Frequency discrimination in the auditory system: Place or periodicity mechanisms?

TLDR: A modification of the structure of an optimum sinusoidal frequency discrimination system is conjectured that seems to be, at least qualitatively, consistent with the available evidence.

Abstract: The pattern of all-or-none firings in many sensory nerves can usefully be modeled as a set of sample functions from non-stationary random processes whose parameters are determined by the sensory stimulus. Given such a description, one can compare the parameter estimation, discrimination, or detection performance of a human observer (as measured behaviorally) with the computed limiting performance of an ideal observer operating on the same neural input. Such a comparison is carried out in detail for the discrimination in frequency of an auditory sinusoidal tone burst. The results suggest that the human listener behaves as if he ignored (at least for this class of sounds) the periodicity information in the auditory nerve pattern (resulting from the phase-locking of the firings to the stimulus), but made full effective use of the place information (resulting primarily, although not entirely, from the mechanical tuning of the inner ear). On the other hand, it is known that the auditory system can discriminate certain more complicated stimuli on the basis of periodicity information if that is the only cue available. A modification of the structure of an optimum sinusoidal frequency discrimination system is conjectured that seems to be, at least qualitatively, consistent with the available evidence.