Learning phonotactic constraints in the production of non-native consonant clusters, as reflected in speech errors

phonotactic constraints;speech errors

Karlinsky, Nadia ; Cole ; Jennifer S.

关键词: phonotactic constraints;    speech errors;   

Others  :  https://www.ideals.illinois.edu/bitstream/handle/2142/42251/Nadia_Karlinsky.pdf?sequence=1&isAllowed=y

美国|英语

来源: The Illinois Digital Environment for Access to Learning and Scholarship

【 摘 要 】

Experiments in which participants produce sounds under artificial phonotactic constraints show that speakers are able to acquire new speech patterns following minimal exposure, even when these patterns involve sequences unattested in the participants’ native languages (e.g., Dell,Reed, Adams, & Meyer, 2000; Taylor and Houghton, 2005; Whalen & Dell, 2006).However, if the phonotactic learning mechanism is robust to such a degree, why do stable linguistic patterns exist—across languages, across time, even across an individual’s language patterns over his lifetime? The literature suggests that phonotactic learning is moderated by universal markedness constraints (e.g., Berent, Steriade, Lennertz, & Vaknin, 2007; Redford, 2008). To that end, this study aims to provide evidence for the idea of an adaptive language production mechanism that is sensitive to recent experience, but subject to a markedness effect, with experiments showing that the relative ease with which speakers learn various phonotactic constraints corresponds to universal markedness preferences. Phonotactic constraint adherence in speech can be assessed by examining movement errors, or errors in which sounds slip within an utterance. An error in which the sound lands in the “legal” syllable edge, or the sound’s position in the intended syllable, demonstrates adherence to the particular constraint involving the moved segment.In the present study, native English speakers recited syllables including unattested marked and unmarked consonant clusters under artificial phonotactics constraints, following training in articulating the unattested segments. If speakers are sensitive to markedness in unfamiliar consonant clusters and capable of learning recently encountered patterns, movement errors involving the unmarked unattested cluster in the production task should be more likely to obey novel positional constraints than movement errors involving the marked unattested cluster. The results showed that participants were more likely to adhere to novel phonotactic constraints involving the unmarked unattested cluster, confirming that markedness constraints on unfamiliar consonant clusters influenced the acquisition of phonotactic constraints.