Gender Assignment to English-Origin Nouns in the Spanish of the Southwestern United States

Abstract

ABSTRACT. This article analyzes the process of assigning gender to English-origin nouns. The two prevailing views are that the process is either done by default where masculine gender is simply assigned to all nouns or a natural process where gender is assigned using the same criteria as that of native Spanish nouns. To test the two hypotheses we used the comparative method to study the gender patterns of Spanish nouns and lone English-origin nouns in the Spanish of bilinguals living in New Mexico. Each of the nouns was analyzed for the factors of biological gender, synonymic gender, and the gender associated with the terminal phoneme. The results indicate that biological gender is categorical with animate nouns. When there is no biological gender, the gender of the terminal phoneme is the most accurate predictor. The results showed that borrowed nouns pattern morphologically like Spanish nouns supporting the Nonce Borrowing Hypothesis (Sankoff et al 1990) and the categorization of these nouns as borrowings rather than code-switches. 1. INTRODUCTION. A defining characteristic of casual speech among bilinguals in the southwest region of the United States is the mixing of Spanish and English. Due to the contact that has existed between these two languages for more than a century in this region, many speakers have similar proficiencies in English and Spanish. As a result, bilingual speakers frequently incorporate both languages within discourse with other bilinguals. Language mixing involves two major processes, borrowing and code-switching. 1.1. LEXICAL BORROWING. As defined by Poplack, 'borrowing is the adaptation of lexical material to the morphological and syntactic (and usually, phonological) patterns of the recipient language' (1993:256). Borrowings (also referred to as loanwords in the literature) are further broken down into two major groups--established loanwords and nonce loanwords. The difference between the two is extralinguistic in nature in that while established loanwords 'typically show full linguistic integration, native-language synonym displacement, and widespread nonce loanwords 'need not satisfy the diffusion requirement' (Poplack 1993:256). In other words, nonce loanwords do not frequently become a permanent part of the lexicon. Consider, for example, the differences between established loanwords (shown in example (1) with the noun troca 'truck') and nonce loanwords (shown in example (2) with norse 'nurse'). (1) Yo la vi, i, una troca verde? (07/A/255/10) 'I saw it, a green truck, right?' (2) A mi me gustaria mucho haber sido una norse, pero no me gusta ver la sangre. (02/A/291/02) 'I would really like to have been a nurse, but I don't like to see blood.' The English-origin loanwords shown in examples (1) and (2) are fully adapted to the patterns of the recipient language, unlike code-switches, which remain unadapted during bilingual speech. Since troca is used in numerous dialects of Spanish in the southwestern United States as well as in Mexico and has gradually become a stable component of the lexicon, it must be considered an established loanword. The word norse, on the other hand, occurs sporadically in local dialects of Spanish and, due to this lack of diffusion, must be considered a nonce loanword. 1.2. CODE-SWITCHING. Unlike loanwords, code-switches remain unadapted to the recipient language. Code-switching has been defined as the 'juxtaposition of sentences or sentence fragments, each of which is internally consistent with the morphological and syntactic (and optionally, phonological) rules of the language of its provenance' (Poplack 1993:255). It can occur either between phrases, referred to as INTERSENTENTIAL SWITCHING, or within a single phrase, known as INTRASENTENTIAL SWITCHING. The following examples, taken from sociolinguistic interviews conducted in Mora, New Mexico, illustrate both types of code-switching: (3) Si I think you're right. …

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 — 18 citations exist upstream.