26 December 1948 - 15 October 2010
Julian Leslie was born in Melbourne, Victoria, Australia, the son of Rupert Thomas Leslie(1916- 2007) and Roberta Julann Leslie (née Gordon) (1921-2006), the youngest of 3children. At the time of his birth his father was attached to the Division of Forest Products,Section Division of Mathematical Statistics, CSIR (which was to become theCommonwealth Scientific and Research Organization in the following year). In 1954 theyoung family moved to England where Rupert was to work with the renowned statisticianRonald Aylmer Fisher. There is a photo on Macquarie University’s website of Fisher withthe young Julian and his sister from that time, surely a portent of things to come. From 1959 the family lived at Artarmon, a suburb of Sydney, and Julian spent a happy adolescence, frequently in the company of his friend Anthony Papallo, on adventurous pursuits often of scientific inclination. These were motivated by Julian’s father, an eminent Australian statistician who became Professor of Statistics (1970-1981) at the University of Strathclyde, U.K. Julian’s grandfather, Percy Robert Leslie, was an engineer who at the end of hiscareer became General Manager of the Spotswood Railway Workshops in Victoria. In hisretirement he wrote a book on the history of railways, which is preserved in the NationalArchives in Canberra.
Julian was educated at North Sydney Boys High School, and went on to the University ofSydney (1966-69) where he completed fourth-year Honours in Mathematical Statistics withFirst Class Honours, in the Department of Professor H.O. Lancaster (1913- 2001). From1970 to July 1972 he was the first Ph D student of C.C. (Chris) Heyde (1939-2008) withinthe Statistics Department of Professor E.J. (Ted) Hannan (1921-1994), in the School ofGeneral Studies, Australian National University (ANU), Canberra. His first paper, jointly withChris, which was published in 1971 in the Bulletin of the Australian Mathematical Society, quickly improved on some classical-type limit theorems obtained within the context ofbranching processes by Chris and one of the undersigned written only months before in thesame department. Julian’s thesis topic was “On rates of convergence to normality”, afavourite Heyde topic of the time, and the dissertation produced a sequence of papers in distinguished journals up to 1976.
Julian interrupted his work on the dissertation for an ANU exchange visit to Moscow StateUniversity until June 1973, when he took up till 1974 a temporary lectureship in Statistics at the University of Lancaster, U.K. During this time he wrote up his PhD thesis and was subsequently awarded the degree.
In the period 1974-1990 he was Lecturer in Statistics, Birkbeck College, University ofLondon. At the time Birkbeck College catered mainly for part-time students who worked during the day, with evening courses. Julian taught undergraduate courses in Applied Statistics and Theoretical Statistics, and for the MSc in Applied Statistics and OperationsResearch taught Stochastic Processes, Probability Theory, Advanced probability Theory,Sequential Analysis, Goodness- of - Fit Tests, Quality Control, Inference and MultivariateAnalysis. Additionally, he carried a substantial administrative load, including computerrepresentative for the department, organizer of internal seminars, and of shared seminarseries between Birkbeck College, Imperial College, London School of Economics andQueen Mary College, a substantial task with no secretarial assistance. He introduced theidea of conducting seminars on several sites simultaneously using a fibre optic video linkbetween 4 of the London University’s colleges. In intercollege work Agnes Herzberg was avalued collaborator.
During his time at Birkbeck, Julian admired the scholarship, wisdom and integrity ofProfessor Philip Holgate (1934-1993), a man of broad cultural interests and unassuming and gentle manner. The following words extracted from a letter Julian wrote about Philip tohis brother, Professor Alan Holgate after Philip’s death, could equally well be said of Julian himself:
“I have never met a person who was more transparently honest than he. He was the ideal head of department, he commanded respect by sheer force of intellect. He would typically take on the awful administrative jobs and not burden us with them. I was particularly struckby the weight he gave to simple words - he didn't use words associated with extremes ofemotion. He somehow rejuvenated the spoken language for me - he was interested intelling you the facts stripped of emotion and innuendo“.
The summers provided some relief and some opportunity for Julian to return to research.The ones of 1983 and 1986 were spent at Simon Fraser University, Vancouver, at theinvitation of M.A. Stephens (an expert in the area of Goodness of-Fit Tests) working on joint research projects. Consequently, the year 1984 saw the first of papers resulting from this collaboration, including one in Annals of Statistics in 1986.
Holgate served as Head of the Department of Statistics at Birkbeck until 1987. The Departments of Statistics and Mathematics there merged in 1989, which may have motivated Julian to think about returning to Australia.
Julian came to Macquarie University, Sydney, as Lecturer in Statistics in July 1990, inProfessor Don McNeil’s Department. He was soon promoted to Senior Lecturer and servedin this capacity 1991- 2003. From 2003 he was appointed Associate Professor in Statistics, the equivalent of full professor in the U.S. system.
At Macquarie University, Julian embraced the most difficult and time-consuming ofadministrative chores, and devised and taught many new subjects, statistical andotherwise, in order to maintain or improve student and therefore staff numbers. Many of the new subjects were outside Julian’s range of experience, including subjects from information technology that he adapted for teaching to Masters students in Business and Commerce. “Computer Applications in Business”, “Market Research”, “Data Mining” and “GeographicInformation Systems” were some of the subjects or topics he introduced and taught. Healso taught extensively both applied and theoretical statistics courses at all undergraduate and postgraduate levels. Julian was a popular supervisor of the thesis component of theMaster of Applied Statistics program, and offered statistical advice to staff and students across the campus. He served on numerous departmental, Faculty and University committees, including more recently the University Staff Consultative committee. He was atwork mostly by 7am, and rarely left before 7pm. In his last application for the “Outside Studies Program” (Sabbatical leave), which he wasunable to take, Julian wrote, when asked to justify in terms of his research activity:
“I have to spend most of my time servicing my teaching activities and associated administrative matters. Research is undertaken out of hours and during holiday periods(much of the latter being eaten into by examination matters, enrolment and otherdepartmental activities).”
Nevertheless, at the onset of his illness, Julian was involved in joint research in ForensicStatistics (distinguishing the type of pencil used in handwriting, and “handedness”) and “Speed Dating”.
Julian applied his strong probabilistic insight intermittently in remarkable instances.Professor Paul Embrechts writes (personal communication) that during his times at theUniversity of London colleges in the early 1980’s, Julian solved one of the open problems on which Paul had been working. This was published by Julian as “On the non-closure of the sub exponential family” in J.Applied Probability, 26 (1989) 58-66. Professor Embrechts describes him as “ … one of those probabilists from the ‘good old’ Australian school … with whom you could discuss any applied problem. Julian would always have something to say in his typically reserved but clear way. “At Macquarie University during Julian’s heavy commitments to non-research directions he nevertheless collaborated in a key way in the remarkable solution of a long-standing problem. This was published as Kozek,A.S.,Leslie,J.R. and Schuster, E.F. “On a universal strong Law of Large Numbers for conditional expectations. “Bernoulli, 4 (1998)143-165. Earlier in the Macquarie period there had been a paper on characterization in another prestigious journal, Annals of Statistics, with the eminent statistician Constance van Eeden in 1993.
Julian was elected to the ISI in 1992. He held office in the Statistical Society of Australia, and was a Member of the Statistical Society of Canada, and of the Royal Statistical Society.
He kept a diary which revealed that the illness/tumour which eventually killed him took holdin late May 2009. A final email to one of us is dated 9th June; Julian had been working onFrancis Galton, and wanted to be informed of the Galton Conference to be held in London in 2011. He was diagnosed with a brain tumour early in the September following, andunderwent radiation therapy and immunotherapy subsequently. Nevertheless, the lastmonths of his life saw the progressive loss of the skills by which a human being defineshimself. He was supported by the loving care of his family, and frequent visits from hismany friends and colleagues, with whose welfare he characteristically continued to beconcerned. Happily, he was able to walk his daughter Annette down the aisle on herwedding day.
Julian felt strongly that it was a human right to die with dignity. When doctors told him hehad only weeks to live, the family took time to travel around Europe together, visiting all theplaces which held special meaning for them. Even in his last days he never ceased tomarvel at the beauty of his surroundings. He was particularly enthralled by the cities and cathedrals which his family visited. At the end his wife Andrea whom he had married on 28 January 1972, his children Annette and Daniel, and his siblings Michael and Elisabeth, were at his death, which was aided by Dignitas, in Zurich, Switzerland.
A very moving Memorial Service for Julian was held at Macquarie University on Friday 12 November 2010. To the background of classical music, poetry and photos, there were tributes by family, friends and colleagues. Julian was a gentle, modest, thoughtful, generous, loyal person, with a strong and principled social conscience who put the welfare of others above his own. One senior former colleague wrote that occasionally there is “one person who is the repository of the soul of a department”, and that was true of Julian. We conclude our tribute with the words which another colleague addressed to him:
“You left the world a better place than you found it.”
Eugene Seneta and Barry Quinn
