This year marks the centenary of the birth of Gordon Foster, one of Ireland's most distinguished computer scientists and statisticians. Gordon was Professor of Statistics in Trinity from 1968 to 1991.

Frederic Gordon Foster was born in Belfast to Robert and Florence Foster. His father owned a garage until the 1930s when the business collapsed, badly affecting the young family's finances. However, in the first of what would be many academic achievements, Gordon won a scholarship to the Royal Belfast Academical Institution (RBAI). In addition to being a gifted student, he was a fine athlete, boxing being one of his many interests; he was captain of the school boxing team in his final year at the RBAI. As a young man he also played the violin. The instrument he played had belonged to his grandfather - a professional orchestral player. Gordon enjoyed sailing, and with the help of a friend, built his own boat. He remained a keen sailor for most of his life.

Gordon Foster 24th February 1921 - 20th December 2010

In 1939 Ohe enrolled in Queen's University in Belfast with, as he later put it, "No idea of what I wanted to study". He decided to try medicine, but found that neither working in laboratories nor cutting up small animals much appealed to him. After investigating other options, he settled on an arts degree, specialising in mathematics. He graduated in 1942 and was recruited almost immediately by the UK foreign intelligence service, better known today as MI6, to work as a code-breaker in the top secret facility at Bletchley Park, the famous station X. He later recalled that, when he turned up for the final interview in London, there were no security checks. At the time, the British government was recruiting mathematicians, languages experts and even people who were good at cryptic crosswords. At Bletchley, he joined a large and rapidly expanding team led by the mathematician and computing pioneer Alan Turing. Gordon would eventually meet Turing, but only after the War ended.

Like thousands of other mathematicians and linguists stationed at Bletchley, Gordon worked on a trestle table in a large room, trying to crack Enigma and Shark, the codes used by the German army and navy respectively. He also worked on Japanese naval codes. The experience, he later recalled, was pretty mundane and not helped by a diet that included horse meat on the canteen menu. The key skills required were the ability to spot patterns and errors made by German soldiers and sailors when they were using encryption machines in the field. Officially a member of the home guard, he had to learn how to polish a rifle, although it seems that he was never taught how to fire one.

After the War, in 1948, Gordon married and had three children. In 1950 he had been awarded a scholarship to pursue postgraduate studies, so he registered for a DPhil at Magdalen College, Oxford which he completed in 1952. His dissertation was entitled: "Some problems in the mathematical theory of probability". In 1951, he attended a lecture by the American mathematician and philosopher Norbert Wiener. Today, Wiener is considered by many to be the founder of cybernetics, the science of communication as it relates to living things and machines. Attending this lecture was to have two important consequences. The first was that Wiener's ideas were to have a recurring influence on Gordon's own life and thinking. The second, and more immediate, was an invitation by the Professor of Statistics at Manchester University to give a lecture there. His host informed him that Alan Turing was in the university at the time, so Gordon finally got to meet his former boss. Turing showed him the Manchester Mark I computer that he (Turing) had worked on developing. Gordon remembered Turing as combining the ability to think in abstract terms with a practical bent for getting things done. It was a description that could be said to have applied to Gordon himself as his own career developed.

This meeting with Turing reignited Gordon's interest in computing. In the years since the war, he had not given much thought to the development of computers or computing. However, the possibilities and the ideas of Turing and Wiener found fertile ground in his inquisitive mind. While at Oxford he had been introduced to Maurice Kendall, then head of Department of Statistics at the London School of Economics (LSE) (whose books will be well known to many a statistics student of a certain era). Kendall invited Gordon to come to the LSE as an assistant lecturer. An attraction of the job was that the LSE had access to an English Electric computer. To test the speed of this machine, Kendall suggested using it to create sets of statistical tables - a computationally intensive task that is tedious to do manually. This was the era of punch cards and machine code; screens and third generation languages were not yet around. To do this job, Gordon and some colleagues developed a higher level language that they called Basic. Unfortunately, somebody else came up with the same name for a different higher level language that went on to become the well-known BASIC language, but at least Gordon and his colleagues had the name first. Gordon was elected a Fellow of the Royal Statistical Society in 1954.

This was a time during which, in addition to computing and statistics, operations research (OR) was emerging as an important field. OR (nowadays often referred to as management science) had been invented and developed during the War to help solve military problems such as logistics and the design of optimal search patterns. During the 1950s, Gordon was involved in consolidating what was in effect a disparate set of techniques into a coherent discipline. He viewed OR as a potential precursor to what he called systems thinking, a term that was later to gain widespread use in the field of information systems. While he contributed to the development of OR, during this period Gordon's main theoretical interest remained probability theory. This led to him being invited by British government to advise them on the regulation of gambling in the UK. In 1964, he was appointed to the chair of Computational Methods at the LSE where he continued his work on developing OR as a new discipline. He also made important contributions to queuing theory and Markov chains, including what became known as Foster's theorem.

Gordon Foster

In 1966 he became involved in a project that is his most famous legacy - one from which many millions of people have benefited without ever having heard of Gordon Foster. Gordon was commissioned by the publishers Wiley to develop a book coding system to help with the computerisation of their UK stock records (having been introduced by the Professor of Economics at the LSE to the Managing Director of Wiley). At about the same time, the bookseller WH Smith was working on a book numbering system for a planned computer warehouse. When other publishers heard of these initiatives, Gordon was consulted by the Publishers Association on the feasibility of designing a book numbering scheme for the UK and Ireland. The challenge was to design a code that was both flexible and that could readily incorporate existing codes (i.e., those used by Wiley and WH Smith and possibly others). Gordon solved this problem in a way that was both simple and ingenious. Following this success, he was asked by the International Standards Office (ISO) whether he could further adapt the scheme for international use. This he did by increasing the number of digits in the code from the original nine to ten. This extended design was approved and Gordon presented his international numbering scheme to the British Library, The Bookseller and the Publishers Association. In this report, he outlined his scheme and how it could be expanded indefinitely. The latter feature, alongside its incorporation of existing codes, was a stroke of genius. His report, Standard Numbering in the Book Trade, was published by The Publishers Association in 1967. In 2007 the code was extended to 13 digits to accommodate an article number. This is why every ISBN today has the prefix '978' or '979' Gordon's design has stood the test of time and the ISBN can be found in the opening pages of every book that you buy or borrow today.

Meanwhile, in Ireland, Trinity College had established a Chair in Statistics in 1965. The College wanted a major figure to head up this new department and in 1966 Gordon was appointed to be the first Chair and the College's first Professor of Statistics. During 1967 and early 1968 he filled these roles as a visiting professor. In 1968 he was appointed full professor.

That same year he was invited by what is now the Economic and Social Research Institute (ESRI) to give the second Roy Geary lecture. He used this as a platform to lay out a vision for the future of computing in Ireland. This was later published under the title: "Computers, statistics and planning: systems or chaos?". He was subsequently commissioned by the ESRI to undertake a study of computing in Ireland. The result of this project, done in conjunction with the Department of Finance, was a report entitled Computers for Ireland. This visionary document recommended, inter alia, the development of a software industry in Ireland and the establishment of a central bureau within the public sector to both promote the use of computing in the sector and to reduce, or better still eliminate, unproductive inter departmental and inter agency rivalries. The latter led to the creation of the government's Central Data Processing Service (CDPS). Fifty years after this report first set out Gordon's vision, software is today one of Ireland's largest and most successful industries.

In 1971, Gordon was elected a Fellow of the College.

Gordon Foster at home in Dundrum

Gordon applied his characteristic vision and energy to in the new Department of Statistics. He built up the department overseeing the development of undergraduate modules for the honours degree in mathematics as well as service courses for several other degree courses. Gordon had always had a practical side to his personality and this led to development of a number of largely taught, application orientated postgraduate programmes. These included a Postgraduate Diploma in Statistics (still on offer today) and masters in Operations Research (1970), in Applied Statistics (1974), and in Statistics and Economics (1973). Under his leadership the department also developed new postgraduate programmes designed for the public sector. The MSc in Economics and Statistics evolved into two further programmes, the MSc in Statistics in Administration (1979) and the MSc in Public Sector Analysis (1983). The value of these programmes was widely recognised, including by external assessors of the department. A further innovation was the creation, in 1972, of the Statistics and Operations Research Laboratory (SORL) to provide services to industry and the public sector and to support academics in other disciplines who needed help with statistics in their own research. SORL would later be spun off to become Insight Statistical Consulting. The original masters courses were run until the early 1980s when they were retired to make way for a number of new programmes including the department's first full undergraduate course, Management Science and Information Systems Studies (MSISS). MSISS produced its first graduates in 1984 and over the following decades was to evolve into one of Trinity's most successful undergraduate programmes. In the midst of all of this activity, Gordon also found the time to supervise 17 PhDs.

Gordon with Dr Christian Agunwanba (on right) Nigerian head of the Nigeria linkage project

Gordon's vision and his ideas for the use of computing were not restricted to Ireland. He saw huge potential for exporting an approach that had worked well in Ireland to developing countries. In 1977 another new programme, the Systems Development Programme was established. This was designed for students from developing countries to come and study in Ireland. It was supported by the Irish Department of Foreign Affairs' Bilateral Aid programme, the (then) EEC and the United Nations International Development Organisation (UNIDO). It attracted students from many countries particularly sub-Saharan Africa (including Nigeria, Ghana, Tanzania, Zambia, and Lesotho). There were also students from Iran, Lebanon, Libya, Malaysia, Mexico, and Guyana. Working relationships were established with several African countries including Nigeria and Uganda. Gordon also worked on a major United Nations University (UNU) development project leading to the UNU publication Information Technology in Selected Countries in 1994.

TRINET: On the roof of Kevin Street Technological Institute

Gordon retired in 1991, but immediately threw himself into a number of new ventures including setting up the non-profit Informatics Development Institute (later renamed the Informatics and Social Development Research Institute). This was a platform for the development of a low cost accessible Internet and email system using low orbit satellites and local radio links for use in developing countries that did not yet have Internet access on a large scale. The network was developed in association with the Dublin Institute of Technology (now Dublin University of Technology) and was christened TRINET. Had Google Earth been available at the time, you could have looked at the numerous aerials on the roof of the Statistics Department and the Dublin Institute of Technology in Kevin Street (see picture). Educational links were established with Ghana, Nigeria, Uganda, Zambia and Zimbabwe. As Gordon said, TRINET not only brought the information superhighway to main centres, but "up a few muddy lanes as well." This infrastructure provided vital communications during the 1995 outbreak of Ebola in Zaire when it enabled doctors at a remote hospital near Kitwit to communicate with the Centers for Disease Control operation in Atlanta.

Gordon left behind an impressive and lasting legacy both in Ireland and internationally. His formidable range of skills, breadth of knowledge and organisational ability contributed to many fields. The world continues to benefit in many ways from his vision and energy.

Gordon died on the 20th of December 2010.

Frank Bannister March 2021

Acknowledgements. I would like to acknowledge the help and advice of several friends and former colleagues of Gordon in putting this together. These include Eileen Drew, John Haslett, Eamonn Mullins, Patrick O'Beirne, Michael Stuart and a number of others who do not wish to be named.