Perception of Sequence in Auditory Events

Abstract

A series of tape-recorded sentences were presented to various groups of listeners, totalling 164. During each sentence an extraneous sound was present on the recording, and the listener had to indicate the exact point in the sentence at which this sound occurred. It was found that errors were made which were large compared with the duration of a single speech sound; which suggests that the listener does not deal with each sound separately but rather with a group of sounds. Errors were reduced if the sentence consisted of a series of digits rather than an ordinary text, or if the listeners were trained in phonetics. Prior knowledge of the content of the sentence did not affect accuracy. The direction of error was usually to refer the extra sound to an early point, but it is affected by the relative position of the extra sound in the sentence. These results can be regarded as an extension, to the case where all stimuli are presented to the same sense, of classic results on prior entry.