1	Comparing Two Approaches to Context: Realism and Constructivism Antti Oulasvirta Helsinki Institute for Information Technology Sakari Tamminen Helsinki University of Technology Kristina Höök DSV at Stockholm University/KTH and SICS
2	Two perspectives Realism Ontological. Context is real, it is structured, and the structure can be modelled by a learner. Contexts share properties (or universals) that exist independent of human interpretation Epistemological. Computers can perceive (i.e., recognize based on sensor data) context universals Pragmatic. If correctly recognized, computers can adapt their behaviour to the requirements of the context Constructivism Ontological. Contexts are constructed socially, in interaction with other agents in the world, and psychologically, in making sense of sense data Epistemological. Interpretation of context is always constituted within a frame of reference Pragmatic. Instead of labelling contexts, computers can provide resources for people themselves to create and maintain contexts in their action
3	Realism - applications Pro-active ubicomp (Tennenhouse) Affective computing (that Paul just talked about) Interruptability applications (Fogarty et al)
4	Constructivism - applications Leaving “surfaces” open for users to fill Social navigation Affective interaction where meaning arises as a consequence of use
5	Realism is winning grounds! Battling over power in the work place is different from battling over power in peoples’ lives Build good, sellable alternatives Use same technology and models, but turn the values around
6	Self-critique Constructivism may put interpretative and appropriation (and reflection!) burden on user If everything is fluid, defined and re-defined over and over by end-users – the role of the designer will be difficult? What do you provide and what to you leave open for end-users to appropriate? Expressions can be too strong (for appropriation) or too weak (becoming cumbersome and hard to interpret) In reflective design the designer has a HUGE ROLE! Too big? Modelling/ontological assumptions are unavoidable in computing
7	Example: seamlessness in mobile networks User always connected – seamlessly Should not have to worry about how and where Seams between networks and different kinds of network technology hidden from users’ view Pro-active network and device behaviour
8	Example: Seamfulness in MobiTip Seams (Bluetooth connectivity) made visible But model is that of characters populating your space affording communication – not true model!
9	Strategies for winning? Better systems that people buy? Pedagogical strategies for convincing our “realism”-colleagues? Picking up on techniques and design-material from the realism-strand and turning the values around? Power struggle between user and designer rather than union and employer?
10	Which values? Christiane Floyd said: “Are we heading for being disembodied cyborgs or rather for acknowledging that we are human beings enacting our lives in unique ways in communication with others and that computing is to be a part of it? [..] the key to critical action is authenticity, making our own values explicit, respecting those of others and reflecting so as to find common steps that we can take.”