Bill Williams was born in Fulham, London, on 18 April 1913, the only child of Thomas and Clara Williams. His father suffered from asthma and so had left Wales, where he had been a coalminer, to work in London but at what has not been ascertained. Whatever it was, his mother found it necessary to work as a midwife and charlady to ensure that Bill received a good education. Having no siblings, Bill spent much of his childhood at the home of his lifelong friend and scientific colleague David Goodall, whose family he sometimes accompanied on their annual holidays.
Bill's mother was vivacious with an enquiring mind and a spirit of adventure. She was deeply interested in religion but, as judged by the number of Christian denominations she espoused and abandoned, none were to her satisfaction. Neither was the Buddhism that she went to India to investigate. In later life, to the amusement of her son, Clara took to holding séances. She followed Bill to Australia in 1966 where, in both Brisbane and Townsville, the two shared houses that had been subdivided so each could lead an independent life. Nonetheless in later life they shared a housekeeper and, following his mother's death in 1976, Bill offered her the now unoccupied flat. The arrangement was mutually satisfactory for as it happens she was obliged to move from her usual accommodation and he still required a housekeeper.
Bill was an excellent cook and in Townsville gave his neighbours Christmas puddings and treated a select group of friends to a very formal traditional Christmas dinner.
With the assistance of scholarships, Bill was educated first at the Stationers' Company's School in north-east London, where he was a brilliant student, and then at the Imperial College of Science and Technology, London, where he graduated BSc with First Class Honours in 1933. He obtained a PhD in 1940 and in 1956 was awarded a DSc by the University of London. He was also an Associate of the Royal College of Science (ARCS 1933) and a Diplomate of the Imperial College (DIC 1940). In 1973 he was awarded a Doctorate of Science (honoris causa) by the University of Queensland in acknowledgement of his unstinting advice to its postgraduate students.
He was a member of the Society for Experimental Biology whose journal he for a short time edited and a member of both the Biometric and the Classification Societies in addition to being a Fellow of the Institute of Biology and of the Linnean Society (London). On moving to Australia he became a member of ANZAAS and in 1966 served as President of Section M. In 1980 Bill was admitted to the Order of the British Empire. The citation accompanying the award noted that he was retiring as a 'Research Fellow, Commonwealth Scientific and Industrial Research Organisation Division of Tropical Crops and Pastures', and 'was a pioneer in the application of computer science to agricultural and biological problems'.
Bill's professional life fell into two separate but overlapping phases. One was played out in England and the other in Australia. His only forays elsewhere as a scientist were a visit to Lahore in 1962 to attend a conference on 'The Role of Science in the Development of Natural Resources with Particular Reference to Pakistan, Iran and Turkey' and in 1968 a short lecture tour to South Africa where he was attached for five weeks to the Department of Botany at the University of Cape Town.
In England, except for a period of four years in the Army, Bill was an academic botanist and taught at Imperial College (1933-36), Sir John Cass Technical College (1936-40), Bedford College for Women (1946-51) and the University of Southampton (1951-65) where he was Professor and Head of Department. His fourteen years in Southampton were busy and happy times. He enjoyed his teaching and built up the Department by attracting active and able staff and postgraduate students of whom several became collaborators. In the early days at Southampton Bill maintained his research interest in plant physiology, but he encouraged others to pursue their individual interests while promoting collaborative projects in which he was often involved. He also supported his staff with their own courses and regularly visited Port Erin as a member of the annual algology excursion. The Department expanded under his leadership and at the date of his resignation he was involved in planning new accommodation. However, his life was not all work and no play for he organized at least one staff-student revue in which his involvement included teaching many of the cast to dance. At about this time he also successfully engaged in competition ballroom dancing for which he won a silver medal.
Concurrently with his teaching and other university duties Bill served on the Agricultural Research Council, being a member of its Potato Marketing Board Research and Development Committee, and of the Governing Body of the Glasshouse Crops Research Institute.
Notwithstanding this busy life, Bill also made time to indulge an interest in logic. He was fortunate in that Anthony Manser of the Philosophy Department at Southampton was a keen supporter of the controversial logician and metaphysician F.H.Bradley whom they both much admired. In his Principles of Logic Bradley had been unorthodox in beginning with a chapter entitled 'The General Nature of Judgement' and then going on to reduce the standard tripartite division of logic into terms, propositions and inferences to two, namely judgement and inference.(1) When Bill turned to the study of numerical classifications he found in Bradley a rationale for abandoning statistical tests as a basis for determining the validity or otherwise of classifications.
Bill also found in Bradley a philosophical ally who was a fellow logical positivist. The two shared many ideas in common including the one expressed by Bradley as follows: 'In the sciences we know, for the most part the end we aim at; and, knowing this end, we are able to test and to measure the means. But in religion it is precisely the chief end on which we are not clear'.(2) Bill expressed similar ideas in several of his Australian radio broadcasts.
Shortly after the outbreak of war in 1939, Bill enlisted. He did not declare his academic qualifications and so began his military career as a private. Not surprisingly, his scholarship could not be disguised indefinitely, and on 20 October 1941 he was commissioned as a 2nd Lieutenant in the Royal Army Ordnance Corps. A year later he was appointed Acting Captain at the Ministry of Supply's Air Defence Research and Development Establishment. Early in 1943 he transferred to a new service, the Royal Electrical and Mechanical Engineers, which was concerned with the development of radar, and in August 1944 he was appointed a war substantive captain and temporary major in what was now known as the Radar Research and Development Establishment.(3) One task of this Establishment was to train radio maintenance officers, many of whose backgrounds were in the biological rather than the physical sciences.(4)
In 1963 Godfrey Lance, who had co-operated with Bill at Southampton from the mid-1950s until 1960, was appointed Chief of CSIRO's Division of Computing Research in Canberra. The Division had sections located in all state capitals but as a matter of policy all computing research was carried out at the Division's headquarters in Canberra. Lance appreciated that Bill's talents would be invaluable to Australia and in 1965 invited him to visit for a few months. During this visit Bill lectured and met many scientists during numerous visits to CSIRO Divisions, including one to the Division of Irrigation at Griffith in New South Wales. There Eric Hoare, also an erstwhile Englishman, 'beered' and dined Bill and took him on an extensive trip into the outback. On waving 'Goodbye' at the end of Bill's visit to Australia, Lance promised to keep in touch but was pleasantly surprised when, within a week, Bill wrote asking if he could come to CSIRO on a permanent basis. The Executive was easily persuaded to offer an appointment at DCR in Canberra as a Principal Research Scientist. Bill accepted this offer and in 1966 he migrated to Australia.
On his appointment to CSIRO Bill signed an affirmation of allegiance, thereby becoming an Australian citizen. This move he never regretted and in a letter to DrVictor Burgmann following his election as a Fellow of the Australian Academy of Science wrote: 'The Fellowship gives me great pleasure, Australia has been very kind to me, and I have always been anxious to repay the debt'.
With Bill's departure from Southampton, the University lost an able professor and the Agricultural Research Council lost a valued member. In response to a request for a reference for the Canberra position, Sir Gordon Cox, Chairman of the Agricultural Research Council, wrote: 'In my work I have to learn whether I can rely on the scientific judgement of others, and I can say without hesitation that I know no one in whose scientific judgement I have greater confidence than Williams'.(5) These words were a signal tribute to the high esteem in which Bill was held in England. In Canberra, however, he found the climate too cold and after two years he transferred to the Division of Tropical Pastures in Brisbane where he worked for five years until his retirement. He then moved to the warmer climes of Townsville. When Les Edye heard of Bill's plans to go north, he kindly offered to help in any way possible. 'Well there is just one thing I'd appreciate', said Bill, 'Could you find me a glasshouse please?' 'Of course', replied Les, 'do you plan to return to experimental botany after this long break?' 'No!', retorted Bill, 'I want to live in it'.
A thorough planner, before moving to Townsville from Brisbane Bill learnt to swim along with a group of children at the early morning classes held at the Toowong Public Pool. However, having acquired the skill, it was never put to use when he travelled north. He claimed that, like cats, he had an aversion to water. Although Bill officially retired at 60, during the following twenty years his research publication rate scarcely declined. This productivity resulted partly from his having been appointed a consultant to the Davies Laboratory in Townsville. This was a branch of the Division of Tropical Pastures and so he was familiar with its research programme. The consultancy supplemented his pension and also enabled him to complete projects begun in Brisbane. In addition he was a consultant to the Australian Institute of Marine Science and informally offered advice to students and staff of the James Cook University of North Queensland.
Although Bill's contribution to most of the publications that appeared in his retirement was methodological, his collaboration with Les Edye was an exception in that it was closely allied to agricultural problems akin to those with which he had been involved in England as a member of the Agricultural Research Council. In a series of seventeen papers dealing with introductions of Glycine and Stylosanthes, they demonstrated the power of clustering analysis for making agronomic sense of the variation observed in field trials. The economic importance of this research was recognised when Les Edye was successfully nominated for the prestigious Sir Ian McLennan Achievement Award and the Cattleman's Union Industry Research Medal for his role in effecting pasture improvement in northern Australia.
Bill's research output as judged from his publications was prodigious, especially when it is realised that he normally worked only office hours. That his opinion was widely respected and sought is indicated by the many papers of which he was co-author. In general Bill and his collaborators were careful to ensure that credit was given where credit was due so, for example, a paper by 'Lance and Williams' would normally be one in which the computational methodology was developed mainly by Lance whereas one by 'Williams and Lance' would be an application of the methods to some scientific subject, Williams being most influential in how the techniques were applied.
Bill's research interests were broad but fall into two quite separate categories, plant physiology and pattern analysis.
Aside from brief encounters with urease metabolism and the role of aerenchyma, Bill's studies in the former discipline were mostly concerned with leaf expansion and stomatal action, topics into which he had been indoctrinated by Professor O.V.S. Heath at Imperial College. Bill's research on these subjects was competent and meticulously quantified but nonetheless somewhat pedestrian by subsequent standards, for at that time the transmission electron microscope was not generally available and the refined biochemical tools available today had not been developed. For one usually so far-sighted, it is surprising that when he proposed a new theory of stomatal mechanism, Bill did not support his views with adequate original data. Instead he raised objections to the classical theory on the grounds that although the carbohydrate and aperture changes were roughly correlated, the correspondence was not exact, a discrepancy he regarded as 'most unanswerable'. As Levitt noted shortly afterwards in a review, carbohydrates include sugars as well as starch and at the time the 'exact measurements of carbohydrate changes in the guard cells, have not as yet, been achieved'.
Bill's analysis of the available data was adequate but reads a little like a Conan Doyle short story. The similarity is not unexpected for Bill was a great admirer of Conan Doyle whom he regarded as a competent scientist. Although listed in several publications as a sometime secretary of the Sherlock Holmes Society, no record of his membership can be located.(6) Therein lies a mystery worthy of the great detective himself.
Once given the opportunity to employ his mathematical as well as his biochemical skills, Bill did so with finesse, as in a paper dealing with the transpiration of Pelargonium leaves undergoing wilting. Noting that the process involved the loss of both heat and mass to the surrounding air, he undertook a series of experiments that measured only the water loss. His neglecting heat loss may indicate that he regarded this as of negligible significance compared with that of water loss, or it may reflect an unconscious but persistent preoccupation with stomatal mechanisms. Nonetheless the experiments performed were elegant and enabled physical meanings to be given to the arbitrary constants in a series of transpiration equations proposed by Hygen.(7) It is a testimony to Bill's unfailing courtesy that his paper, in which he was mildly critical of Hygen, was published with that author's knowledge and approval.
Notwithstanding Bill's appreciation of the physical sciences, it is clear that by the mid-1950s he had become concerned that modern technology 'had persuaded many botanists that all science is measurement and that nothing else was respectable [and] that botany should in Whitehead's words "ape the manner of physics", in that we should first study the simple system and move on to complex systems only after their simple counterparts were fully understood'. This he regarded as the latest but by no means the least of the great fallacies of science. Bill also became disillusioned with 'the increasing tendency to insist that teachers spend a period undergoing diploma courses in Education, and the occasional suggestions that lecturers should be taught how to lecture'.
After ten years of studying stomatal behaviour, Bill was growing tired of the subject. Therefore when approached by his colleague Joyce Lambert for advice on how to 'take unbiased samples of vegetation and sort them without any preconceived ideas as to how they should be grouped', he readily agreed to co-operate. So began the second stage of his scientific career.
As an undergraduate Bill had attended Eric Ashby's lectures on statistical ecology(8) and so was familiar with the subject as it had been grappled with up to that date. Furthermore, because he exchanged reprints with David Goodall, a botanist inclined to statistics, he had a copy of his friend's brilliant paper(9) in which a solution to the problem raised by Joyce Lambert was offered.
Random sampling within an area selected for study will take care of bias in the choice of sites from which to collect data. Such sampling was commonly practised at the time but was used largely as a basis for the description of vegetation that had been selected visually for its apparent uniformity. A few ecologists had counted the numbers of individuals of species in their quadrats and shown that, with rare exceptions, they were not randomly distributed. In general the lack of randomness was ascribed to the biology of the species rather than the diversity of the environment, which by definition was assumed to be uniform.
However, non-randomness of species distributions, especially if considered in pairs, would confirm that the environment was non-uniform. This theory had been tested by a number of ecologists, especially in Scandinavia, and was employed by Goodall in his study of mallee vegetation in Victoria. Using chi-squared as a measure of association between species, based on their presence or absence in randomly placed quadrats, he demonstrated that the area was vegetationally diverse since more significant correlations were present than would be expected by chance alone, were the species randomly distributed. Having established that the area under study was heterogeneous, Goodall then subdivided the total number of quadrats into subsets on the basis of a number of objective criteria. Here was an answer to the question that Joyce Lambert had raised.
Bill, however, although in sympathy with the general principle proposed by Goodall for subdividing the total quadrat number into subsets, found it to be inadequate because the results were not clear-cut. Whereas Goodall generally employed the presence or absence of the commonest species as the basis for constructing subsets, Bill hit upon the idea of using the species for which the chi-squared values summed over all other species was a maximum. Such a decision-making procedure involved all rather than one species. Initially he was unaware of the additive property of chi-square and employed it only on intuition, prompted by his familiarity with the methodology of factoranalysis.
Bill rapidly mastered the art of writing programs for the then recently installed Ferranti Pegasus Computer at Southampton. Furthermore, he was highly skilled in the art, as was shown by his Association Analysis program which, written in machine language, had only one bug and this one that appeared only when more than 38 species were to be considered.
In Godfrey Lance, at that time the Director of the Computer Centre, Bill found a congenial colleague. The two went on to collaborate fruitfully for nearly thirty years and to co-author many joint papers. The last of these was a note in which they commented upon the success of their 'mixed data classificatory program' which had been published in the first volume of the Australian Computer Journal and had become a 'citation classic', having been quoted more than 145 times in a period of twenty years.(10)
As a logical positivist, Bill found in classificatory problems a perfect outlet for his interest in logic and an opportunity to apply his considerable mathematical skills. He soon expanded from the manipulation of binary to continuous and multistate variables, and became immersed in multivariate analysis where the psychologists had already developed a formidable array of techniques centred on factor analysis and principal components (eigen vectors). Information measures soon joined his armoury of techniques.
Quite early in his studies, Bill suspected that several of the clustering strategies then in common use were related in a simple fashion. This suspicion proved to be true and with the co-operation of Godfrey Lance he showed that five of the strategiesnearest and furthest neighbour, median and group average, and median (but only when similarities were based on distance-squared)were variants of a simple linear model. This model involved four variables, one of which was the difference between a single pair and one of the other three. Each variable was weighted by a constant (a1,a2,b,g). Varying the magnitude of the constants altered the clustering strategy and if their values were restricted to ± ½, ± ¼ or 0, a suitable choice of these constants would generate one of the five strategies mentioned above.
As each of the strategies results in a different intensity of clustering, the linear model was later amended to allow the user to obtain clustering intensities intermediate between those commonly employed. Such 'flexible sorting' was achieved largely by fixing the values of a1 and a2 whilst allowing b to assume a value between -1 and +1 (Lance & Williams 1967). It was later shown that when b was positive, the space in which the clusters formed contracted during the clustering process, when zero the space was conserved, and when negative it was dilated as in clustering strategies based upon information measures. Because 'flexible sorting' leaves the choice of clustering intensity to the operator, it removes some of the objectivity from the methodology although the results are still reproducible provided b is defined.
From the beginning, Bill and his collaborators appreciated that the classificatory strategies being developed were applicable not only to ecological data but to almost any set of objects for each member of which a series of comparable observations was available. Accordingly, he and Joyce Lambert wrote on the application of multivariate methods in taxonomy and shortly afterwards, when he visited Australia, Bill undertook a numerical classification of the algal genus Chlorodesmis. Bill's enthusiasm was infectious and extended to others in the department at Southampton. Amongst these was Leslie Watson who was interested in higher-level classifications where it was easy to become overwhelmed with data. The two collaborated in a study concerning angiosperm classification. Shortly after Bill's migration to Australia, Watson took up a position at the Australian National University where he considerably extended the application of numerical methods to the classification of the world's grasses.(11)
It is perhaps surprising that, as one who was a pioneer of the subject, Bill did not write a substantial text. However, that he did not do so is explained by his response when asked if he would care to be listed as a co-author of An Introduction to Numerical Classification,(12) during the writing of which he had given unstintingly of his time and expertise to the authors. He declined on the grounds that he had contracted to write a larger treatise for Wiley, the prestigious American publisher.
That book was never written for, as he once said to me, he was incapable of writing more than ten pages on any subject. Instead, he edited a set of conference papers that were published as a book.(13) To this he contributed twelve chapters dealing with methodology. They varied in length from five to twelve (7.33 ± 2.19) pages, thereby falling within his self-assessed limit of ten pages on any topic. Published shortly after his retirement, the book is an excellent summary of his views on the use of pattern analysis for investigating multivariate data.
Seeking patterns in other people's data suited Bill's temperament and became central to his research in Australia. It led to the production of many co-authored papers, few of which he initiated unless the topic involved extending the application of the methodology of classificatory strategies that he had helped to devise.
In general, Bill was little interested in the source of the data. According to Len Webb, who once inveigled him into the field to become personally acquainted with the vegetation they were jointly describing, Bill said the scene meant nothing to him and it was only the data that were meaningful. This indifference may have resulted from his being interested solely in numbers, or it may have stemmed from his colour-blindness. However, the latter disability is not necessarily a serious hindrance to field work as shown by Professor Desmond Herbert who, although colour-blind, was for many years an active field botanist in Queensland.(14) Bill's indifference to the source of data was almost certainly associated with his interest in logic as expressed in algebraic formulations: he once told Norm Duke at the Australian Institute of Marine Science that he was almost blind to diagrams.
The modesty and willingness with which Bill applied himself to data collected but not analysed by others is remarkable. Through his ability to detect patterns in a very wide range of data, especially that collected by agriculturalists, he rescued from oblivion a vast amount of valuable research not amenable to standard statistical analysis. His appointment to the then Division of Tropical Pastures of CSIRO was largely predicated on the understanding that he would look through their unpublished files and extract anything publishable. So successful was he in this regard that on any cost-benefit analysis his appointment must be regarded as akin to a bargain.
Bill's analyses were not always appreciated, however, by those whose view of research was to set up testable hypotheses. Whilst there is some justification for such criticism, his claim was that it was better that patterns, no matter how weak, be detected than that the data remain unresolved. Bill regularly stressed that computer classifications are in no sense absolute and carry no authority but suggest 'to a user, what boundaries between groups might repay further study'. That is, the classifications should be seen primarily as 'hypothesis generating'. Since it is possible to provide a limitless number of classifications for the same set of data, it is important that there be some guide as to which are useful. The decision must always be made by the user but nonetheless an attempt was made to evaluate selected classifications in terms of their profitability. The concept was criticized by Goodall(15) who suggested, as an alternative, the utility of a classification.
Although neither of these concepts has persisted in a formal sense, that of utility is still recognised, albeit unwittingly, for once patterns have been detected it is often possible to collect further data suitable for statistical analysis. The ability of a program to locate groups in a set of data depends primarily upon the similarity measure chosen and the clustering strategy employed. Standardization of the data as when combining attributes into a single index automatically leads to the weighting of characters and in some circumstances the numbers of closely similar individuals in the sample may influence the order in which the groups unite. Both of these phenomena are basically properties of the data rather than of the clustering strategy.
There is a classic example that illustrates the over-riding role of the data. It involved the classification of quadrat data collected in the Norfolk Broads in England and recording the presence or absence of species at a number of localities. Surprisingly, the results highlighted the parish boundaries rather than any of the more obvious environmental variables. No-one believed that the Pegasus computer was divinely inspired and so Bill asked Joyce Lambert to review her data. It transpired that each parish had treated its land in a unique way. For example, some had done drainage work and others had used fertilizer. These differences in historical land treatment had influenced the environment sufficiently to affect where the species were growing. The present pattern of species distributions therefore provided the demarcations that showed the parish boundaries. Pegasus was not, after all, divinely inspired! Here was an example of the computer producing a classification that led to the testable hypothesis that the observed species distributions reflected different patterns of historical land usage.
Although many of the earlier practitioners of numerical classification stressed that the methodology was free of bias in that they did not weight the characters employed, Bill, because of his mathematical insight, appreciated that bias could arise as a consequence of the similarity measure employed and the clustering strategy adopted. To both of these problems he brought new insights. The 'Canberra Metric' was developed for combining quantitative characters into a single index in such a way that no single character could completely dominate the others. The Metric is particularly useful for handling data such as annual rainfalls where a record for a single year may differ markedly from the remainder. The concept of 'group-size dependence' contributed to a clearer understanding as to how it came about that composition of a group depended not only on the similarities of its members but also on the numbers in the groups. The matter was particularly significant when information measures were involved and made it clear that the branching pattern of the dendrogram was often as important as the composition of its terminal branches. Having found groups in the data it is important to determine their relationships, and here, too, Bill displayed inventiveness in devising methods for comparing the pathways by which the groups are linked. These pathways are usually presented as rooted tree structures, where the root is the sum total of objects to be classified and the tips of the branches are the individual objects. The number of groups recognised is usually determined arbitrarily by truncating the tree. The relationships between the tips of the branches (individuals) may be expressed in terms of the number of nodes passed in travelling from one tip to another. If the individuals are taxa it is tempting to treat the dendrogram as if it reflected an evolutionary tree. That it does not is clear for the tree root represents the total population under investigation.
However, when the individuals are taxa the differences between them may be treated as phylogenetic distances and the minimum spanning tree linking them could be regarded as reflecting their evolutionary relationships, taxa near to one another being more closely related than those further apart. The choice of root for such a tree is problematical for any branch tip may serve this role. The problem is resolvable only in terms of evolutionary concepts and, in particular, which taxa or characters are accepted as primitive.
Here there is an opportunity to unite numerical methods of classification with evolutionary theory. However, although aware of the development of cladistics as a taxonomic tool, Bill remained an onlooker rather than a player in this new field. That he stood on the sidelines rather than enter the fray supports the view that, at least in later life, he was basically interested in the detection of patterns rather than the mechanisms by which they were generated.
This interest in the search for patterns, together with his genial personality and penetrating insight, made him an attractive colleague to a very diverse group of scientists. The range of topics on which he wrote is astonishingly broad. It includes papers on temperate and tropical ecology, benthos, bird and foram distribution, the taxonomy of grasses, algae and monocotyledons, grazing and fertilizer trials, crop and silage chemistry, the ripening and packaging of fruit, the behavioural outcome of parental deprivation, and the qualities of travel agents. This list does not include the numerous methodological papers he wrote.
An alternative explanation of his indifference to cladistics may be that the development of the subject came at a time when he was committed to the very time-consuming process of acquiring the qualifications necessary to practise as a professional musician.
Bill's passion for music manifested itself at an early age. He taught himself to play the piano and enjoyed singing, as did his friend David Goodall. The two used often to sing duets to Bill's accompaniment on the Goodall family piano. Bill also taught himself to read and write music. David's sister Joyce remembers him writing a simplified version of 'Rhapsody in Blue' for her twelve-year-old friend Connie, who still plays the tune seventy years later at meetings of her local 'Pensioner's Athletic Club'.
However, it was not until he came to Australia that Bill studied the piano seriously, electing to take lessons from Larry Sitsky of the then Canberra College of Music. From then on he was a dedicated musician. After transferring to Brisbane, Bill continued his keyboard studies with Alan Lane of the Queensland Conservatorium. The two quickly found common interests, particularly in their regard for twentieth-century music, and their weekly meetings became more than just instrumental instruction. Alan says: 'Bill's mixture of talent, intelligence, character and open self-criticism was an ideal platform for rapid progress and the development of professional awareness. His wealth of experience and understanding allowed his piano studies to be absorbed within a much wider framework than is usual.' It is interesting to note that the one style of music that was quite alien to his personal outlook was that of the early Romantics, while he never quite came to terms with the particular performance demands of Chopin.
Although in Queensland there are no government requirements or registration for setting up as a private music teacher or performer, Bill was well aware of the usefulness of publicly recognised credentials and he gained both his AMusA and LMusA performance diplomas with very high results. Once resident in Townsville, he took pupils and was for a time chairman of the Townsville Music Teachers' Association. As a result of these activities he discovered a number of local pianists who would be delighted to play a concerto movement in public but could not afford the costs associated with entering the nearest competitions, which were held in Brisbane. In 1980 Bill overcame their problem by organizing the first North Queensland Piano Competition, choosing Alan Lane as the adjudicator. The concept was popular and in 1988 it expanded into the North Queensland Concerto and Vocal Competition, with Bill as patron.
He also fostered the musical life of the city by being deeply involved in the affairs of the Townsville Community Music Centre, by serving as a music presenter on Radio 4TTT, and by arranging musical evenings at his home. In appreciation of these services, Bill in 1991 received a Townsville City Council Arts, Culture and Entertainment Award. Furthermore, he was generous with his extensive library of books, scores and records and gave many people a key to his house so they could use the collection in his absence. Fortunately the collection has been preserved, for Bill bequeathed it to the James Cook University of North Queensland where it is available in the Department of Music and Fine Arts.
Because it was Bill's custom to play mezzo piano, he was in demand as an accompanist. This preference for mezzo piano was in stark contrast to that of his friend and near neighbour, the distinguished concert pianist Nancy Weir, who preferred the forte piano. In consequence they rarely played duets though they were often soloists on the same program.
In addition to music, Bill in his younger days was active in the theatre. In about 1930 he and David Goodall played the roles of Bottom (DWG) and Oberon (WTW) in a North London production of 'A Midsummer Night's Dream'. The two were members of the Imperial College Musical and Dramatic Society of which Bill, who had now abandoned the name 'Willie' by which he was known until entering university, was President. As always he took the position seriously and not only played roles, including that of Hahalaba in Lord Dunsany's 'The Jest of Hahalaba', but also directed Patrick Hamilton's 'Rope' in which David Goodall played a role.
The Thespian skills acquired during these years were often employed, for example at Southampton where he and Tony Manser jointly produced 'Age Cannot Wither', a student-staff revue. The book and the lyrics on which the revue was based were written by Bill and David Cook, while the music was written by Bill and Kenneth Brooks. Unfortunately the book has not been located but a recording was produced of which there is a copy in Townsville. Thereon Bill is recorded as a performer in three items. These same Thespian skills contributed to his fine lecturing style and stood him in good stead whenever he appeared on television shows or performed impromptu on the piano in the bars of innumerable pubs.
Given the great diversity of Bill's activities, it would be easy to overlook that his was an ordered life in which he assumed the role of critic and communicator. Whether the scene was set on the concert platform, stage, lecture podium, television or radio studio, the pages of a scholarly journal, the columns of a newspaper or the public bar in one of his favourite pubs, Bill had an opinion to offer. Stomates, pattern analysis, science education or fiction and politics were all grist to the mill.
In England, Bill for many years took part in BBC television and radio programmes including the 'Brains Trust'. In Australia, the ABC radio series 'Insight' and 'Ockam's Razor' gave him the opportunity to display his wit and penetrating insights into many controversial issues on which a scientist might be expected to have an informed opinion. He was at his best when producing short pithy articles such as those published in The Listener, and the Australian radio talks reproduced in his book, The Four Prisons of Man, and Other Insights.
In one of his last broadcasts, 'The Tape of Many Colours', he tackled two controversial topical issues where he felt emotion rather than logic had gained the upper hand. Logic, his lifelong companion, compelled him to be critical of both the 'Green Lobby' and those concerned about the preservation of 'sacred sites'. The former did not appreciate his view that, without management, the character of vegetation will change, his comments being based upon English experiences where strict conservation had led to the extinction of the endangered species. The latter were offended by his remark, 'I am not convinced that development should ever be held up by religious scruples'. In taking these positions Bill realised he was being contentious, but he stuck to his belief in the power of logic. The word 'belief' is appropriate here, even though it carries a religious connotation. While Bill regarded 'all religious beliefs to be irrational'(16) and therefore the antithesis of the logical positivism he espoused, his attachment to logic was almost dogmatic. However his views were never presented dogmatically.
Amiable at all times, Bill was nonetheless eccentric in several aspects of his behaviour. His reluctance to wear shoes or a tie in Queensland and elsewhere could be regarded variously as an affectation, an indifference to dress codes, or a conscious desire to be unconventionalto name a few of the possibilities. On one occasion he was not allowed to eat breakfast at the Hotel Windsor in Melbourne because he was deemed to be improperly attired. At the 'Royal Exchange' in Brisbane, however, he always wore sandals and so conformed with the pub culture he so much admired. Such conformity was in marked contrast to his behaviour in Mareeba where, after leaving the pub, he repaired to the median nature strip, divested himself of his shirt and lay on the lawn to sleep in the sun. He always feigned surprise that he was questioned by the police for this action, although a person of his education and the holder of a commission in the British Army is unlikely to have been ignorant of the law appertaining to vagrancy.
Bill's eccentricity may have reflected a personality in which living and acting unconsciously intermingled so that all the world became a stage. When living in Southampton, he breakfasted at a Truckies' Cafe where he kept his own jar of marmalade. In Townsville he shunned the RSL Club of which he was entitled to be a member in favour of a local pub, and in later life took to leading his dog around the streets clad in shorts and a singlet. These very visible actions are befitting of an actor. A love of the theatrical was also reflected in his habit of crawling down the corridor of the Botany Department in Southampton so as to go to lunch without being seen by Joyce Lambert whose half-glass door he had to pass.(17)
An intensely private person, Bill revealed little of himself in letters or conversation. Still, he was not reticent, as may be seen from his writing to fellow dog-lover Nancy Weir on the death of his dog, and his sending Joyce Lambert transcripts of his Australian radio broadcasts and detailed accounts of his dogs. That he had difficulty in personal relationships is nevertheless suggested by his going to the pub rather than to the wake after his father's funeral.(18)
Such behaviour was surprising considering his close relationship with his mother, but reflected his deep attachment to the 'pub culture' that he discussed so eloquently in the third of his talks in the ABC Insight series, 'The Three Cultures'. Therein he described 'the direct personal culture of the working man's public bar. It is a culture that has been much reviled, but little understood.' Bill saw it as a 'culture of great honesty' and 'great kindness to all frail and helpless things: to small children, dogs and especially aged parents'. He also described it as a 'gladiatorial culture' but one 'of fierce loyalties'. Furthermore, he said, 'It is a unisexual culture, such as would have been understood in ancient Greece. It is in no sense homosexualthough I suppose the Freudians would try to make it sobut it is understanding of such things and is tolerant.' It was perhaps this tolerance that he so much appreciated, for in the public bar he would have found few people with whom to share anything of his professional life.
It is unfortunate that Bill's account of 'The Bernie-and-Bill Pub Pilgrimage' was distributed privately and then to only a select few. Therein is 'The Record of a Remarkable Journey by B. McMullen and W.T. Williams from Brisbane to Cooktown and Return in a Morris 850 Mini-Minor, covering Twenty-nine Days (12 June 1971 to 10 July 1971), Three Thousand Four Hundred and Seven Miles, and Two Hundred and Sixty-Three Pubs'. No hotel was included in the pilgrimage unless it had a genuine public bar, at which the travellers consumed at least one five-ounce glass of beer or one half-Scotch.
In contrast to his apparent feelings of insecurity with people, Bill had no hesitancy in accepting the companionship of dogs. He spoke at length about this in a broadcast entitled 'A Man and His Dog' wherein he wrote: 'But a dog offers silent companionship; and in that gracious silence there need be no more than a gentle scratch behind the ear, acknowledged by an affectionate lick. No more is asked, and no more is needed.' Later in the broadcast he gave a poignant account of his feelings after agreeing to have his dog put down on account of its infirmity: 'There comes the dreadful day when the vet shakes his head, and says he's sorry, but there is nothing more he can do. And so, fighting back the tears, you bow to the inevitable, give a last caress and murmur of farewell, as, desolate, you watch an important part of your life being led away. I suppose it might help if you were religious; for then the possibility of reunion would not be quite inconceivable.'
In the middle of the afternoon a few weeks after this broadcast, Bill tripped over a panel of wire fence lying on a pathway in the grounds of the Causeway Hotel. In falling he sustained serious injuries and died five days later. Joyce Ashby remembers the teenaged Bill as a 'kind, thoughtful, amusing person'.(19) These admirable qualities he retained throughout a long and highly productive life. Bill never married.
This memoir was originally published in Historical Records of Australian Science, Vol.12, No.1, 1998. It was written by H. Trevor Clifford, Department of Geology, Queensland Museum.
I am grateful to the following for the advice and information they gave so cheerfully during the writing of this memoir: Mike Dale, Les Edye, David Goodall, Merv Hegarty, Brian Hopkins, Jiro Kikkawa, Joyce Lambert, Godfrey Lance, Alan Lane, Pat Newman, Mary-Lou Schönfeldt, Kathy Stephens, Anne Tuppack, Les Watson, Len Webb, Nancy Weir, Robyn Williams. The photograph was taken by Patti Holden in Townsville on 14 July 1988.
© 2020 Australian Academy of Science