Can Gold " Cope " with Wals? Retrofitting an Ontology onto the World Atlas of Language Structures

0. Introduction The World Atlas of Language Structures (WALS, Haspelmath et al. 2005) is a large-scale “database of databases” consisting of 141 typological databases, covering a wide range of grammatical features, joined into one composite resource through the use of a common metadata scheme. While this metadata scheme ensures interoperability among databases across some dimensions (e.g., language names and families), it falls far short of allowing complete database interoperability. At present, a project is underway to “retrofit” an ontology onto this existing resource. Two broad questions being addressed by the project are: (i) What conceptual and design problems need to be solved in order to build an ontology “internal” to WALS which can allow for a high degree of interoperability among the WALS databases? and (ii) How can the WALS categories be related to a general ontology? Or, to put it another way, we are interested in determining (i) how we can build a worthwhile Community of Practice Extension (or COPE) for WALS and (ii) how the categories in this COPE can be related to categories in the General Ontology for Linguistic Description (henceforth, GOLD ontology; Farrar and Langendoen (2003)). This paper is structured as follows. Section 1 gives some background information on WALS, including discussion of its history and its overall design. Section 2 discusses the basic methodology adopted by the WALS ontology project. This section also discusses some of the research results of the WALS ontology project which we believe may be relevant to the conceptual relationship between a typological COPE and the GOLD ontology. Section 3 discusses some ways in which we believe GOLD could be extended to better relate to typological concepts and some design desiderata for ontological tools which would facilitate the creation of an ontology like the WALS ontology. Finally, section 4 summarizes the current findings of this ongoing project.