Tiwi revisited: a reanalysis of traditional Tiwi verb morphology

© 2013 Aidan WilsonTraditional Tiwi is a language isolate within the Australian language group, traditionally spoken on the Tiwi Islands, north of Darwin. This language exhibits the most complex verb structure of any Australian language. Altogether there are 18 distinct verb slots; 14 prefixes and 4 suffixes. They encode subject, object and oblique arguments, they inflect for tense, aspect and mood, the location and direction of events with respect to the speaker, and the time of day that an event takes place. They also take prefixes and suffixes denoting associated motion, can be argument-raised by a causative or detransitivised by derivational morphology, and can take incorporated nominals, incorporated verbs, and incorporated comitative or privative arguments. Traditional Tiwi has not been adequately described. Previous descriptions are limited and do not cover verb morphology with enough detail. This thesis brings together previous descriptions, early recorded data, and adds newly collected data and findings to produce an updated description of the language, with special reference to the verb morphology. I focus in particular on two aspects of the verb morphology: agreement and incorporation. The Traditional Tiwi agreement system of inflecting verbs shows a high degree of complexity due to the interactions between subject, object and tense marking. I argue for the occurrence of an otherwise unreported phenomenon by which agreement affixes can shift between various controllers depending on the morphosyntactic context. Incorporation is also highly complex, as with other northern Australian languages that exhibit this feature. There are four distinct types of incorporation including verb incorporation, comitative and privative constructions, body part incorporation and regular nominal incorporation. I describe these with reference to incorporation phenomena in other Australian languages. Traditional Tiwi however, is no longer spoken; the last two speakers died in 2012. The majority of the data on which this thesis is based was collected with one of these two last speakers, and therefore represents possibly the last documentary linguistic record of this important language