RNID Comments on WCAG Working Draft 30 June 2005

Dear Wendy,

Thanks for inviting RNID to comment on the latest draft of WCAG 2, the 30 June draft. In general, we are grateful for the effort that all those involved clearly have put into the document. RNID is committed to a vision of a world where individual abilities and preferences do not create obstacles to opportunity and fulfilment in our modern Information Society.

If any of these comments would require further clarification, please do not hesitate to contact me directly.

Best wishes,

Guido

Guido Gybels
Director of New Technologies

RNID, 19-23 Featherstone Street
London EC1Y 8SL, UK
Tel +44(0)207 294 3713
Fax +44(0)207 296 8069

==========================================================================
General comment

There are numerous references to how WCAG 2 "helps people with disabilities". There is explicit reference to people with disabilities as specific constituents of AAA level success criteria. This concept of dividing the world in two groups, people with and people without disabilities is a rather archaic concept. In reality, many of the features that are labelled as "helping people with disabilities" result in a better user experience for non-disabled individuals as well. WCAG 2 fails to make that case sufficiently, at least in our opinion, and as such is ignoring one of the prime advantages for content providers and designers: the fact that when you design a site for accessibility, the user experience will be improved for every user, not just people with disabilities. The largest formal investigation into web accessibility ever undertaken, by DRC and City University in the UK, provides relevant data to support this claim . It would be a missed opportunity if WCAG would fail to state that case explicitly. 

Specific comments

1) Under the heading "The Four Principles of Accessibility", it is stated:
"The success criteria are written as statements that will be either true or false when specific Web content is tested against the success criteria"

However, this gives a rather misleading perception that anyone answering these questions will actually provide the same answer (true or false). This is clearly not the case and subjective perception might provoke different answers from different people.

Therefore, I do not believe the following statement is correct:
"When multiple people who understand WCAG 2.0 test the same content using the same success criteria, the same results should be obtained."
Furthermore, statements only make scientific sense if they are backed up by relevant data demonstrating their correctness. Thus, if it is asserted that multiple people would come to the same results, this should be demonstrated to be true, otherwise it remains a mere assertion. Obviously, the qualification "people who *understand* WCAG 2" is itself a rather subjective one.
This problem that the success criteria which should offer objectively "testable" conditions often fall short of that definition is acknowledged in other parts of the document or in editorial comments, where it states for example that "The use of "clearly" and "concisely" makes this untestable".
In terms of credibility, it would be better if the guidelines would simply admit the subjective nature of certain criteria. Part of the problem is the reluctance to go beyond accessibility and embrace usability as the other side of the same coin. In practice, there is quite a bit of usability stuff in the guidelines, but "officially" accessibility is the concept. As with all other disciplines of science, it should be acknowledge by the working group that theories should be made to fit reality, the other way round is not a possibility. Hence, if the reality is that accessibility and usability are two sides of the same coin, and that human subjectivity in whatever definition is key to understanding, evaluating and using that coin, then once can only accept that as part of the theory. With the present state of technology, the clear consequence would be that user testing and other techniques to assess these usability aspects will have to become integral part of designing, building and maintaining accessible websites. There is clear reluctance amongst some to accept that model, mainly because it steers away from a "clean" and "well ordered", almost mathematical perception of accessibility. However, the theory is a means to and end, not an end in itself. If the aim is to create websites that actually work for users in the real world, one cannot ignore these truths.

2) "Conformance".
Despite the intentions, I doubt that the perception of the importance of level 1 versus level 3 success criteria will be any different than was the case for WCAG 1 priorities. The very system of numbering in this way will lead - justified or not - to people attaching levels of importance to them.

3) The Guide to Guideline 1.1 Level 1 Success Criterion 1, as well as other parts of he document refers to silent videos and discusses text alternatives. It is important to note that the absence of sound might not be clear to deaf and hard of hearing people and an indication in text of the fact that this is a silent video, or stream, would be an important feature, otherwise such users might be under the impression that there is sound and they are missing part of the information.

4) Level 3 Success Criteria for Guideline 1.1 mentions: "For prerecorded multimedia content, a combined transcript of captions...". There is an inconsistency here (and reference is made to this question in the editorial note under Guideline 1.2): "transcript" and "caption" are definitions that usually indicate two different things: a caption is a textual representation of the content that appears in close synchronisation on top of, or side by side with the multimedia content. A transcript means usually a separate delivery unit that contains the full textual representation of the multimedia content, from beginning to end and that is not necessarily used at the same time as the multimedia content. Deaf and hard of hearing people commonly distinguish between the two in this way. In addition, both provisions do not have to be mutually exclusive: one can have captions as well as having transcript available. Not all multimedia players and formats lend themselves for captioning, so offering a transcript in addition to captions might actually be of great benefit for those users who were unable to use the captions - as well as, of course, for non-disabled people. Note that when transcripts and captions are provided, these are features often used by hearing people as well.

Based on the above, the answers to the editorial note under Guideline 1.2, are as follows::
a)	Should a text transcript be required in addition to captions at Level 1?
Answer: YES, they are neither mutually exclusive, nor do they meet the same need
b)	There is concern by some that the captioning requirement is too high a bar at Level 1. Should it be moved out of level 1: or should it remain at Level 1 and use the scoping mechanism to scope uncaptioned multimedia out of the conformance claim? Answer: it should remain at level 1.

5) Under the heading "Who Benefits from Guideline 1.1 (Informative)", a reference is made to translation of text into sign language. There is a persistent, yet quite erroneous, assumption that sign language is by definition a gesture based version of spoken and written languages, and that as such simple translation from the spoken or written into the gesture version is possible. That is simply untrue for most sign languages used in the world today. Languages such as British Sign Language (BSL) are languages in their own right, NOT based on English and with no way of automatically deriving the BSL version from English text. The same is true for many more sign languages. There are some sign languages, or sign supported versions of spoken languages - which is not the same thing - that do follow fairly precisely the semantics, grammar and syntax of spoken language and for which such transcoding might be possible and relevant. However, the text offered by WCAG under this heading gives, inappropriately, the impression that providing a text equivalent by definition allows for rendering of that content into sign language, which is factually untrue and should therefore be revised.
Another reference to providing a sign language equivalent of some written content is made under the Technology-Specific Techniques for Guideline 3.1 L3 SC1, where it states under the "Advisory techniques": "Provide access to a dictionary of Sign for the language of the content, to support users who are Deaf". This is the same erroneous notion that a given word or utterance in a spoken/written language can be represented by a corresponding sign in sign language and that this correspondence is an unambiguous one to one correspondence. 

6) Example 5 of Guideline 1.1 is about an animation with no audio. Some indication should be given to inform deaf and hard of hearing people that no audio is present.

7) Level 3 Success Criteria for Guideline 1.2 mentions "Sign language interpretation is provided for multimedia". However, there is no such thing as universal sign language. Just as much as we have different spoken languages in different parts of the world, there are different sign languages too. British Sign Language is not the same as American Sign Language, and neither one of those is similar to German Sign Language for example. To provide sign language interpretation for multimedia content therefore must be qualified as to apply for the "native" region for that content. For globally relevant content, there is no real solution, but this reinforces the need for lower level success criteria alternatives like transcripts.

8) One of the Level 3 Success Criteria for Guideline 1.4 is: "Audio content does not contain background sounds or the background sounds are at least 20 decibels lower than the foreground audio content, with the exception of occasional sound effects. [V]". This is not just of benefit to a narrow group of hard of hearing users, rather it is a feature that makes audio content much more accessible and usable to a huge part of the population (hearing in the average adult rarely shows an optimal profile). As such, it should probably be a Level 2 Success Criteria. It basically means the same as offering sufficient contrast at a visual level.

9) Under "Who Benefits from Guideline 2.1" reference is made to people with disabilities. Yet, for many other categories of users this success criteria is of great relevance: people using for example mobile handsets like the Nokia Communicator or the Blackberry.

10) For the Level 3 Success Criteria for Guideline 2.2, the last two of those ("Non-emergency interruptions, such as the availability of updated content, can be postponed or suppressed by the user. [V]" and "When an authenticated session has an inactivity timeout, the user can continue the activity without loss of data after re-authenticating.") are relevant to a very broad range of users, as such are not really specific to people with disabilities and thus should be either level 1 or level 2 success criteria.

11) The first two of the Level 3 Success Criteria for Guideline 3.1 ("A mechanism is available for finding definitions for all words in text content. [I]" and "A mechanism is available for identifying specific definitions of words used in an unusual or restricted way, including idioms and jargon. [I]") seem both meaningless and inappropriate. Surely, it is purely a personal decision and a personal choice of implementation for a user on whether or not to look up a word (jargon) or not what source to use for that. When I am reading an English book, I might as an individual choose to consult a dictionary for a certain term or expression I encounter. I fail to see why it would be the book's author's responsibility to provide me with a reference to such a dictionary or a mechanism for finding one.
If the purpose is to merely suggest that a glossary of terms and definitions should be included, then the wording needs to be amended to reflect that.

12) Under the same heading "Level 3 Success Criteria for Guideline 3.1", it mentions under number 5: "When text requires reading ability at or above the upper secondary education level...". What constitutes "upper secondary education level" for a given language is a purely regional and culturally determined criterion. It will mean different things to different people and as such is quite inappropriate in a document that aspires global relevance. While the editorial note states that definitions of primary and secondary level education will be provided, the use of these culturally laden terms themselves seems quite inappropriate. It would be far better to refer to some neutral term that then can be specifically defined.

13) Under the "Examples of Guideline 4.2", example 2 ("an intranet site that transforms gracefully") seems rather badly worded. It mentions a large organisation with different baselines at different locations. Surely the answer is to implement a baseline version usable by all sites, with ADDITIONAL content and/or functionality provided to those sites that have more advanced technology available - rather than suggesting two different versions, a concept that is in practice most likely to lead to both versions becoming out of sync with one another.

14) The document refers to "ITC Guidelines" in a number of places. It is important to note that the ITC has recently been incorporated in a new regulator, called Ofcom, and as such no longer exists as such. Ofcom has adopted most of its guidelines, but as these are being updated and replaced the reference to "ITC Guidelines" will become invalid and they will become "Ofcom Guidelines".

15) The "Appendix A Glossary" refers to "Captions" with a definition that sustains the confusion with "transcripts", potentially leading to confusion with users and implementers (see above). The word "Subtitles" should also be included as an alternative in many parts of the world to the word "Captions" as used in the US.

16) Also in the glossary, it would be useful to include a definition of "Sign Language". Such definition should make clear to the reader that sign languages are often languages in their own right and that they consist not just of manual features but often include non-manual features that are of relevance to the meaning of the signs.

Guido Gybels, RNID Director of New Technologies,
August 2005.


*******************************************************************
This email and attachments (if any) must be swept for viruses before opening. Their contents may be confidential or privileged and are intended solely for the named recipient. If you are not the intended recipient and you have received this email in error you must not read or use this email and should notify RNID on: +44 (0) 20 7296 8282.

 

RNID, Registered Office 19-23 Featherstone Street, London EC1Y 8SL No. 454169 (England, Registered Charity No. 207720)

********************************************************************

Received on Wednesday, 10 August 2005 14:01:55 UTC