Clause types in Gooniyandi

Abstract

Systemic theory generally abhors universal claims about language. However, Halliday has suggested in various places that all languages will have three major clause types: material process, mental process, and relational process. This seems to be borne out by some systemic analyses of languages other than English. In this paper I dispute this claim using evidence from Gooniyandi, a language from Western Australia. To begin with, as I have shown elsewhere, relational clauses in Gooniyandi are not “process” clauses. Thus, there is a primary distinction between process (or situation, as I prefer to call them), and relational clauses. The former refer to situations — occurrences, processes, happenings, and so on, whereas the latter establish logical type relationships such as attribution and identification, without mediating these relationships through processes of being, as is the case in English. In this paper I turn attention to the situation clauses, demonstrating that there is no language internal evidence supporting a distinction between mental and material processes in Gooniyandi. Instead, I suggest that there is a primary ternary distinction within situation clauses between active, behavioural, and receptive situations. The former, active clauses, include what are, from the point of view of English grammar, material and mental processes. Close attention is paid to argumentation and methods of justification. These are particularly important in the present context, given the fact that not all sytemicists accept Halliday's analysis, even for English, and that linguists from other traditions have suggested conflicting universals (e.g. Dik 1989), using somewhat similar arguments and criteria. In raising the general issue of justification of linguistic analyses, I also address problems of indeterminacy and conflicting evidence.