In the insightful pasta and vinegar blog is an interesting discussion about the role of theory in interaction design based on the relevant discussion in a book by Nardi & Kaptelinin.
The problem is the one between reductionist, analytical approaches to practice and theory heavy, ethnographic approaches. The first translate easier to design and can be communicated to programmers. The latter are better at coping with the complexities of the real world, but cannot easily lead to designs.
For my phd I tried to base the problem on the notion of typicality. Models attempt to remove the particularities of a situation and transfer its solution space to a typical space. The more typical a model, the more generalizable it is. As an extreme example of typicality, we have the modelling of a worksystem through code: the variable referring to the worksystem “hospital” could also refer to the worksystem “bank”. In such an extreme case theory itself is extinct since “the theory is no longer a system of meaningful propositions, but one of sentences as sequences of words, which are in turn sequences of letters. We can tell [say] by references to the form alone which combinations of the words are sentences, which sentences are axioms, and which sentences follow as immediate consequences of others” (S.C. Kline cited in Rosen, 1988). The ultimate goal is a kind of algebra, where the symbols used are devoid of meaning: the ideal of positivism.
Methodologies according to their approach lead the analyst to model early (positivist, typical) or later (or not at all in the case of ethnography). I think there is a relationship between the typicality of a method/approach and its descriptive power (how much of the real world’s complexities are included in the model), which I attempted to represent in the following chart.
Nardi & Kaptelinin believe that Activity Theory combines the best of both worlds. I believe that AT leans heavily towards the ethnographic and is difficult to guide new designs (see for example Kaptelinin et al (1999)). The conclusion I reached, is that in order to reach an adequate analysis of a worksystem we have to combine methodologies even if they are contradictory.
References
- Kaptelinin, V., Nardi, B., & Macaulay, C. (1999). The activity checklist: a tool for representing the “space” of context. Interactions, 6(4), 27-39.
- Rosen, R. (1988). Effective Processes and Natural Law. In R. Herken (Ed.), The Universal Turing Machine, A Half Century Survey. Oxford: Oxford University Press.
Leave a Reply