The linguistic competence of second-generation bilinguals

Abstract

This paper discusses the native linguistic competence of second-generation bilinguals born of immigrant, refugee, expatriate, or otherwise dislocated parents, concentrating on the grammars of second-generation Hispanics in the U.S. Scholarly opinion has gravitated toward the position that the Spanish of these speakers reflects a process of incomplete acquisition. This paper invites examination of the alternative view, namely: what we observe in second-generation bilingual Latinos is not errors, as they are frequently described in the literature, but rather points of divergence between their Spanish and that of the previous generation, due to normal intergenerational language change accelerated by conditions of language contact. The notion of incomplete acquisition rests on an incorrect view of child language acquisition as a process of perfect reproduction of parental grammars. But the process is one where children engage in grammar construction through hypothesis testing. Consequently, all next-generation grammars end up somewhat different from parental ones, paving the way for language change. The grammars of U.S.-born Latinos are thus, like all next-generation grammars, different, not incomplete. Examining the use of subjunctives for a brief illustration, grammarians regularly note variability in cases like Quizás venga (subjunctive) ~ Quizás viene (indicative) but obligatoriness in Quiero que lo llames (subjunctive), with the indicative alternative occurring seldom or never, and analyzable as ungrammatical, *Quiero que lo llamas (indicative). Second generation bilinguals have extended variability, so that for them the latter is usable, and analyzable as grammatical, a fact that disables these bilinguals from successful participation in experiments centered on somebody else’s grammaticality judgments, leading to conclusions of incompleteness.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 48 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.