L2 Comprehension: a discussion of some influencing factors  

Robert McAllister, James E. Flege* and Thorsten Piske^

*Department of Rehabilitation Sciences, Speech and Hearing, University of Alabama at Birmingham
^Department of English, Kiel University

To appear in the proceedings of the 9th Annual Conference on the European Second Langugage Association (EUROSLA 9)
(Department of Linguistics, Lund University, Sweden)  


Recent research has given us reason to believe that several factors are likely to influence how well an L2 user understands the spoken L2. Experimental data on L2 performance will be presented as a basis for a discussion of these factors and their relation to the Speech Learning Model (Flege, 1995). Results are presented from two recent studies which have focused on the effect of age of L2 learning, and amount of L1 use, on the perception of English vowels and consonants by Italians who are highly experienced in English (average length of residence in Canada = 35 years). In the first study the perception (categorial discrimination) of English vowels by 3 groups of 18 native Italian subjects who had been living in Canada for over 30 years was examined. As their AOA to Canada increased, their accuracy in perceiving English vowels decreased. The Ss' production of English vowels was assessed auditorily, by English speaking listeners. As AOA increased, vowel production accuracy declined. There was a significant correlation between vowel perception and vowel production accuracy. In the second study the identification of English consonants in the initial and final position of non-words presented in noise was assessed. The native Italian Ss (see above) identified English consonants less accurately than did native English monolinguals, even though they had been speaking English for 35 years. We compared two groups of early bilinguals differing in amount of L1 use. The subjects who seldom spoke Italian -- but not those who often spoke Italian -- resembled the native English Ss in identifying consonants better in initial than final position. There was an effect of amount of L1 use for final position (where Italian has few consonants), but not initial position. Another ongoing study addresses the problem of the role of contrasts in the L1 in the learning of L2 contrasts. Specifically the so called "feature hypothesis" is tested. The learning of the perception of the durational contrast in Swedish by experienced adult speakers of English and Spanish, who have no durational contrast in their L1 was found to be more difficult than for Estonians whose L1 has a complex system of durational contrasts. Finally experiments with globlal L2 comprehension and the concept of "perceptual foreign accent" will be briefly discussed in light of the above mentioned factors.

View pdf-version of paper

PERILUS | All issues | 1999 issue | Previous | Next | Phonetics at Stockholm University