Following on my last post about open culture versus open content, I wanted to expand a bit on the topic of the distributed system of sharing that I mentioned. I argued that sustainable models of sharing must first be embedded in culture and practice before they will truly make a difference; and that this model is most likely going to be a distributed one, since open education in practice is largely distributed as well.
The challenge in the distributed model, though, is how to quickly locate openly published materials and information. In the case of some high profile services like Flickr, there is now the capacity to filter search results to items that have been released under an open license, however this isn’t necessarily a standard feature with many sites. So you might say that openly licensed materials are currently cast to the four corners of the Earth.
While sites like Creative Commons give people the opportunity to catalogue their materials so they can be easily located, in my experience with the LRC, this is rarely done in practice. People release their materials when it is contextually relevant to them. Only a fraction of the people will actually take the time to visit an unrelated site and enter metadata describing their resource. So in order to gain a more holistic picture of what is being shared, where, and in what context, we have to go to the spaces where the users are working.
Feed Oriented Architectures
In my view we can learn a whole lot from the efforts of people like Jim Groom at UMW Blogs. If you’ve ever heard Jim speak, you will know how passionate he is about each individual having their own space. Therefore rather than establishing a single space where everyone is forced to work, you use the power of RSS to aggregate content from across the web.
All that you need is a common tag, which is referenced and queried by the central portal. When an author assigns the tag to a piece of information, such as a blog post, a video, and image, or a document, the portal is able to locate the material, aggregate it and then incorporate it into its own database, where it can be located by others. The only requirement for the author then is to use the tag when publishing the content to their chosen space. The fact this model is tagged-based or feed-based also means that one item can be pushed out to many different portals simply by using a collection of tags.
Instead of adopting a project- or content-centric view that pulls in people, the distributed, portal-based model adopts a people-centric view that pulls in content. This leaves the individual completely to their own devices, while enabling the organising of large collections of content into an architecture that can be browsed and reused by others.
Why is this better?
The fact this model slots in with the way people are already working is really important because it doesn’t introduce a new requirement, therefore it’s likely to gain a far more comprehensive view of the sorts of activities that are taking place in disparate spaces.
As such it also inherently focuses more on on-going practice, by providing far more of a sense of the stream of activity that are taking place, rather than snapshots of the major milestones.
So the sharing and reuse becomes as much about analysing and supporting practice, and influencing the change of practice, as it does about indexing a de-contextualised set of objects.
I’ve got more to add, but I’m out of time. More on this later…
A common tag, in a single language, that we all agree to use? How, Mike, how?
No, not necessarily a single language – it's far more folksonomic than that. Think of Diigo and Delicious, and the way a single link can be tagged, and re-tagged according to different criteria.
With a distributed model it would have to be folksonomic and flexible; not only because you'd be relying on the inputs/feeds of a web of people who each control what terminology is used, but also because it would rely on the context in which it was originally created.
This is part of where it has to be embedded in practice I think. Feed-based architectures are extremely flexible. They can be used as broadly as tag = education, for specific courses (tag = cck09) or even smaller, and everywhere in between. Part of the process then becomes the negotiation around practice, what tags are used, and in what context. (You know this already of course, I'm just saying this for the benefit of others who may stumble across this conversation).
Additionally, if you look at some of the stuff that people like Jim Groom are doing, and Mat Wall-Smith at New South Blogs here at UNSW (http://newsouthblogs.org), they're syndicating at a site-level as well as aggregating. They'll pull in feeds, re-filter them against a different set of criteria, and then re-syndicate them out again. So conceivably we could begin to see sites syndicating and aggregating with one another.
So really what's needed is not only the technical architectures that are able to support folksonomic organisation, but a culture that recognises the need to negotiate terms and co-operate in their implementation.
I want to expand this post to include an educational use case – perhaps my thoughts on this will be a bit more clear at that point. I'll try to get to this today
I'm not sure that I really do know a lot of this, or at least I'm still having trouble with it. I get the principle of crowd-based aggregation, but the idea of re-syndication adds another layer. I'm still not seeing how my use of my tag will necessarily lead me where I want to be, though I do see how it could possibly do so. I look forward to your expansion to help me understand.
Pingback: Use Case for the Distributed Model | TechTicker
I started responding to this as a comment but it quickly grew into something else so I've devoted a new post to it: “Use Case for the Distributed Model.” Comments, skepticism and debate welcome
Mike and Lisa,
I see the possibilities and limitations of this model, and while we have been using it regularly at UMW, there are still a few things we need to streamline. Have students tag their posts with a unique tag for a specific class was a bit of a challenge at first, but over time it seems like it has gotten a bit easier. If they tag it right, it does republish in the course blog. A benefit of this is that everyone in the course can subscribe to one singl feed in the course blog for all the distributed updates, and the eprmalinks will actually revert back to the student's blog, not the course blog. Another possible beenefit is that everyone's post is not only their own, but also republished within a course space that traces and archives the specific posts for that course.
That said, it has a major weakness, and that is syndicating in comments. So while the commentary happens on the students' blogs, it cannot be republished on the appropriate posts on the course blog which is still an issue, especially when we recognize the commentary can be just as powerful, if not more so, than the original post. That said, it still leaves control of the comments and original posts in the hands of the students, not the course blog. This is a tricky slope that something like Disqus or Intense Debate may help solve eventually—but their requisite login and the like is a barrier to commenting.
One think I like about this model at UMW, is that it gives us a series of spaces that represent course nodes in the larger publishing network that we can highlight and use to filter all the work going on. And I think that is the purpose of the tag, a way to simply and intelligently filter syndicated content that is relative to a specific group of students, while at the same time letting them use their own spaces for a variety of other course, personal posts, reflections, movies reviews, etc.
While far from a perfct model I do like the way it pushes us to think about ways to imagine aggregation, and in is running parallel to developments like Google Wave nd the like, but just not as integrated and easy. Google will perfect that, and we may very well be turning to engines like that for ease of use, but I still think the model, when aggregated through a twitter hashtag through a twitter plugin as CoGDog points out here, offers a lot of ways to approach this through a series of broadcasting tools.
And Mike, you frame it here brilliantly, so thank you.
Jim and Mike,
I found myself nodding at Jim's remarks about problems with commenting syndication and log-ins.
Blogs are what we have, but they really seem to be the wrong format for discussion. They are post and reply. Their main advantage is subscription, but you're right that commenting subscription is inconsistent (I almost missed Jim's reply here because I thought I had subscribed to this discussion, and it didn't go through — when it does go through I have to confirm, which is annoying, proving his point). In some ways, a subscribable forum might be a better way to have a discussion. But, as Jim notes, then the participant isn't participating from his own space. It seems we haven't built the good bridges yet between spaces, even when they're public.
Students having to tag, and do it correctly, is cumbersome. And how can tags filter effectively unless they are being used consistently, which may (or may not) be possible in an institutional system, but is impossible on the open web, where I'd prefer to work? I don't buy the theory that massive participation will result in the information I want — that's why I still search Google still as well as Delicious. Little tools like the one Alan discovered for Twitter are highly geeky, a problem I recently criticized in my own blog (do let me know if the comment subscriptions are working…).
Perhaps the real difficulty is simply asynchronicity. We're trying to do as a group what we are used to doing in person as a group, or by writing individually. It could be as basic as this: we don't have the skills yet to envision what we want. Thank goodness we have visionaries like you two.
More on Mike's new post coming up, 'cause otherwise I'll have to do all this in my space instead!
RE: @LisaMLane I started responding to this as a comment but it quickly grew into something else so I’ve devoted a new … http://disq.us/cv71
This comment was originally posted on Twitter