Maximizing the explanatory power of LHQ  data
for the study of heritage populations

There is a need for scholars in the field of heritage language research to provide more detailed and meaningful characterizations of their heritage speaker samples. While Language History Questionnaires are usually used to characterize sample, they often produce a number of highly correlated variables, which leads researchers to “cherry pick” variables based on their own judgement. Here, I present Principal Component Analysis as an empirical method to generate evidence-based characterizations of heritage speakers by examining which language-experience related factors best explain the variability inherent to a sample.

​

The questionnaire created for this study consisted of a total of 70 questions; however, we will only focus on those relevant to the analyses carried out. Based on the extant literature, there were 40 relevant questions divided into four dimensions:

 

1. Heritage Language Background – 15 questions. 

2. Contexts of Use – 9 questions.

3. Interactional Network – 7 questions

4. Attitude, Motivation and Identification – 9 questions

​

A group of 45 New Mexican heritage speakers of Spanish participated in this study (27 females, mean age 20.13 years old, SD= 1.67, range= 4). All participants were in the process of completing a bachelor’s degree at the University of New Mexico at the time of data collection; participants were recruited from a range of upper-level courses. The PCA conducted yielded the following results:

​

​

​

 

 

 

 

 

 

 

 

 

 

 

 

To complement this, I also proposed a theoretical approach from which to dynamically understand how variation along language-experience related factors shape heritage grammars:

​

Achieving a more meaningful characterization of heritage speakers and their grammars.

Scholars within the field of heritage language studies have made substantial efforts to produce a definition of the term heritage speaker that is flexible yet concrete enough to have descriptive power. A thorough survey of the literature shows that definitions of the term heritage speaker pivot around two main dimensions: proficiency and cultural identification. Definitions based on proficiency, also often include notions of language dominance. For example, Valdés (1997: 38) defines a heritage speaker as “someone […] who speaks or merely understands the heritage language and who is to some degree bilingual in English and the heritage language” (see also: Benmamoun, Montrul & Polinsky, 2013; Carreira & Kagan 2011; Wiley & Valdés 2000). In contrast, definitions based on cultural identification describe heritage speakers in terms of their cultural ties to the heritage language. For instance, McCarty et. al., (1997: 20) define heritage speakers as “individuals who are members of a community with linguistic roots in a language other than English”(see also: Carreira, 2004; Cho 2000; Fishman 2000; Van Deusen-Scholl 2003).  

​

The variety and lack of concreteness of these definitions raises the possibility that defining heritage speakers based solely on their linguistic knowledge and cultural identification may not be very useful. In fact, the incredible variability of heritage grammars and the extraordinary diversity of heritage speakers’ linguistic experience challenge the idea that heritage speakers can be meaningfully categorized solely in terms of their grammars at all. The terms heritage language, heritage grammar and heritage speaker actually refer to a wide range of realities, defined not so much through characteristics of the constructs themselves as through those of their environments and circumstances. For this reason, a better understanding of the socially-situated process of heritage language acquisition and use would allow for more meaningful characterizations of heritage grammars and their speakers. Here, I propose a usage-based dynamic approach according to which acquisition and use of the heritage language vary along the following dimensions (López-Beltrán & Carlson, 2020):

​

• When: this dimension is comprised by factors related to the age of acquisition and years of use of the heritage language.

• Where: this dimension is comprised by factors related to the contexts within which the heritage language is used and acquired. It also includes aspects pertaining to literacy.

• Who: this dimension is comprised by factors related to the interactional networks within which the heritage language is used and acquired.

​​

Under this approach, a heritage speaker is but an individual who acquires a heritage language and the linguistic knowledge resulting from their acquisition process is a product of variation along the dimensions explained above. Variation along the when, where and who dimensions of the heritage language will determine the variety of constructions and vocabulary that heritage speakers are exposed, directly impacting their grammatical representations and processing strategies.

 

Conclusion

There is a need for scholars in the field of heritage language research to provide more detailed and meaningful characterizations of their heritage speaker samples (Benmamoun, Montrul & Polinsky, 2013; Meisel, 2019). More nuanced descriptions and differentiations of the different heritage populations studied will directly impact the interpretation and extrapolation of results across studies, as well as the theories derived from them. Here, I have presented PCA as an empirical method to generate evidence-based characterizations of heritage speakers by examining which language-experience related factors best explain the variability inherent to a sample. To complement this, I have also proposed a theoretical approach from which to dynamically understand how variation along language-experience related factors shape heritage grammars. Hopefully, these tools can contribute towards a more unified and accurate account of language acquisition and processing in heritage contexts.

​