associated with different patterns of activity (Yu and Margoliash 1996). This
results even if the syllable types share some or many notes. RA neurons,
which receive from HVc, exhibit patterns of physiological activity that contrast with those of HVc neurons. Premotor activity of RA neurons is organized by note type, independent of the syllable in which the note is embedded
(Yu and Margoliash 1996).
The syllable is probably the largest unit of vocalization that represents a
motor program. When singing birds are startled, they tend to complete the
current syllable before ceasing to sing (Cynx 1990). Upon electrical stimulation of HVc during singing, birds tend to restart the song at the beginning
of motifs (i.e., at large segmental boundaries representing repeated
sequences of syllables). In contrast, stimulation of RA during singing
disrupts the syllable morphology but not the sequence of syllables (Vu et
al. 1994). Thus, information about the large-scale organization of song
(sequences of motor programs) could be available to the VMP and the AFP
through the two distinct classes of HVc projection neurons; HVc projection
neurons target exclusively either RA or area X (Fig. 7.3).
The different classes of HVc projection neurons, and HVc interneurons,
exhibit differences in their intrinsic properties and subthreshold auditoryresponse properties (Mooney 2000). Furthermore, recent data from electrical-stimulation studies designed to establish identities of extracellularly
recorded classes of HVc neurons suggest that most of the chronic recordings were of HVc interneurons (Shea et al. 2001). Thus, it remains to be
seen whether the mapping from syllable-level to note-level representations
occurs within HVc (at the level of the RA-projecting HVc neurons) or in
the projection of HVc to RA. The HVc to RA mapping potentially represents a transition from categories of behavior to motor output. It is likely,
in any case, that the VMP and the AFP receive different mixes of premotor activity and sensory feedback during singing.
The physiological properties of other VMP nuclei contrast with those
of HVc and extend the concept of a hierarchical arrangement of sensorimotor control. Neurons in NIf and Uva (Fig. 7.3), which project to HVc,
are active during singing (McCasland 1987; Williams and Vicario 1993).
Multiunit recordings from Uva suggest that some activity is more closely
associated with the timing of entire motifs, and bilateral lesions of Uva
disrupt the suprasyllabic organization of song but do not abolish singing
(Williams and Vicario 1993). Recent data from zebra finches suggest that
NIf is a major source of auditory input to HVc (Janata and Margoliash 1999;
Boco and Margoliash 2001). Bilateral lesions of NIf have phrase-level
effects on singing in Bengalese finches (Hosino and Okanoya 2000) and
apparently only transient effects on singing, with full recovery, in zebra
finches (Vu et al. 1995). Thus, lesions of VMP nuclei afferent to HVc affect
but do not abolish singing whereas lesions of HVc or RA abolish all singing
behavior.
7. Neuroethology of Vocal Communication
347
results even if the syllable types share some or many notes. RA neurons,
which receive from HVc, exhibit patterns of physiological activity that contrast with those of HVc neurons. Premotor activity of RA neurons is organized by note type, independent of the syllable in which the note is embedded
(Yu and Margoliash 1996).
The syllable is probably the largest unit of vocalization that represents a
motor program. When singing birds are startled, they tend to complete the
current syllable before ceasing to sing (Cynx 1990). Upon electrical stimulation of HVc during singing, birds tend to restart the song at the beginning
of motifs (i.e., at large segmental boundaries representing repeated
sequences of syllables). In contrast, stimulation of RA during singing
disrupts the syllable morphology but not the sequence of syllables (Vu et
al. 1994). Thus, information about the large-scale organization of song
(sequences of motor programs) could be available to the VMP and the AFP
through the two distinct classes of HVc projection neurons; HVc projection
neurons target exclusively either RA or area X (Fig. 7.3).
The different classes of HVc projection neurons, and HVc interneurons,
exhibit differences in their intrinsic properties and subthreshold auditoryresponse properties (Mooney 2000). Furthermore, recent data from electrical-stimulation studies designed to establish identities of extracellularly
recorded classes of HVc neurons suggest that most of the chronic recordings were of HVc interneurons (Shea et al. 2001). Thus, it remains to be
seen whether the mapping from syllable-level to note-level representations
occurs within HVc (at the level of the RA-projecting HVc neurons) or in
the projection of HVc to RA. The HVc to RA mapping potentially represents a transition from categories of behavior to motor output. It is likely,
in any case, that the VMP and the AFP receive different mixes of premotor activity and sensory feedback during singing.
The physiological properties of other VMP nuclei contrast with those
of HVc and extend the concept of a hierarchical arrangement of sensorimotor control. Neurons in NIf and Uva (Fig. 7.3), which project to HVc,
are active during singing (McCasland 1987; Williams and Vicario 1993).
Multiunit recordings from Uva suggest that some activity is more closely
associated with the timing of entire motifs, and bilateral lesions of Uva
disrupt the suprasyllabic organization of song but do not abolish singing
(Williams and Vicario 1993). Recent data from zebra finches suggest that
NIf is a major source of auditory input to HVc (Janata and Margoliash 1999;
Boco and Margoliash 2001). Bilateral lesions of NIf have phrase-level
effects on singing in Bengalese finches (Hosino and Okanoya 2000) and
apparently only transient effects on singing, with full recovery, in zebra
finches (Vu et al. 1995). Thus, lesions of VMP nuclei afferent to HVc affect
but do not abolish singing whereas lesions of HVc or RA abolish all singing
behavior.
7. Neuroethology of Vocal Communication
347