Laboratory Phonology 10 / Phonology and Phonetics Bd.4-4 (PDF)

Prosodic structure and tongue twister errors (p. 433-434)


Karen Croot, Claudia Au and Amy Harper

Two experiments investigated whether segmental speech errors in tongue twister production were associated with utterance-level prominence and prosodic phrase-initial position as described by an autosegmental metrical model of intonational phonology. In Experiment One, 40 undergraduate students produced words with confusable onset consonants (“tongue twister words”) in four-word lists under prominence conditions that varied the number and location of emphatically-produced words in the list. There was a three-way interaction between prominence condition, word position in the list, and format (ABBA versus ABAB ordering of onset consonants). In Experiment Two, 38 undergraduate students produced tongue-twister words with and without narrow informational focus in sentences. Onsets of tongue twister words with narrow focus were produced more accurately than onsets in unfocussed tongue twister words. Neighbourhood density of tongue twister words had no significant effect on error rate. Phrase-initial position yielded more errors than non-initial position, but was confounded with format (all tongue twister words occurred in ABBA format). Results suggest that information about utterance-level prominence and phrasing is available at the time of segment-to-frame association, consistent with the prosody-first account of phonological encoding proposed by Keating and Shattuck-Hufnagel (2002).

1. Introduction


An account of the psycholinguistic processes that occur during speech production must specify how the segmental content and prosodic structure of an utterance are integrated to produce the acoustic-phonetic properties of the speech signal (Shattuck-Hufnagel andTurk 1996). For several decades, researchers have analysed speech errors as a source of information about the subprocesses occurring within word-form

Cécile Fougeron and Barbara Kühnert, Laboratoire de Phonetique et Phonologie, Paris, France; Mariapaola D'Imperio, Laboratoire Parole et Langage, Aix-en-Provence, France; Nathalie Vallée, Gipsa-Lab, Grenoble, France.