Graham Nudd became the Chair of Computer Science at Warwick in 1984 at a highly significant point in the department's history. Though the department had - and to this day enjoys - an international reputation for its research in theoretical computer science, there was at that time a danger of Computer Science becoming marginalised as larger influential departments such as Engineering and Business Studies were impatient to develop their own courses and resources based on a pragmatic approach to computing education tailored to their needs. 

Though Graham was originally recruited as a professor in Engineering, he had a specialist interest in architectures for high-performance computing and appreciated the importance of taking the science of computing and its implications for computing education seriously; he found Computer Science to be a more congenial environment in that respect. Throughout his tenure as Chair, which spanned two decades, he not only encouraged all academics to develop and champion computing as a broad independent discipline but also helped to establish the appropriate culture and infrastructure needed to address this objective. His vision embraced both the creation of a new purpose-built building with enhanced facilities for research and innovations in practices for teaching and administration (including joint undergraduate degrees with Engineering and with Business Studies) that enabled the department to act with a greater degree of autonomy and influence than a relatively small department might typically command. 

In his chairing role, he inevitably had to contend with internal controversy over policy and at the same time defend the department from sometimes unwelcome intervention by external agencies in the university. His long period of office is itself a testimony to his personal qualities and skills in leadership, management and diplomacy. The fact that he had come from a major US company at the forefront of technology gave him a holistic view of computing as a complex practice and discipline with both theoretical and empirical roots. This made him naturally wary of aspirations to make 'world-class' contributions to computing whilst working in isolation within academia - a pretension that, when appropriate, he would memorably parody with his mischievous sense of humour. He encouraged academics to build teams and to cooperate with industrial partners, setting an example in this respect that had an important positive impact on the research culture in the department; this not only benefited staff in their career development but also led to much richer engagement with the many exceptionally gifted computer science undergraduates, attracted during his period of leadership, who could contribute to research through projects that often drew them into postgraduate studies.

From my personal perspective - looking back - I recognise the day when Graham first came over to discuss the possibility of joining our department as one of the most significant moments in my 50 year association with Computer Science at Warwick. Though I never worked in Graham's group, and he and I came to the discipline from quite different directions respectively informed by engineering and mathematics, we shared a perception that the foundations of computer science have to be broader than the classical theoretical mathematical-logical framework alone affords - we bonded as kindred spirits through this conviction. The encouragement Graham gave me to follow my own research interests throughout his Chairmanship had a major impact on my subsequent career, as did his initiative in recruiting Steve Russ as a sympathetic colleague with a deep understanding of foundational issues in mathematics. It was fortunate for me that the programme of research into 'Empirical Modelling (EM)' as a plausible broader framework for computing that I had initiated with Graham's support in the 1980s was acknowledged by the eminent theoretical computer scientist Robin Milner to be well-aligned with his injunction that 'we must make an organised attempt to unite the development of theory with its experimental application'. In retrospect, the development of an introductory module on EM for the first departmental MSc programme on 'Parallel Computers and Computation' was an important turning point in my career. At the time, we were scarcely prepared for such a bold innovation in teaching - the topic itself was far too immature to have been approved as a conventional undergraduate module, we had only the most basic home-grown prototype environment to sustain practical work and had exceedingly limited teaching and technical support. In undertaking this improbable venture, we were indebted to Graham in two aspects: firstly, he had conceived the unconventional format for the MSc modules (taught over a period of five days with one week of practical work to follow - in our case typically during the Christmas vacation) in such a way as to foster experiment and expose - and invite - genuine research contributions; secondly, but for the enthusiasm with which he cajoled us into developing a module and lent his support and authority to its incorporation into the MSc programme, a successful outcome would have been inconceivable. As it was, our MSc module proved to be an invaluable vehicle through which we could consolidate on EM research that was very well-received and supported by many students: a large proportion typically opted to devote their dissertations to EM themes each year, and of these several proceeded to doctoral studies. Our MSc experience established the pattern for rich subsequent collaborations with research and project students that identified EM with a practice of sense-making ('making digital construals') that is complementary to 'computational thinking' and became the basis for a fourth year undergraduate module that was taught from 2004 to 2014. 

The importance of Graham's endorsement only became fully apparent when EM ceased to be recognised as an official Computer Science research and teaching programme a few years after his retirement; the rich set of tools, associated teaching resources and interested former and present researchers that it has left behind are a part of his legacy. As for Graham's concern to bring coherence to the theoretical and practical elements of computer science, this could hardly be more topical in a world where contemporary advances in AI highlight the potential and pitfalls of empirical techniques that more closely integrate engineering practices with abstract formal approaches. Above all, it is credit to Graham's vision that the current Computer Science department, with the blend of leading research and inspirational teaching that he helped to forge, is still well-suited to grappling with Milner’s challenge of 'uniting the development of theory with its experimental application'. In that respect, we are all - and myself especially - in his debt.