cult to interpret. Boughman and Moss (Chapter 4) examine the role of vocal
learning (both learned acquisition and social modification) in the communication systems of birds and mammals, citing examples from a wide variety
of species and a wide variety of different types of acoustic signals. They
provide experimental guidelines to help distinguish between the roles of
maturation and learning in the development of communication behavior.
2.3. Sensory Mechanisms
Sensory mechanisms constrain both the structure and the function of vocalizations. An influential model of a signal-processing strategy for effective
communication in a noisy environment was proposed by Capranica and
Moffat (1983). According to this formulation, an ideal auditory system
should be selectively biased, in terms of either peripheral sensitivity or of
central recognition, to emphasize distinguishing relevant characteristics of
communication sounds. Such a receiver would appear to be specialized for
reception of those sounds to the exclusion of others. Because many examples of animal acoustic communication involve transmission of specific frequencies of sound, it is important to consider whether selective reception
of just these frequencies is an adequate strategy for the receiver to interpret the message. Using the relationship between the frequency content of
advertisement calls and the tuning of eighth-nerve fibers in several different species of anuran amphibians, Capranica and Moffat argued in favor of
a “matched-filter” model of receiver processing in the frequency domain.
In this formulation of an ideal receiver, the frequency response of the auditory system exactly mirrors the energy spectrum of the sender’s vocalizations. Such matching would maximize the signal-to-noise ratio for reception
by excluding masking due to frequencies not actually present in the calls.
This would make the communication channel more private in the sense that
sounds at other frequencies would be selectively rejected at the moment of
reception. At more central levels, matched filtering would be manifested by
the presence of “mating call detectors” or auditory “grandmother cells”
capable of strongly responding only to sounds containing the correct combination of frequencies. Gentner and Margoliash (Chapter 7) discuss recent
advances in the search for matched filters and call detectors. As they point
out, combination-sensitive cells having properties consistent with these
have been found in central auditory nuclei of different species of frogs,
birds, and mammals. In a different approach, Ryan and Kime (Chapter 5)
present a neural network that allows sensory biasing to a conspecific signal
without the explicit need for dedicated mating-call detectors.
2.4. Production Mechanisms
Because perception and production of sounds have coevolved (Ryan and
Kime, Chapter 5), selective biasing in sensory perception should be
8
A.M. Simmons
learning (both learned acquisition and social modification) in the communication systems of birds and mammals, citing examples from a wide variety
of species and a wide variety of different types of acoustic signals. They
provide experimental guidelines to help distinguish between the roles of
maturation and learning in the development of communication behavior.
2.3. Sensory Mechanisms
Sensory mechanisms constrain both the structure and the function of vocalizations. An influential model of a signal-processing strategy for effective
communication in a noisy environment was proposed by Capranica and
Moffat (1983). According to this formulation, an ideal auditory system
should be selectively biased, in terms of either peripheral sensitivity or of
central recognition, to emphasize distinguishing relevant characteristics of
communication sounds. Such a receiver would appear to be specialized for
reception of those sounds to the exclusion of others. Because many examples of animal acoustic communication involve transmission of specific frequencies of sound, it is important to consider whether selective reception
of just these frequencies is an adequate strategy for the receiver to interpret the message. Using the relationship between the frequency content of
advertisement calls and the tuning of eighth-nerve fibers in several different species of anuran amphibians, Capranica and Moffat argued in favor of
a “matched-filter” model of receiver processing in the frequency domain.
In this formulation of an ideal receiver, the frequency response of the auditory system exactly mirrors the energy spectrum of the sender’s vocalizations. Such matching would maximize the signal-to-noise ratio for reception
by excluding masking due to frequencies not actually present in the calls.
This would make the communication channel more private in the sense that
sounds at other frequencies would be selectively rejected at the moment of
reception. At more central levels, matched filtering would be manifested by
the presence of “mating call detectors” or auditory “grandmother cells”
capable of strongly responding only to sounds containing the correct combination of frequencies. Gentner and Margoliash (Chapter 7) discuss recent
advances in the search for matched filters and call detectors. As they point
out, combination-sensitive cells having properties consistent with these
have been found in central auditory nuclei of different species of frogs,
birds, and mammals. In a different approach, Ryan and Kime (Chapter 5)
present a neural network that allows sensory biasing to a conspecific signal
without the explicit need for dedicated mating-call detectors.
2.4. Production Mechanisms
Because perception and production of sounds have coevolved (Ryan and
Kime, Chapter 5), selective biasing in sensory perception should be
8
A.M. Simmons