Te reo Māori is the indigenous language of New Zealand, and is unusual from a linguistic perspective in that it contains an unusually high proportion of both passive constructions and passive alternations (Bauer et al., 1993). Passive constructions occurr much more frequently in te reo Māori than in many other languages, including English (Harlow, 2007); indeed, passive forms of verbs commonly appear more often than their active forms, particularly historically (Sanders, 1990). The te reo Māori passive morpheme also contains an unusually high number of passive alternations (Bauer et al., 1993), so it is therefore not surprising that this aspect of the language has attracted significant interest from linguists (Bauer et al., 1993; De Lacy, 2004; Hale, 1968; Harlow, 2007; Parker Jones, 2008; Sanders, 1990).

The extensive variation in the passive allomorphs can make the selection of an appropriate passive suffix more difficult for speakers, particularly when speakers are under-exposed to passive forms through low-frequency verbs, the declining use of passive constructions, or novel words (Sanders, 1990). This uncertainty in the face of suffix selection can in turn lead an increase in free variation across passive allomorphs, and can also lead to speakers being more likely to revert to a more rule-based approach to suffixation through the use of ‘default’ suffix (Sanders, 1990). These issues conspire to make it difficult to predict which passive allomorph will surface in a given environment.

This thesis explores how proficient re reo Māori speakers approach passivisation, testing the predictions made by previous explorations of passive variation, through both corpus analysis and in an experimental setting. A review of the relevant literature is presented in chapter 2, which identifies the predictions made by previous work. The testing of these predictions on corpus data is presented in chapter 3, the results of which suggested that a number of outstanding research questions still remained. The exploration of these questions is the primary focus of this study, and the methods by which this was conducted are outlined in chapter 4. Chapter 5 presents the results of this exploration, which found that a number of different factors are implicated in the use of different passive alternations, operating differently at different degrees of novelty. The results are discussed in respect to the literature in chapter 6, finding that effects are complex and varied, and show a high degree of interdependence. The conclusion of this thesis is presented in Chapter 7.