The role of processing instruction in the acquisition of English restrictive relative clauses by L1 Saudi Arabic speakers

Abstract

Recent literature has extensively examined the acquisition of English restrictive relative clauses (RRCs), indicating that Arabic learners encounter particular difficulties (Zagood 2012; Shaheen, 2013; Alroudhan 2016; Abumelhah, 2016). English and Arabic differ in their use of overt versus covert relative markers and agreement features. However, instruction in English rarely emphasises the form-focused and processing aspects of relative clause constructions. This study draws on insights from current generative second language acquisition research on the linguistic properties of English RRCs to examine the effectiveness of Processing Instruction (PI) in the classroom.
VanPatten’s Input Processing (IP) model (VanPatten and Oikkenon, 1996; VanPatten, 2004; VanPatten and Williams, 2007) addresses the acquisition of morphosyntax by exploring how linguistic form and cognitive processing interact during comprehension, proposing that L2 morphosyntactic difficulties may arise from learners’ allocation of attention to input. To mitigate these difficulties, VanPatten (2004) describes PI as comprising two types of structured input activities: referential activities, which require learners to focus on a particular form and its meaning, and affective activities, which present multiple examples of the target form while directing learners’ attention to the overall meaning of the sentences. Until now, these two PI activity types have typically been combined into a single instructional approach, and few empirical studies have separately examined their distinct effects. Consequently, this experimental study aims to isolate and compares the effects of referential and affective PI activities on the acquisition of English RRCs by beginner Saudi speakers of English.
Four experimental groups and a control group are included in the study. It begins by identifying the specific difficulties faced by L2 learners in acquiring English RRCs and by pinpointing features believed to involve re-assembly difficulties, such as syntactic constraints on the use of the definite article when a noun phrase is modified by a relative clause. Three research instruments are employed to assess participants’ performance: a Grammaticality Judgment Task, a Picture-Cued Task, and a Translation Task, all of which are administered as pre-tests, immediate post-tests and delayed post-tests (seven weeks later). A quantitative approach is used, comparing scores from the pre-tests, post-tests and delayed post-tests to measure learners’ development. The data are analysed using descriptive statistics, repeated measures ANOVAs and t-tests.
Initial results indicate that only the referential activities condition produced improvement in learners’ accurate use of English RRCs, including a sustained long-term effect that was not observed for the affective activities, Traditional Instruction or the control group. Subsequent findings indicate that an ERA intervention also leads to improvement across all four conditions examined: a definite RRC with an overt complementiser; a definite RRC with a null complementiser; an indefinite RRC with an overt complementiser; and an indefinite RRC with a null complementiser. ERA also produces a sustained long-term effect. Therefore, this study suggests that in such challenging contexts, ERA instruction – and especially its referential activities component – is the type of instruction recommended to help learners decrease the L1 crosslinguistic influence, thereby developing the target L2 knowledge. Moreover, an important implication is that claims of previous PI studies about the main factors driving its effectiveness require a more nuanced explanation. The findings of the present study thus contribute to both theoretical understanding and empirical knowledge, particularly in the areas of Processing Instruction and generative SLA research. Finally, the study adopts the Modular Online Growth and Use of Language (MOGUL) model, part of the Modular Cognition Framework (MCF), a language development model to bridge the gap between PI and GenSLA research. MCF also accounts to tease apart modular and extra-modular types of L2 knowledge.

More information

Identifiers

ORCID for Faraj Ahmed F Alhamami: orcid.org/0009-0006-8329-1553

ORCID for Roumyana Slabakova: orcid.org/0000-0002-5839-460X

Catalogue record

Export record