Geographic, Cultural, and Ecological Correlations with Indigenous Language Vitality in North America

Abstract

Over the last several decades, linguists and community language activists have expressed concerns about language loss in the modern era, citing threats to linguistic and cultural diversity and to knowledge systems that may never be recovered once they disappear. Some point out the human rights violations and the impact on families and communities who no longer have access to language-related components of their heritage. Some also argue that documenting languages that are likely to stop being used in the near future is important to the field of linguistics as a science, as it allows linguists to capture information about languages and language use that may otherwise be lost. These arguments are true, and yet there is a component missing from the picture: there is not yet a clear and developed science that integrates the causes and contexts for language loss with interventional measures undertaken by communities into a cohesive explanatory framework that can make accurate language vitality predictions and possibly even recommendations. This dissertation has been undertaken with that eventual goal in mind. This dissertation seeks to better understand language vitality in North America by examining cultural, geographic, and ecological factors that may be causally related using quantitative methods, specifically ordinal regression, with the hopes of contributing to the body of research already under way that applies a similar set of questions to the global context. In particular, it includes six factors—historical land area, elevation, subsistence strategy, displacement, accessibility, and revitalization efforts—in an ordinal regression model, with language vitality as the dependent measure. Two new databases were created to facilitate this exploration. Due to the limited availability of data for some variables, twelve models were run on six data subsets to include as many data points as possible for each variable. An interaction was included between two of the variables. When the interaction effect is taken into account, all variables that were included as predictors turned out to have statistically significant correlations with language vitality in the top-performing models in which they were included. Larger historical land areas were correlated with higher language vitality, as were traditional agricultural practices in places where the historical land area had a greater elevation range. In the continental United States, languages with modern-day reservations or Native land areas covering both part of the historical land area and land area outside of that have higher vitality than languages with no modern-day reservations. Highway presence in these modern-day land areas is associated with higher language vitality levels, and all types of community-driven language revitalization efforts show a strong correlation with higher language vitality compared to languages that have none. Bilingual education and language immersion programs oriented toward children show a particularly strong correlation with increased language vitality. Part of the goal of this study was to determine whether a quantitative approach could provide evidence of the influence of these types of factors on language vitality. While some language-external factors may increase the risk of language loss, others may help to protect against it. The results of this study corroborate other studies using similar methodology applied to global language samples, suggesting that the methodology does indeed work to produce results that capture a component of the wide range of factors that may be influential. However, this dissertation only examined six independent variables, and there are many more that might be important to consider. Some of these shortcomings are discussed in the conclusion, along with suggestions for future avenues of research.