During the week there were several demonstrations of current generation LD tools and it is certainly true that these are an order of magnitude more usable than the first round of tools. These days you don’t have to know or understand the IMS specification, either its conceptual model or the technical details, to use the tools. Gilbert Paquette showed us TELOS, which is impressive and takes a graphical approach to visualising the workflows. Dai Griffiths and Paul Sharples showed two products of the TENCompetence project: ReCourse, a mixed graphical and tabular LD authoring tool that hides complexity and Wookie, a widget-based approach to providing the “services” (forum, chat, voting etc). Fabrizio Giorgini showed work from the PROLIX project, oriented towards work-place staff development, where Giunti have extended their eXact Packager to include a graphical LD editor.

In spite of the substantial progress demonstrated by the above software, we will still hear even tech-savvy academics exclaim “impressive but I can’t see how I’d ever use it” (not a real quote) or “I keep feeling that IMS LD was a solution looking for a problem and havent yet seen anything that solves any problems I have in learning & teaching.” (David Davies). I don’t think we can address this by talking about LD. Rather we need to talk about talking-about LD.

Dai Griffiths made some observations about LD that, if you will indulge me in continuing my dot-joining metaphor, I think pointed out which way up the paper is. He said: “The history and multiple uses of the specification mean that it is a complex artefact with many perspectives on it.” He produced a diagram (following) to expand on this point.

For me, the diagram did more than expand on the point: it gave me an indication of a profitable way of reducing the complexity of our discussions by being clear that there is more than one way to perceive LD. Unless we can move discourse onto a more differentiated set of conversations, I believe we will not be able to really get anywhere with LD or, indeed, make much progress in dealing with the challenges the creators and proponents of LD believe it can address.

The situation of LD is not unique and there has been some interesting work exploring the concept of “enactment” in relation to Ecological Modelling Language conducted under the Comparative Interoperabilty Project. In this work Miller and Bowker say: “Jane Fountain invites us to distinguish between an ˜objective technology “ that is to say a set of technical, material and computing components such as the Internet “ and an ˜enacted technology “ that is to say the technology on the ground as it is perceived, conceived and used in practice, in a particular context.”

So, what are we to do? Where should we start?I speculate that we should begin by clarifying what LD is. Bill Olivier had opened the “summit” with some reflections on the work of IMS and JISC in support of interoperability and thoughts on where we as a community could profitably work in the future. He described the work of IMS on data models as being more akin to domain modelling and I think this may be a good insight. Domain models are necessarily rather more abstract and application agnostic than most people care to deal with. I think they are fundamentally models of “objective” rather than “enacted” technology. I believe we should accept and embrace this and conclude that LD is a language for technologists to coordinate the creation of artefacts that are the subjects of the different differentiated discourse I referred to earlier.

There are two parallels with the eFramework to be made here, but it would be a diversion to go into detail. It too has a “history and multiple uses” and consequentially there are many perspectives. The second parallel is that one of the purposes of the eFramework initiative is to enable dialogue within and across domains through the emergence of an explicit vocabulary appropriate to a service-oriented approach.

As a candidate for one differentiated conversation, I suggest picking up on another pearl from Dai Griffiths: you can consider LD to be about provisioning a learning environment. “Provisioning” is a bit of a jargon term for setting-up-what-you-need. If you start a new job, you expect a number of facilities to be provided: desk, computer, security card, payroll, user id, staff handbook …. The equivalent provisioning of a learning environment entails the marshalling of resources, conversation (forum/chat) and other “tools”, assignment to groups etc. It is online classroom management of a sort. Let us now have a conversation about this “thing” that you can drop into a virtual learning environment that magically does all of the provisioning for e.g. a 2 way online debate. Its a “wizard”: just add a few Word docs, choose how groups are assigned and its done. This isn’t a new use case; I discussed something very similar with Bill Olivier 5 years ago. I do think, though, that this is closer to the language of enactment.

Is this an application profile? i.e. is it a definition of which data elements to use, vocabularies and extensions. Not exactly: an application profile may emerge as a necessity but it would be prudent to be clear what the application is first and that entails discourse in the language of enactment not the domain model. As a closing aside, I would like to stress that I see an “application profile” as being a quite opinionated work; it should be for a purpose.

I am conscious that this is a some-what under-developed argument, probably with numerous errors and certainly with leaps of faith but I think it is time to expose my thinking out loud for criticism and to leave this piece definitely un-concluded…

The presentations referred to above are available on the web, linked from the agenda.