highly productive and, if anything, reflect an expanding scope of issues successfully being addressed.
Finally, an additional question concerns the extent to which the neuroethological approach to acoustic communication can give insight into
human acoustic perception. These concerns rest largely on the assumption
that human perceptual processes, particularly those involved in speech
and language, are unique. Nevertheless, determining which phenotypes of
speech are unique and which are conserved requires comparative analysis.
For example, although early work on speech processing emphasized what
appeared to be unique properties of the human system, many of these initial
claims evaporated in the face of later animal research (Kuhl and Miller
1975; Kluender et al. 1987). More recent theoretical perspectives emphasize that features of the speech signal, and associated neural processing
mechanisms, are elaborations of general features of the mammalian and
even vertebrate auditory system (see Fowler 1996; Lotto et al. 1997). Comparative studies have also demonstrated specialized sensory representations for autocommunication signals (see Popper and Fay 1995; Brenowitz
et al. 1997). Such representations are likely to exist in humans because
humans are well-known to be sensitive to the acoustic structure of their
own vocalizations, but such representations have hardly been studied at
all in humans. Furthermore, where there are specific theories of vocal learning in animals, they can lead to specific predictions regarding speech learning in humans (Margoliash 2001). These examples give confidence that
insight into human speech perception can be gained by studying the perception of acoustically complex signals from a comparative perspective
(Doupe and Kuhl 1999; Fitch 2000).
1.2. Structure of Animal Communication Systems
Arriving at explicit definitions for animal communication is a notoriously
difficult problem. However, most will agree on a minimum description of
communication as a process involving the transmission of information, via
a signal, from a sender to a receiver. Where debate arises is in the extent
to which various researchers attribute intent to the sender and in the degree
to which the various fitness benefits for either the sender or the receiver
are emphasized. Although such considerations remain a topic of continued
debate among theoreticians (Dawkins and Krebs 1978; Beer 1982; Smith
1997; Bradbury and Vehrencamp 1998, 2000), from a proximate standpoint,
it is the transmission of information that is of interest.
For information to be transmitted via some signal, there must be parity
between the sender of that signal and the receiver. That is, the receiver must
interpret at least some of the variability (i.e., information) in the signal in
a predictable manner. Thus, the structure of a communication system can
be considered as a behavioral feedback loop in which information flows
7. Neuroethology of Vocal Communication
327
Finally, an additional question concerns the extent to which the neuroethological approach to acoustic communication can give insight into
human acoustic perception. These concerns rest largely on the assumption
that human perceptual processes, particularly those involved in speech
and language, are unique. Nevertheless, determining which phenotypes of
speech are unique and which are conserved requires comparative analysis.
For example, although early work on speech processing emphasized what
appeared to be unique properties of the human system, many of these initial
claims evaporated in the face of later animal research (Kuhl and Miller
1975; Kluender et al. 1987). More recent theoretical perspectives emphasize that features of the speech signal, and associated neural processing
mechanisms, are elaborations of general features of the mammalian and
even vertebrate auditory system (see Fowler 1996; Lotto et al. 1997). Comparative studies have also demonstrated specialized sensory representations for autocommunication signals (see Popper and Fay 1995; Brenowitz
et al. 1997). Such representations are likely to exist in humans because
humans are well-known to be sensitive to the acoustic structure of their
own vocalizations, but such representations have hardly been studied at
all in humans. Furthermore, where there are specific theories of vocal learning in animals, they can lead to specific predictions regarding speech learning in humans (Margoliash 2001). These examples give confidence that
insight into human speech perception can be gained by studying the perception of acoustically complex signals from a comparative perspective
(Doupe and Kuhl 1999; Fitch 2000).
1.2. Structure of Animal Communication Systems
Arriving at explicit definitions for animal communication is a notoriously
difficult problem. However, most will agree on a minimum description of
communication as a process involving the transmission of information, via
a signal, from a sender to a receiver. Where debate arises is in the extent
to which various researchers attribute intent to the sender and in the degree
to which the various fitness benefits for either the sender or the receiver
are emphasized. Although such considerations remain a topic of continued
debate among theoreticians (Dawkins and Krebs 1978; Beer 1982; Smith
1997; Bradbury and Vehrencamp 1998, 2000), from a proximate standpoint,
it is the transmission of information that is of interest.
For information to be transmitted via some signal, there must be parity
between the sender of that signal and the receiver. That is, the receiver must
interpret at least some of the variability (i.e., information) in the signal in
a predictable manner. Thus, the structure of a communication system can
be considered as a behavioral feedback loop in which information flows
7. Neuroethology of Vocal Communication
327