The Locus of L1 Constraints on L2 Phonological Acquisition

Charles B. Chang, University of California, Berkeley

How is a late learner's acquisition of novel phonological categories in a second language (L2) constrained by the phonological structure of their first language (L1)? I address this question by examining the acquisition of the three-way Korean laryngeal contrast among lenis, fortis, and aspirated stops by 27 native English speakers, novice learners with no prior exposure to Korean. In this paper I report results from an imitation experiment (cf. Flege and Eefting 1988) in which learners heard and repeated a two-dimensional continuum of Korean syllables differing in the primary cues to the Korean laryngeal contrast in initial position: voice onset time (VOT) and fundamental frequency (f0) onset in the following vowel (cf. Figure 1). How successful are late learners at restructuring a phonetic space of two laryngeal categories ("voiced" and "voiceless") into a phonetic space of three laryngeal categories?

Acoustic analyses indicate that while native Korean speakers' imitations bunch into three clusters corresponding to the three laryngeal categories of Korean, late learners' imitations tend to bunch into two clusters corresponding to the two laryngeal categories of English (cf. Figure 2), even after three weeks (= 39 hours) of immersion instruction. However, there is wide variation among late learners, who divide into two groups that show either good or bad within-category discrimination, as reflected in the degree to which they accurately imitate the acoustic parameters of the stimuli. For instance, in the case of late learners LF21 and LF40, LF21 imitates the stimulus parameters relatively poorly, while LF40 imitates them quite well. This inter-speaker difference is reflected statistically in the much better correlations of stimulus and response acoustics for LF40 as compared to LF21 (cf. Table 1).

What these results suggest is that some learners (e.g. LF21) perceive the L2 signal in terms of L1 categories and thus produce as their L2 imitation just an exemplar of the L1 category they classified the stimulus as, without much regard for the original stimulus parameters. On the other hand, other learners (e.g. LF40) perceive the L2 signal faithfully; their troubles seem to arise, then, not in perception, but in production: they can only reproduce the detail they hear with the articulatory programs that are at the time reliable for them, and these are likely limited to those for L1 categories at such an early stage of L2 acquisition.

Thus, these findings are generally consistent with one of the fundamental postulates of Flege's (1995) Speech Learning Model - that adults retain, rather than lose, the perceptual mechanisms used in learning their L1 sound system - contra the notion of a "Critical Period" for language (Lenneberg 1967). Although some learners appear indeed to be held back by L1-biased perception, the problem for many learners is not in perception, but in the move from perception of phonetic details to production of a new abstract category. In short, late learners' early acquisition of a novel L2 category seems to be considerably constrained by their L1, but the influence of L1 may intervene either in perception or in production.

Please click here for diagrams

References:
Flege, James Emil. 1995. Second language speech learning: Theory, findings, and problems. In W. Strange (ed.), Speech Perception and Linguistic Experience: Theoretical and Methodological Issues in Cross-Language Speech Perception, 233-272. Baltimore: York.
Flege, James Emil, and Wieke Eefting. 1988. Imitation of a VOT continuum by native speakers of English and Spanish: Evidence for phonetic category formation. Journal of the Acoustical Society of America 83(2): 729-740.
Lenneberg, Eric. 1967. Biological Foundations of Language. New York: John Wiley & Sons, Inc.