Socio-psychological factors in the attainment of L2 native-like accent of Kurdish origin young people learning Turkish in Turkey.

다국어 초록 (Multilingual Abstract)

Second language acquisition research has sought to identify socio-psychological factors underlying language learners' degrees and rates of acquisition. Studies have shown that learners with autonomous motivation orientations and positive attitudes towards the L2 community (Donitsa-Schmidt et. al., 2004; Schumann, 1978; Spolsky, 2000) acquire the target language better than those without such orientations and attitudes. This study utilizes social network theory (Milroy, 1987), identity theory (LePage & Tabouret Keller, 1985; van Dijk, 1998) and self determination theory (Deci & Ryan, 1987) to explore how L2 learners' socially constructed identities, external, introjection, identification and integration motivational orientations, and exchange, interactive, and passive family and non-family networks relate to the attainment of the regional Turkish accent by young Kurds. Using a cross-sectional research design, this study addresses the following questions: (1) How native-like is the participant's accent when speaking Turkish as rated on a 1-5 scale? (2) What are the identity patterns found in the Kurdish-speaking community, and how do these patterns relate to their Turkish accent? (3) Do different motivational orientations significantly relate to attainment of native-like accents? (4) What are the social networks of the Kurdish-speaking community, and how do these networks relate to accent native-likeness? Data collected from 120 middle and high school students included speech samples from a read-aloud accent test and four questionnaires regarding their motivation to learn Turkish, their identification patterns, and social networks. Global accent ratings revealed significant degrees of variation in participants' accents varying from 1.1 to 4.7. Findings suggested that the degree of identification with the Turkish-speaking community was a positive predictor (.31, p < 0.01), and the degree of identification with the Kurdish-speaking community was a negative predictor (-.34 p < 0.01) of accent native-likeness. Data also showed that among four motivational orientations, integration orientation was a positive (.32, p < 0.01), and introjection was a negative (-.20; p < 0.01) predictor of accent native-likeness. Results indicated that participants with a more native-like accent also had more Turkish-speaking family and non-family networks that were exchange and multiplex in nature than the networks of those participants with less native-like accents. Results also suggested several significant gender and age effects.