Audiovisual Training Effects on L2 Speech Perception and Production

Abstract

This study investigated whether audiovisual perception training can improve learners' auditory perception of L2 speech sounds. 29 subjects (experimental group) who had difficulty in the perception of English /θ/-/s/ and /ð/-/z/ were recruited to attend a 9-session audiovisual perception training programme with identification tasks on the target contrasts. Their perception performance was tested before, during and at the end of the training with an AXB task. A control group was tested with the same AXB task and intervals as that of the experimental group. The results showed that the experimental group's perception accuracy improved substantially during and by the end of the training programme. The control group also showed perception improvement across the pre-test and post-test. Their degree of improvement, however, was significantly lower than that of the experimental group. These results therefore confirm the value of the audiovisual modality in L2 speech perception training.

The effects of articulatory gestures on L2 pronunciation learning: A classroom-based study

“…The materials employed in the pilot study and the pre-test were both adopted from Li (2016): a short text used for the pilot study to detect which sound(s) were difficult for the students to pronounce, thus determining the target sounds for teaching (/θ/–/s/ and /ð/–/z/ were selected as a result); and 12 English sentences that embedded the target contrasts in initial, medial, and final positions were employed for the read-aloud tasks in the pronunciation tests (/θ, s, ð, z/ occurred 15 times each).…”

“…One of the most extensively discussed examples would be Japanese speakers’ difficulty differentiating between the English /ɹ/ and /l/ (Best & Strange, 1992). Another example would be the production of fricative dental /θ/–/ð/; both Mandarin and Japanese speakers were found to realize them as alveolar /s/–/z/ (Picard, 2002; Li, 2016a), while Italian and Russian speakers tended to substitute /θ/–/ð/ with either /s/–/z/ or /t/–/d/ depending on the speakers’ English proficiency (Flege, 1995a, 1995b). Moreover, L2 learners of different first languages (L1s) were found to have different difficulties in pronouncing L2 sounds of different syllable positions.…”

“…Hazan et al (2006) also agree with this view and identified three types of visual category based on their occurrence in learners’ L1 and L2: a visual category that (1) occurs both in the L1 and L2; (2) exists in the L2 but not in the L1; and (3) exists both in the L1 and L2 but is used in different phonetic distinctions in the L1 and L2. L2 learners are predicted to be more likely to have difficulties in the acquisition of L2 sounds in the (2) and (3) situations than in the (1) situation (see also Li, 2016a; Wang et al, 2009).…”

“…In fact, some previous studies have provided us evidence in support of this view. Li (2016a), for example, carried out audio-visual training among a group of Mandarin speakers who had difficulty differentiating the English /θ/–/s/ and /ð/–/z/, the articulatory gestures of which are visually distinguishable. The participants undertook a nine-session audio-visual training programme in which they were trained with audio-visual materials that demonstrated the articulatory gestures of the target sounds.…”

“…Compared with approach (3), approaches (1) and (2) may be more likely to meet the needs of segmental teaching, particularly for the teaching of accurate pronunciation of specific phonemes; given that the present study aimed to examine whether previous phonetic training experiments relating to audio-visual modality could yield similar results in classroom based settings, studies using similar procedures were followed (e.g. Li, 2016a; Hazan et al, 2005). As such, it comprised asking the students to listen to/observe the recordings without intervention.…”