Many sociolinguistic studies of morphosyntactic variation primarily focus on highly socioindexical variants
(ideologically linked to social meanings), and it is rarer to see comparison across variants of differing degrees of social salience and discourse-pragmatic utility (cf. Moore, forthcoming; Cornips & Corrigan, 2005: 87). Consequently, the present study is interested in the acceptance (ie. perception) of the following morphosyntactic constructions: Argument movement (raising-to-subject with the verb seem); ii. Optional discourse-based movement (left dislocation, right dislocation, topicalisation), and iii. Nonstandard agreement (past-tense BE: nonstandard was, nonstandard were, and nonstandard weren’t).
Using an Acceptability Judgement Task, acceptance of these constructions is measured in Polish-born
migrants - newer members to the population, compared with English-born participants. An additional dimension of comparison is by membership to a specific non-hegemonic community - the LGBTQ+ community. Factors on acceptance are considered: i. macro-social demographic factors; ii. L2 constraints on migrants' acquisition; and iii. meso-social (community embeddedness) factors.
Results from ordinal logistic regression models reveal a clear pattern in acceptability: Argument movement > optional discourse-based movement > nonstandard agreement. Migrants also follow this pattern, though have lower acceptance than English-born participants, with acceptance of nonstandard agreement diverging more than optional discourse-based movement. Higher English proficiency and education lower acceptance while other aspects of migrants' identities, such as age, and acculturation level (embeddedness in British culture) increase acceptance. Being LGBTQ+ increases acceptance, particularly in migrant participants. However, level of LGBTQ+ community embeddedness lowers acceptance of optional discourse-based movement.
This study challenges us to consider which speakers and communities 'matter' for variation, given the
social diversity and varied lived experiences across both migrants and non-migrants. Methodologically, this study is an exemplar for accessing information about lived experiences while operating at a wide-angle 'macro' scale (without access to participants' instantiated use of language, or to participants themselves).
