Standardization process – ISO SC36

In mid March I spent several days at the ISO SC36 meeting in Strasbourg, an experience which was … how can I say this … frustrating. Let me give some background, explain my frustration, then offer ideas about causes and possible remedies.

ISO/IEC JTC1 SC36 (official ISO page; own web site) is the International Organization for Standardization’s committee on ITLET — information technologies for learning education and training. It currently meets twice a year at points round the globe, and as it was meeting relatively close by, and was dealing with an item of considerable importance to my work (e-portfolio reference model) I chose to attend: not, however, for the full week and more, but just for the WG3 meetings, which spanned 4 days. These meetings are attended by representatives of national standards bodies — in our case, BSI.

What happens at these meetings is governed by procedural rules that participants explain are necessary, and to which they are resigned, rather than interpreting creatively. The problem is not just the process itself, which would not be too bad if everything ran according to plan, but the way in which it is applied inflexibly. Even if it is clear that a draft presented to a committee needs a lot of revision, participants have to stick to wading through as many comments as national bodies have provided. As each national body provides comments independently, many of the comments conflict. Comments can be general, technical, or editorial, and the committee has to find a resolution on each one, fitting in to just one of the few acceptable formulae. This might be fine if everything was going the way the procedure writers imagined, but when it isn’t, it can be excruciating. The only way I found to make it tolerable is to multi-task, and to do other work at the same time as the committee proceedings when the matter in hand was of no great importance (though this does depend on having a good Internet connection available).

The result is great inconvenience, and either an inefficient process where people are only giving part of their attention, or a frustrating waste of time if you try to give the whole of your attention. It’s not that these processes don’t need to be gone through — they do — but better ways must be possible, and surely need to be implemented. For instance, what if the editors received feedback from national bodies, and produced an integrated redraft, dealing as well as they can with all the comments? No physical meeting is needed for this. This inability to change the process to suit the actual situation seems to feed the whole procedural problem.

The consequence of inefficiency could be that busy people are less likely to engage with the process than they might otherwise have been. Or perhaps the people who are really influential in large organisations will just stay at home, and send some minion to do the negotiation for them. That would compromise the nature of any agreement that is able to be reached. Many of the people who do appear at such meetings are either academics, with their own research agenda, or independent or semi-independent professionals, who use the opportunity to network, and in the worst cases (happily not evident at all among the people I met) such people can even use the processes to advance their own business interests at the expense of others. This is hardly the ideal recipe for an effective, meaningful, significant standards-setting body.

Sometimes, committee output is “just a technical report”, with no official status as a standard. Now in CEN, the processes is usefully differentiated: in our area there is an informal Workshop (WS-LT) to discuss things and come to agreements, and a formal Technical Committee (TC353), made up of national body representatives, to take decisions on standardisation. But in ISO, there is no similar division of role.

It seems to me that, particularly in the current economic climate, the SC36 mode of operation may not be sustainable, as people come to apply some kind of cost-benefit analysis. If we want such standardisation bodies to continue in the future, I’d say we need to review the processes deeply, coming to a new understanding of what structures and practices are helpful towards what ends. (We can then ask those who care about those ends to finance the process.)

Of course the following speculation is not of itself going to make changes happen. However, it is just possible that conversations about the issues may help towards a consensus about how to move things forward. (Co-incidentally, see my private blog about the piece by Theodore Zeldin on the Pont de l’Europe in Strasbourg.)

First, I would move most of the ISO process away from face-to-face. We do need to get to know others personally, but this could be done more effectively through something more like an annual conference, with the majority of time devoted to networking, and some presentations of the live issues that most need discussion, in preparation for the consensus process.

Second, I would adapt the processes so that they are better tuned to producing durable consensus. How to do this is too large a topic to address here.

Third, I would put in several checks at different stages of the work to confirm that whatever was being discussed was genuinely of importance to significant stakeholders. When a work item failed such a test, it would be dismissed. This would probably reduce the workload very significantly.

Fourth, I would try (though I don’t know how) to ensure that all participants

  • properly understand consensus process
  • are committed to acting transparently
  • come to the proceedings with good will

Even if bodies like ISO don’t get round to it, it would be good for those who care to formalise some set of principles such as the ones I am suggesting above, resulting in what could be seen as agreed standards for standards bodies. If we had a list of criteria by which to judge standards bodies and standardisation process, we could agree to support and attend only bodies that conformed. This would apply not only to official “de jure” standardisation bodies, but also to the many other bodies (including all those we know in CETIS) that prepare and publish interoperability specifications.

If anyone knows of any such existing guidelines, I’d be grateful to learn of them.

3 thoughts on “Standardization process – ISO SC36

  1. Sorry, not able to provide an answer to your last question.

    But then, you certainly have a point! I read through all of your blog post and, indeed, you couldn’t have described it better! I could add another category of people who are frustrated by the approach and give-up following-up to it: IT developers who try to keep up in the field and suggest solution to existing problems and struggle for standardized, open, approaches for their software systems. They just don’t have the luxury to go through this process, when they face the reality of diminishing “web years” (quote from 1996, Tim Berners-Lee: “What is a Web year now, about three months?”)

    Then again, great post, would love to see your writings being used in a more “efficient” social / research community and supportive virtual communication environment. But that’s another story: the Web started to support researchers, it has strayed away to support all other teams and communities, what needs to be done for the original goal now? And what about ‘online’ reading? What are the affordances of books and papers?

    Anyway, thanks for a provoking reading!

    PS : The first, second, …. section could deserve a (concept) map.

  2. Tasos – thanks for all the deeply thoughtful comments here – you know I always want to spend more time with you talking these things through!

    Luk – you are most welcome – we will have to snatch some time from our over-busy schedules…