The “Tongue Twister” Paradigm as a Technique for Studying Phonological Encoding

Publisher Website
Google Scholar
Abstract
The “tongue twister” paradigm is underutilizedas a research tool because so little is known abouthow it induces errors. The two experiments reported here explore this paradigm in detail using a task variationwhich minimizes articulatory and mnemonic load. This task was found to elicit good rates of apparently “pure” prearticulatory errors. Two of its features had a significant error-inducing effect: a) repeated reiteration; and b) the use of similar phonemes in targets (e.g., moss knife noose muff). The presenceof phoneme repeats (e.g., palm neck name pack) had no reliable overall effect, but did influence error distribution. Performance on the task differed in several ways from that observed on a control task with similar output demands, but no reiterative component. A model of the task is proposed, in which phoneme similarity and reiteration are seen as independent contributors to the task's error-inducing potential. Wider theoretical implications of certain results are also discussed.