How to Save Web Accessibility from Itself

by Joe Clark

26 Reader Comments

Back to the Article
  1. I agree, but surely (and like the British Standards Institute in the UK) the W3C are at least making vested efforts to bring about standards.

    The interpretation of ‘standards’ is an issue which can’t be resolved unless specific lines of code are defined (which is basically impossible!).

    So hats off to W3C, and lets see if some ‘standards’ can be forged using best practices by the wider web community…

    Copy & paste the code below to embed this comment.
  2. Joe’s arguments make a lot of sense; it certainly would seem that the WAI is making similar mistakes with WCAG 2 as those made when drawing up WCAG 1.

    I provide training on Web accessibility and focus on practical subjects like navigation, images, forms and tables, rather than attempting to follow the 65 (!) WCAG 1 checkpoints. WCAG 2 unfortunately fails to simplify the requirements by reducing the numbers of checkpoints and compliance levels but increasing their scope and, perhaps, their ambiguity.

    Perhaps part of the problem is that public authorities and organisations are adopting WCAG 1 (and soon, presumably, WCAG 2) as ‘standards’, forgetting that they are guidelines. A site may not reach Level AA (to be replaced with Core+) compliance, but its designer might be well-versed enough in Web accessibility techniques that the site is more accessible than another that purports to achieve Level AAA.

    Yes, more examples drawn from the day-to-day practicalities of designing and using Web sites are needed to illustrate the principles of Web accessibility — but also to inform the WAI about what users with disabilities are using the Web for and how they do it, so that we can make Web sites that they can and like to use.

    Copy & paste the code below to embed this comment.
  3. This is the most important article ALA has run in a long time and I hope it will get the desired response. You’ve done an ace job of breaking down a long, incomprehensible document into red flags and action items. There’s a huge disconnect between the scientists at the W3C and those of us who create Web sites. Many of us get our standards and accessibility information not from the W3C but from its interpreters who speak English and common sense (you, Eric Meyer, Zeldman, WSP, Pilgrim, Scott Andrew, etc.).

    There’s another disconnect between the Usual Suspects who incorporate standards and accessibility into design (Bowman, Shea, Hicks, Pick, Cederholm, Zeldman, etc.) and other strong designers who focus on pushing the envelope. I’d love to see the W3C sit down with Matt Owens, Trevor Van Meter, Cuban Council, etc., look at some of the award winning work being done, and figure out how leading edge graphic design and accessibility can merge. Maybe a summit meeting?

    Anyway, the longest journey begins with the smallest step, and this article is a big step in the right direction. Thanks.

    Copy & paste the code below to embed this comment.
  4. <<<<If you choose to make standards-compliant websites, inevitably you will have to follow the guidelines. It’s foreseeable that you could be legally required to follow WCAG 2.0. You could opt into following the guidelines or they could be foisted upon you.>>>

    1. I seriously doubt that will ever happen.

    2. No need to scare developers with some vague “in-the-future” threat.

    This reminds me of the situation a couple of years ago when all the pressure was on designers to completely abandon tables for layout (the proponents have since “recanted”). Accessibility is all fine and dandy, but people should make their own decision as to what is best for the sites they are creating.

    The dream of a site accessible TO ALL is just that: a dream.

    Copy & paste the code below to embed this comment.
  5. >>>I seriously doubt that will ever happen.

    Already many sites are legally mandated to observe WAI-derived guidelines such as Section 508 of the Workforce Investment Act in the U.S.

    When the next generation of guidelines is written, it’s not improbable that laws governing the accessibility of Web sites will be updated accordingly.

    On that level alone it is in the interest of people who create sites to help ensure that the guidelines actually make sense and are achievable.

    That said, even if all accessibility-related laws stayed as they are (or even if they went away, which won’t happen) it would still make sense for designers to involve themselves in the process of creating clear, usable, achievable accessibility guidelines.

    Why? Because many of us want to make more accessible sites, and good, usable guidelines can help us do that. Unclear, confusing, or unrealistic guidelines will not help.

    You may not ever be legally mandated to make your site(s) accessible. You may never have clients who care about it. You may never, not even in the next 20 years, personally have an interest in it or a change of feeling about it. But you might!

    Copy & paste the code below to embed this comment.
  6. Great article.  I intend to re-read it and then head over to see what I can contribute.  Accessibility is becoming a requirement with more and more clients, and I would like to be involved with this process.

    Copy & paste the code below to embed this comment.
  7. Joe Clark is a seriously intelligent and well-informed commentator. From what I can see, this is better and wider-ranging than his FIR article. Like Ben, I’d have to read it more than once.

    I’m not sure about the re-wording task. It’s a nice parlour game, and I’d like to try it. But my understanding of translation is that this approach is less than ideal. They tried this with the New English Bible and the end result does not rival Tyndale.

    Tyndale translated well because he was properly at home in both languages. What is needed is someone who understands the Web languages involved – and all the issues around their use – and who can write simple clear direct, properly colloquial, English.

    But, as Joe says, contributing to the discussions would be good.

    Copy & paste the code below to embed this comment.
  8. If Browser vendors were required to release a ‘accessibility view’ function with all of their browsers you could switch easily to a text only, properly alligned, etc, interface. This would cause developers and designers the world over to take notice.

    I propose that browsers be made to integrate a standard ‘accessibility view’, perhaps even designed as an open module of code by the W3C, the WAI or the development community at large.

    This would create a snowball effect of develpers and designers knowing that their sites were, exactly, designed to be accessible. Without browser support, we will have no common design. And without a common accessibility design goal, we can have no common accessibility.

    Copy & paste the code below to embed this comment.
  9. Interesting article :-)

    The WAI has stepped away from a rigid definition of what web content might be and has instead started trying to describe ‘all’ content in the broadest and most abstract terms. My feeling is they are now lost in the territory of linguists and utterance interpretation, the murky academic provinces where sense is sought of the “how” and the “why” of what we say, mean or do—seriously confusing stuff (one of these maniacs should be able to help though: http://linguablogs.fieldmethods.net/)

    Perhaps an interim solution would be to narrow the scope of what content bits need to be made accessible?

    If we assume the key issue is equal access to information—or at least lack of overly privileged access to information the by non-disabled—then the question becomes

    “Does the non-disabled person have privileged access to this information beyond that inherent to his lack of disability?”

    The graphic “/cognima_diag.gif” on page http://www.cognima.com/products/ conveys information about what a piece of software does.

    The alt tag says “diagram: The Cognima Server at the network operator mediates between content on the one hand and subscribers on the other”

    The sighted user gets the info about how the product is used, but the graphic also conveys unspoken branding messages about the company “This company wants to make complex things clear and simple to understand” “This company’s products are simple to use, fresh, cool, upbeat” and so on.

    My point is that the WAI should not ask us to convey these ‘indirect’ messages in writing—that’s just silly. Instead, by keeping the alt tag to the factual information the site’s designer does not disadvantage the disabled user but also, does not restrict the sighted user’s experience.

    Similar common sense stuff could be used for structure.

    Test your site in Lynx. Can you use it? Does it make sense, provide all the functionality and information it does in Internet Explorer?

    If not, then the same question applies:

    “Does the non-disabled person have privileged access to this information beyond that inherent to his lack of disability?”

    If your site is a game, then of course not—a sighted person is always going to be better at aiming a gun.

    If your site is an online bank, then you are giving the non-disabled an unfair leg up.

    In a nutshell, it comes down to effort and discipline. It can be a pain getting sites to work properly without hacks but if you don’t, you’re being lazy :-)

    Copy & paste the code below to embed this comment.
  10. I personally have never liked W3C’s explanations of anything.

    Back in the day when I started learning HTML I took a copy of the zombieweb.com homepage (when it was still active) which was a BIG three column page and essentially table hell, but I learned more about the code there then anywhere on W3C. They haven’t improved much since then which was 1998. Their specs are difficult to understand. Their site has nearly no navigation besides the link to the homepage. Directory navigation sort of comprehendible but not memorable (Just like the w3c not to use plain english in there directory structure also! Ex. Using http://www.w3.org/TR/WCAG20/ when something like http://www.w3.org/WorkingGroup/WebAccessibility/ will be more memorable to visitors and make use of a directory structure that follows the page title ‘Web Content Accessibility Guidelines 2.0’)

    Semantic code, separation of content and design are both good ideas that have origins in the W3C’s Recommendations. The W3C have done a lot to improve the web’s accessibility especially with XML, its variants, and stylesheets. They just don’t teach it very well.

    I do agree with this article though. A severe rewording/reworking of the Accessiblity Guidelines is a must. I have one major complaint so far about the guidelines. Why do they have to rule out images so often when making an accessible site? There are more than just blind/visually disabled people on the net. I think they might learn a thing or two from (http://www.webaim.org) WebAIM and this interesting article on (http://www.webaim.org/techniques/articles/vis_vs_cog) accessibility.

    OK, now I’m done venting, thanks ALA for sparking up this conversation though.

    Copy & paste the code below to embed this comment.
  11. Classic ALA article.  Thanks.

    Copy & paste the code below to embed this comment.
  12. “Abbreviations and acronyms are clearly identified each time they occur.”

    Wasn’t that the FIRST time they occur before? Doesn’t anyone else find including acronyms for everything often makes a mess of the text?

    I’ve started using a Glossary of common acronyms instead down the side of my site. Not sure if it’s a good idea or not.

    Copy & paste the code below to embed this comment.
  13. “If Browser vendors were required to release a ‘accessibility view’ function with all of their browsers you could switch easily to a text only, properly alligned, etc, interface.”

    Your solution is unworkable. Accessible websites are not necessarily text only. This is a common straw man argument. An accessible website is one that doesn’t prevent or obstruct a visitor from accessing the content. Text-only websites can also be inaccessible.

    Accessibility is a two stage process: 1.) Remove barriers to accessibility. 2.) Add in accessibility features.

    A browser could possibly handle 1.), but it can never satisfactorily accomplish 2.) since it is missing the information you need to have given it. Making websites accessible isn’t a browsers job alone. It isn’t a web designers job alone. It is the job of both groups together to deliver an accessible website.

    Offloading an impossible job to the browser is just not on.

    Copy & paste the code below to embed this comment.
  14. I have to start off by saying that I only had time to skim Joe Clark’s article before writing the following comments. So, don’t attribute anything of what I am about to say to him. Thanks.

    I have concerns about how accessible WAI material is to people (such as myself) who are interested in accessibility issues but are not technical.

    To combat this problem, I have set up a monthly group to study the Web Content Accessibility Guidelines (version 1.0). The first meeting is tonight (11/17). I am amazed that I had to start a group like this, that no other ones exist.

    As for participating in a working group with the WAI, I’ve been told that I need to participate in a two-hour conference call every week, which is impossible for me because I can’t take that much time away from my work day AND because I am hearing-impaired. It’s ironic that a group that promotes accessibility would set up an accessibility barrier for a potential volunteer with writing experience, when writing experience is sorely needed by the WAI to make it’s material more accessible to those who are not technical.

    Copy & paste the code below to embed this comment.
  15. Lori (or anyone else with a vested interest in the development of wider accessible standards for the web), would you be interested in contributing ideas and findings from WAI Groups to a new venture called GAWDS (Guild of Accessible Web Designers)?
    If so, pop over to http://www.accessifyforum.com/forum15/ and say hello.

    Copy & paste the code below to embed this comment.
  16. Meaning of title: Any game that is too hard will have few if any players, and thus no one watching (or benefiting). It will be only unactualize theory.

      After seven (7!) months of trying to upgrade my service-oriented site to full CSS (done), and WAI Level II standards (done), while getting it to work in the zoo of browsers in use (done), I’m so tired and peeved that I can’t find the words to describe it. I’ve been programming for 30 years, but I’m a counselor in real life. I do my site to help my peers. I’m not a web design/production professional, and I don’t have a budget. My revised web site still hasn’t seen the light of day, and I personally haven’t seen much of that light, either, for months. This is nuts.
      I find it inexplicable and inexcusable that existing supported browsers still can’t meet standards acceptably. So we get to hack away. But I can’t. Can’t afford the time costs.
      I find also it inexplicable and inexcusable that the W3C and the WAI, along with supposed leaders like Microsoft, cannot seem to understand the needs of the common woman and man. Computers in general, operating systems, the bloody WAI accessibility guidelines aren’t accessible. They’re not even understandable. So…the majority of the population, being left out, turn their back. I’d give a lot for fewer tricks and MUCH more reliability and ease of use. It’s about priorities, and I now set my own.
      Where this all takes me is in the direction of SIMPLE. I now do as little as possible. I’m not waiting for any new light. If it comes, good, but meanwhile, I have work to get out. When you have water to carry, you may not have time to wait for stainless steel buckets. I don’t.

    Copy & paste the code below to embed this comment.
  17. Hard to care greatly about accessiblity when ALA and how many others insist on styling text in a marginally legible color. Like, what? You’re so right! My problems with reading your text don’t exist! How could I not know that my use of your site doesn’t count. If only I were visually impared.

    Copy & paste the code below to embed this comment.
  18. Bob:

    Text color is #444, a web-smart medium-dark grey. Background is #fff, pure white.

    While #444 over #fff is not as strong in contrast as #000/#fff (the ultimate contrast), #444 is pretty darned dark and the contrast is fairly strong and quite legible unless your monitor is wildly out of step with the wide range of normal gamma settings.

    When I speak of normal gamma settings I’m not talking about bothering to calibrate your monitor and set it to the accepted sRGB standard. I’m talking about a wide range of fairly random factory settings, since most users (including some web professionals) don’t calibrate and don’t set their monitors to the standard sRGB baseline.

    Even on a Macintosh laptop set to the traditional Mac screen gamma (which is considerably brighter than most PC settings), I can’t reproduce the problem of marginal legibility you describe. I see good contrast between foreground and background.

    I hope you’re not saying that any site that doesn’t use 100% black text on a 100% white background is illegible to you.

    If you are saying that, you may want to visit your control panels and adjust your gamma to bring it closer to the norm. Not only will doing so make it easier to read text on the web, if you’re a designer, developer or content person it will also help you get a more accurate idea of what your site looks like to the “average” reader. (Admittedly, gamma variants are so all over the place that there is norm.)

    Copy & paste the code below to embed this comment.
  19. apartness—You say, “I can’t reproduce the problem of marginal legibility you describe. I see good contrast between foreground and background.” That’s a judgment call. The whole point of accessiblity is that one person’s judgement doesn’t substitute for another’s experience.

    As it turns out, I find #555 over #fff hard to read.

    What’s hard mean? Squinting. Concentrating. Feeling the eye muscles contract and re-contract, to compensate for a text that’s at the (upper) limits of legibility … Leaving the site early, before I’m done.

    Having said all that, I wonder if ALA’s legibiltity issues are compounded by a text color that’s similar to the background color framing the content: DARK background – light background and DARK text – DARK. Maybe creates a visual conundrum that some eyes struggle to understand.

    Copy & paste the code below to embed this comment.
  20. See if it’s any better now. I’ve darkened it a bit more. You make an excellent point that one person’s judgement doesn’t substitute for another’s experience. I absolutely agree. At the same time, to some extent, judgement and experience are all any designer has to work with.

    We know that if we put black text on a white background and use blue links, almost nobody will have a hard time with that color scheme, regardless of vision, gamma settings, monitor age and quality, etc. Yet we can’t limit the entire web to black text on a white background with blue links.

    We know that some color combinations represent a particular problem for people with particular types of color-blindness, and we try to avoid those color combinations. We also try to avoid combinations that will shift severely in 16-bit mode. But there are many variables. What is orange on a Mac might be brown or vaguely purple on a Dell monitor left at its factory settings.

    (Similarly: what to you is a dark background at ALA is medium-bright on laptop and Cinema screen tests here. The laptop is set to Apple Standard gamma; the Cinema screen is set to sRGB in the Windows gamma space. In both cases the background looks almost washed out; yet to you, it is dark. I don’t dispute your experience, I’m merely saying there’s a lot of variation out there and it’s impossible to predict how a color will look from one viewing situation to another.)

    Sometimes your judgement tells you something works—and testing seems to verify it. Yet a reader’s experience will tell you otherwise. So you go back in and tweak. Which is what we’re constantly doing.

    Thanks very much for your valuable feedback.

    Copy & paste the code below to embed this comment.
  21. Andreas K. Bittner has translated the original article into German, and Tomas Casper has HTMLified it.

    http://www.einfach-fuer-alle.de/artikel/rettet-die-wcag/

    Of course, you have not experienced me until you have read me in the original German.

    Copy & paste the code below to embed this comment.
  22. OK, the way I see it, Jakob Nielsen is over-confident in his reliance on feedforward analysis. It’s not surprising that his encyclia miss as much as they net. Like, when was the last time JN & Group saw that separating content from design makes the real-world usability problems of a site rectifiable in real-time, based on real-world user experience. Based on feedback. Dialogue between site users and site designers—A dimension of usability and accessibility as key to site success as the guideline on where to place the search box … Which is to say, thanks for increased contrast between text and background. Makes a difference to my eyes.

    Copy & paste the code below to embed this comment.
  23. “Foreground/background distinctions are, as they say, a personal thing. WAI has provided no evidence that authors can reasonably anticipate how a person with a visual impairment or learning disability will be confused by figure–ground combinations. Further, who runs in “256 grayscale” anymore, let alone black and white (i.e., one-bit colour, like the original Macintosh)?”

    I found this rather amusing. I do not believe that the recommendation of 256 greyscale and black and white legibility is because of people actually running in that mode. I guess you haven’t heard of the design technique in converting to greyscale to check contrast.
    It’s a pretty reasonable way to anticipate confusion if you ask me.

    Copy & paste the code below to embed this comment.
  24. http://alistapart.com/articles/saveaccessibility/#unclear

    (BTW, I see “unclear” is actually an ID on an H3 tag. Have named anchors been deprecated?  I acknowledge this is probably more maintainable than having the separate anchor; on the other hand it is a choice that makes same-page links inaccessible to some browsers.)

    It is no longer possible on many video cards to test in 256 grayscale; they don’t make grayscale available as an option.  It’s a pity, as some people do see the world in grayscale.

    http://alistapart.com/articles/saveaccessibility/#impossible

    “live description has been attempted a mere five times in the broadcast sphere”

    Not to minimize the difficulty or expense of this, but I would put the number of attempts in the thousands.  It’s called sportscasting.

    Copy & paste the code below to embed this comment.
  25. In reference to my parenthetical comment on breaking same-page links in my preceding post, I see the relevant W3C recommendation here:

    http://www.w3.org/TR/xhtml1/#C_8

    (which is on a page which is XHTML 1.0 strict and which works)

    This looks like future XHTML plans will conflict with the backward compatibility requirement.

    And I withdraw my “some browsers” comment and replace it with “Netscape 4”, as I see IE 5.1/Mac, iCab 2.9 and Lynx accept the same-page link to an ID.

    Copy & paste the code below to embed this comment.
  26. That is the anchor C_8

    (which is on a page which is XHTML 1.0 strict and which works in Netscape 4)

    so I can’t use same-page links in Netscape 4 on ALA. this is a shame, as I find the text much easier to read in Netscape 4 where it is black on white and not gray on white, which I, like Bob, find tiring to read.

    I have a problem with the idea of having to adjust my monitor, which is set to Mac standard gamma but which is aging, for a particular site.  But if you put an id of “alistapart” on your body tag, I could override the text colors for your site in my personal style sheet (in the aware browsers) in the compliant browsers, according to some article or other I read.

    Copy & paste the code below to embed this comment.