Retour vers liste

Détail de la contribution

Auteur: Dorottya GYARMATHY

Co-Auteur(s): Tilda NEUBERGER, Research Institute for Linguistics, Hungarian Academy of Sciences, Hungary

Self-monitoring strategies: the factor of age

Abstract/Résumé: The mental lexicon is the central mechanism of the speech production process as it contains indispensable information for speaking. Its structure, the number of stored items and the activation strategies change with age. During the activation of the mental lexicon, difficulties of lexical access may encounter possibly resulting disfluencies, e.g. false starts and false words, which can be corrected by the self-monitoring system. False starts occur when the speaker realizes before articulating the complete word that it does not correspond with the target word, thus the utterance is stopped before the end of the complete production of the given word. On the contrary, in the case of false words the correction takes place when the articulation of the word is already complete. The aim of the present research is to examine the self-monitoring mechanism by analyzing two types of disfluency phenomena: false starts and false words. Our corpus (22 hours) consists of narratives of 70 children of three age groups (6-year-olds, lower and upper primary school pupils) and 70 adults (aged between 20 and 76 years) from the BEA Hungarian Spontaneous Speech Database. The relation between the reparandum and the repair, the type of the erroneous word, the duration and structure of the editing phases (interval between the interruption point and the onset of the repair that can be filled with pause, hesitation and/or speech) were analyzed by means of Praat. The research question was addressed: what kind of age-related differences can be detected in the course of lexical access and self-monitoring process. According to our hypotheses a) age has a great effect both on the frequency of occurrence of the phenomena and on the strategy of correction and b) the duration of the editing phase depends on its structure, the relation between the reparandum and the repair, and the type of the word (content/function word). Statistical analysis was carried out using SPSS software. The main results show that a) false starts occur more frequently in the speech samples of all age groups than false words do (in total 366 false starts and 134 false words), however, the correction strategies showed differences depending on the age groups (e.g. adults can correct their false starts significantly faster than children). b) The duration of the editing phase is affected by its structure the same way in the case of both types of disfluencies. This research contributes to the closer exploration of self-monitoring by age and highlights the factor of language development in this respect.