The influence of written input on encoding difficult phonological contrasts for L1 Arabic learners of L2 English

Research investigating the relationship between orthography and second language (L2) phonology has grown considerably over the past 20 years, in both size and complexity (Hayes-Harb & Barrios, 2021). However, few studies explore the influence of L2 written input across writing systems and varying proficiency levels (Hao & Yang, 2021; Mok et al., 2018; Showalter, 2020). Further, participants are rarely invited to reflect on their beliefs and strategies related to written input. The present study draws together behavioural and reflective insights to investigate the influence of written input on the lexical encoding of confusable phonological contrasts by adult first language (L1) Arabic-speaking learners of
L2 English.

An L1 Arabic group (n=114), with varying English proficiency, and a control L1 English group (n=117), with no Arabic experience, completed an online word learning study. Word learning involved 12 English pseudowords presented auditorily, accompanied by an
image and written input. Words were minimal pairs differing by either /m-n/ (easy) or /f-v/ (difficult) contrasts. Minimal pairs from each contrast were taught with Arabic spelling, English spelling, or audio only. Lexical encoding was then tested in an audio-visual matching task, followed by a reflective post-test questionnaire.

Mixed-effects modelling of L1 Arabic responses revealed an inhibitory effect of any written input on accuracy matching /f-v/ words, while written input facilitated encoding of /m-n/ words. In contrast to the behavioural findings, participant reflections on the influence of written input were overwhelmingly positive. Participants valued written input for a variety of reasons and reported distinct influences of Arabic and English script input. Further, a range of strategies were reported, drawing on phonological and orthographic knowledge; however, these strategies did not improve accuracy. Both quantitative and qualitative analysis converged on the pivotal role of individual differences, such as proficiency, in modulating the relation between written input and phonological learning.