I think perhaps the most important problem is that we are trying to understand the fundamental workings of the universe via a language devised for telling one another when the best fruit is. Terry Pratchett on alt.fan.pratchett
WCAG 2.0 And Testability
I’ve just been reading Gian Sampson-Wild‘s article Testability Costs Too Much on A List Apart,=. she has raised an issue regarding the impact of testability upon the decision to include, or omit, some success criteria. It certainly makes for slightly disturbing reading.
I’ve had some vague feelings on unease about WCAG2.0 for a while now. Whilst I was able to highlight some issues in the earlier draft and there are signs of an improvement in the current draft, I’ve still been unsure about the overall approach for reasons I couldn’t specify. Having read what Gian has to say, I think she’s hit the proverbial nail.
Some years ago, I was discussing WCAG 1.0 with a colleague of mine who could quite accurately be described as a standardista. Since this was an mindset we shared, I asked him why he wasn’t similarly motivated about WCAG1.0. His response was that, unlike a
DOCTYPE or CSS specification, he couldn’t “validate” a page against WCAG. He felt that the WAI checkpoints were “woolly” and “nebulous” and said that he would become interested when there were checkpoints that he could test against that gave him black or white answers. Given his programming background, I wasn’t surprised at this response. He feels most comfortable in an environment where the answers come out as 0 or 1 and where judgement calls are infrequently required.
Reconsidering WCAG 2.0, I can see that the Working Group may be populated by a similar group of people who either feel most comfortable working within strict identifiable parameters or who feel they must answer a demand for this kind of approach. In all fairness, it’s a perfectly valid and natural standpoint for people working with machines. The root problem is that web accessibility is only partially about machines. A big part of it is also about people and their interaction with machines.
It’s simply not possible to categorise people, their behaviours or their perceptions into 0′s and 1′s. I think that web accessibility will always involve personal judgement calls on the part of developers. It won’t be effective if it doesn’t. And my programming friend is on a hiding to nothing if he really expects it to ever boil down to simple “yes” or “no” compliancy answers.
Testability can be a guiding principle and something to aim for. However, since we cannot exclude the human element from web accessibility, it seems to me that we can never expect to apply testability in the same way that we would to, say, an program algorithm. Testability can still be a good tool at times. But it cannot be allowed to govern decisions with regard to accessible web design criteria.
A good craftsman uses the tool. The tool must never dictate to the craftsman.