A number of recent experiments have shed light on the neural basis of the upper limit of rate discrimination. One pertinent finding is that the limitation also applies to tasks that, in bilaterally implanted listeners, involve inter-aural timing judgements, rather than estimates of the pitch of sounds presented to a single ear. Evidence for this comes from two paradigms, one of which investigated whether discrimination between a lower- and a higher-pulse rate, presented to one ear, could be improved by presenting a copy of the lower-rate pulse train to an electrode in the opposite ear, in all intervals of each trial (van Hoesel and Clark
1997; van Hoesel
2007; Carlyon et al.
2008). A benefit occurred when low-rates stimuli were presented contralaterally, thus providing the listener with a binaural cue; the percept of the lower-rate stimulus was reported as fused and was heard in the centre of the head, whereas the higher-rate stimulus was reported to sound diffuse. Conversely, no benefit was observed when a 300-pps stimulus was presented contralaterally to create a binaural cue, showing that limitation in temporal processing, as possibly due to the existence of an “upper limit”, are not restricted to pitch-based tasks. In a second approach, Ihlefeld et al. (
2015) measured detection of rate differences as a function of the standard rate for three electrodes in each ear of eight bilaterally implanted listeners. In each ear, the three electrodes were in the base, middle and apex of the array and had a place pitch that was matched to the corresponding electrode in the opposite ear. Ihlefeld et al. (
2015) also measured sensitivity to an ITD difference between each matched pair of electrodes as a function of baseline rate. Over the 100–500 pps range studied, both the monaural rate discrimination and the ITD detection became worse at higher rates, also in agreement with previous research (Majdak et al.
2006; Laback et al.
2007; van Hoesel
2007). Importantly, once these general trends were removed, it was possible, to some extent, to predict ITD sensitivity from the worse of the corresponding rate discrimination scores in the two ears. Ihlefeld et al. (
2015) concluded that the processing of fine timing differences at high repetition rates is limited by a factor that is not restricted to tasks requiring binaural processing. Interestingly, a study that combined measurements of the electrically evoked compound action potential with rate discrimination tasks, using the same subjects and stimuli, concluded that this limitation lies central to the auditory nerve (Carlyon and Deeks
2015).
The present study adds to this body of knowledge by showing that rate discrimination at high rates is marginally correlated across electrodes, and highly correlated across subjects with another task, gap detection, which involves quite different stimuli. Whereas the rate discrimination and binaural tasks described above used pulse trains having nearly identical rates and levels, the pulse rates in the gap detection stimuli were more than 2.5 times greater than the stimuli in the rate discrimination experiment. Taken together, the results of these studies are consistent with a limitation central to the auditory nerve that is common to tasks that require accurate encoding of short temporal intervals in high-rate stimuli.
An interesting question is why weaker or absent correlations were found between RDR
100 and either GDTs or RDR
400. The degradation in temporal processing beyond a certain rate could, in principle, be linked to the existence of a rate-independent temporal jitter in the neural response to each pulse; at high rates, when the jitter period is comparable to the inter-pulse interval, the performance in tasks requiring fine temporal discrimination may deteriorate. A simple model of this type would predict a strong correlation between RDR
100 and RDR
400, which in our study was not observed across electrodes or across subjects. A trivial explanation could be that measurements of RDR
100 were not as reliable as the other measures in this study. However, the test-retest correlation was highly significant, and the RDR measure was sensitive enough to correlate strongly (
r = 0.89), across subjects, with the low-rate sensitivity obtained from the pitch ranking study. Hence, the absence of a correlation between RDR
100 and either RDR
400 or GDT does not seem to be due to inaccuracy in our measurements of low-rate discrimination. Rat her, from the data in this study, it seems low and high rate processing are subject to different sources of limitation, and that the upper limit is due to something specific to the processing of high-rate stimuli rather than to rate-independent jitter. A physiological basis for this limitation is suggested by the finding that neurons in the cat inferior colliculus (IC) produce sustained time-locked responses only at low pulse rates (e.g. Hancock et al. (
2012)), with pulse trains having a rate exceeding some limit resulting in only an onset response. Furthermore, the finding that this “upper limit”, as measured in the IC, is influenced by auditory deprivation (Vollmer et al.
2007) is consistent with the correlation between duration of deafness and both RDR
400 and GDT found in the present study. However, there are at least two reasons for caution when speculating further as to the precise physiological basis for the upper limit. First, although the correlation between deafness duration and RDR
100 was not significant, it was not significantly smaller than those between deafness duration and either RDR
400 or GDT. Hence, we do not have strong evidence for a correlation with DoD that is
specific to high-rate stimuli, and so cannot rule out the possibility of a non-sensory basis for the correlations observed. Second, there is now evidence that the upper limit to which IC responses phase lock is influenced, in animal experiments, by the anaesthesia used to obtain those recordings (Chung et al.
2014). Such limitations will not, of course, apply to the human subjects performing psychophysical tasks.