Abstract/Summary

Orthographic input has been the focus of much investigation in the search for the second language (L2) phonological development. The focus of studies investigating this issue has been directed on to the role of experience factors in modulating this effect, e.g., congruency level between the first language (L1) and L2 orthographic systems, and level of orthographic depth of L1 and L2. However, to the best knowledge of the researcher, few studies have linked the role of internal factors with experience factors (e.g., Bassetti, Mairano, Masterson, & Cerni, 2020). This study seeks to understand the role of orthographic input in terms of availability and script familiarity in discriminating between three pairs of Arabic phonemes in word-level (phonemes embedded in words) and phoneme-level (phonemes uttered in isolation) discrimination. This role is investigated in relation to the participants’ phonemic coding ability and L1 (English) /L2 (Arabic) phoneme similarity. The pairs of Arabic phonemes investigated in this study were (/θ/ and /ð/), (/s/ and /sˤ/), and (/χ/ and /ʁ/). These three pairs were selected based on Flege’s (1986) equivalence classification in which the first is identical to English, the second is similar to English, and the last is novel to English. After taking a phonemic coding ability test, participants were allocated into three orthographic groups: 1) unfamiliar script (Arabic), 2) a combination of familiar and unfamiliar scripts (Roman and Arabic), and 3) no orthographic input. Generally, easily perceived contrasts (identical and novel phonemes) were found to be discriminated better than hard-to-distinguish contrasts (similar phonemes). Notwithstanding that all groups did well overall in both phoneme discrimination levels, there were some differences in these two levels. In the word-level discrimination, findings indicated that the poorest performance was reported by those who had Roman along with Arabic script. Groups who had unfamiliar script or had no script at all did not show any significant difference from each other in their performance. Participants performed best when they were asked to discriminate between two new phonemes, whereas there was no significant difference in their performance when they were asked to discriminate between two identical or two similar phonemes, which might be caused by impaired learning in the case of identical phonemes and perceived difficulty in the case of similar phonemes. In the phoneme-level discrimination, orthographic input type did not have any significant effect on participants’ performance. On the other hand, Participants did not show any significant difference in their performance when they were asked to discriminate between two new or two identical phonemes, whereas their poorest performance was when they were asked to discriminate between two similar phonemes, which might be caused by the perceived difficulty of the two phonemes. There was no significant interaction between L1/L2 phoneme similarity level and orthographic input type in both levels of discrimination. Phonemic coding ability, however, was found to not affecting participants’ performance in the following discrimination tasks. The study contributes to our understanding of models such as the Speech Learning Model (SLM) (Flege, 1995), Perceptual Assimilation Model (PAM) (Best, 1995), and Second Language Perceptual Assimilation Model (PAM-L2) (Best & Taylor, 2007), which argue that similar phonemes tend to be more difficult to discriminate than different ones. The findings also are supported by Cognitive Load Theory (Sweller, 2011) which encourages taking the capacity and duration limits of working memory into consideration when devising instructional procedures to avoid hindering learning.