Stephen Downes

Blog post from Terry Anderson, and a companion blog post from Jon Dron, discussing the distinction between groups and networks. So far so good. But they include a third type of entity, 'the collective', with which I'm less comfortable. "Collectives are what happens to you while you are busy making other plans. Collectives are about aggregation. The logic of the collective is that of the set, not the network. A classic piece of collective software would be the recommender system or the tag cloud."

I can see the reasoning here, but the term 'collective' - with all its connotations to the epitome of group formation, communism, is a very bad choice. And to say 'collectives are about aggregation' and then to compare it to a recommender system is to (in my view) misunderstand how recommender systems work. That said - yes, I can see that there is this type of organization, too much like a group (being based on properties and aggregation) to be a network, too much like a network (having no internal cohesion or order) to be a group. To Terry and Jon: can you find some other name that doesn't have so much baggage? 'Collection', maybe? Finally, the chart of teaching and learning implications needs some work, mostly technical, making sure the right things are put in the right category - data mining, for example, has as much (more! on my account) to do with networks and connections than aggregations or collections; the same with 'search and query' (and especially Google's link-based sorting algorithm).