I'd written a draft grammar of Dyirbal during 1966, but was now dissatisfied with it. It tried to make a contribution to linguistic theory, using all sorts of neologistic terminology – in effect putting the language second to the theory, referring to aspects of the grammar only as needed for linguistic model-making. When I got back, in May 1967, I decided to start all over again. I now had a better understanding of how the language worked, and my aim would be to try to describe it as clearly as possible, with the minimum of pseudo-scientific jargon. The ergative structure of Dyirbal was different enough from the grammars of familiar languages, without the added complication of an obscure presentation.

I've never been fond of collecting degrees and had up to this time subtly managed to avoid doing a doctorate. But it was becoming increasingly clear that it would be a useful thing to have – a Ph.D. would make getting a visa to work in the U.S.A. very much easier to obtain, for one thing. I was writing a grammar of Dyirbal anyway so – why not submit it for the degree?

As I worked through the mass of material from that productive month in the field, and wrote the thesis, some points emerged that really required further clarification or checking.