The effect of multilingualism on acquisition of genericity in the L2

 

Generic statements convey generalizations about the world (Krifka et al., 1995). Generic meanings are expressed in all human languages. In addition to the two main generic meanings (kind and characterizing), we explore type-denoting (TD) generics, (Borthen 2003) (all three examples in Table 1).

Acquiring genericity expressions in additional languages presents difficulties, as learners must reconfigure form-to-meaning mappings (Table 2). Previous empirical studies have demonstrated that the L1 influences the understanding of genericity in L2 (Ionin et al., 2011; Snape, 2013).

In the current study we sought to determine if multilingualism impacts the comprehension of nuanced differences in L2 English generic meanings. Our participants were Norwegian native speakers (n = 23) and two distinct groups of Polish speakers: bilinguals (n = 22) and multilinguals (n = 24). The multilingual participants were enrolled in L3/Ln Norwegian classes. Three singular forms of nominals (NP) were tested: the definite, indefinite, and bare singular. Norwegian uses all three forms for expressing different generics, while Polish only employs the bare singular form (Table 2).

The task was a contextualised acceptability judgment task (AJT) distributed online, with the target sentence expressed with one of the three singular NPs. The Polish multilinguals also completed the task in Norwegian; we found that the form-to-meaning mapping had not yet been acquired in the L3. Proficiency in the target L2 was assessed through filler items (grammatical/ungrammatical and comprehension questions). All L2 groups differed significantly from native speakers; however, the multilinguals scored the lowest.

Acceptability choices on each NP form (Figure 1) were analysed using glmer models (Bates et al., 2015) with response as the dependent variables (group and condition as  independent variables). Participant and test item were set as random effects. We observed minimal differences in how the NP forms were accepted across conditions by the Polish bilingual group, but more pronounced differences in the Polish multilingual group. The Norwegian group demonstrated strong target-like choices, signifying a comprehensive understanding of L2 genericity marking.

The analyses reveal that the Norwegian speakers largely pattern with the native English controls. This can be attributed to L1 influence, as Norwegian has genericity form-to-meaning mappings similar to English. Conversely, the Polish speakers differ from the native controls in their acceptance of forms across all generic conditions. These differences are more pronounced for the bilinguals, suggesting that the multilinguals had a better understanding of the subtle meaning differences. The observed differences cannot be attributed to proficiency (since the bilingual group was stronger on the filler items), nor to transfer from the L3 in the multilingual group, as the Norwegian generic mappings had not yet been acquired.

We conclude that multilingualism may positively influence L2 acquisition, when knowledge of an additional language enables learners to better recognize complex genericity marking. This finding has implications for educators and how we perceive the influence of all languages in the mind: influence is not excerpted chronologically (earlier acquired  later acquired), but multilingualism as a whole influences acquisition, and can be seen as an advantage.