The paper is aimed at investigating the sources of cross-linguistic influence in the third language (L3) phonology, and, particularly, the impact of the second language (L2) on the phonological acquisition of another foreign language. The study consisted in foreign accent judgements performed by a group of expert judges who were presented with recorded samples of L3 English. The findings confirm the results of some previous studies on the phenomenon of 'L2 status' and demonstrate a tendency for the L2 phonological transfer in L3 production at the initial stages of acquisition that decreases with growing third language proficiency. IntroductionPhonological aspects of third language acquisition (TLA) constitute a relatively unexplored research area that reaches beyond Second Language Acquisition (SLA) by drawing attention to the complexity of cross-linguistic influence while learning more than one foreign language, a situation which has become a commonplace in today's multilingual world. The present study aims at exploring the phenomenon of L2 status or a 'foreign language effect', i.e. the impact of previously learnt foreign language(s), rather than only the mother tongue, on the phonetic performance in a third language (cf. Meisel 1983, Cenoz 2001, Hammarberg 2001, Hammarberg & Hammarberg 2005, Fernandes-Boëchat 2007. The paper aims to provide further evidence for the claim that L2 outweighs the transfer from L1 at the initial stage of L3 acquisition leading to L2-accented speech, however, this interference diminishes with the gradual approximation to the L3 target norms. This paper is a part of an on-going investigation into the patterns of interference between first, second and third language, focussing particularly on the articulatory resetting in the process of third language phonological acquisition. The ultimate goal of the research is to assess the influence of L1 and L2 in third language oral productions by comparing perceptual judgements of a foreign accent with an auditory and acoustic analysis of articulatory setting of L3 English 1 . Theoretical backgroundIt is generally acknowledged that foreign language acquisition, particularly in the area of phonology, is subject to interference from the first language (L1). Moreover, the process of acquiring a third language is rather commonly believed to parallel that of the second language learning. Only recently, attention has increasingly been drawn to the specific and complex nature of phonological acquisition in a multiple language setting (cf. e.g. Cenoz et al. 2001, Bannert 2005. The process of acquiring a third language is more complicated in the sense that phonological transfer is not limited only to the mother tongue, but other earlier acquired 1 The present contribution constitutes an extended version of a selected part of the research presented in Wrembel (2008).

The present study is a continuation of previous investigations into the nature of sound−colour associations in a non-synaesthetic population conducted on English and Polish vowel sound systems and it aims at providing further evidence for the non-arbitrary nature of cross-modal mappings. The experiment 1 was run on a specially designed computer program and involved 90 participants who were asked to match randomised auditory stimuli (12 English vowel sounds recorded in 2 conditions: in isolation and in a CVC context) with one of 11 basic colours (red, yellow, green, blue, brown, purple, pink, orange, black, white and grey) presented as coloured rectangles on a computer screen. The program kept record of the colour choice and reaction time of the participants, who fell into 2 groups with respect to the level of their language proficiency and phonetic awareness.An analysis of the results revealed statistically significant interactions between specific colours and individual vowel sounds for all 12 English vowels examined in the combined analysis; for 10 vowels in Condition 1 (isolated auditory stimuli) and for 7 vowels in Condition 2 (stimuli in the CVC context). A group effect was not found to be significant as far as the quality of mappings was concerned; however, in the case of reaction times the less advanced learners took significantly longer to assign colours to sounds in context. The findings indicate that vowel−sound mappings in non-synaesthetic perception appear non-arbitrary and follow the general tendencies in which bright colours (yellow, green) are associated with high front vowel sounds, whereas dark colours (brown, blue, black) are attributed to back vowels, while open sounds tend to be perceived as red and central vowels are mapped onto achromatic grey.

No abstract

No abstract

No abstract