ABSTRACT. This article presents preliminary results from two long-term, mixed design qualitative and quantitative studies of Spanish language usage in the Lower Rio Grande Valley of South Texas. In these two research series, interview data were collected and the use of Spanish was analyzed using critical discourse analysis. In order to better understand the patterns of language maintenance and shift in this region, an expanded model of the GENERATION construct is presented. The analyzed data are then discussed employing the expanded generation model. The Spanish of first through fifth generation consultants is described, as they move away from the immigration experience. Although results are preliminary, clear distinctions among the five generations are observed. Further research and analyses are subsequently suggested. *
INTRODUCTION. Research on the contact situation between Spanish and English in the U.S. has consistently documented an intergenerational shift from the former to the latter (see e.g. Lopez 1978, Veltman 1988, Bills 1989, Sole 1990, Pease-Alvarez 1993, Bills, Hernandez-Chavez, & Hudson 1995 and Rivera-Mills 2001, inter alia). In general terms, researchers have documented the unidirectional shift to English in this nation for many populations from various regions around the world, often between the second and third generation. However, societal language shift away from Spanish to English does not appear to follow the traditional three generation pattern in the Lower Rio Grande Valley of Texas (LRGV), perhaps due in part to the large Spanish speaking population encountered there, which continually increases as a result of the arrival of first generation immigrants (cf. 2000 U.S. Census and updates at www.census.gov). Studies such as those of Mejias, Anderson-Mejias and Carlson (2002, 2003) have supported this change in the pattern of shift. Thus, the continuum paradigm suggested in works by Kouritzin (1999) and Wong Fillmore (1991), in which intergenerational language loss occurs subtly over a long period of time, may account for this slowed loss.
In order to further explore this phenomenon, the research offered here presents preliminary results from a long-term, mixed design qualitative and quantitative study of language usage in the LRGV. Works by Milroy (1980), as well as the standards in sociolinguistic language data collection established by scholars such as Labov (1972) and Hymes (1974), informed the practical techniques utilized in the data elicitation for this study. In the initial phase of the project, trained Hispanic research assistants under the direction of the principal investigator (PI) began collecting information on the topic of DICHOS and CORRIDOS, and included personal narratives. A subsequent study built on the original data and elicited speech samples employing an expanded instrument. Recorded language generated from the two data sets were analyzed using various procedures of critical discourse analysis (Fairclough 1995), including text and grammatical analysis. Results suggest that the variable of generation in South Texas clearly impacts Spanish language use there, and is the topic of this article.
1. DEFINITION OF GENERATION. Discussions in the literature on language shift across generations frequently present the concept of GENERATION as a homogenous construct. That is, the 1st generation uniformly represents foreign born immigrants, with subsequent generations of the children and grandchildren of the first generation born in the U.S. However, for the purposes of this study I define generation on a broader scope; the term does not refer to a uniform concept but rather to a possibly variable order of birth among family members, which may be non-uniform in nature. Thus, for this investigation, generation is defined as follows:
1st generation: consultant was born outside the U.S. 2nd generation: consultant was born in the U.S. with either one parent or both parents born outside the U. …