On the perception of English high vowels by Polish learners of English

The experiment examined how two pairs of English vowels (/i:/-/I/ and /u:/-/U/) are perceived by Polish learners of English. The underlying assumption was that non-native vowel perception might be influenced by listeners’ L1 vowel system (Flege 1995, Best 1995, Major 2001) or by universal factors (Major 2001). Two research questions were formulated. How much attention do Polish learners of English pay to spectral cues for /i:/-/I/ and /u:/-/U/ distinctions in English? Are temporal cues equally important for Polish learners of English when they distinguish between /i:/-/I/ and /u:/-/U/ vowel pairs? Predictions concerned the influence of L1 vowel system (Polish does not use duration differences and has only one vowel in the high back vowel region), salience of temporal cues and identification of the vowels in question. 37 Polish subjects and a control group of English native speakers identified stimuli from /hi:d/-/hId/ and /hu:d/-/hUd/ continua varying in spectral and durational steps. English subjects strongly relied on spectral cues (typical categorical perception) and almost did not react to temporal cues. Polish subjects relied strongly on temporal cues for both continua, but showed a reversed pattern of identification of /u:/-/U/ contrast. Their reliance on spectral cues was weak and had a reversed pattern for /i:/-/I/ contrast. The discussion of the results focused on duration as a readily available cue to L2 vowel contrast, assimilation of L2 vowels to the existing vowel system, and the influence of orthography and L1 reading habits on the perception of L2 vowels