No sharpening? a challenge for cochlear mechanics.

Recent data on mechanical movements of the basilar membrane (BM) suggest that the part played in cochlear physiology by a sharpening mechanism is much less important than hitherto has been thought. In an extreme view, one could dispense with a sharpening mechanism completely and assume that (near the threshold) hair-cell excitation is proportional to BM velocity, or a very simple linear transform of it. In the present paper the consequences of this idea are worked out. A theoretical cochlear movement pattern is constructed that shows the same frequency selectivity as an average reverse-correlation function of an auditory nerve fiber. This response is called a revcor-spectrumlike response. Cochlear mechanics is then simplified to a pure shortwave model. It is shown that, if the cochlea model should present a revcor-spectrumlike response, this can only be achieved when the resistance component of the BM impedance is negative over a part of the length of the cochlea. This result is refined in several respects, and it is shown that a model equipped with the right kind of BM impedance function can have a response of the required type. It remains difficult to conceive of a physiological mechanism that would cause the desired effect on the BM impedance.