Documenting Raising and Control in Moro

Himmelmann (2006) defines language documentation as the creation of “a lasting, multipurpose record of a language” (p.1). With such a broad definition, description and analysis of linguistic structure form vital components of language documentation, and this process is in turn informed and enriched by linguistic theory (Sells 2010). In this paper, we show that the semantic and syntactic diagnostics from formal theories of grammar enable a nuanced understanding of embedded clauses in Moro that would not be possible without such tests. Specifically, these tests detect the distribution of various types of raising and control predicates in Moro. Such tests are especially important, as the identification of control and raising predicates are often completely neglected in grammatical descriptions and documentary work,1 even while they have played a central role in the development of syntactic theory for over forty years (Davies and Dubinsky 2004). By supplying theorists with additional data, documentarians can help shape future developments in theoretical topics like raising and control, which have sometimes been focused on a restricted set of easily accessible languages (Polinsky and Potsdam 2006). In this paper we show that the superficial morphological categories that mark embedded clauses in some cases correspond quite neatly to the complements of raising and control predicates. In other cases, however, the connection is less direct. Distinguishing these cases, and the connections between complement types and raising and control, must make use of data which can most reliably gathered via elicitation, including negative judgments. Such in-depth elicitation is only possible due to a solid analytical foundation in the phonology, morphology and basic syntax of the language, information which has been accumulated through extensive prior documentation and description. Thus, there is no doubt that the lines between documentation, description and formal theoretical analysis are blurred and these different enterprises are complementary. While documentary approaches have rightly encouraged theorists to look beyond elicitation as a basic tool of data collection, elicitation guided by linguistic theory also permits insights which are difficult to establish solely on the basis of texts. Thus, the study of endangered languages is most effectively conducted with the benefit of a solid foundation in linguistic theory. This paper is part of the research of the Moro Language Project hosted by UC San Diego, a project which was funded by the National Science Foundation (Grant No. 0745973) from 20082013, and has focused on in-depth data analysis of all aspects of the language. Moro is a Kordofanian (Niger-Congo) language spoken in the Nuba Mountains area in South Kordofan State in the