Table 1 (a) The C5I (in seconds per syllable) for the Three Experimental Conditions: Nasal Voice, Normal Voice (reduced and maximal rate), where N Represents the Number of Ss, (b) The Probability of Stuttering (per syllable) for the Nasal and Nonnal Voice Conditions.EXPERIMENT I : THE EFFECT OF ASSUMING AN UNUSUAL ACCENT The first experiment was an attempt to demonstrate a parallel between normal and pathological stuttering. Bryngelson (1958) reported that when pathological stutterers put on or assume an unusual accent such as a Louisiana drawl, their stuttering is reduced. We were interested in whether normal Ss also stutter less under DAF when they speak with an accent.Experimental delays in auditory feedback are known to cause stuttering in normal individuals (Lee, 1951). The present study of DAF investigated the following questions about stuttering under delayed auditory feedback (DAF): Do we stutter less under DAF when we speak with an unusual accent? Is the effect of speaking with an accent due simply to the change in the acoustic characteristics of the returning feedback? What is the relation between the intelligibility of the returning feedback and the extent of DAF interference? Can the effect of DAF on foreign speakers be attributed to their accents? Does DAF stuttering reflect an attempt to correct a distortion in feedback resulting from delaying the acoustic signal as postulated in servosystem theories? What is the relationship between DAF and pathological stuttering?
The main question in this study was whether
ProcedureFor one condition the procedure was as follows: S was presented with IO IS-syllable sentences on index cards and instructed to read each sentence as fast as he could in his normal voice regardless of what happened to his speech under DAF.The other condition was identical except that the Ss were instructed to read 10 sentences with a nasal accent. The E showed the S how to nasalize speech (the result of contracting the velar muscles, Heffner, 1964). Each S was given several sentences without DAF so that he could practice his nasal accent.
ResultsRate of speech. The time to produce each sentence was determined with a standard-stop watch. The speech rate was estimated from a measure of syllable duration known as the correct syllable interval (CSI). The CSI is defined as Tn/S, where Tn is the time to correctly produce some proportion n of the S syllables in the sentence (after Fairbanks, 1955).The rates of speech for the three experimental conditions are compared in Table Ia. As might be expected, the mean CSI was smaller when the Ss were speaking at their maximum as compared to their normal rates. When speaking with a nasal accent the Ss spoke at about their normal rates (see Table I), allowing valid comparison of DAF stuttering in the nasal and normal rate conditions.
ControlsFeedback intensity. Since Ss in a pilot study tended to speak more softly than normally through their noses, the level of the nasal feedback was amplified to match that of the non-nasal feedback. Subsequent meas...