The Impact <strong>of</strong> <strong>Applied</strong> <strong>Linguistics</strong>: <strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> 44th Annual Meeting <strong>of</strong> BAAL University <strong>of</strong> <strong>the</strong> West <strong>of</strong> England 16
5 Introduction Identifying criterial aspects <strong>of</strong> pronunciation in L2 English across CEFR levels: Implications <strong>for</strong> language learning Fiona Barker, Brechtje Post, Elaine Schmidt and Mike McCarthy Identifying criterial aspects <strong>of</strong> pronunciation in L2 English across CEFR levels: Implications <strong>for</strong> language learning Fiona Barker 1, Brechtje Post 2, Elaine Schmidt 2 and Mike McCarthy 3 1 University <strong>of</strong> Cambridge ESOL Examinations, 2 University <strong>of</strong> Cambridge Department <strong>of</strong> Theoretical and <strong>Applied</strong> <strong>Linguistics</strong>, 3 University <strong>of</strong> Nottingham While teachers agree that pronunciation is a crucial aspect <strong>of</strong> L2 English competence, it still remains an under-researched area. Much <strong>of</strong> <strong>the</strong> previous research has looked at vowel or consonant inventories alone, describing which sounds learners can or cannot produce at different stages <strong>of</strong> language learning and which effect this might have <strong>for</strong> pronouncing specific words (Hansen Edwards & Zampini, 2008). However, relatively little is known about o<strong>the</strong>r components <strong>of</strong> pronunciation, such as rhythm, and how <strong>the</strong>se factors interact with, <strong>for</strong> instance, syllable structure, in L2 learning (Derwing & Munro, 2005). Languages have traditionally been divided into stress-timed and syllabletimed languages, where <strong>the</strong> <strong>for</strong>mer are supposed to be characterised by time intervals with roughly equal length between prominent syllables (e.g. Dutch, English and German), while <strong>the</strong> latter have successive syllables <strong>of</strong> roughly equal length (e.g. Cantonese, Korean and Spanish). The different perceptions <strong>of</strong> <strong>the</strong>se types <strong>of</strong> rhythm have been claimed to be crucially dependent on a number <strong>of</strong> phonological factors which can differ crosslinguistically, including consonant clusters, vowel reduction, pre-boundary leng<strong>the</strong>ning and accentuation (Dauer, 1983, Prieto et al. 2012). In syllabletimed languages, <strong>the</strong>re are typically predominantly CV (consonant vowel) sequences while stress-timed languages tend to have complex consonant clusters. This means that some syllables might consist <strong>of</strong> 4 or 5 segments, while o<strong>the</strong>rs only have 2. Moreover, <strong>the</strong> reduction <strong>of</strong> unstressed vowels, which is typical <strong>of</strong> stress- but not syllable-timed languages, reduces <strong>the</strong> duration <strong>of</strong> vocalic timing in unstressed syllables, and thus makes unstressed syllables shorter than stressed ones. Additionally, accented syllables, i.e. syllables which carry <strong>the</strong> main sentence accent, tend to be significantly longer in stress-timed languages but in syllable-timed 17