Some comments on the proposed perception of phase and nanosecond time disparities by echolocating bats.

In a series of recent reports, Simmons and his colleagues propose that bats are able to accurately encode the spectral, temporal and phase information of their emitted calls and echoes. The information so encoded is then extracted by the networks of the auditory system with specialized processing. They propose that bats use this information to determine the distance to their target by crosscorrelating the entire structure of the emitted call with the structure of the echo. The idea is that slight deviations in the correlation function can be detected by the bat and the degree of mismatch provides an accurate measure of temporal disparity and hence range. The data in the reports purport to show that bats perceive the phase of ultrasonic signals and that they can resolve temporal disparities of about 10 ns, and thus can distinguish range differences as small as 2 microns. The hypothesis also attempts to explain how a variety of acoustic cues are processed and represented in the auditory system and how they are combined to form a unitary percept of space and fine structure. The theory incorporates some time honored processes of extracting information, such as crosscorrelations. The implications of the hypothesis, however, go far beyond a theory of neural processing and representation of information by ensembles of cells. The hypothesis requires some remarkable abilities, such as the phase coding of ultrasonic signals and a temporal acuity on the order of 10 ns. These features have never been seen in any neurophysiological study of any animal nor has its existence been implied in behavioral studies of other animals.(ABSTRACT TRUNCATED AT 250 WORDS)