Patterns for HCI and Cognitive Dimensions: Two Halves of the Same Story?

This paper firstly surveys the search for Patterns and Pattern Languages (PL) in HCI, and examines some of the problems that the search has encountered. Secondly, some aspects of the Cognitive Dimensions (CD) framework are examined and I suggest that there may be a relationship between the two endeavours, to the probable enhancement of the pattern endeavour and the possible enhancement of the expression of the CD framework. The first half of the story To make a single narrative of the two endeavours of patterns in HCI and CDs, I shall have to start by telling two separate stories. However, these stories are not equal. I am an “insider” when it comes to patterns and an “outsider” with regard to CDs. So the patterns story is told from what I know ... What are Patterns and Pattern Languages? Patterns (as such) were defined and named by Christopher Alexander in his two works A Timeless Way of Building (Alexander, 1979) and A Pattern Language (Alexander, Ishikawa, & Silverstein, 1977) within the domain of architecture specifically and the built environment generally. They espouse an approach to design—which is codified in the patterns—that focuses on the interactions between the physical form of buildings and the way in which that form inhibits or facilitates various sorts of personal and social behaviour (Bayle et al., 1998). Important aspects of Alexander’s patterns are: • they were devised with the express intention of providing a common vocabulary between users and architects, as well among architects themselves • patterns are not created or invented; they are identified via an invariant principle (of good design) as manifest across different places and cultures (several examples are given in each pattern). • they are structured around the problems that designers face, and those problems are addressed by the provision of a “solution statement”. “Each pattern describes a problem which occurs over and over again in our environment, and then describes the core of the solution to that problem, in such a way that you can use this solution a million times over, without ever doing it the same way twice” (Alexander et al., 1977. p.x) Each “pattern” follows a prescribed form that is based on evidence for, and examples of, the use of the pattern, together with instructions for how to achieve its effect. Each pattern is named, and illustrated with both photograph and a diagram. NAME (usually describes the effect of using the pattern) A PHOTOGRAPH showing an archetypal example of the pattern in use AN INTRODUCTORY PARAGRAPH which sets the pattern in the context of other, larger scale patterns THE HEADLINE an encapsulation of the problem (one or two sentences) THE BODY of the problem (this can be many paragraphs long) THE SOLUTION the heart of the pattern, always stated in the form of an instruction A DIAGRAM shows the solution in the form of a diagram A CLOSING PARAGRAPH shows how this pattern fits with other, smaller patterns Figure 1. Alexandrian Pattern Format. The sections and descriptions are from Alexander (Alexander et al., 1977), the comments in brackets are mine The 253 patterns that Alexander identifies are collected together into a pattern “language”, which allows them to be used in combination with other patterns (often at different scales) so that whole environments can be constructed along these principles. Since Alexander’s specific construction, other groups have tried to replicate the approach in other domains (notably software, see: (Gamma, Helm, Johnson, & Vlissides, 1994),(Patterns Home Page, 2001) latterly pedagogy (The Pedagogical Patterns Project, 2001) and more recently, for “shaping the network society” (DIAC, 2002))