Concurrent Repetition of a Continuous Flow of Words

Abstract

This study was designed to evaluate the effect of rate (68, 127, 180 words/min) and intensity (20, 40, 60 SRT) of stimulus presentation, and of practice, upon concurrent repetition of a continuous flow of heard words. Subjects were 33 enlisted men. A three-part tape was made to obtain an SRT, to present a pretest trial, and to present three PB-50 lists read as continuous speech. Intelligibility scores were assigned to each subject's taped repetitions. Twelve subjects had an 8-day practice with other PB lists. Analysis of variance yielded an F associated with practice significant beyond the 0.5% level. The F's for rate and intensity were significant beyond the 0.1% level. Pearson r's were calculated between performance level and six psychological and auditory tests. Only the r of +0.54 with the Seashore Rhythm test was significant. It was concluded that: (a) Performance improves with increased intensity of heard speech, and with practice; (b) Performance increases when the rate of stimulus presentation is decreased; (c) A positive relationship exists between performance and scores on the Seashore Rhythm test; (d) Under proper conditions of rate and intensity, some persons can be trained to perform with more than 90% intelligibility.

A New Approach to Evaluating the Intelligibility and Naturalness of Shadowed Speech

“…For decades there has been discrepancies regarding the naturalness and intelligibility of shadowed speech. Some research has suggested that shadowed speech is distorted and unnatural (Cherry, 1953;Wingate, 1976;Sergeant, 1961), while others have reported shadowed speech as clearly intelligible and containing prosodic contour (Chistovich, Aliakrinskii & Abulian, 1960;Marslen-Wilson, 1985;Schwitzgebel & Taylor, 1980). Cherry (1953) reported that participants found shadowing surprisingly easy; however, there were apparent effects of the task on the naturalness of the subsequent speech output.…”

“…Cherry (1953, p. 978) asserted that one noticeable characteristic of shadowed speech was its monotony and described shadowed speech as having "…very little emotional content or stressing of the words…" Wingate (1976) reported that shadowed speech is often described as lacking normal prosody and is poorly articulated. Sergeant (1961) investigated the intelligibility of shadowed speech. Twentyfour participants shadowed prose presented at 115 words per minute (WPM) or less.…”

Exploring Verbal Shadowing as an Effective Method of Enrolling in a Speech Recognition System

“…One concern with using verbal shadowing as a method of enrollment is whether or not the speech output during shadowing would be natural and intelligible enough to be effective as an enrollment method. For example, some research has suggested that shadowed speech is distorted and unnatural (Cherry, 1953;Wingate, 1976;Sergeant, 1961), while others report shadowed speech as clearly intelligible and containing prosodic contour (Chistovich, Aliakrinskii & Abulian, 1960;Marslen-Wilson, 1985;Schwitzgebel & Taylor, 1980).…”