et al. 1998). Vocal recognition is also apparent in the referential call systems of many primates (e.g., Seyfarth et al. 1980a, 1980b;
Hauser 1998; Rendall et al. 1999; Fitch and Hauser, Chapter 3). In these
cases, rather than having signals that are associated with specific individuals or groups of individuals, particular calls are used to refer, for example,
to different classes of predators, food types, or various other behavioral
events. This aspect of primate vocal communication has been reviewed elsewhere (Cheney and Seyfarth 1990; Ghazanfar and Hauser 1999), and we do
not address it further except to point out that such behaviors provide
another set of natural categories for vocalizations that is amenable to future
neuroethological study.
A third class of vocal-recognition behavior has been studied extensively
in Japanese macaques, following the observation that several subtypes of
the “coo” vocalization in this species can be defined on the basis of acoustic
364
T.Q. Gentner and D. Margoliash