Uniting software tools for the archival, management and analysis of linguistic data: lessons from developing the Language Data Repository

Abstract

The complex tasks involved in both the production and use of linguistic data tend to be highly repetitive and tedious. These are the tasks at which computers excel, and at which humans perform very poorly. In this thesis I will begin by describing the current state of the art in computational support for the documentation and description of language. To do this I will outline the three major areas in which computers are currently being used to support linguistic research: data archiving, data analysis and data management, discuss some of the current projects and tools in each of these areas, and present a summary of work currently being conducted by the Open Language Archive Community to support open web based access to the resources available to the linguistic community. Following this I will describe some of the limitations presented by the current approach to this field and present a vision for system that will support analysis, management and archival of linguistic data in a single "universally" accessible system, providing specific examples from the system I have been working to develop, the Language Data Repository (LDR). I will conclude by laying out areas of future work, both for the development of the LDR and for the community in general.