The time course of listening bands

It has been known for decades that frequency analysis in the auditory system reveals the existence of bandpass filter-like channels—the Critical Bands—a finite number of which cover the whole range of audible frequencies, with the consequence that nearby frequencies are not resolved individually. Because critical bands are formed already in the cochlea, they appear with no delay: their contours are formed the moment an acoustic wave reaches the inner ear and disappear when the wave goes silent. Psychophysical measurement of the width of critical bands has been pursued by a many investigators (for a summary, see Chapter 3 in Moore 2003) and this behavioral indicator of frequency selectivity has been often compared to biophysical and physiological measures of frequency selectivity, both excitatory and inhibitory, at various stages along the auditory pathway (see e.g., Evans 2001). Yet, when moving away from the psychophysics of what the listener is capable of doing with the aim of investigating what he/she actually does, one comes upon phenomena that critical band analysis alone cannot explain. One such phenomenon was studied by the early proponents of signal detection theory in audition (Swets 1963), long before the concept of attention surreptitiously escaped the watchful eyes of orthodox behaviorism and gradually settled in experimental psychology. These investigators were wondering whether detection of signals of uncertain frequency occurs by the system switching between different bands or shifting a unique band from one frequency to another. They were careful not to refer to these bands as critical bands—to avoid confusion they named them “listening bands.” Among the many properties of listening bands reported, an important one was their capability to keep the listener’s attention tuned to a particular frequency even after a pure-tone signal was turned off, and to maintain it for a rather long duration in the absence of any signal or in the presence of broad-band random noise (Greenberg and Larkin 1968). Other investigators demonstrated that the listening band sluggishly remained centered at the frequency of the last signal heard until an audible tone of a different frequency was presented (Pastore and Sorkin 1971). Still others showed that it is possible to simultaneously tune several listening bands