The Phonetics Laboratory

CASL: Computer-Assisted Sound Learning for Greek learners of English: Effects on perception and articulation of conversational speech

Dr. Angelos Lengeris

The aim of this project is to examine the effect of computer-based phonetic training on Greek speakers’ perception and pronunciation of English vowels and consonants. We will train one group of Greek speakers on English vowels and another on English consonants using the same training stimuli. This will allow testing whether the two groups (vowel-attending group and consonant-attending group) will improve not only in the type of sounds they attend to (vowels and consonants respectively) but also in the type of sounds they do not attend to (consonants and vowels respectively). Participants’ perception will be tested on English sentences instead of isolated words used in previous research. Participants’ production will be assessed using acoustic, perceptual and articulatory measures (Electropalatography) in sentences and in conversational speech (using a variant of a Map Task). The research findings will explore the role of attention in phonetic learning and, importantly, will provide important information on how academic knowledge can be put into practice to improve L2 perception and pronunciation in EFL settings.    

Objectives and methods

Study 1
Study 1 will identify those English consonants that pose difficulties for Greek learners of English and will be appropriate for the computer-based training since, apart from impressionistic data and general predictions based on a phonemic comparison of the two consonant systems (e.g. English has both alveolar and postalveolar fricatives while Greek has only alveolar ones which makes it difficult for Greek speakers to differentiate and produce the two places of articulation), there are no experimental studies in the literature examining the identification and production of the full set of English consonants by Greek speakers. In addition, our participants will complete a battery of perception tasks (e.g. frequency discrimination, Phonological Short-term Memory, perceptual assimilation to L1 categories) in an attempt to explore sources of individual differences in their performance.

Study 2
Study 2 will train Greek speakers on English vowels and consonants using the High-Variability Phonetic Training approach. Pre- and post-training tasks will examine the perception and production of English consonants and vowels. L2 perception will be tested using an identification task with English sentences containing the target vowels and consonants. L2 production will be tested by recording participants (a) reading sentences containing the target vowels and consonants and (b) producing spontaneous speech containing the target vowels and consonants. To elicit spontaneous speech, we will use a variant of a recently-developed method for obtaining spontaneous laboratory-quality speech containing specific words of interest whereby participants perform a ‘spot the difference task’ (Van Engen et al., 2010; Baker and Hazan, 2009). Participants’ production will be assessed using acoustic, perceptual and articulatory (EPG) measures.