Convergence via code-switching? Yo and I in New Mexico

New Mexican Spanish is the oldest surviving Spanish dialect in the US, where it has been spoken for approximately four centuries, for the past 150 years, in intense contact with English. The effect of this contact in terms of widespread borrowing and code-switching has been well documented, but the grammatical effects have yet to be studied in detail. In this paper we investigate whether convergence between Spanish and English is taking place through an analysis of variable first-person singular subject expression in a corpus of spoken New Mexican Spanish. We further address the question of whether code-switching between the two languages promotes convergence, as has been claimed by some scholars of language contact (e.g. Gumperz and Wilson 1971; Thomason 2001, inter alia). We use the comparative variationist method (Poplack and Meechan 1998; Tagliamonte 2002), comparing bilinguals' patterns of variable first-person singular subject expression with those of Spanish-dominant speakers, as well as those reported for monolingual Spanish varieties (Silva-Corvalán 2001; Travis 2007). We make a further comparison within the bilingual data across contexts where code-switching is, and is not, present in the preceding discourse. Our results demonstrate that first-person singular subject expression in New Mexican Spanish follows the same patterning as monolingual varieties, and that this is the case regardless of the degree of bilingualism. We do find a slightly higher rate of expressed subjects in the presence of code-switching (41% vs. 35%), however analysis reveals that this is attributable not to the code-switching per se, but to the presence of an expressed first person singular subject I in the preceding English, which we interpret as a cross-linguistic priming effect. We thus find that the data do not support Spanish convergence with English in this variety, nor code-switching as a mechanism of contact-induced change. We conclude that the study of the linguistic consequences of contact must rely on comparison of variable patterns, including consideration of priming effects.