La Sémantique du prédicat en Mwotlap (Vanuatu) (review)

Mwotlap (better known as Motlav) is an Austronesian language spoken by some 1,800 people on the island of Motalava in the Banks Islands, northern Vanuatu. It was first described by Codrington in The Melanesian languages (1885:310-22), where the neighboring language Mota, the lingua franca of the Anglican Church in the southeast Solomons and northeast Vanuatu until 1931, is the focus of attention (253-310). Mwotlap is also mentioned briefly in Ray (1926:427-40). François discovered Motalava for the first time in 1997 and has worked constantly on the language ever since, producing, inter alia, a series of children's books in Mwotlap.

In this book, François has developed and considerably expanded chapter seven of his huge doctoral dissertation (Le Mwotlap, University of Paris IV-Sorbonne, 2001, 1,078 pages). He himself admits to a fascination with the semantics of predicates, and particularly the mechanics of tense-aspect-mode marking, influenced by the theoretical orientation instilled in him by Antoine Culioli. Chapter 1 (1-26) describes Mwotlap structures in summary form, sufficient to follow the analysis that follows in chapters 3 to 8, which constitute the bulk of the book. Chapter 1 presents the basic phonology and morphology of Mwotlap, and the morphosyntax of simple and complex sentences.

In chapter 2 (27-43) François discusses the general mechanics of the TAM (tense-aspect-modal) system in Mwotlap, stating (28) that he will adopt an empirical approach rather than presenting his data in the framework of any particular linguistic theory. As he sees it, there are both single and discontinuous TAM markers. In addition he decides that it is necessary to set up a zero morpheme to cover the case of the aorist. François lists the twenty-six TAM markers in the language as shown in table 1 overleaf.

Chapter 3 (45-75) treats grammatical categories and aspectualify. François demonstrates that in Mwotlap, as in most Austronesian languages, almost any grammatical category can act as a predicate. However, the main question he asks is why certain categories and not others are compatible with aspect and mode markers. Chapter 4 (77-164) exhaustively describes seriatim all the TAM markers listed above that are associated with realis, that is, the real, whether that be present or past in relation to the situation described. These markers are: stative, completive, preterite, completed action, continuous, and presentative.

Chapter 5 (165-216) deals with two TAM markers that are not intrinsically associated with either the realis or irrealis, since they may cooccur with both, according to context and the structures with which they are associated. The first of these is the aorist, the most commonly encountered in discourse. The second marks temporal focus. Both of these markers oscillate between the past, the present, and the future, unlike any of the other Mwotlap TAM markers. Chapter 6 (217-312) discusses a series of TAM markers that are intrinsically associated with irrealis, situated in the virtual rather than the real world. These markers are more charcateristically modals (e.g., wishing, desiring, seeming). François is adamant that it would be erroneous to match realis/irrealis with past/future, since Mwotlap is tenseless (a point to which we will return presently). No TAM marker can be unambiguously situated in time; so, for example a "future" could well be used as a "future in the past" or "future perfect."

Chapter 7 (313-40) describes the negative TAM markers in Mwotlap. In this language there is not a single marker that is applied to form all realis and irrealis negatives. As indicated in the table above, also, there is no one-to-one match between affirmatives and negatives (indicated by the gaps in the table). Chapter 8 (341-76), the final chapter, sums up the principal characteristics of the Mwotlap TAM system, namely that there are twenty-six TAM markers in Mwotlap, 19 affirmative amd 7 negative, forming a single paradigm...