More Temporally Segmented Speech

The intelligibility of alternated spech passes through a minimum at an alternation rate of about 3 Hz. The critical parameter defining the minimum is not the duration of the bursts of speech reaching the listener, but rather the amount of speech each burst contains [J. Acoust. Soc. Amer. 36, 1055 (1964)]. Recent work with temporally segmented speech [J. Acoust. Soc. Amer. 52, 176(A), (1972)] showed the decline and recovery of intelligibility, as switching rate is increased, to be due to different effects. The decline depends on the decreasing duration of the speech bursts, but the recovery depends on the decreasing duration of the silent intervals. Since the minimum is determined by both these effects, the amount of speech contained in the speech bursts cannot be the only critical parameter, as stated above. To resolve this conflict, an experiment was performed with temporally segmented speech. Intelligibility was measured as a function of silent interval duration, speech interval duration, and speech interval content, the latter two being varied independently by varying the playback speed of the speech. [This research was supported by NIH grant.]