Ethnoclassification, Pattern Languages, and MERLOT

Something clicked for me when I saw the reference on elearningpost to Peter Merholtz’s piece on the bag of keywords approach. The word Maish highlighted was “ethnoclassification.” This made me think of how you could unearth pattern languages in a learning object repository like MERLOT, MLX, or Pachyderm.Let’s use MERLOT as an example. Currently, MERLOT employs a taxonomy of subjects that is centrally managed. If you want them to add a new subcategory (e.g., ESL) you have to ask them to do it and hope they get around to it. But what if they supplemented the managed taxonomy with a user-driven distributed categorization? Further, what would happen if this distributed categorization were faceted? Imagine, for example, that there was a category of user-generated keywords related to learning activity type. We might start to see a pattern emerge of different educational activity types. To turn our list of patterns (which corresponds to our list of learning activity types) into a pattern language, we need only construct some guides about sequencing the activity types. Layering a rudimentary sequencing environment on top of a system like this would be simple; it could even be something as basic as a wiki.

Tracks in the mud.