Lost—or liberated?—without theory Jan Gulliksen, Inger Boivie Deptartment for Information Technology/HCI, Uppsala University, Sweden Jan.Gulliksen@hci.uu.se , inger.boivie@hci.uu.se Liam Bannon Deptartment of Computer Science & Information Systems, University of Limerick, Ireland liam.bannon@ul.ie Lidia Oshlyansky UCL Interaction Centre, University College London, England lidiaosh@yahoo.com Harold Thimbleby Department of Computer Science, University of Wales Swansea, Wales h.thimbleby@swansea.ac.uk HCI is diverse and rapidly developing. Should we celebrate its diversity, or should we require it to have a clear theoretical foundation? And, whatever our view of the discipline itself, how can we support each other (and especially our students) to work to high standards as the discipline develops? Keywords: HCI theory, interdisciplinarity, consensus, paradigm. 1. THE WORKSHOP What is your conceptual framework? Why is your work valid? What is your theory? Many PhD students, whose work is most sensitive to these questions, start out with a fairly well-defined research problem, often provided by the project funding the research. Finding a suitable theoretical or conceptual framework (and methods consistent with that framework) with which to address the research problem is, however, a very complex and difficult issue. Indeed, the relations between different theories, and between theory and application within HCI has long been a source of frustration, debate and problems [1–6]. This workshop will focus on some of the problems related to theory and the application of theory within HCI research. The workshop will make authoritative and positive recommendations to the community. The workshop particularly addresses the difficulties facing PhD students as well as researchers trying to navigate in a multi- disciplinary and multi-theoretical research area. There are very many theories and theoretical frameworks within HCI with different origins and traditions. In other areas like biology that have longer traditions (and a more stable world to research!) there are pretty clear subdisciplinary boundaries, and people can be clear they are botanists, geneticists, or biochemists without difficulty. It is rare for a competent botanist to be told that they ought to be studying cetaceans and that their work is bad until they do so; there may certainly be some interesting botanical issues with cetaceans that might be worthy of study — but that isn’t all there is in botany. Fanciful as that criticism seems, it or something similar happens all the time for both young and experienced HCI researchers. Many successful HCI researchers often learnt the rules of how to be successful within a single discipline (e.g., as undergraduates) before entering HCI; possibly they now impose these traditional criteria on others. In HCI, the boundaries seem to change continually: in the 1970s we were after principles and organisational issues; in the 80s we were after model human processors; in the 90s, phenomenology and situated action reigned; in the 2000s we have moved to activity theory and grounded theory — and the field has started to split, represented by the very different flavours of HCI, CHI, DIS, DSVIS, and so on. Grudin [3] is a mature and recent review of the main issues; Thomas and Thimbleby [7] is an example of alternative proposals, in their case suggesting there needs to be a “new usability” since the present one has problems. HCI includes designers, computer scientists, anthropologists, and psychologists as some of the major stakeholders. How do the different community traditions impact on HCI in practice? How do we address the problem of the incommensurability of different theoretical frameworks? Can theory itself be interdisciplinary? Few would argue that the theoretical framework underpinning software engineering approaches can be seamlessly merged with the social constructivism underpinning, for instance, anthropology. Yet HCI borrows extensively from both areas, and in a practical research or development project the different approaches often have to cohabit.