The interaction-design.org Encyclopedia of Human-Computer Interaction recently released a new chapter on Activity Theory by Victor Kaptelinin, one of the foremost experts on Activity Theory in the field of HCI, and I was asked to provide a response, which was published alongside the chapter. Victor's article does a terrific job of summarizing the framework's historical roots and its trajectory of adoption by the HCI domain, and he was also kind enough to highlight Giornata as one of his illustrations of how Activity Theory has been used to guide the design of information systems.
I used my response to talk about some of the challenges that I believe are inherent in creating activity-based computing systems that remain true to the nuances and complexities of Leontiev's framework (and Engeström's subsequent extensions). At the end of my response, I noted that one of the things that really jumped out at me when reading Victor's article was how unevenly the HCI field has supported one of the key elements of Activity Theory in both our activity-based analytic techniques and in our system designs: development. Activities are dynamic; mediating tools and the subjects both change as a side-effect of working towards a particular outcome (see also Alison Kidd's CHI paper on knowledge work for another way of framing this idea). But how well do our activity representations -- and the systems that we construct -- account for changes throughout an activity system over time, rather than encoding a bunch of (more-or-less static) relationships between applications, documents, events, and people in our electronic tools? Generally speaking, not very well.