The second language acquisition of Spanish stress: Derivational, analogical or lexical?

다국어 초록 (Multilingual Abstract)

This research investigates English speakers' acquisition of stress placement in Spanish. There has been a lack of research into the acquisition of suprasegmental features of language, although Spanish teachers will be aware of the fact that beginning students do make a number of errors in stress assignment. This is thus is an area of Spanish acquisition that merits further research.
This acquisition process is investigated by analyzing the data of 56 learners of Spanish, at three proficiency levels, as well as those of a native control group (14 Spanish speakers). Participants took part in two experiments: a production experiment and a perception experiment. The production study incorporated both real and invented words to examine what learners do with words that are not familiar to them, and included a variety of syllable structures in testing. The perception study investigated learners' ability to perceive stress location on different syllables, as there is debate as to whether perception precedes production, or vice versa. Results show that stress is largely lexical, contrary to much of the recent literature in the area, which claims that rules are employed by learners to compute the stress placement. Under the present analysis, the learner's familiarity with a word allows for correct stress placement through memorization and lexical recall In the absence of lexical knowledge, however, learners recur to an internal system of stress output checking. In initial stages of language learning this system is influenced by a variety of factors, including overgeneralizations generated about the target language based on limited data as well as interference from the first language. With increased proficiency the learner system comes to resemble that of the native language, thanks to greater lexical knowledge and greater experience with patterns encountered in the language. The most advanced learners' performance is indistinguishable from that of native speakers.
These results bring to light interesting insights into the process of learning this under-investigated aspect of foreign language, and provide novel information regarding the processes at work in second language phonological learning.