It took me a long time to realise I wanted to become a linguist. I had a resolutely monoglot childhood in 1960s Canberra, Australia's 'bush capital', and can't even remember hearing any language spoken but English. Nor was school any better when it came to languages: the teachers were dogmatic types who obviously couldn't speak the languages well and were irritated by why-questions. So like my fellow nerds at Campbell High School I took science subjects and when it came to uni studied biology and psychology.
During my third year at the Australian National University (ANU) I had a spare slot in my degree and decided to take first year Russian - mainly so I could read Russian literature in the original. For the first time I experienced what it could be like to learn a foreign language from a teacher (Prof. de Bray) who could lay out its individual logic. He placed great stock on students getting a good pronunciation, and in week two we had to hand in an account of all the phonetic realisations of Russian vowel letters under different stress conditions. This wasn't the last time I benefited from a teacher with what many would think were unrealistic expectations - I don't think any of us had studied phonology before, but somehow that planted a seed in me.
In my fourth year, as I was finishing off my honours in psychology and feeling less and less that I wanted to continue with that, two things made me change course.
One was a fourth-year course I took on the psychology of language. Another demanding lecturer (Dr Trotter) set us one whole linguistic classic book per week - I lapped up John Lyons' Theoretical Linguistics, and Whorf's Language, Thought and Reality crystallised my feeling that studying languages could offer a much more direct route into other thought-worlds than the methodologically overwrought and conceptually oversimplified approaches of psychology. (That was what I thought then - now psychology has become much more interesting.) But the most decisive reading was Chomsky's review of B.F. Skinner's Verbal Behavior. I'd had far too much behaviorism during my psychology courses and hated its corny reductionism. Reading Chomsky's deft disembowelling of Skinner's position made me want to get seriously into linguistics.
Around the same time was that a friend of mine, Cliff Goddard, suggested I come along to one of the classes on Aboriginal languages taught by Prof. Bob Dixon. The students had a page of Yidiñ text to translate and analyse before the class with the help of a glossary and the grammar taught so far. Bob stood up the front, playing little snatches of Yidiñ on a reel-to-reel recorder, while a story of two brothers and the origins of death, order and chaos gradually took shape in the process of analysis.
Hearing the Yidiñ-speaking voice come out of that recorder was a decisive moment in my life. I had never heard an Aboriginal language before and was completely hypnotised by its sound. I think most Australians grow up with an aching sense of unconnectedness to their land, stemming from the invisibility and inaudibility of Aboriginal culture and the brutal way it was swept aside by the British colonisation process. So there was something deeply and indefinably fascinating about the conjunction of the storyteller, his language, Yidiñ culture, and the whole process of making sense of what he was saying. That one-hour class decided me on the spot that I wanted to study linguistics, and focus on Aboriginal languages.
I went on to take a coursework Master's in linguistics with some brilliant teachers - Bob Dixon, Anna Wierzbicka, Bill Foley, Tim Shopen. Lecturers often came to each other's courses, and most classes had a broad mixture of participants ranging from first years to distinguished visitors. This lack of status-consciousness was an educational blessing: it's very hard to write linguistics until you know how to argue in a more oral setting. Learning from the cut-and-thrust between the more seasoned members of the class was about the best way imaginable of getting into the subject.
We got a lot of Australian and Papuan problem material - often composed straight from the lecturer's field-notes. A key part of the training was the range of language-focussed courses on offer - especially for Aboriginal languages, which had three semesters' worth of offerings, and which really trained us to put together what we had learned and apply it to a real but exotic language. But the part of the departmental ethos which I most appreciate, in retrospect, was the confidence to find our own solutions to crack new codes in our own way, without being bound too much about the latest theoretical fads and how they said languages should work.
For some perverse reason, after finishing my thoroughly enjoyable Master's training, I decided I wouldn't keep studying. I think I was just sick of being a poor student. Meanwhile I had been offered a job teaching English in Beijing, accepted it, and bought a suit in readiness.
But that's not what life intended for me. I was two days off flying to China, when Bob Dixon rang and called me in for a chat. He'd received a letter from Mornington Island in northern Australia, asking him to find a PhD student to work on the Kayardild language. Ken Hale had recently been on Mornington to present a draft Lardil dictionary, the fruit of fieldwork he'd done there. Kayardild speakers, concerned at the fate of their own language, had wanted him to work with them on their language. He'd told them he wouldn't have time himself, but that through Bob Dixon he'd find a PhD student to do it. I thought it over: Mandarin was hardly in danger of disappearing, while Kayardild was on the ropes. Pretty much on the spur of the moment I told Bob I'd take up the offer, pulled out of the Beijing job at the last minute, and applied for a PhD scholarship.
In the Australian system then you just dived straight into your thesis without coursework. So a couple of months later I took a series of flights from Canberra to Mornington Island, flying in on a vintage DC3. It was a strange place. Somehow I had expected a lush tropical island. But it was more like a bit of savannah plonked into the sea, not far offshore from the unpromising saltpans running back from the Gulf of Carpentaria. There was a strong odour of inadequate sanitation, mixed with mangrove-wood smoke from cooking fires and what I later came to recognise as the smell of turtle and dugong fat. A winter wind blew off the sea, but it smelt of the desert not far to the south, rather than like a sea wind.
Late that afternoon on the beach I finally got to meet the Kayardild men, who had been out spearing fish until then. Traditionally monolingual, they were lousy language teachers. (Five years later when I began work in the polyglot universe of Arnhem Land I realised how unusual that was in Aboriginal Australia.) They fired words and sentences at me at machine-gun speed, all talking at once, working their way through my body-parts which they grabbed one by one. It was hard to write any sort of transcription in the circumstances. So much sand and ash was blowing around in the wind that I didn't want to take my Uher reel-to-reel recorder out of its case.
Despite their lack of pedagogical technique I warmed to them immediately. Their tribal leader Darwin Moodoonuthi and his wife May had missed out on having children of their own. Like many in their generation the trauma of their forced relocation from Bentinck to Mornington Islands had taken its toll on their fertility and they welcomed me into their lives as a son. They became my second father and mother.
In this way I was immediately welcomed into the Kaiadilt world, without emotional reserve. That open-hearted reception was crucial in getting me hooked as a field linguist. I think it's important for linguists to learn the languages they study on as many fronts as possible - simultaneously trying to figure out the structure through elicitation, building up a corpus to replay and trawl through endlessly while it slowly sinks in, but also trying to speak and hear it as much as possible. Usually the speaking part lags far behind - especially when the language is on the retreat and there are few people to speak it to - but if you don't work at that, your teachers don't get the pride and encouragement that always brings out the best in a teacher. And you won't get inside the language, or notice all the unofficial things which show you how it really works. Yet if I don't feel close to people emotionally, I find it very hard to learn their language: it starts to feel coldly technical, or even like satirical acting, rather than like the natural and loving emulation that makes learning a language from our family so easy.
Kayardild is probably the most unusual and interesting language I have ever worked on, from a grammatical point of view. But my Kayardild teachers weren't very good at explaining what things meant. Mostly this was because there were very few bilinguals. The whole population had gone over very quickly from speaking Kayardild to speaking English, so there were only a few people who could speak both. But it was also because they just weren't traditionally interested in language difference. Later, when I've worked in Arnhem Land and in New Guinea I've encountered a very different situation: people who were very used to learning many languages, and fascinated by fine differences of meaning that made their explanations a delight.
Nor did Kayardild people like doing very structured work - they were just too anarchical. So in some ways I developed bad habits for a while, of not pushing for every combinatoric possibility or asking for acceptability judgments. It just never worked with them, or else they just assented to everything I said, so that I didn't trust the results. This made me something of an extreme inductivist during my Kayardild research, an approach that only worked because the social setting was so conducive to language learning. Perhaps because it was my first field experience and had the open identity you have in your twenties, I developed a very strong secondary ethnicity as a sort of Kayardild person. But you have to beware of taking that too seriously and tread a tightrope there, accepting people's invitations into their lives and doing what you can to help from your outside perspective, but accepting that indigenous people have enough outsiders trying to usurp their rights. So even while trying to fit in linguistically and culturally as much as possible, you should never kid yourself that you're more than a privileged outsider who doesn't have your family hurt by alcoholism or early deaths and who hasn't had to deal with generations of disenfranchisement.
Since my PhD I've often gone back to Mornington, working up a dictionary and revising my grammar for publication, following up on other more subtle questions. But the number of speakers has just gone on dwindling to the point where most are now in the graveyard by the sea. When I first went there in 1982 there were about forty speakers, aged forty and up. The language was exuberantly bellowed out every day, especially in the women's camp on the beach. Hearing it spoken has increasingly required special orchestration, and now even the old ladies who spoke it well in 1982 use a pretty attenuated version, by habit of having to communicate with younger people who don't understand the full form of the language. It's still possible to get the odd new word, but impossible to get at complex grammatical patterns.
I've just given one part of my fieldwork experiences and one part of my career - working on Kayardild has led out in all sorts of other interesting directions (typology, historical linguistics, semantics...). My other fieldwork experiences in Arnhem Land and Papua New Guinea have each had their own stories. And working on little-known languages is only one part of the total challenge of our field. Linguistics has a unique but unfulfilled destiny, as the most scientific of the humanities and the most humanistic of the sciences. Our quest for meaning within and across languages is intimately tied in with our quest for meaning in life more generally. Linguistics is also a field with a thoroughly democratic appreciation of human creativity - as well as an ability to give voice to every one of the world's cultures. With the right balance of humanistic insight and scientific rigor, it has the potential to reach what Ortega y Gasset referred to as the revelation of the secrets that peoples and epochs keep from each other and which contribute so much to their dispersion and hostility - in sum, an audacious integration of humanity.
When I started into my PhD, or even when I finished it, I don't think I was conscious of where it would lead me, or how much my encounters with my teachers of Kayardild and later of other languages would change my ways of looking at language and at life. Nor was I even thinking about where it would lead, professionally or intellectually. I was just pulling a thread which I found intriguing, and the more I listened the more interesting things were there to be discovered.
 Ortega y Gasset (1937), Miseria y esplendor de la traducción. La Nación (Buenos Aires) May-June 1937; translation mine.