Breaking up is hard to do: the ‘atomisation’ of ELT content

Is the linear ELT course on the way out? There’s a growing trend towards the provision of modular mix and match content, or even completely ‘atomised’ content. There are four key things driving this trend:

• The need to make content development faster and more efficient
• The potential for huge improvements in production efficiency
• The rise of adaptive learning which relies on content being broken into small chunks
• Demand from schools and teachers who want more flexibility in their materials without needing to do the slicing and dicing themselves

The problem of the linear ELT coursebook

The standard coursebook follows a carefully sequenced syllabus: units build and recycle language; students are carefully introduced to language which assumes they’ve covered and understood the language from earlier units; each lesson is designed to flow seamlessly from language input and presentation to practice to communicative tasks;  when new grammar is presented, the presentation doesn’t use vocabulary not already covered in the course. Level 1 of the course leads nicely on to level 2. Each level is probably mapped to the CEFR. There are tests available which assume that students have reached unit 6 (or whatever). And a vast array of supplements – online resources, DVD-ROMs, class audio, video, and so on.

All of this makes for a nice integrated course series, where you can be pretty sure all of the component parts will work together properly. But how does all of that fit with the dream of endlessly remixable content? It certainly doesn’t sound like a great match.

Developing materials faster and more efficiently

There’s huge inefficiency in the development of ELT content. That’s one of the reasons the whole process takes so frustratingly long. Every course has its own unique supplementary content, even though it all does the same job – extra practice exercises, lesson plans. As an obvious starting point, why can’t we just have a single big set of supplementary materials which would work with all courses? Just map them to the course grammar syllabus, and you’re good to go, right? But what if different courses cover grammar areas at different points? And what if they teach different vocab, pron and functional language as the grammar syllabus progresses? What’s the likelihood of this happening – 100% or thereabouts? That would mean none of the supplementary material would work very well, right? You could create more material in order to be able to cater for these variations, but then you’re starting to see that efficiency slipping through your fingers.

Despite these problems, this is a growing trend – it just requires more thought going into both courses and supplements, and maybe an acceptance that the core course and the supplementary materials aren’t going to be quite so tightly integrated. If you know that you’re going to have a single large bank of supplementary materials, that could influence how you design your courses. Gradually, the whole process could become better – if not ever actually easy.

Technology and production

At a conference a few weeks ago, I watched a demo of an online content management tool which would allow a publisher to create a single set of content and then – at the click of a few buttons – ‘push it out’ to a website, the InDesign file for a print book, and anywhere else you could think of. Metadata could easily be added to the content and to each output format. All of the assets were stored in a neatly organised and easily accessible library. You could search for a photo, view possible options, check their permissions status, format and other metadata, and then insert them into your publication. It was very impressive. If you work at a publisher in a production role, this is the kind of stuff you dream of being able to do in your everyday working life. The possible improvements to efficiency are pretty obvious. The potential for saving money is great. The scope to speed up the whole publishing process is huge. Even better, what if you then want to create new products derived from your beautifully organised library of content and assets? Now it’s the sales people and CEO who are getting excited, not just the production people.

In this context, a coursebook that exists as a single unbreakable entity just isn’t going to cut it. Sure, this kind of technology can still improve the production process of coursebooks as we know them right now, but for the really exciting stuff, you need that bank of granular materials – a tagged database of content chunks, each of which which presents or practises a specific element of the language.

Adaptive learning

Tim posted on adaptive learning recently, and we’ve talked about Knewton on more than one occasion. We’re going to see a steady stream of adaptive learning products over the next two or three years. A fully integrated course which relies on content being worked through in a particular sequence isn’t great news from an adaptive learning point of view – how can it adapt in response to what the student is doing? If you’ve got a piece of technology analysing your every move in order to be able to recommend the best possible path through a set of materials, then you really want those materials to be in the form of a library of small chunks.

A standard coursebook might aim to get a student from A1 to A2 level over the course of 100 hours of classroom time using the Student’s Book, plus a few dozen hours of self-study using the Workbook and other supplementary components. The adaptive equivalent would still aim to get the student from A1 to A2, but would have no opinion on how many hours it should take to make that journey, or in what order the grammar, vocab etc should be covered. If it turns out that the student already has a good grasp of a lot of the vocab and most of the grammar covered in the first half of the coursebook, but is struggling with other bits of the grammar and has real problems understanding spoken English, then it would be a waste of time for them to wade through all of the early grammar and vocab at the expense of spending time on the areas they need to do more work on. For full-on adaptive learning, you need to be able to flex the course syllabus in response to the student’s progress. That can only work if the course has flexibility built into its structure. And that means granular chunks of content which the adaptive software can get its teeth into. “If the student has grasped concept A very quickly, then they should try concept D. If they mess that up, then try sending them back to concept C” and so on.

But can language learning be covered as a collection of concepts that the student has to navigate through in the most efficient possible way? That’s a challenge, to put it mildly. You can easily see it working well for purely supplementary materials, but much trickier for a four skills course series. And what happens to your classroom lessons if the content is modular and adaptive? We’ll see soon how well this works out, as products begin to hit the market. Someone’s going to find a way to make it work – attach adaptive learning tech to a tagged database of content chunks, find a way to deal effectively with the fact that language learning isn’t just about understanding concepts, and you might have something pretty exciting.

Demand from schools and teachers

Lots of teachers have always wanted to customise their materials, so what I’ve described should sound quite appealing to them. Well, not all of them (I was generally pretty happy to just plough through English File when I was a teacher…). The trends in ELT publishing that I’ve discussed suggest that for those teachers who do want much more flexibility without having to do all of the hard work themselves, things might be looking up. It’s interesting though that, while there certainly is demand out there for this kind of thing, it feels like the stronger pressures are internal within the publishing companies.

Despite the caveats, there are surely benefits in content becoming more modular and flexible, just as long as it can be done in a way that doesn’t add a whole load of complexity into the lives of teachers who mostly don’t want to have design every course they teach by pulling together all of the elements from an a la carte menu. For self-study, it feels like adaptive learning is the future, and that maybe there’s less of a possible downside in breaking content into smaller self-contained chunks. The issue there is whether language itself can be broken down in that way.

Higgs Boson photo from http://www.flickr.com/photos/jurvetson/