The acquisition of variation: Arab migrants' acquisition of (ING) and Coronal Stop Deletion in Wellington

<p>This dissertation investigates the patterns of acquisition found among Arab migrants to Wellington for two stable variables: coronal stop deletion (CSD) and (ING). CSD is the alternation between retained and deleted final consonant clusters, i.e. /wɛst/ vs. /wɛs/ and (ING) is the realisation of the final nasal in unstressed word-final syllables i.e. /dɹaivɪŋ/ vs. /dɹaivɪn/. CSD is a phonological variable that is mainly conditioned by articulatory constraints while (ING) is a morpho-phonemic variable with syntactic conditioning as well. An emerging trend in variationist sociolinguistics is to study variation in non-native varieties by analysing how far non-native speaker (NNS) patterns of variation replicate constraints on variation found among native speakers (NS) of a target variety. This study applies variationist methods to investigate the following questions: 1. What are the linguistic and the non-linguistic constraints that condition variation in the production of (ING) and CSD among NS in the New Zealand Spoken English Database (NZSED) in Wellington? 2. What are the linguistic and the non-linguistic constraints that condition variation in the production of (ING) and CSD among Arab migrants in Wellington (AM)? 3. Based on the results for (1) and (2), is there any evidence for “transformation under transfer” (Meyerhoff, 2009a). Interpretation of the results is done in line with the so-called “three lines of evidence”, and considers significant and non-significant constraints, constraint hierarchies and rank ordering of constraints (Tagliamonte & Temple, 2005). I consider the proposition that AMs, of all ages, are prone to transformation under transfer of NS constraints on the variables CSD and (ING), mainly illustrating strong and weak transfer. It is expected that old and middle-aged AMs will have patterns different from those found among young AMs. I also consider the possibility that articulatory constraints may be more readily transformed by AMs into ethnolectal marking, whereas grammatical constraints may be more likely to be strongly transferred by AMs. Old and middle-aged AMs seem to be more likely to display strong transfer of NS constraints, but they do not seem to be using variation in the L2 stylistically. By contrast, young AMs stylistically use articulatory constraints to convey important social indexicalities. The results suggest that oold and middle-aged AMs with developing grammars are like NS children acquiring variation of their L1, in the sense that old and middle-aged AMs are sensitive, in both CSD and (ING), to dialect-specific constraints on variation as they display strong transfer of the highest ranked NS constraint, be it articulatory or grammatical in nature; they also seem to perceive NS frequencies of occurrence of variables. Old and middle-aged AMs have an advantage over NS-children in their cognitive abilities that enable them to apply global constraints on variation by filtering their previous exposure to English, to replicate grammatical constraints of the L2 variables. Old and middle-aged AMs also seem to replicate the articulatory constraints that are perceptually salient, or that can host L1 transfer. They sometimes innovate articulatory constraints that are meaningful to them probably because of the influence of their L1. Young AM, who have arrived in New Zealand at an age of six years or younger, would be expected to illustrate strong transfer for stable variables like CSD and (ING). The results, nevertheless, illustrate that although young AMs share the same significant constraints found among NS of NZSED, they have different rank orderings, internal hierarchies and frequency of variants. Young AM, unexpectedly, diverge from NS norms and exhibit weak transfer of NS articulatory constraints on CSD, while they show strong transfer of NS grammatical constraints for the variable (ING). I suggest that young AMs seem to be using articulatory constraints in the L2 stylistically, to convey important social indexicalities. In addition, young AMs seem to hold an intermediate status between NS of NZSED and first-generation AM. Like old and middle-aged AMs, they replicate global-grammatical constraints on (ING) with an internal hierarchy that follows the Labovian nominal-verbal continuum, rather than the local, internal hierarchy. This suggests that (ING), as a morphophonemic variable with syntactic interfaces, has less room for the stylistic use of variation patterns as a reflection of identity marking. Articulatory constraints may be more subject to L1 transfer and these may become a marker of ethnicity among a Second-generation of migrants.</p>