W3C home > Mailing lists > Public > w3c-wai-gl@w3.org > April to June 2005

Re: Responses to comments on GL 3.1 proposal

From: Jason White <jasonw@ariel.its.unimelb.edu.au>
Date: Thu, 26 May 2005 12:48:48 +1000 (EST)
To: Web Content Accessibility Guidelines <w3c-wai-gl@w3.org>
Message-ID: <Pine.GSO.4.58.0505261238130.3570@ariel.its.unimelb.edu.au>

For the most part I agree with John's responses to the comments received
on guideline 3.1.

I have commented before on a number of the proposed success criteria, and
John recognizes that there are still unresolved issues, for example how to
reconcile the "measure of educational level" requirement with the fact
that WCAG can't impose reporting obligations on implementors for legal and
policy reasons.

It is also true that the educational levels specified in the l2 success
criteria are somewhat arbitrary. John suggests this could be dealt with by
requiring the educational level of the content to be reduced by a certan
percentage (say 30%) at level 2, and even more (say 50%) at level 3. While
this is a move in the right direction to the extent that it seeks to
eliminate the arbitrary numbers, it doesn't work, for a rather fundamental
reason. Suppose I draft a document and measure its educational level.
However readable I have made it, and however successful I have been, I
still in principle can't satisfy a reduction requirement, for it demands
that I reduce the reading level by antoehr 30%, whatever it is currently.
Now if I reduce the required educational level of the content by 30% and
re-evaluate, I again fail the requirement and have to reduce it again, and
on it goes - an "infinite regress" as we call it in philosophy.

If we require instead that the level must be reduced by 30% in comparison
with the first draft, then I can render the requirement moot by doing a
particularly bad job in writing a readable first draft; and again, however
good the first draft is, it can never satisfy the requirement.

So I think we're left with arbitrary numbers unless there is some way of
determining the education level required to process whatever the content
is supposed to communicate. Either way we can impose stronger requirements
on content judged to be of particular public importance, as John's
response to Lisa's comments acknowledges.
Received on Thursday, 26 May 2005 02:48:52 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 5 February 2014 23:39:37 UTC