Still Guessing on Accessibility
I have no illusions that accessibility on the web can be tricky. It's primarily tricky because of the way developers choose to implement it. Web Axe nicely sums it up:
Most people don't realize that the web IS accessible. The problem is that designers and developers break it.
— Web Axe (@webaxe) January 28, 2013
Part of the reason developers break it is because they make assumptions about disabled users. I feel there is a trend toward assuming all disabled users are blind, all blind users surf on JAWS, all JAWS users have the same version, and this version is configured in the same way for all users.
But that's nowhere near true. The spectrum of disabilities runs from cognitive impairments through physical impairments, from dyslexia to no motor control. The spectrum of vision impairment alone runs the gamut from colorblindness, to full blindness, from age-related vision loss to at-birth loss. Even from normally-sighted users in ideal lighting conditions to normally-sighted users at an awkward viewing angle with the sun on their LCD screen.
When we think of only one disability we treat all disabled users the same. When we get feedback from one user with an issue we ourselves cannot comprehend, we tend to apply it to all users. You can learn plenty one user at a time, as I did from a web developer with dyslexia and dyscalculia. But just as a focus group of one is not adequate to extrapolate to all your users, finding one user with one variant of an impairment doesn't translate to all users with all forms of disabilities.
All of this swirled around in my head as I read two well-meaning posts from two respected web developers that, both in the original post and in the comments, show a whole lot of guesswork:
- Navigation in Lists: To Be or Not To Be, January 29, 2013 by Chris Coyier.
- On Alt Text, January 30, 2013 by Jeffrey Zeldman.
I think it's great these posts were written, even if I think we should be beyond this. While I disagree with some of the arguments in the posts, and many of the arguments in the comments, you'll find that both posts lack direct feedback from the users they purport to be supporting. The comments are littered with people guessing and asserting what disabled users need, but with little to back it up.
As you read through you will find a lot of justification built around how easy it is for the developer or how the developer thinks the user wants to "view" the content.
— Adrian Roselli (@aardrian) January 30, 2013
As commenters with experience working with accessibility or using accessibility tools weighed in, it still felt like the other commenters and the original authors didn't quite like the answers they were hearing.
And so we are, some 20 years after the rise of the web, still guessing how to implement accessibility. As a community, we're happy to build device labs where we can let people test on every conceivable mobile device, but we aren't building the same thing for accessibility. I am fortunate to have had a good deal of experience with disabled users both in my community, among my clients, and in my personal life. I am not so sure we as a web development community can claim that.
If you do wade into the comments on both articles, I recommend you read John Foliot's comment on the A List Apart article, even if others seem to dismiss it. In the CSS Tricks article, Louis Lazaris collects a nice set of references and Léonie Watson provides a quick straw poll of users. Ultimately Chris Coyier gathers some of this feedback.
People are talking about these issues. These two posts provide far more handy, contextual discussion in the comments than most web developers will bother to do on their own.
I've had plenty to say about the
alt attribute in the past (for example, let's not call it an
alt tag). But with some people asserting that
alt is no longer required in HTML5, it's easier to just link to my own retelling of why that's the case in HTML5 and why I think it's a terrible idea.
- Image alt Exception Change Re-Re-Requested, June 11, 2012.
- More on Image alt Requirement in HTML5, May 2, 2011
- Image alt Attributes Not Always Required in HTML5, April 19, 2011
Update: February 1, 2013
There is a comment on the A List Apart article from Marc Drummond that outlines cases where a description in the
alt attribute, in this context, is useful to users
Dennis Lembree (Web Axe) has added to the meta discussion with his post Leave Accessibility to the Experts Please. It's worth reading the whole post and the ensuing comments to get the whole picture instead of just reacting to the title.
Update: February 4, 2013
Turning the proper application of
alt text into a popularity contest (ok, I know that's not quite fair, but that's how it feels to me), CSS Tricks has a poll for alt Text Usage. With no context (what else is on the page, does the graph get explained later, why use a graph if a
table can do the job, etc.), it's not exactly a fair question. There are, however, a few good comments among the rest that try to set people straight.