DMLX · Demolinguistics and Language Revitalisation

Final Activity Report Summary - DMLX (Demolinguistics and Language Revitalisation)

The future of bilingual communities depends on many factors, from the histories and politics of entire societies, through their sociology and demography, to the psycholinguistics and sociolinguistics of individuals and interactions. No single approach can have much predictive value. Demographic projections can be rendered worthless by changes in educational policy or by political or attitudinal change. Projections based on language attitudes and current language practices can be nullified by large population movements. And the targeted projects of language planners are notorious for falling short of their goals because of budget changes, uncertainties in implementing educational policies, and the difficulty of monitoring short-term progress. Thus multiple approaches are required to achieve a coherent vision of the future dynamics of language shift, including technical, historical, sociolinguistic and political aspects of language shift and language revitalization around the world. I have been working at bringing a range of new tools from different disciplines to the study of language knowledge and use, including new statistical protocols for complex types of data. This has involved obtaining, accessing, extracting data from, and analysing large data banks, including census data and linguistic survey data, interacting with national census bureaus and language planning offices in six Autonomous Communities in Spain as well as the central government, in Wales, Scotland, England, Ireland and France, as well as with Eurostat, and have developed working relationships with academic colleagues in the field in Valencia, Catalonia, Wales, Ireland and Canada. There being no accepted protocol for projecting language shift from census data, other than extrapolation of existing trends and allowing for worst case and best case scenarios. I developed and validated an entirely new projection method based on mathematical models for L1 transmission in the home, L2 acquisition in the school and linguistic integration of immigrants, based entirely on census and other official data, grafted onto elementary demographic projections. I have tested and refined this in eight different bilingual contexts in Spain and the UK. At the same time I have been exploiting the various surveys of context-determined language use carried out by the community language authorities. I developed techniques to render these surveys comparable and to situate the various communities in a two or three-dimensional space, with axes measuring at least the familiar versus unknown status of participants and the official versus commercial nature of the interaction. I have studied how to confront patterns of competence with patterns of use in order to incorporate both into projections. Neither predicts the other. A community with high levels of competence may be in the process of rapidly losing its proportion of speakers through loss of L1 transmission or L2 acquisition. Two communities with superficially the same patterns of competence may be heading in entirely opposite directions as far as future generations are concerned. Some of this may be explained by usage patterns, especially with respect to the familiar versus unknown participants dimension. But the key problem for future research that emerges from my study is to model the feedback between usage and competence. The solution to this is outside both types of data, in the political and linguistic attitudes of the population. These are unpredictable, in the mid- and long term, but incorporating them in the projection model will allow testing their effects on usage and competence patterns.

Data: CORDIS, © European Union