Second Language Acquisition of Stress Patterns: A Study on the Effects of L1 Tamil Transfer on L2 English

Linguistic studies have observed that during second language acquisition (SLA), L2 learners tend to produce non-target like structures, resembling grammars/patterns of the learner’s L1. Schwartz and Sprouse (1996) have accounted for these observations by proposing the Full Transfer Full Access Hypothesis. This hypothesis assumes that the initial state of L2 acquisition is the L1 grammar (Full Transfer). When L2 input data cannot be assigned to L1 representations, this will force re-structuring and L2 learners will have access to the same properties of Universal Grammar that were available to them during L1 acquisition (Full Access) and therefore target-like acquisition should be possible.

This study looked at the L2 acquisition of English stress patterns by L1 Tamil speakers. While English stress patterns can be predicted to some extent by rules that rely on syllable weight, Tamil has fixed stress, occurring in word-initial position. This study examined whether L1 Tamil transfer effects are observed during the acquisition of stress patterns in English and whether these effects weakened as proficiency and exposure to the L2 input data increased. The study consisted of three tasks. In Task 1, participants were asked to read aloud a list of real words in English, while in Task 2, the list consisted of pseudowords. For Task 3, participants were given two pronunciations of pseudowords, with either penultimate and antepenultimate stress, and asked to choose their preferred stress placement. All target words were tri-syllabic. The study was piloted with 13 native English speakers (control group) to account for any variability across the target words. The study then tested 14 L1 Tamil L2 English speakers, with differing levels of English proficiency – beginner, intermediate and advanced. When the results of the L2 learners were compared to the native speakers, an effect of L1 Tamil transfer was found, as L2 learners predicted stress as expected for APU words (with stress on the word-initial syllable) but not for PU words (where the second syllable was stressed) and therefore preferred word initial stress on all words. Furthermore, results showed that within the L2 groups, there was an effect of proficiency – advanced L2 learners performed most target-like, followed by intermediate and beginner speakers. These results were in line with the Full Transfer Full Access Hypothesis, showing that while effects of L1 transfer are stronger during earlier stages of proficiency, this effect weakens, and target-like performance is possible with increased proficiency and exposure to the L2 input.