of natural stimuli in neurophysiological experiments is very valuable.
Although the theory underlying their use is not yet well-developed in a
formal sense, natural stimuli realistically constrain the search space to a
range of input parameters that, because of their behavioral relevance, are
likely to be represented by central neurons.
The use of natural stimuli to characterize neuronal responses is somewhat analogous to the application of psychophysical paradigms to characterize neuronal responses. Just as a psychophysical illusion may identify new
neuronal response properties (von der Heydt et al. 1984), so may presentation of a new category of natural sounds identify new components of
receptive fields, even in cases where neurons have previously been extensively studied. For example, although bat cortical neurons had been extensively studied in the context of echolocation, entirely new excitatory regions
of their receptive fields that had previously been missed were uncovered
when social communication calls were first presented to those neurons
(Ohlemiller et al. 1996). The receptive field domain suitable for communication calls may represent the plesiomorphic condition. The existence of
two separate receptive field domains highlights the degree of nonlinearity
or high dimensionality (complexity) of the parameter space and emphasizes
that the choice of stimulus repertoire can be a subtle decision that shapes
the limitations of the experimental paradigm. Beyond this obvious caveat,
procedures have been established for using natural stimuli in neurophysiological experiments.
2.2.1. Selectivity and Specificity for Conspecific Vocalizations
Two strategies have been employed to assess the responses of neurons to
species-typical vocalizations. The first is to use natural vocalizations and
derivative sounds to identify the potential behavioral relevance of neural
responses. This is assessed in terms of a neuron’s selective responses to a
subset or category of natural vocalizations (i.e., its selectivity). The second
strategy is to identify the specific acoustic features underlying a neuron’s
selective responses (i.e., its specificity) by using artificial stimuli to characterize its response properties. These two strategies are complementary, and
both may be necessary to adequately describe a neuron’s responses to
natural stimuli. Testing for selectivity rarely unambiguously identifies the
specific acoustic features that result in the selective response. Conversely,
exploration of specific features of a neuron’s response properties may give
only limited biological insight into the significance of those features in the
absence of a well-delineated behavioral context. Absent knowledge of the
acoustic behaviors, neural analyses may ultimately emphasize response
parameters that are not central to the behavioral decisions associated with
real-world tasks such as vocal recognition.
In cases of complex vocal behavior (i.e., where multiple acoustic parameters affect the behaviors under study), an experimental approach is man7. Neuroethology of Vocal Communication
341
Although the theory underlying their use is not yet well-developed in a
formal sense, natural stimuli realistically constrain the search space to a
range of input parameters that, because of their behavioral relevance, are
likely to be represented by central neurons.
The use of natural stimuli to characterize neuronal responses is somewhat analogous to the application of psychophysical paradigms to characterize neuronal responses. Just as a psychophysical illusion may identify new
neuronal response properties (von der Heydt et al. 1984), so may presentation of a new category of natural sounds identify new components of
receptive fields, even in cases where neurons have previously been extensively studied. For example, although bat cortical neurons had been extensively studied in the context of echolocation, entirely new excitatory regions
of their receptive fields that had previously been missed were uncovered
when social communication calls were first presented to those neurons
(Ohlemiller et al. 1996). The receptive field domain suitable for communication calls may represent the plesiomorphic condition. The existence of
two separate receptive field domains highlights the degree of nonlinearity
or high dimensionality (complexity) of the parameter space and emphasizes
that the choice of stimulus repertoire can be a subtle decision that shapes
the limitations of the experimental paradigm. Beyond this obvious caveat,
procedures have been established for using natural stimuli in neurophysiological experiments.
2.2.1. Selectivity and Specificity for Conspecific Vocalizations
Two strategies have been employed to assess the responses of neurons to
species-typical vocalizations. The first is to use natural vocalizations and
derivative sounds to identify the potential behavioral relevance of neural
responses. This is assessed in terms of a neuron’s selective responses to a
subset or category of natural vocalizations (i.e., its selectivity). The second
strategy is to identify the specific acoustic features underlying a neuron’s
selective responses (i.e., its specificity) by using artificial stimuli to characterize its response properties. These two strategies are complementary, and
both may be necessary to adequately describe a neuron’s responses to
natural stimuli. Testing for selectivity rarely unambiguously identifies the
specific acoustic features that result in the selective response. Conversely,
exploration of specific features of a neuron’s response properties may give
only limited biological insight into the significance of those features in the
absence of a well-delineated behavioral context. Absent knowledge of the
acoustic behaviors, neural analyses may ultimately emphasize response
parameters that are not central to the behavioral decisions associated with
real-world tasks such as vocal recognition.
In cases of complex vocal behavior (i.e., where multiple acoustic parameters affect the behaviors under study), an experimental approach is man7. Neuroethology of Vocal Communication
341