Sociolinguistic variation (SLV) entails that language is affected by social context (i.e. register, pragmatics). Interest in the acquisition of sociolinguistic variation in a second language (L2-SLV), as a key component of communicative competence, has grown exponentially over the past thirty years (Geeslin & Long, 2014). Researchers have primarily framed L2-SLV as a skill reserved for advanced learners (Geeslin, 2018). However, many language functions at even the novice proficiency level necessitate SLV awareness (ACTFL, 2012b). This is particularly true for learners of diglossic languages such as Arabic. In this study, I expand on L2-SLV research by exploring the acquisition of SLV at the initial stages of learning in a tightly controlled experimental setting. Novice participants studied Mini-Arabii, a miniature language (e.g., Cross et al., 2020; Mueller, 2006) which mimics lexical and morphosyntactic SLV between Modern Standard Arabic (MSA) and Egyptian Colloquial Arabic (ECA). The experiment was designed to explore two fundamental questions about the acquisition of L2-SLV: 1) is learning two registers (i.e., in a curriculum that teaches SLV; “+SLV” ) more difficult than learning in one register (i.e., in a traditional curriculum that ignores SLV; “-SLV”); and 2) for curricula that include SLV, is it more difficult to learn through an integrated approach (where SLV registers are taught side-by-side), or through a sequential approach (where one register is taught first, and an additional register is taught second)? The training conditions I designed for my study were intended to operationalize these L2-SLV curricular approaches. For the first comparison, +SLV vs. -SLV curriculum training, participants (novice learners of Arabic) received six exposures to the -SLV vocabulary and grammar set as well as six exposures to the +SLV vocabulary and grammar set (for +SLV, these were divided into three exposures in MSA, three in ECA). This within-participants manipulation mirrored the real-world trade-off of choosing to cover one vs. two SLV registers within a set period of time. As for the second comparison, participants were assigned to either a +SLV-Integrated or +SLV-Sequential approach as part of their +SLV training above. Participants who received the +SLV-Integrated approach were exposed to MSA and ECA side by side on each day. Those following the +SLV-Sequential approach were exposed to MSA on the first day and ECA on the second day. Learning was measured as progress on the training tasks throughout the three-day, web-based experiment in terms of mean accuracy (both as a binary 0/1 measure and as an approximate measure through the Levenshtein Distance([Levenshtein, 1966]) and mean processing speed (log RT). Vocabulary and grammar knowledge were assessed via production and reception post-tests. For the first comparison (+SLV vs. -SLV), participants were marginally less accurate but equally fast when they studied in two registers rather than one. As for the second comparison (+SLV-Integrated vs. +SLV-Sequential), participants also responded with comparable speed regardless of which approach they trained in. However, neither approach seemed to give participants a clear advantage in terms of accuracy. As a whole, the current study sheds nuanced light on how L2-SLV can be processed and acquired. For teachers, incorporating SLV at the initial stages of learning does not seem to come at a noticeable cost in learners’ overall performance. For researchers, the results invite further investigation into why the acquisition of multiple variants affects accuracy but not processing speed, and what this means for theoretical models of L2-SLV.
Read
