First, I want to apologize to the participants of the Accessibility Guidelines Working Group (AGWG) at W3C, as this blog post will criticize their work. I know everyone in W3C WGs is passionate and well-meaning. But from my work at W3C, I also recognize that there are group dynamics which reinforce the opinion inside the group and sometimes devalue outside input. It’s a group confirmation bias situation, if you will.

My goal with this article is to hopefully help to change some fundamental assumptions that lead to a situation that I find unfortunate and that might hurt WCAG 3 in the long run.

Earlier this week, a new draft of WCAG 3 was published, and it got the usual fanfare, including a very coarse change log and an announcement by the group chairs. You should read both.

The latter article points at the Motivation section in the WCAG 3 Requirements document:

The Guidelines motivate organizations to go beyond minimal accessibility requirements with a conformance model that shows which organizations demonstrate a greater effort to improve accessibility.

Support Eric’s independent work

I'm a web accessibility professional who cares deeply about inclusion and an open web for everyone. I work with Axess Lab as an accessibility specialist. Previously, I worked with Knowbility, the World Wide Web Consortium, and Aktion Mensch. In this blog I publish my own thoughts and research about the web industry.

Sign up for a €5/Month Membership Subscribe to the Infrequent Newsletter Follow me on Mastodon

Or, as the group chair announcement puts it, the requirements mean:

• facilitating adoption into law, regulation, or policy, and
• motivating organizations to go beyond minimal accessibility requirements.

The good

WCAG 3 combines roughly the requirements of WCAG 2 A and AA into a base level of conformance. The two levels were never really clear and assigning them was more of a gut feeling than exact science (nothing wrong with that), but for the last 15 years or so, it took explaining that there is no meaningful separation between the levels, especially as both were usually required by law.

In the current WCAG 3 draft, you are conformant with WCAG 3 when you meet all Core Requirements.

The bad

To allow organizations to show that they are going beyond basic accessibility, there are additional requirements:

• Supplemental Requirements – These are just more requirements, like AAA requirements in WCAG 2.
• Assertions – These are promises to do certain activities, like a visual aids review or that you have a focus indicator style guide.

In short: Nobody will give anything about those. Just like they currently do with AAA requirements. The working group seems to think by calling the levels Bronze, Silver, and Gold, depending on the number of additional Supplemental Requirements or Assertions, companies will be motivated to exceed baseline conformance.

This is a fever dream.

Conforming to anything remotely optional just won’t happen. Quite the contrary: Being asked if all the issues to meet WCAG 2 AA conformance need to be fixed is an ongoing concern.

Why this is a problem?

Mixing up required core requirements and non-required supplemental requirements¹ and assertions makes the document long and difficult to read. I have argued before to remove the AAA requirements from WCAG 2, and I think WCAG 3 non-required parts should also live somewhere else. I have built the WCAG 2 QuickRef when working at WAI, but it is really there to band-aid a systematic problem with the specification,

These supplemental requirements and assertions make WCAG 3 difficult to scan, difficult to understand, and difficult to meet.

Even if some websites meet "WCAG 3 Bronze" that means nothing to anyone. Websites might meet wildly different supplemental requirements, there is no predictability on how accessible a website is for a user. It’s a designation without user value.

In addition, AAA criteria in WCAG 2 have been used as a pressure valve to put requirements without enough consensus for the AA designation. I wrote about the Focus Appearance success criteria which was unceremoniously punted to AAA. I am sure if that meant not including it at all, the WG would have found a better compromise. AAA is the place where the WG can have their cake and eat it, too. There, even the most strong accessibility requirements can be published without consequences, unfortunately also without positive consequences for disabled people.

What to do instead?

As much as I am not a fan of including optional requirements in this way, I don’t like the stasis in the current WCAG 2 development either. I think to make websites (yeah, yeah, and apps, and agents, and mindreading devices) more accessible, there needs to be a mechanism to make the standard stronger and add requirements over time. Things that raise the baseline for meeting the standard. (Or in WCAG 3 language, supplemental requirements must eventually become core requirements.)

The best way to address this is: Modules. (Yes, I suggested this for WCAG 2, and I am suggesting this here. At least I’m consistent.)

Instead of reaching "WCAG 3 Bronze", modules for different types of contents could publish their own requirements in a neat modulous bundle. Let’s say the current supplemental requirements for captions are bundled into a BetterCaptions module. This will bring multiple opportunities:

1. The requirements that are not core requirements are not in the core of the specification.
2. Experts for different aspects of WCAG 3 can work in smaller subgroups where their expertise can shine instead of sitting through meeting after meeting about unrelated information. (IYKYK)
3. Companies who want to exceed core WCAG’s requirements can adopt the BetterCaptions module or indeed any of the modules. They can claim WCAG 3+BetterCaptions conformance when they do. Clear information about the requirements that are met and clear information to users on what to expect. (Think of it as a creative commons type string of information: CC-BY-ND = Creative Commons license, attribution, no derivatives. Easy peasy.)
4. Policy makers have the opportunity to fine-grain required compliance based on the area of certain organizations. Public organizations might have a WCAG 3+BetterCaptions+SignLanguage requirement. If a policy maker wants to require Easy Language for home pages, they can specify that: WCAG 3+BetterCaptions+SignLanguage+EasyLang.

I think assertions are interesting ideas, but they feel unnecessary. Take the focus indicator style guide one. If the requirements for focus indicators are clear, why would it need the assertion that people are trained on it. I don’t know how this makes products actually more accessible. Many people go through accessibility courses every year and it does not seem to help a lot.

If at all, these assertions feel more at home in something like the Accessibility Maturity Model. And there again, the maturity model could be a module of WCAG 3 or a model could refer to it. Or there could be a module that is called OngoingTraining.

Where do we go from here?

I cannot stress it enough: The Working Group has made a lot of good progress in the last 10 years since the Silver² project started.

But before we’re getting into the home stretch of a new standard, the working group should determine if they think this is the path they want to go down before more cost is sunk.

I believe simpler, easier to understand guidelines with fewer, clearer instructions are more effective and will be easier to adopt. WCAG 3 already feels bloated with a lot of small, itemized requirements, and we are years away from publication.

The standard will surely be refined and improved and I hope that more of the supplemental requirements become core requirements. But three years since writing about this, I’m still not convinced that a new WCAG standard that breaks with convention is necessary. If the core of requirements of WCAG 3 is equivalent to WCAG 2 AA, with a few additions, changing everything feels hardly worth the time and effort needed to make the switch. Worst case scenario would be some jurisdictions switching quickly and other lagging behind, requiring conformance to two standards for many organizations.

In any case, a standard seems to be not the right place to codify good intentions.

Support Eric’s independent work

I'm a web accessibility professional who cares deeply about inclusion and an open web for everyone. I work with Axess Lab as an accessibility specialist. Previously, I worked with Knowbility, the World Wide Web Consortium, and Aktion Mensch. In this blog I publish my own thoughts and research about the web industry.

Sign up for a €5/Month Membership Subscribe to the Infrequent Newsletter Follow me on Mastodon