Stephen Downes

This is a set of 364 photos from my four week cycling trip to Iceland, the land of ice and fire (and wind and cold and rain and hills). I hope you enjoy them.

This part of the ongoing effort by Creative Commons to redefine what it calls "the social contract of sharing." There is no such social contract; sharing doesn't involve any sort of quid pro quo, except perhaps from a commercial perspective. But more interesting than this ongoing media campaign is the idea of provenance itself, which Creative Commons is labeling 'attribution'. We never fully report provenance; the author does not cite his first grade English teacher, though without learning the language he could never have completed his book. The ideas, the turns of phrase, the cultural influences: these are all typically unremarked. Indeed, it's a social and political question that determines what should count as a citation and what should not (there's an explicit parallel in the logic of explanation, where we search for 'causes' in a complex soup of interactions). Why does Wikipedia cite newspapers and journal articles and not, say, my blog? Why is the person who publishes an article considered the discoverer of a phenomenon, or the coiner of a term, even when the actual act of discovery or coinage took place elsewhere? Is Creative Commons protecting creators with this focus on attribution, or is it entrenching privilege?

Colin Beer criticizes learning analytics. "While the GPS-style data may tell you that a student is slowing down or taking longer routes, genuine educational insight requires talking to the driver, understanding the changing landscape, and sometimes redesigning the road itself, not just tagging them 'at risk' because their current speed deviates from last year's average." All this is true, as is the bulk of the argument as a whole, but it strikes me as an argument saying "don't make predictions on limited data," which makes sense too, but doesn't really get at the question of whether predictions would be useful at all.

I don't want to link to any of these in particular, but there's a theme here which is hard to pin down in words, but is unmistakable. I commented a few days ago: "the value proposition for the poor is 'skills'. The value proposition for the rich is network and connections." That's why, eg., The Chronicle says legacy admissions are so hard to eliminate. "Universities do not build wealth solely by diversifying their endowments or managing academic budgets. A university builds a portfolio of people." Exactly. But it makes universities "a city for sale and doomed to perish if it finds a buyer!" We see the pattern elsewhere. "The dominant mechanism for distributing professional opportunity is no longer slow reputation, it is algorithmic visibility." Related: "windows of increased plasticity" "a happy society is more than just a collection of happy individuals," and this: "Every era's turnaround began the same way: some unmeasurable number of people, each acting alone, declined to sell — and then made the single most subversive move available to a member of a collapsing civilization. They found the people who could not be bought." Reform isn't a political process; it's a social one.

Network Weaver came out with such a good guide to how to think like a network a few weeks ago, then delivers exactly the opposite of that thinking with this item today. And so we're back to the same old tired story involving influencers and followers, dragging everyone along into the same objective, working together toward a single (leader-defined) goal, and collaboration. You can download the primer here. Instead of thinking like a network, the author is thinking like a system - "a relatively bounded configuration of actors, institutions, rules, practices, resources, and decision-making arrangements that interact to produce recurring outcomes." So many things a network is not, so many points of command and control, and of failure. A society isn't a system, and solving social problems through systems analysis is to use the wrong tool for the job.

This looks useful: "Federate xAPI is a low-code HLA federate that can be added directly to an existing Runtime Infrastructure (RTI) to begin transforming simulation activity into xAPI statements... Federate xAPI, a new open source software solution that enables High Level Architecture (HLA) simulations to emit Experience API (xAPI) data with minimal integration effort." Something to add to my CList functionality when I get back to it.