39 8 A metrical system that defies description by ordinary means ALAN RUMSEY One of the first times that I met Andy Pawley — and probably the first time that my wife and co-worker Francesca Merlan and I met him together — was in Brisbane in May of 1981 at a meeting of the linguistics section of the Australia and New Zealand Association for the Advancement of Science. We had stopped off for that meeting while on the way to Papua New Guinea for our first fieldwork there. We had planned to do that work somewhere in the Nebilyer Valley west of Mount Hagen, but had little idea what we would find there. Andy was full of sound advice based on his already long experience in the Kalam region some 100 km to the northeast, and with his help and others’ we were able to get started within the next few months on what has turned out to a long series of research projects based in the Ku Waru (‘Steep Stone’) dialect area which straddles the Tambul Range along the western edge of the Nebilyer Valley and north-eastern reaches of the Kaugel Valley on the other side of the range. 1 After returning from our first extended fieldwork there Francesca and I continued to benefit from Andy’s work, especially from his insights into the grammar of Papuan languages. As many readers of this volume will know, over a quarter-century ago Andy wrote a paper which became famous among Papuanists long before it was published, entitled ‘A language that defies description by ordinary means’ (Pawley 1993). The language was Kalam — a typical Highland New Guinea language in most respects, but one which in those same respects presented a radical challenge to linguists’ understanding of what is ordinary, especially with regard to verbal semantics, and the question of where the grammar stops and the lexicon begins. Here I wish to honour Andy’s work by considering some aspects of a Ku Waru genre of verbal art which seems to present a similar challenge to current linguistic theory in another domain, namely with respect to the nature of poetic 1 For their helpful comments on earlier versions of this paper I would like to thank Francesca Merlan and Don Niles. I am grateful to Don Niles also for his extensive research collaboration with me on tom yaya kange, for sharing the results of his insightful work on their musical aspects, and for allowing me to reproduce his musical transcription in Figure 4. Many thanks to Paul Kiparsky for his sponsorship on a period of a sabbatical leave that I spent in 2005 in the Department of Linguistics at Stanford University, and for inviting me to participate in a series of workshops on Language and Poetic Form at Stanford, including a day-long ‘Poetics Fest’ on 14 May, 2005, at which I gave a presentation from which the present paper has arisen. Thanks also to Kristin Hanson and others who discussed that presentation with me. For funding my research on tom yaya kange and related matters since 1981, I acknowledge the Australian Research Council, the University of Sydney and The Australian National University. In John Bowden, Nikolaus Himmelmann, and Malcolm Ross, (eds.) 2010. A Journey through Austronesian and Papuan Linguistic and Cultural Space: Papers in Honour of Andrew K. Pawley. Canberra: Pacific Linguistics. Pp. 39-56.