Differences of CT (2f1 - f2) phase in psychophysical and physiological experiments

Neural and psychophysical studies of combination tones (CT) give highly correlated evidence for the existence of stimulus-like intracochlear distortion products. However, a large systematic difference was found between psychophysical and neural measurements of the phase of the cubic CT, 2f1 - f2. The psychophysical phase, as measured by monaural concellation, decreases typically at 6-12 degrees per decibel increase in stimulus amplitudes, while the physiological phase, measured both by neural phase locking or by cancellation of the locking, is nearly independent of stimulus amplitudes. Through new psychophysical studies of combination tone interactions monaurally as well as their lateralization binaurally, we examined whether the phase of the psychophysical cancellation tone directly measures the phase of the intracochlear CT. We found evidence for amplitude dependent biases in the cancellation measurement, but the biases were generally far too small to account for the amplitude dependence of the cancellation phase. On the other hand, binaural lateralization of the CT showed a similar amplitude dependence of CT phase as found in monaural cancellation. No evidence exists in neural data either for an amplitude dependent bias of the cancellation measurement or for systematic amplitude dependence of the neural phase. Therefore, we conclude that a real difference exists in the intracochlear nonlinearity for alert human subjects and anaesthetized laboratory animals. We model the human cochlea nonlinearities by modifying Hall's nonlinear transmission line model through the addition of nonlinear stiffness in the nonlinear mechanical loading of the basilar membrane.