A couple of things have gotten me musing about social lately. The first was Dave Cormier’s thought-provoking blog post about how PLEs are supposed to disaggregate power, not people. The second was a private conversation with a friend who is thinking hard about how to add a social layer to an existing LMS. And the third was a blog post by Union Square Ventures VC Fred Wilson expressing skepticism about Google’s plan to add a social layer to its existing services. I have been struggling to articulate why I am ambivalent about both the attempts I have seen so far to create a Personal Learning Environment and attempts to layer on social capabilities to existing LMSs. Wilson says it better than I can:
Q. Why did Twitter succeed and FriendFeed fail?
A. Because FriendFeed was largely a social aggregator whereas Twitter is a service with specific social intent.
I think this is an important distinction. I have not seen any breakout social layers. The social services that have broken out to date have been services where a user has a very specific intent.
Social engagements are weird out of context. Comments to this blog make perfect sense on this blog but less sense when they are tweeted out into a Twitter stream or show up in FriendFeed.
There is value in social aggregation but not huge value.
Twitter was constructed with an idea of facilitating certain types of conversation. You can use Twitter as an aggregator by tweeting links, but that function is contextualized by and subordinate to the conversations—as staccato and shorthand as they may be—and relationships that Twitter supports. More often than not, when you just dump data from a feed into a social environment, it doesn’t work. Sure, some of the items may prompt conversation, but more often than not, they don’t. (How often depends on what kind of stuff is in the feed and what kind of community it’s feeding into.) In and of itself, aggregation has limited value as a social (learning) lubricant.
I firmly believe that a social platform for learning, whether it is a PLE, an LMS, or whatever, needs to be designed with a “specific social intent in mind,” as Wilson puts it. What kinds of social interactions are people looking to have in your learning environment? What kinds of social interactions are you looking to facilitate? How and from where do those interactions start? Where do the conversations go? What do people do with the outcomes of those conversations? Putting feeds everywhere and then aggregating some of them into some kind of sharing/discussion environment isn’t enough to reliably facilitate social learning. Don’t get me wrong. Feeds are good. Aggregation is good. Wrapping conversational capabilities around your aggregator has some value. But it’s not enough to create a truly social learning environment. To do that, we have to start by understanding and articulating the participants’ specific intention to share, and then build an environment that supports that intention.
That’s hard to do. And it’s particularly hard to do in an environment in which you assume very loose and light coupling (as is often the case with early designs of PLEs) or in which you’re working within and around the structure of existing software that has strong and relatively rigid definitions of roles, groups, and permissions (as is often the case with current-generation LMSs). I’m not saying that it can’t be done, but it requires a lot of hard work, both conceptually and technologically. My personal belief is that, to the degree that you succeed in creating such an environment (regardless of where you start from), you will substantially blur the line between a PLE and an LMS.
Update: I just thought of an example that might make this at least slightly more concrete. Today I signed up with Ping.fm. It’s been bothering me that I haven’t been able to share items that I like from my Google Reader, which is my RSS aggregator, with my Twitter feed, which is the social network I use the most at the moment. Ping.fm lets me do that. But already, after less than a day, I’m discovering that I’m not sure I can continue to share the same items I’ve been sharing through Google Reader and, by extension, Buzz. I read a fair number of political blogs and often share posts that I think are interesting or provocative. But I do much less sharing of political links in my Twitter feed because I’m aware that a lot of people follow me for my e-Learning stuff and aren’t necessarily interested in my politics. So what do I do? Do I continue to share the same number of political links and let people on my Twitter feed decide if they want to keep following? Do I change my RSS sharing habits? Do I decouple the two? And what would happen if I decided to also update my LinkedIn status with my shared RSS items? How would that affect what I want to share? Or what if I added a feed to Facebook? For the same reason, I no longer automatically feed my tweets into Facebook or Buzz, and I never fed my tweets into a widget on my blog. Each of these social contexts has a huge impact on what I want to share and how I expect it to be received. If we don’t start our designs by understanding the initial social intention behind each act of sharing, then we are not going to end up with designs that facilitate social learning.
Getting a bit philosophical without much concrete to say, but …
It might that the big challenge to social networking and aggregation, and by extension to integrate some of this in learning environments, is the issue you note in your update: that we are contextual beings. Beyond just broad distinctions between work and play, politics and whatever, I even notice that I tend to want to present different kinds of aspects of, say, what I’m reading to different students in different courses. I can manage some of this using tags and a bit of simple programming, but certainly part of the problem is all of these services pretty much assume a monolithic online persona. Maybe we need to start from the beginning and recognize that each person has multiple personas, depending on context?
Context is definitely one of the hard problems, but it’s not the only one. My analogy is limited because I’m already starting with items in an RSS feed reader that were mostly designed to be atomic, e.g., blog posts and news articles. But what if the “item” you want to share is a discussion thread, collaboratively created by a handful of people, each of whom “owns” one or more posts on the thread, and each of whom may have different expectations and needs for privacy? To the degree that your learning content is collaboratively created, you have a huge ownership issue to settle before you can share.
Thanks for this thought provoking post. At the University of Florida, College of Education we’ve been working on building a platform that blurs the lines of the PLE and LMS by adding a social layer to our LMS. We’re calling it a Social Learning Environment. We’ve just launched the social layer in May of this year. Results have been good. We’re attempting to build communities of inquiry and practice around shared intent, by fostering collaboration and communication in a user interface that makes sense to our users. In our environment, context is king. What are our users doing, and with whom? Aggregation is only part of the picture.
Daniel, are there any public screencasts of your work available? I’d be interested in seeing it. Some work I saw over the weekend (which will hopefully be public in the not-too-distant future) shows some early promise; as I say, it’s not impossible to do this on top of a traditional LMS infrastructure. It’s just really hard. The details matter.
I think one of the ways that feeds can be useful is when you filter them in interesting ways. For instance, I have a WordPress site with a Delicious widget, but it doesn’t send everything from my Delicious feed to the site, which wouldn’t be relevant. Instead, it filters on a certain tag, so that only certain things go through.
Carrying this idea further, you could then filter the material you are dealing with so that one medium contained both political and e-learning material, but it gets filtered so that two distinct twitter accounts are drawing from that source: one for politics and one for e-learning.
In the classroom setting, it would be great to have a filter that sent appropriate material to be visible by the class (such as Sakai’s profile2 twitter feed, which I have been too squirrelly to hook up) while other material was left out.
Filterless aggregation isn’t particularly useful for most purposes, but intelligently filtered feeds could do wonders. Even within one medium, like Facebook, this would be very useful. A vision for making this less complex: next to the submit box you could have 5 buttons in rainbow colors. Click purple to send to everyone, blue for classmates, orange for family and friends, red for close friends, etc. You pick what groups/levels you want beforehand, and assign them colors. (I hereby welcome developers to steal this idea!)
I think one of the issues to consider in this is how does traditional brick and mortar learning deal with these issues? One of the short comings of the lecture method is that people feel they can’t connect what they hear with real life. That is why technical education, internships and other methods become popular. People want to be working along at their job and learn how to use a concept where they need it – not in a separate walled off area called a classroom.
In online learning, I think that you want to figure out how to do the same. You are right – you don’t want to just add a social layer on top or just blindly aggregate feeds that might end up bringing in information that is irrelevant.
What you want is for the learners to keep using their online tools – Twitter, blogs, whatever comes down the line – as they would every day normally. When they think of a way to use those tools to demonstrate a concept they are learning in a course, they tag that part of their daily life as being the “proof of learning”. The Social Learning Environment is watching each learners’ feeds for a tag that indicates that a certain part of that feed is for course, and then connects that part with the assignment the instructor has created so that students can prove they have learned the concept. This goes a bit more granular than just aggregation – the SLE would still aggregation, but would comb through the aggregation looking for tags that tell it where the parts of the feed go, if at all. Many parts wouldn’t be tagged and would therefore be ignored.
My colleagues and I have been tossing around the concept of Social Learning Environment for a few years – but I think it is different than how it was previously used in the comments. We don’t see it as a method to combine the LMS and social networking – we see it as an idea that dumps both and does what you are talking about here: to create something with a “specific social intent in mind.”
We wrote out a basic outline of this a while back: http://www.edugeekjournal.com/2010/03/18/social-learning-environment-manifesto/
I think using tags could also be akin to the filter idea that Brian is speaking of.
Brian, I agree that filtering would be helpful. But it’s also hard, unless you rely on the human whose feed it is to tag the individual posts for filtering. That kinda defeats the purpose of automated data streams.
This sort of Matt’s point, I think. We can get further when the user acts as a filter through tagging or some other mechanism. That said, even aggregating tagged entries from public feeds probably only gets you partway there. It is a kind of social activity, but me tagging a post for sharing with classmates is not the same, for example, as initiating a conversation with them. How much social glue you need in your environment depends a lot on the nature of the collaborative effort, the comfort level of the individuals, and a bunch of other factors, but I tend to think that a substantial dedicated collaborative learning environment is necessary more often than not, even if it is highly permeable and able to take feeds (whether filtered or not).