Integrating Articulatory Constraints into Models of Second Language Phonological Acquisition

• Published in: Applied psycholinguistics Vol. 29; no. 3; pp. 489 - 534
• Main Authors: Colantoni, Laura, Steele, Jeffrey
• Format: Journal Article
• Language: English
• Published: Cambridge University Press 01-07-2008
• Subjects: Acoustics; Applied Linguistics; Articulation (Speech); Auditory Perception; English; French; Language Processing; Phonetics; Predictive Validity; Psycholinguistics; Second Language Instruction; Second Language Learning; Spanish
• ISSN: 0142-7164
• DOI: 10.1017/S0142716408080223

Models such as Eckman's markedness differential hypothesis, Flege's speech learning model, and Brown's feature-based theory of perception seek to explain and predict the relative difficulty second language (L2) learners face when acquiring new or similar sounds. In this paper, we test their predictive adequacy as concerns native English speakers' mastery of French /[image omitted]/ and Spanish /[image omitted]/. Based on an acoustic analysis of the learner data, we demonstrate that these three models do not account for the full range of variability nor for the developmental sequences attested, because they do not consider the degree of difficulty involved in the simultaneous mastery of multiple phonetic parameters across prosodic positions. Consequently, models of L2 phonological acquisition must not only integrate findings from markedness theory and speech perception but also incorporate phonetic constraints on production.