Various academics have been getting excited about MOOCs (massive open online courses) recently: especially Clay Shirky who's claiming that it will "disrupt" higher education. As someone who's old enough to have lecturers in library school telling us about how CDs would replace books, I'm sceptical. But given that some prominent UK universities, including my own, are now getting together with the Open University to create Futurelearn, it looks like we should be thinking more seriously about what MOOCS might mean for universities.

The involvement of the Open University reminds us that distance learning isn't new; the delivery mechanisms may have changed since the day of taping TV programmes, but the principles haven't. And there are many other organisations doing similar things: my husband has been working in legal distance learning for more than a decade. So what is it that makes MOOCS different from such courses? One basic answer: they're free. So why are universities, which do not have primarily philanthropic objectives, getting involved with them? How do they think they're going to get something of value from them?

The most useful piece of Shirky's article, amid the over-excitement, is in his reference to "unbundling" of education. If you look at the educational side of a university, it currently does three main things for its undergraduate students:

1) Provides teaching materials

2) Provides support for learners

3) Provides credentials.

Shirky is perfectly right that these three functions can potentially be separated out. You can buy a university textbook or Great Courses videos whether or not you're enrolled on a course. There are some university programmes (such as the University of London International Programme which have learner support provided by other bodies. And (at a lower level), examination boards such as OCR provide exams and credentialing of students without getting involved in their teaching at all.

The problem, however, remains the same: doing these things well costs money. (The Open University, for example, now charges 15,000 GBP in fees for a full undergraduate degree). It takes a lot of time and effort both to create good online content and to keep it updated. Learner support is more and more done by cheap postgrad labour, but it often can't easily be scaled up. It may not quite take 10 times as long to mark 100 essays as it does to mark 10, but there's very little economy of scale. And providing effective credentialing needs a lot of organisation. The dream of MOOCs is that they are available all over the globe. But in order to ensure quality, the University of London has to have a string of local exam centres. You can't simply use online testing for serious certification, because on the internet nobody knows you're a dog. More specifically, how do you know that the person taking the exam is who they claim to be and not a professional exam-taker? There may be ways to get round physical presence at exams, but they're not going to be cheap or easy. And any subject that can't be tested purely by exams, but needs practicals or some kind of coursework/portfolio is going to be even trickier and more expensive to administer.

So how can free courses work? Or perhaps a better question, what kind of free courses can be made to work? Firstly, in terms of teaching materials, Shirky's big idea (it might be slightly harsh to say his only idea) is that the internet makes videoed lectures easy and almost free to distribute. Let's ignore for the moment all the discussion about whether lectures are actually an effective form of teaching. What courses can be taught largely by lectures or books and what can't be?

Anything with a practical component (which means most of science and technology subjects) is hard to teach online effectively, and so are modern languages. The more feasible subjects are mathematics and IT, business studies, law, some of the social sciences and some of the humanities. Introductory courses generally tend to be easier to teach online than more specialist ones: more basic knowledge to be learnt, less sophisticated analysis required. The other advantage for introductory courses is that they're often more standardised: the ideal course is one for which there is global demand and which doesn't need frequent updating (which promptly makes law courses look a lot less appealing for the MOOC treatment, given different jurisdictions).

The next issue is the support that can be offered to students. The main thing that MOOCs can potentially offer is user forums and other opportunities for students to get together and help each other. Study groups can be very useful for motivation and assistance with basic problems. For example, I taught myself Latin via an internet study group back around 1997. At a point where I wasn't living anywhere with real-life access to such a course, an online group of people who shared the same interests was wonderful. That kind of easy interaction with fellow-students is an important change that the internet does make possible.

The question for universities, however, is whether there is also a market for paid student support services provided by them for free courses and if so how it could be priced. This would presumably work something like the support services provided by tech companies: if you want help from an actual human being, you either pay a subscription or you pay per enquiry. This would probably work better with subjects which had more right/wrong answers, i.e. possibly quite well for law and maths, but less so for the humanities. It's not quick and easy to explain how to improve a history essay.

My husband also suggests, probably rightly, that it's a lot harder to sell support services (where you're not getting anything terrible tangible for your money) than it is to sell course manuals, where his students get ring-binders full of printed detail and feel they've got their money's worth. He reckons that the time when people are most likely to want support (and be prepared to pay for it) is when they've got an exam looming: that's when his firm sells revision courses, for example.

So it might well make sense for universities to provide a combined (re-bundled) support/credentialing service. I suspect most students would expect to pay universities for credentialing, if they're going to get useful qualifications; the question is how much they would be willing to pay. But there are also important markets worth pointing out of students who don't need credentialing. One is the pure hobbyists. The other is fields where formal credentialing is not particularly important and the emphasis is on practical tests of competence.

If you put all these factors together, what sort of MOOCs have you got for which it makes sense for universities to invest time and money? Basic mathematics and statistics courses look like one of the obvious targets: taught by lectures, relatively easy to mark exercises automatically, global market and the content doesn't need to be cutting-edge stuff. The other big area in MOOCs so far has been IT courses, which again can be taught almost entirely via online teaching and exercises, and where everyone's used to getting help from online forums. There's also the advantage that external certification of IT skills is already available (as well as companies that are prepared to take on people without qualifications if they can display programming skills). The big problem here is going to be keeping courses up to date.

There are a couple of other departments that I think should be looking very hard at what MOOCs might mean for them. A lot of business and management education takes place via one form or another of distance learning (especially the self-help book). If anyone is going to work out how to make a profit via support services for MOOCs, I'd expect business departments to be able to do it. The other academics who I think should be taking MOOCS very seriously are those who teach "old" languages (and I'd include not only Greek and Latin, but also classical Hebrew, Chinese, Sanskrit, Old English etc). These are courses where it's often hard to collect enough students in one physical place to make a conventional course viable and which don't have the problem of students needing to learn to speak the language (which makes modern languages hard to study via the web). In particular, well-designed MOOCs might be a way of encouraging students to make a start on these languages for free and thus intrigue them enough that they want to study them further via a paid course.

In other subjects, I honestly can't see MOOCs being a disruptive force. I suspect that all most universities are going to end up doing is putting online a few short taster courses/lectures to encourage people to sign up for paid courses. I can imagine, for example, a slightly more interactive version of the Nottingham Theology video series coming, but that's a long way from a full distance learning experience. But, as usual, I'd be interested to hear from anyone who thinks that there are ways in which MOOCs can actually make sense for universities.