University Archives, ANU.

What made Joe Moyal decide to move from such an illustrious Department of Mathematics, brimming with intellectual challenge, at Manchester University and move across the world to a young national university rising slowly in Australia’s `bush capital’?

Personal choice lies at the heart of a life in science. Frequently, this is dominated by the social context of science and its disciplines, the imperative to choose a track and remain with it; to adhere to collaborative team work; or to remain in a secure and congenial setting with the reward system of promotion and scientific accolades in sight. Alternatively there may be an impetus to the ‘existential choice’ of Söderqvist’s vocabulary and the ‘continuous renewal of oneself’.^[1] Joe’s choice, no doubt, owed much to the latter and it marked a significant character trait of individualism and independence, the maverick streak that had already shaped his course in science.

Nonetheless, there were identifiable influences.

During his Readership at the University of Sydney, he had paid a visit to The Australian National University (ANU), a unique institution conceived in war as a potential powerhouse for national talent and nation-building and set up in Canberra as the `national’ university by the Commonwealth Government in 1946. Founded as an Institute of Advanced Studies, it had academic research and the training of postgraduates at its core.

Its planners, the eminent Australian expatriates, medical scientist Sir Howard Florey, physicist Professor Mark Oliphant and historian Sir Keith Hancock, together with the New Zealand anthropologist Raymond Firth, had seen it as a venture that offered new opportunities to place Australian research on the international map and to train gifted graduates from Australia and overseas.^[2] It was a viewpoint regarded with unveiled distaste and envy by the six poorly-funded universities in the Australian States. Nevertheless, The Australian National University had assembled its four Research Schools in Medical Research, the Social Sciences, Physics and the Earth Sciences, and Pacific Studies and, by the mid-1950s, interdisciplinary connections and a spirit of forward thinking had taken root among its small staff and its sprinkling of postgraduate students.

During 1955, Joe spent a week with the Department of Statistics. It was planned by the founding Director of the Research School of Social Sciences, Sir Keith Hancock, within his School and outside the Science departments and was headed by Professor P.A.P. Moran, who was appointed there in 1952. A graduate of Sydney and Cambridge Universities, Pat Moran’s professional career in the post-war had been as Senior Research Officer at the Institute of Statistics at Oxford from which he had moved briefly to a post as lecturer in statistics at Oxford University when he received the call to Canberra. Moran was on a learning curve when Joe visited his Department in 1955, but he hoped to stimulate fundamental work in the theory of statistics and already had two excellent postgraduate researchers, Joe Gani and E.J. Hannan, both of whom would go on to achieve high reputation. Although Moran was at that time little known in the world of statistics, the challenge of his appointment was broadening his thinking and opening up promising directions in his research.

A new well-endowed research university with national purposes in a young country which he had come to admire, clearly made a direct appeal to Joe. It had some psychological resonance with the Weizmann Institute. There was also a professional motivation. Although in demand as a Visiting Professor in American universities, he was still a Senior Lecturer in his home department at Manchester. His visits to Columbia and Berkeley had prompted invitations from several American universities to elevate him to high tenured posts abroad and a similar prospect awaited him at Sydney University. Harry Messel was articulate on the subject. ‘We could,’ he said in interview, ‘have produced a lot of good work together and numerous papers, if he had come back to work with me. I was a good foil for Joe.’^[3]

Joe’s impulse for leaving Manchester also had a personal root. He was in an unhappy marriage. He had been married to Suse for some 22 years but, as his long visits to other countries suggested, it was no longer a close and rewarding union. His daughter was now a young woman in her early 20s and his son was a teenager. Joe’s appointment to Australia heralded divorce proceedings and the migration of Suse, Orah and David Moyal to permanent residence in California.

In this sense, Joe’s choice of Canberra offered him distance and a new piece in the mosaic of his life. It also offered him what he most desired — a unique opportunity for research — and he sent off a notably brief and unembroidered Curriculum Vitae (the days of the `big sell’ CV were yet to come) in application for the Readership in the Statistics Department at the ANU. The commentary of his distinguished referees was, however, decidedly more telling.

Sir Harold Jeffreys, FRS, at Cambridge, a Royal Society medallist and author of numerous updated editions of his classic works and his ongoing Methods of Mathematical Physics, set down:

Herbert Robbins at Columbia University offered his strong personal support.

It remained for Maurice Bartlett to add his pertinent and straightforward words:

In Canberra, as Sir Keith Hancock and Pat Moran agreed with satisfaction, Joe Moyal was `a good catch’.^[7]

Joe arrived in Australia by sea in August 1958, and disembarked at Fremantle to attend an ANZAAS (Australian and New Zealand Association for the Advancement of Science) Conference at the University of Western Australia. From there, he flew on to Adelaide to present a paper at a conference of the Australian Mathematical Society. His advent in Canberra with trunk-loads of books and a smart new Honda scooter was preceded by an apologetic letter from the Vice-Chancellor of the University of Western Australia. Addressed to the ANU’s Vice-Chancellor, Sir Leslie Melville, it deplored a physical assault made on Dr Moyal by a member of his staff at the Conference.^[8]

Private collection.

Joe Moyal was clearly different! For his part, he looked forward to his new environment. ‘I have retained a very pleasant memory of my stay at University House’, he wrote the Registrar before sailing, and he was soon ensconced in one of its spacious apartments looking out upon the University’s rural grounds where the pink-plumed galahs and white sulphur-crested cockatoos gathered at twilight in colourful, chattering groups.

Joe was rapidly drawn into the University’s life. It would prove a pivotal experience. Multidisciplinary and purposeful in this early heyday, The Australian National University presented a diverse range of academics attracted to the university for its focus on research. Here in a rudimentary wooden building, a mix of people jostled at morning and afternoon tea, demographers talked with mathematicians, geographers with anthropologists, political scientists with statisticians, historians and sociologists with economists and physical scientists, in a rare collegiate life that centred for several years around the ‘Old Hospital Building’ with its open courts and verandas as plans for new university structures took shape.

Joe was embarked on work on a general theory of point processes and on what was to become one of his most significant papers, `A general theory of population processes’. The Australian National University Report for 1959 announced that he was completing `a long original monograph on the stochastic theory of populations, point processes and counting processes’, which was to be published by the University of California Press. But, baulked, apparently, by its size, and what Harold Jeffreys had also noted in his reference as Joe’s reluctance to publish before he has `done everything’, he set this composite venture aside to work — as the successive Report indicated — on ‘the asymptotic theory of multiplicative processes, and on the completeness of axiomatic systems in mathematical logic’.

It was illustrative of the gap in the diffusion of knowledge of statistics and stochastic processes in Australian universities and of its pioneering nature at the time that, over and above their research activities, both Pat Moran and Joe were busy offering courses of lectures throughout 1959 for their Department and elsewhere, Joe specializing on point processes and information theory while Pat Moran concentrated on statistical methods in medical research. Joe was also invited to extend a tradition inaugurated by Moran of visiting the Department of Mathematics at the University of Western Australia to give sets of courses on such subjects to postgraduate students and third year honours students, and to lecture to students and staff members in Sydney at the University of New South Wales.

He was also supervising two graduate students, Chip Heathcote and S.R. Adke with a focus on random processes, with both of whom he published papers. Professor Moran’s work was moving into an expanding area in mathematical genetics, but it was soon apparent that Joe’s wider knowledge and experience and a characteristic generosity with time and ideas, had a strong pull on the graduate students enrolling in the Department. A number sought him as their supervisor, transferring at times from a less accessible Department head, and his mentorship helped a widening circle of younger academics into key and influential teaching and research posts in statistics in Australian universities and overseas.

Chris Heathcote graduated with his doctoral thesis on the theory of queues and moved to an appointment at Stanford University to return across the years 1971–96 as Professor of Statistics in the Faculties at the ANU. Christopher Heyde, whose work Joe fostered in the classical theory of the determination of probability distributions by their moments, would shape a distinguished career that brought him back as Professor of Statistics to the ANU Research School of Mathematical Sciences. Peter Brockwell, who joined the Department in 1964 to study for a Ph.D., followed closely in Joe’s footsteps, becoming something of a ‘scientific son’ and collaborating on three papers as his career developed in 1964–67.

The presence of a man of Joe’s originality and stature in the ANU’s young and evolving Department of Statistics, had, as his referees had predicted, a defining effect. His influence spread. He was an open and readily available source of knowledge to colleagues and students from other parts. John Corbett recalls how, beginning his Ph.D. studies under H.S. Green, Professor of Theoretical Physics at the University of Adelaide, he thought that Joe’s phase space methods might help him confront a particular problem in his research. As Green was to visit Joe in Canberra, he invited Corbett to go with him. ‘We went into this building and knocked on his door,’ Corbett remembers. `I was a little nervous because I wasn’t sure that I knew what I was talking about, and I opened a door and saw a roomful of smoke and there on the other side of the room as the smoke began to clear a little, I saw this rather pleasant looking man who invited us in and I asked a couple of questions. And I think I was satisfied. Here I was a very young and naive student and Joe was ready to share his knowledge with anyone who came along at any time.’^[9]

Joe served as Acting Head of the Department of Statistics during Moran’s 12-month study leave at Oxford during 1960 and as a member of the Academic Board. Sitting also on a number of Appointment Boards, he seized the opportunity — important in the formative days of the University and no doubt coloured by his own experience with the Weizmann Institute - to present a strong case for bringing back brilliant young Australian scientists from abroad to fill the emerging Chairs of Science at the ANU. In this his attitude was conspicuously at variance with the university’s most dominant scientific figure, the Director of the Research School of Physical Science, Sir Mark Oliphant, who considered that the university was better served by importing older Fellows of the Royal Society drawn from the ‘old boy’s network’ from Britain than attracting home grown younger men from eminent positions overseas. On this point, the maverick, Moyal, with his respect for originality and Australian talent, and the elitist Oliphant, remained on an intellectual collision course.

Herbert Robbins, refereeing, had sketched Joe Moyal as ‘a friendly and co-operative person, popular with his associates and students’ and likely to be ‘a great asset to any university community’. At University House, he stood out — part European, part Israeli, a cultivated man deeply read in philosophy and history who forged friendships with the younger postgraduate and postdoctoral scientists and humanists and the university staff members who resided in the House.

Joe lived at University House for several years. ‘He had the air’, one resident observed, ‘of one who belonged to no particular nationality, and his deep, faintly accented voice puzzled interlocutors.’ Deeply engaged on what would become his foundation paper on stochastic population processes and given to lengthy rumination in his evening bath, he developed a habit (much admired) of arriving late for the rather early House dinner hour and appearing, damp and hastily attired, to open the locked Dining Hall door with a quick backward flip of his foot.

Scientific visitors came and went. The Hungarian mathematician and relentless traveller, Paul Erdös, was a Visitor in the Statistics Department during 1960, adding number theory to the well-attended lecture series that drew both insiders and outsiders from the Research School of Physical Sciences, Canberra University College and the CSIRO, to the Department’s research talks. Joe himself in these years offered a twice-weekly three-term course on functional analysis, semi-groups and spectral theory, and a first term course on random processes in physics.^[10]

Joe had arrived at the ANU with a considerable body of original work behind him, foundation papers that would endure. In 1961, he was in the United States on study leave — a singular enfranchisement at the ANU for senior scholars to compensate for the tyranny of distance — developing his work on stochastic population processes and ‘trying it out’ at Stanford and at the Rand Corporation in Los Angeles, where he was a visitor for several months.

‘This last month,’ he wrote to Australia in August, ‘has been extremely profitable and rewarding for me. Not only did I meet and talk with a whole lot of people I wanted to see, and attend a conference here on functional analysis which was of particular interest to me, but I have several general offers of American posts … I feel all wound up again and there’s a whole lot of things I want to do.’^[11]

‘The general theory of stochastic population processes’, appeared in Acta Mathematica in 1962, and was subsequently republished in an anthology of mathematics. ‘Multiplicative population processes’ also emerged that year.

Writing of Joe’s Acta Mathematica paper after his death, Gani characterized its scope and girth. It provides, he noted, `the foundations of a general theory of population processes in which both the number of individuals in a population and the states characterizing each of them are traced. Moyal considered point and counting processes, and develops the concept of the probability generating functional in the population context … [He] then offers as examples of his methods, cluster processes, counting processes with independent elements, time-dependent Markov population processes, and multiplicative population processes.’ It was, he adds, a paper that promoted a wealth of citations and other papers.^[12]

Private collection.

While professional plaudits mounted, Joe’s personal life also changed. At 48, he was an attractive man, serious and active, and popular with women. Among a stimulating array of colleagues at University House, he met the historian, Ann Mozley, some 16 years his junior. Mozley, a graduate of Sydney University, had been working in Britain for nine years, latterly as personal research assistant to the powerful press baron, Lord Beaverbrook, with whom she travelled the world and mingled with some of the great political figures of Britain and the USA of the time while assisting Beaverbrook to write his political history of World War I, Men and Power. She arrived at the Research School of Social Sciences at the ANU late in 1958 to work with Sir Keith Hancock in founding the Australian Dictionary of Biography and, four years later, launched her career in the history of Australian science and technology.

From this background it was not surprising that she found Joe Moyal to be the most cosmopolitan member of University House. ‘For a long time after he was introduced,’ she wrote of her first encounter with him in her autobiography, Breakfast with Beaverbrook, ‘he sat carefully away from me protected by several chairs … He must have improved on this exchange for I soon became an unconfident passenger on the back of his Honda motorbike and was thrust through the fancy jitterbugging he affected on a dance floor. There were, it seemed, two men inside the scholarly Moyal.’ Yet, she added, `compelling qualities drew me in. In that company Joe was a civilized man, deeply read in history and philosophy as few scientists are, a true intellectual "betrothed to thought".’^[13] They were married at a registry office in Sydney in September 1963.

He subsequently became Professor of Statistics at the University of Sheffield (1965–1974), and Chief of the CSIRO Division of Mathematics and Statistics (1974–1981).  He eventually retired from the University of California, Santa Barbara in 1994 and is currently a Visiting Fellow in the ANU Mathematical Sciences Institute.

Courtesy of the Mathematical Sciences Institute (http://www.maths.anu.edu.au/~gani/)

During 1963, Dr Joe Gani also joined the Department of Statistics as Senior Fellow. Following his year as a postdoctoral fellow at Manchester University in 1955, he had moved through appointments at the Universities of Western Australia and Columbia before returning to the Department in which Joe had first encountered him. In these fertile years, Gani had developed well-defined ideas of the balance required for mathematical statisticians which, in his view, lay between a strong mathematical background and a philosophical and empirical approach to statistics and its varied applications. The spirited and lively Gani brought a breath of fresh and, at times, controversial air to the Department which Joe found most congenial. He gladly supported his younger colleague in his founding of the international Journal of Applied Probability in 1963-4 and joined its editorial board.

1963 also brought the eclectic philosopher of science, Karl Popper, to the ANU as a distinguished visitor at the Unit of the History of Ideas. Lodged at University House, Popper spent much time in Joe’s company. Their interests linked. Popper had been a student of philosophy, training to become a schoolteacher when the discoveries of Heisenberg and Schrödinger leapt to view, and, while fired and excited by them, he acknowledged that original research in physics and mathematics was beyond his reach. Nonetheless, he had allied his theory of scientific discovery with a critical interpretation of quantum mechanics in his book The Logic of Scientific Discovery. Now a Professor at the London School of Economics, he had made his first entry into academia in 1937 as a lecturer in philosophy at Canterbury College, New Zealand, where, as a refugee from Nazism, he wrote his other landmark book, The Open Society and its Enemies. There, he also met his lifetime friend, neuroscientist and future medical Nobel Laureate, John Eccles, who had moved subsequently to the John Curtin School of Medical Research at the ANU. An intellectual elitist, openly intolerant of social concourse, Popper confided privately that, while at the ANU, there `were only two academics, Eccles and Moyal, with whom he could discuss and share ideas’.^[14]

For Joe, these years at The Australian National University set a seal on his own sense of assured creativity and his role as a pioneering mentor in the statistical field. During them he added to his publications on stochastic population processes and multiple population chains with `Multiple population processes’ in 1963 and completed work on his 1965 paper `Incomplete discontinuous Markov processes’.

Yet the young Department itself was not without tensions. As Chris Heyde summed up thoughtfully in his obituary of Pat Moran, ‘he rarely sought to exercise power but he was reluctant to share it’.^[15] Inevitably, this had negative repercussions for his high profile deputy, yet in this period the solid foundations for Australia’s major school of statistics were laid. Joe was also happy in his marriage and in the society of congenial friends, as well as in the healthy open-air life spent among the mountains and rivers of the Australian Capital Territory.

Private collection.

For someone long displaced from his native country, Australia provided a `second soil’. A strong and venturesome swimmer, he took up snorkelling and diving and, cajoling Ann into these vigorous sports, set in train their long pattern of travelling to Queensland for holidays among the brilliant corals and marine life of the Great Barrier Reef. They also explored the Pacific to find the dazzling coral fringes of the islands where (in less populous tourist times) Joe would descend alone into the depth of the sea while Ann, no diver herself, hovered on the surface snorkelling and watching anxiously. So great was his pleasure in drifting absorbed in the brilliant underwater world, that Joe asked for his ashes to be scattered in the waters of the Great Barrier Reef.

Even so, by 1964, Joe had begun to harbour some broad misgivings about the ANU. Despite pockets of marked intellectual vigour and an attractive emphasis on research, there were, in his view, prevailing tendencies of comfortable privilege at Australia’s national university. Sir Keith Hancock had put a finger on it when he observed there was `too much frittering, pottering and gadding’, and ‘privileged people needed to watch their steps’.^[16]

Aware of the stunted funds of the State universities and the need for The Australian National University to show that it was producing great quality research, Joe believed that the ANU was ‘less accountable’ than it ought and he wondered in private if ‘it might be found out’. An increasing critic, he articulated the view that the great international research centres of the Institute for Advanced Study at Princeton and the Rockefeller Institute in New York, with their few outstanding tenured researchers and a flow of brilliant, short-term visitors, were better models for scholarship than the hierarchical, departmental and School structures and their maze of tenured appointments that had grown at the ANU.

Joe’s instinct proved correct and over the next 40 years the ANU has moved in a direction more consistent with his vision than that of his contemporaries.

It was, however, while in this frame of mind that Joe received an invitation from the Director of the Applied Mathematics Division of America's leading Atomic Energy Laboratory, Argonne National Laboratory, in Illinois, to consider joining the Laboratory as a Senior Scientist. This headhunting approach, with its offer of a significant salary increase and participation at a major centre of international research, was a response to Joe’s cumulative research as a mathematician that spanned quantum physics, stochastic and population processes and their wide applications in nuclear fields. For the Laboratory also, the generalizations of his general theory of stochastic population processes had particular significance for populations of biological organisms and subatomic particles.

With some demurring from Ann, then building a centre for the study of the history of Australian science at the Australian Academy of Science in Canberra, Joe accepted the position, comforted by the knowledge that there were some excellent historians of science in the USA.

As a British citizen, he faced a lengthy waiting queue to enter the USA or a special Act of Congress to secure his appointment. His entry, in the event, proved well in character. As US immigration quotas were fixed by birthplace, he was identified, conveniently, as having been born in Old Jerusalem on the city’s eastern side, a geographical division agreed upon by Israel and Jordan in 1968. Thus he was swiftly removed from the bulging British lists and given permanent immigration status in the United States as a Jordanian. He left The Australian National University in September 1964 and went ahead to join Argonne in November that year. Science and its beckoning international extension became the trigger for change.