Conducting research on the effects of intoxication on speech

Abstract

Most types of behavioural research are difficult to carry out, at least with acceptable rigour; investigation of the effects of ethanol upon speech is a good example of such a challenge. To date bur little research on the issue has been reported; the problem here appears to be the difficulty in developing and carrying our appropriate experiments. In response, the thrust of the present project was two-fold: ( 1) to address the robustness/accuracy of traditional investigational procedures and (2) to generate controlled pilot data on speech-intoxication relationships. First, both established and new methodologies were assessed and attempts were made to correlate the physiological approaches with the behavioural. Second, initial data was gathered on the suprasegmentals of speech. To accomplish these goals, carefully selected subjects were studied at a number of highly controlled levels of intoxication. It was found that traditional approaches resulted in difficulties with experimental precision whereas modified procedures could be used to overcome these problems. Second, while the aural-perceptual data did not correlate very well with the physiological, some information was gained as to how to assess motor speech deficits under controlled procedures. Finally, measures of FO shift exhibited an upward trend for most speakers, and duration was increased, as a function of increments in intoxication.