An Ontological Approach to Canonical Typology: Laying the foundations for e-Linguistics

With the rise of the Web as an important medium for displaying and curating typological data, the next generation of linguists will have access to thousands of data sets on many of the world’s remaining languages. It is not out of the question that many or most of the grammars ever published and possibly many sets of field notes will be digitised and stored in a Web accessible format (cf. Google Book Search). As more and more resources are digitised, there is an ever-increasing need for interoperability of data and transparency regarding the intended meaning of annotation elements. Standardisation of terminology is certainly one approach; however, there are more basic challenges, including a clear articulation of a conceptual framework for specific areas of linguistics including, among many others, canonical typology. Put another way, there is currently no broadly accepted logic or computer language that would allow arbitrary descriptive data to be automatically published, searched, and visualised, i.e., no possibility of an e-linguistics. In this paper, we explore laying the foundations for such an enterprise by drawing on recent and emerging Web technologies, as well as from applied formal ontology. In particular, we will address work within the GOLD Community of Practice whose aim is the creation of a General Ontology for Linguistic Description. We will argue that the resulting conceptual framework can be employed to solve some basic issues concerning an implementation of canonical typology. An OWL-DL implementation of GOLD is presented with relevant examples for canonical typology.