the sample corroborates the common belief that that ergative languages have antipassives with greater chances than nominative-accusative languages, the author suggests that this is not a direct correlation, but rather a consequence of the fact that both antipassives and ergativity are favoured in languages with rigid transitivity classes. In sum, this is a very comprehensive study, both in breadth and in depth, which offers a wealth of new data and insights and should become a standard reference on antipassives. 4 Dana Louagie. 2017. A typological study of noun phrase structures in Australian languages Katholieke Universiteit Leuven Supervisor: Jean-Christophe Verstraete This dissertation presents a study of noun phrase structures in Australian languages based on a sample of 100 languages. The analysis is developed in two main parts. The first part of the dissertation presents a general survey of NP features, developing a synthesis of the available Australianist literature, testing some of its ideas on the languages of the sample, and showing where Australian languages stand in relation to other languages in the world. Chapter 1 deals with nominal classification, which is the best-described aspect of NP structure for Australian languages. Chapter 2 discusses the domains of qualification and quantification, which have received less attention in the literature, and chapter 3 introduces the domains of determination and NP constituency, which are most poorly understood. The second part of the dissertation presents a more detailed analysis of the last two aspects, determination and NP constituency, in the languages of the sample. In Chapter 4, on NP constituency, Louagie concludes that there is in fact no strong evidence against constituency, contrary to what has been traditionally claimed in the Australianist literature. More generally, it is shown that constituency is not an absolute value that can be applied to languages as unitary wholes, but rather a matter of degree. Chapter 5, on determiners, likewise challenges the received view that Australian languages lack determiners. Interestingly, Louagie shows that a determiner slot can be filled by a range of structurally different elements, which share the functional feature of identifiability. This approach is cross-linguistically applicable to languages with and without ‘classic’ determiner systems. This thesis is very clearly structured and reads easily. The analysis and presentation of the data is very transparent and conscientious, including possible limitations of the research due to scarce or inconclusive data. An important merit of this thesis is that in addition to providing a detailed overview of NP structure in 100 Australian languages it also draws on and extrapolates to general typological work. CHAIR’S REPORT ON THE FINALISTS FOR THE FOURTH PĀṆINI AWARD, 2019, HILARY CHAPPELL, CHAIR: (i) THE 2019 WINNER Nadine Grimm. 2015. A grammar of Gyeli Humboldt University, Berlin Supervisors: Tom Güldemann and Maarten Mous This thesis presents a remarkable and comprehensive grammar of Gyeli, a Bantu language whose description is based on the Ngolo speech community in southern Cameroon, West Africa. The research draws on 19 months of fieldwork, some of which Nadine Grimm carried out as part of a DoBeS (Documentation of Endangered Languages) team project between 2010 and 2014. The analysis is firmly anchored in a multimodal corpus, which includes texts of diverse genres such 5 as traditional stories, narratives, multi-party conversations and dialogues, descriptions of everyday activities, procedural texts and songs. This rich documentation has been supplemented by data from elicitation work, questionnaires, and experiments. As to be expected of a winning grammar, it covers all levels of language, ranging from Gyeli phonology to its information structure. In her analysis, Nadine Grimm has chosen to use an approach which explicitly privileges form over function in her presentation so that each successive chapter topic neatly mirrors its role in a hierarchy of structures that she has established. Crucially, the description reveals itself as one that is well-entrenched in Bantu linguistics, providing a wealth of in-depth comparative and typological information and supplemented by observations on reconstructed forms for proto- Bantu. Some more specific comments follow below. An important reason for singling out Grimm’s grammar among the sixteen submitted to the Pāṇini Award are its in-depth analyses and discussions on a range of topics that will appeal to a wider typological audience, not just Bantuists. What is particularly laudable is that these analyses are clearly argued as to the reasons for favouring one theoretical solution over another. This is not just an occasional instance of good argumentation; it is evident in every chapter, and many sections within chapters. A few examples follow to illustrate this from different parts of the grammar.