H809: Week 9 - Activity Theory

April 2009

This should be last H809 blog post for a while!

Week 9 had just one reading that I rather enjoyed: Jonassen, D. and Rohrer-Murphy, L. (1999) ‘Activity theory as a framework for designing constructivist learning environments’, Educational Technology Research and Development, vol. 47, no. 1, pp. 61–79.

The paper started with a long summary of activity theory, discussing how it is based on the assumption that there is a 'unity of consciousness and activity' or 'a dynamic relationship between consciousness and activity' so that conscious learning emerges from activity rather than leading to it. I am struggling to figure out exactly what this means. For starters, I don't think the paper makes it clear what counts as an activity. Is reading a book or solving a mathematical problem in your head an activity? I am also wondering if it is actually trying to say that the only sensible way to define consciousness is as the activity of thinking? Or maybe that we only learn when we strive to attain goals? And does it imply that knowledge states do not exist or just that we shouldn't try and study them because there is no sensible way to?

Nevertheless, I think activity theory is still useful even if we ignore the assumptions that it is based on. It essentially says that to do research, you should study activities and their context. Activities consist of a hierarchy of actions which in turn are chains of operations. Activities become actions and then operations as they become more automatic. We can then examine the context of an activity by looking at the following factors and the relationships between them, usually illustrated in a triangular diagram:

Together all of these are known as an activity system. I see the factors as things that you should make sure that you look at if you want to understand the context in which an activity happens, and that if you haven't examined each of these, you probably haven't grasped the context properly. As well as these factors, there is also an emphasis in activity theory on understanding the history of the situation and how it has evolved over time. I think activity theory can perhaps be seen as an extension of social-cultural theory from the previous week which essentially treated the subject and tools as the sole object of study.

The paper goes on to describe an approach to building constructivist learning environments based on activity theory. The rough idea is that you base the environment on an authentic activity system in which the goal relates to what you want the students to learn. You then analyse that authentic activity in six stages - clarifying the purpose of the activity system, analysing the activity system using each of the factors above apart from tools, analysing the activity structure as activities/actions/operations, analysing the tools and mediators, analysing contextual bounds and finally analysing activity system dynamics (i.e. how the components affect eachother). For each of these stages, the paper gives the type of questions that you are trying to answer and the outcomes that you might expect. The authors emphasise that this is a slow process!

Once you have this information, you design your learning environment with the following components:

I think this is quite an interesting approach to designing the learning environment, with the caveat that I don't see it working in all contexts.

For instance, trying to think how you would apply this to pure mathematics, your first major challenge would be to pick a suitable problem. Your bigger goals as a mathematician are things like 'advance knowledge in a particular area of mathematics', 'get more publications' or 'find a good research problem' and I'm not convinced that problems like that will work because they are so difficult. A more precise problem that nobody has managed to solve yet isn't likely to get you far, whereas if you give students an already-solved problem, they will stumble on the solution as they research the area.

The other major issue is that the information resources would be massive and take you several years and lots of support to work through. Although if you are willing to assume that your students already have a high level of mathematical knowledge and ability, and don't mind if most folk get nowhere with the problems (which might be difficult if you need to assess the students in some way), then this type of approach might be possible. I had two summer jobs as a student at different institutions after my second and third years as an undergraduate, and both threw me into trying to do genuine pure mathematics research.