Towards standardizing and sharing bilingual data

We describe a project (the LIPPS project) whose purpose is to set up a computerized database of bilingual texts to be used by researchers in the field of "language interaction" (i.e., codeswitching, borrowing, and other outcomes of contact between varieties). Current work includes an adaptation of the CHILDES system (MacWhinney,1995) to take account of the different needs of researchers in this area, for example, solving the problems of distinguishing, coding, and representing language interaction phenomena; providing a gloss / translation in user-friendly format; and developing qualitative as well as quantitative tools for comparing data-sets. Problems and some advantages of such a comparison are illustrated through a pilot project conducted by Gardner-Chloros using two sets of data: Greek-Cypriot/English data and Punjabi /English. Practical problems included giving transcribers/coders the necessary training in using the system and ensuring a consistent approach. Coding problems derived from interlinguistic differences were also highlighted. Clear advantages ensued, however, from being able to compare the type/quantity of codeswitches and so forth across data-sets, as this allows an assessment of the relative weight of linguistic and sociolinguistic factors in determining the form of language interaction.