Cover Page »
Feature article
The wide open learning world: sea, land, and ice views
In my opinion
Potential implications of the online learning innovation fund
Conference reviews
@#altc2009: impressions of ALT 2009
Future of technology in education conference
Project updates
Interdisciplinary design to support learning design
Global resources for on-line education
Back to basics: the web, academic values, and Sakai
Education in a changing world: teachers and technology
Case studies
Twittering the student experience
Online 'student common rooms'
Introducing a radiofrequency clicker system: staff perceptions
ALT news
Chief Executive's report
All of a twitter
Innovating e-Learning 2009
ALT-C Keynote and invited speaker sessions
Older issues »
Back to basics: the web, academic values, and Sakai
by Clay Fenlason

Academic Values and Sakai 3
One could be forgiven for confusing virtual learning environment (VLE) debates with those of theology. Patent seekers have constructed tortured definitions like medieval scholastics, straining at logical proofs of existence. Teachers wrestle, in the dark night of the soul at term's end, with the question of whether something could be good that seems to be a source of so much evil in the world. The freethinkers of Edupunk [1] chafe at the mental shackles of the VLE, while some scorn it as a crutch for the old and weak. Among them a growing number of voices has taken up a Nietzschean cry, declaring the VLE dead. What now? What will take its place, and on what grounds? It's a useful provocation, no matter one's stake in the existential debate.

Technologists who labour in this area are in a period of soul-searching. The forms of VLE we've known, however useful they have been, now plainly represent an intermediate stage that will soon be superseded. The world has moved on, and the form of the VLE must shift with it. Call this shift a "death" if you like, but we're still left with the business of working out its consequences.

Sakai has always held labels such as 'VLE' or 'learning management system' at arm's length, and if at one point this seemed like a bit of waffling to outside observers, it is now in these circumstances a decided strength. The Sakai community took as a central insight the notion that many of the kinds of collaboration effective in learning were also central to good research and to some informal aspects of academic life more difficult to circumscribe. Sakai ventured a new acronym: CLE ("collaboration and learning environment") just to drive the point home. As a result Sakai need not have a stake in the survival of the VLE as such: we take no particular pedagogy as our organizing principle, nor are we committed to a business plan that requires growth in the form of new rings of walled garden. We're even happy to do less, if that's what's needed.

During the last year a great deal of development effort in the Sakai community has gone toward a next generation of the software, which we're calling Sakai 3. We are now approaching the point where early adopter institutions will start to deploy this new framework on their campuses. Yet while there is understandable excitement about what's new, it seems just as worthwhile to emphasize how it represents Sakai getting back to basics. The days of naive enthusiasm for the VLE are behind us; it's not enough to simply accrete new features, we need to trace with care how the next advance will support the academic mission more deeply. Let's attempt this with a couple key examples.

Scholarly Networking
Social networking has often been confused with its fads, with a recent report even suggesting it is a fashion on the wane. [2] Nevertheless, there is something solid and enduring at its core. It is no new insight that relationships frame and support our knowledge of the world, not to mention how deeply the sociality of scholarship runs. This is not a discovery of Web 2.0, it is simply the technology catching up to where people have always been. A backlash against the Facebooks and Twitters of the world can miss the mark as widely as a critique of the VLE based only on recent incarnations.

Merely incorporating current social networking applications is insufficient in our domain for the simple reason that academic relationships differ both in character and activity. Peer-to-peer relationships and flat friend lists alone can't satisfy the needs of an arena that adds mentorship, guidance and assessment to its practices of peer review. What's more, academic relationships tend to grow up around shared work and interests, only more rarely focusing on the relationship for its own sake. We might therefore gain insight from more generic social networking tools, but they should not lead our investigation. Instead the problem should be re-framed in terms of support for networking effects already operative in academic societies: call it 'scholarly networking' instead.

Social networks were a seminal force long before there was any such thing as a web application, it's true, but what's at stake is more than just bringing technological support to what universities are doing already. It's also about finding new ways to support more robustly something that universities have often done poorly for their constituents: helping them connect with people who will advance their academic success. Individuals have mainly done this on their own through serendipitous encounters at conferences and seminars, or the simple expedient of a common department or alma mater. What's more, the top-down data models employed by student information systems and administrative hierarchies offer very poor representations of how people actually organize learn from each other and get their work done. This is the sort of thing that universities should care about doing better. If social networking hadn't yet taken off as a web technology, we should be trying to invent it.

User research into the opportunities has been conducted through a JISC-funded project [3], cementing the suggestion that in academic settings the important relationships will be less about friends and more about people who can help. As a technical support to what will likely prove to be an ongoing investigation, Sakai 3 incorporates an OpenSocial [4] container (Apache Shindig), which in addition to offering a standard for handling user profile and group data also allows us to bring OpenSocial gadgets into Sakai (or vice versa, allowing Sakai gadgets to be rendered in other OpenSocial containers). It is a way to reach beyond the confines of the institutional boundary. If we pay attention to how our academic societies organize themselves, we know that no other form of solution would suffice.

From Wikis to Waves
Wikis have been successful in illustrating the power and pedagogical import of collaborative authoring, but their uptake in our arena has been impeded. The first hindrance has been wiki mark-up itself which, while a welcome simplification over HTML for basic formatting, does not lower the bar nearly enough. It can still be abstruse, varying with each wiki product, while mastering any particular dialect does not entirely reward the effort. There is also the static, textual nature of authored wiki pages. Despite advances allowing the embedding of media objects and programmatic macros, the strengths of the wiki have remained grounded in collaborative text editing. The wiki offers powerful support for wordsmiths, but as a general authoring environment it leaves something to be desired. At the end of the day it still hasn't engaged a large fraction of our academics.

The earliest Sakai 3 work grew out of an attempt to address these challenges without losing what was so powerful about the wiki. A WYSIWYG authoring environment is now commonplace, and we've followed suit, but it was not enough to simply remove the barrier of wiki markup. Among the traps VLEs have fallen into in the past -Sakai included- was a structure built upon siloed tools while the actual workflows of our learners and researchers often cut across those boundaries. It was essential to add the ability to easily embed functional, interactive regions within the page so that text, media and activities could interleave one another in context. As a result, authoring becomes a matter of arranging content and activities into potentially new academic workflows. To this is added a template mechanism for storing and sharing such structures, allowing them to proliferate as concrete best practices which have emerged organically from the activity of teachers, learners and researchers rather than being imposed by the constraints of the technology.

Enter now Google Wave [5], and the conversation as shared, living document. In its basic structure it matches many of Sakai's ambitions for easy collaborative authoring with interactive widgets, and then goes a step or two further with synchronous editing and federation. What's more, lightweight integrations with Waves seem a relatively simple matter. Deeper integrations seem also achievable with judicious use of gadgets and 'robots' to exchange data between a Wave and Sakai, and some exploratory work has already been done in this area in the Sakai community. [6] Wave's Open Source (OS) licence should also allow us to incorporate and adapt its core capabilities.

While Wave's advances are promising, they also serve to illustrate the point that powerful but more general technologies do not tend to anticipate academic use cases. For example, Wave has only a very basic notion of roles: you are essentially just a participant or not [7]. Nor does it support the OpenSocial application programming interface (API) as yet. Peer-to-peer activities are its sole focus and, although such activities should arguably have a stronger place than they do now in learning contexts, a reliance on this simple relationship model may not adequately support mentoring activities so central to academia.

Academic use cases seem likely to always require an elaboration of the basic model, most often by adding new workflows and roles. This is a significant point for the future of e-learning technology, and further confirmation of the importance of OS licences and Open APIs [8]: each new level of sophistication of web tools still leaves a gap which academically oriented technologies must span. The future of the VLE may well be found in a superstructure of roles, responsibilities, and data riding on top of core, generic capabilities from third parties. We should work to identify the unique supports that academic activities call for, and devote our energies to that slice of effort for which we have particular insight, rather than recreate the technological stack in our image.

Technology Catching Up
It's not uncommon to view the VLE, or institutional technology in general, as behind the times, always lagging and needing to catch up. But the history of this product domain suggests nearly the opposite course: the VLE got ahead of itself by bundling together a set of existing web tools not designed for its core mission. The dangers of tool-driven development are twofold. First, that the practice is forced to conform to the tool rather than the technology supporting the practice. This is not always a bad thing -having to adjust one's practice can provoke constructive reflection, no matter the shortcomings of the result- but it's nevertheless an inversion of technology's supporting role. Second, the tool boundaries often serve as impedances to intuitive workflows. If we uncritically mash up the latest Web tools we  simply repeat this error on a new level. We should instead get back to the fundamentals of what our users need, and what activities we are trying to support. Of course the latest web developments are themselves better grounded in the fundamentals: relationships matter; freedom and peer review produce finer products; individuals order their experiences in diverse ways, and so forth. The technology needs to be robust to do all that, but we still need to take these basic lessons and apply them to academic contexts.

The Edupunk movement [1], for example, has done valuable R&D work in exploring the utility of diverse web tools, but by its very nature tends to be locked into a tool-driven mode. By their generic nature the tools it employs have limited ability to support the variety of roles and relationships operative within societies of scholars. They have little ability to provide the sort of security and control that is sometimes called for, and little ability to capture their successes within an infrastructure that can support even the technophobic.

We in Sakai think this is our role: bringing the insight of our community of educators to bear, manifesting academic values in a technical infrastructure as well as a community of practice. This infrastructure needn't be a VLE as we've come to know it. As far as the future of the VLE is concerned, pick the metaphor that suits your rhetorical need -death, revolution, metamorphosis- we don't need to participate in that debate. We're focusing on the wants and needs of scholarly life, the slice of academic support that more general tools are unlikely to ever treat adequately. That work will remain even when the VLE goes away.

Clay Fenlason
Director of Educational Technology, Georgia Tech
Product Manager, Sakai Foundation


[1] [Accessed 24/9/2009]
[2] Upon further study, earlier reports appear to have misconstrued the data: [Accessed 16/9/2009]
[3] [Accessed 18/9/2009]
[4] [Accessed 24/9/2009]
[5] [Accessed 24/9/2009]
[6] [Accessed 18/9/2009]
[7] [Accessed 18/9/2009]
[8] At first glance the Google licence does not appear to be unequivocally open. Should you decide to charge users any type of fees in order to access to your Wave API Implementation (perhaps including technology fees at a campus?), you must first get written consent from Google, and you may have to enter into a separate written agreement with Google before Google grants you permission to establish any such fee-based restricted access to your Wave API Implementation. See
Created with Newsweaver