In this talk I focus on the instrumental function commonly attributed to heritage languages (HL) as a positive factor in literacy skills development in local languages (often school languages, SL). Many scholars and policy makers evoke interdependence and/or transfer between languages when arguing for the importance of heritage language development. Such transfer-oriented rationales usually involve claims of beneficial effects of skills in the HL on skills in the language of the host country.
I discuss theoretical and methodological questions that arise in the empirical investigation of interdependence. Whereas most variants of the interdependence framework presuppose causal influence between language skills, the nature of the data collected in typical inquiries into heritage language speakers’ repertoires does not easily allow us to draw causal inferences.
I discuss the example of a longitudinal study carried out in Switzerland focusing on the acquisition of reading and writing skills in bilingual Portuguese-German and Portuguese-French primary school children (3rd – 4th grade). I discuss possible hypotheses that are derived from the interdependence idea and ways of testing them empirically. The dependent variables analyzed are global assessment scores for written texts, vocabulary assessment of these texts (lexical diversity and sophistication) and reading comprehension scores. Descriptive and inferential analyses of the linear associations of these variables across languages are discussed, as well the problems that emerge when the goal is to disentangle manifestations of transfer from the impact of other variables affecting the skills both in HL and SL.
The talk ends with a discussion of the conceptual vagueness of commonly used notions of interdependence and ways to clarify the notion and put it to the test more rigorously using experimental approaches and better statistical modelling techniques.