Phonological Representations of Stop Sounds in L2A: The Perception of Korean Stops by English and Finnish Speakers

This study investigates how the phonological representations of stop phonemes in the learner's first language (L1) affects the acquisition of stop sounds in the second language (L2). In order to investigate this issue, perception data was collected from 13 British English speakers and 15 Finnish speakers learning Korean. These three languages were chosen because their stops exploit distinctive features in different ways; in Korean three types of stop phonemes called Tense, Plain and Aspirated exploit [spread glottis] (Steriade 1991, Silva 1992 and Ahn & Iverson 2003), in English voiced and voiceless stops [aspirated], and in Finnish unaspirated voiceless stops and voiced stops [voice]. Based on these facts, two hypotheses are tested in this study: (i) Both English speakers and Finnish speakers will have difficulty in acquiring the distinctions of three different types of Korean stops. (ii) Despite the fact that their L1 feature representations for stops are different from each other, both English and Finnish speakers may show a similar pattern of difficulty in discerning Korean stops. This is because they are facing the same challenge of acquiring the new distinctive feature [spread glottis] in the target language. The results support the hypotheses, showing that (i) both English and Finnish learners were not successful in identifying the three different types of Korean stops and (ii) English and Finnish speakers showed a similar pattern of difficulties in discerning Korean stops regarding the feature [sg]. However, beyond the factor of the feature [sg], the two language groups performed differently from each other in perceiving Korean stops. I claim that the different patterns of error types have been caused by the absence or presence of the geminate in the learner's L1.