Barrier-free Web design, a.k.a. Web accessibility 2.0

This article is co-authored with Tommy Olsson.

Recently, the old debate about the definition of “Web accessibility” has flared up again, despite attempts to make peace. There are two camps in this debate: Camp 1 saying that “accessibility” means “possible to access” and Camp 2 saying that it means “possible to access for people with disabilities”.

The debate has been heated at times, with ad hominem attacks and proliferation of FUD from both sides. This internecine strife is potentially a greater threat to Web accessibility than ignorant and reluctant clients and Web professionals.

Both authors of this article belong firmly in Camp 1. We believe in building websites with no unnecessary barriers, thereby making the Web accessible to as many people as possible. In this article we attempt to explain why we believe that is important and why we do not think including everybody risks excluding people with disabilities.

Accessibility, universality, or both?

One fundamental premise for Camp 2 advocates is that accessibility deals with circumstances over which a person has no choice. That is usually translated as having a disability which affects a person’s ability to use the Web. There are, however, other situations where people have no real choice. As an example, many people live in rural areas where it is impossible to get a fast broadband connection. Arguing that they do have a choice (moving to a city) is beyond ridiculous.

For both camps, people with disabilities are a very important group of users. In the brick and mortar world, a physical disability can make it difficult or impossible to fully enjoy public or commercial services. There is no reason whatsoever to put up the same barriers in the online world. The only real difference between the two camps is that Camp 1 advocates want to take this principle one step further and remove all unnecessary barriers.

We do not comprehend the rationale of making a website that is only accessible to people with disabilities as long as they use a particular operating system and browser, with ActiveX, JavaScript and Flash enabled, and have an 8 Mbit/s connection to the Internet. We feel that a much better strategy is to build the site so that the content is accessible to everybody. Unobtrusive enhancements can then be added to increase aesthetics and usability.

This concept, often referred to as progressive enhancement, means starting with the actual content, which normally consists of text and sometimes images. Once that is in place you add styling to make the content aesthetically attractive to those who can perceive the design. After that you add usability enhancements with JavaScript, Flash or whichever technology suits the purpose. The important thing is to add these enhancements without becoming dependent on them.

It is occasionally argued that “universality” (Camp 2’s term for what we mean by “accessibility”) means you cannot use JavaScript, Flash, or anything but text marked up with HTML. That is completely wrong, and displays a fundamental misunderstanding of what Camp 1 means by accessibility.

People come first

No matter which camp one chooses to adhere to, the single most important thing is people: the human beings that visit our websites. Accessibility (or “universality”) is not about browsers or assistive technologies; it is about the people who use those browsers or that technology.

Making a website accessible to people who are blind, for instance, is to a large degree a question of making it work with screen readers and keyboards, because that is what most blind users use to interface with their computers. That doesn’t make the screen reader more important than the person using it, but it does make it something we have to consider when attempting to make our sites accessible to the people who use that screen reader.

Camp 2 advocates often say that confusing “universality” with accessibility (for people with disabilities) causes the concept of Web accessibility to somehow become diluted. That is a valid and relevant statement, but we do not share those concerns.

Selling accessibility

We, as accessibility proponents and Web developers, have to somehow “sell” the concept of accessibility to our clients and employers. That can be relatively easy for government sites, where legislation often prohibits discrimination, but it is often a real challenge for commercial sites. What is the ROI of accessibility? How does it affect our bottom line? How will it help our next quarterly report?

In cases like these, pleading on behalf of an anonymous group of “people with disabilities” can be like talking to a brick wall. It’s not that the clients are callous or cruel; they simply balance the perceived cost of accessibility against the expected increase in revenue. If the scales tip the wrong way, they’re willing to write off the “disabled” group as an acceptable loss. Money talks.

The end result is what matters

If we eschew the narrow definition of accessibility that some members of Camp 2 so vociferously stand for, we have a better chance of convincing the sceptics about the benefits of an accessible site. Losing 15% of the visitors because they use Firefox has a larger financial impact than losing 1% because they are blind. Reductions in server load and bandwidth usage can be translated directly into monetary numbers in an Excel sheet and show the gain. SEO benefits are instantly understood. And if this is what it takes to convince the client; if this is what it takes to make them “let” us build an accessible site, does it really matter all that much?

The result is a site that is accessible to as many people as possible, including people with disabilities. How can that be a bad thing?


This article has been translated into the following languages:

Posted on October 24, 2006 in Accessibility


  1. Well said, Roger and Tommy. I agree with you wholehearteedly. Although I can’t imagine that’s a surprise to you, really…

    Also, it is worth emphasising now again that the purpose of accessibility is “to allow people to access your site” rather than “to comply with a set of checkpoints”, as one of my own personal bugbears is that too many people look to do the second, because that’s easier than actually thinking about what’s necessary to do the first!

    NB it’s nice to see that my Camp 1 and Camp 2 terminology has really taken off!

  2. I don’t think that comparing compatibility with Firefox and ensuring access for the blind is fair. I see that as two separate issues, as neither issue is dependent upon the other. Making surethat a float is cleared or that a border is dotted or dashed cannot be compared to whether or not a “skip to content” link is given or that labels are added to form elements. Can you truly equate these issues?

    These issues both must be addressed, but not compared.

  3. It seems a bit unnecessary that people write at such length about this, imho. Turning words into idealogical jousting sticks doesn’t serve anyone; although I realise this particular article is merely the authors defending themselves against some rather unsavoury comments they’ve received.

    I don’t think it matters much how you use the words, as long as the people reading/listening understand what you mean. A bit like how people use all other words, oddly enough!

    The semantic arguments don’t really matter; what matters is what web developers do.

    As long as we make sites which work in a wide range of devices and are usable by a wide range of users (obviously including disabled users), then there isn’t much to cricise each other about. :)

  4. October 24, 2006 by Roger Johansson (Author comment)

    JackP: Nope, no surprise ;-). Agreed about the checkpoints.


    I don’t think that comparing compatibility with Firefox and ensuring access for the blind is fair.

    No it isn’t, and we’re not doing that - we simply used that to illustrate the mindset of many businesses.

  5. I can’t trackback you, because this website of mine is not a blog, but I’ve linked this wunderful article anyway from this page, which explains my views on what a website should look like, or rather, what it should not look like. I share your views on accessibility fully.

    Keep up the good work, Roger.

  6. I totally agree that accessibility is about more than disability and a far easier sell when put forward to clients as such, but am left wondering where you and Tommy see disabled access fitting into the universality workflow?

    Drawing from the progressive enhancement technique for the addition of javascript or flash to a site, do you first ensure accessibility for those with disabilities, then move on to tackle dialup users and minority browsers/mobile devices etc? Or some other approach?

    I guess I’m asking if you prioritise disabled access above other facets of universality, or is it equality for all? :)

  7. If I were to classify myself under one of these Camps, I would say that I fall more under Camp 1 in an ideal setting. However, this is not always the case and the real world (a.k.a. hardcore business requirements) has the tendency to set limitations to what can and should be technically achieved with respect to available resources for any given project. Hence, the logic behind Camp 2.

    Having said that, without blind fully assigning oneself to any Camp it is important to treat each case independently from one another and then apply the most suitable method.

    I have written a coverage on the intentions of a site (2005/01); namely it being a form of communication and a technical achievement and how to have the most impact as far as reaching site’s intended audience. The tone is set that of Camp 1, but it still highlights the importance of Camp 2 thought process.

  8. Web accessibility means for everyone, Web accessibility for people with disabilities means… well the obvious.

  9. Roger,

    great post.

    I gave a talk last week at our iQ Bootcamp. I’m firmly in the universality or inclusive camp. Our core mantra at iQ Content “Accessibility: now everyone’s invited”.

    Mark Magennis, from the NCBI (National Council for the Blind in Ireland) Centre for Inclusive Technology gave a great speech at our “guru” session. He first started in the camp by talking about helping disabilities. The first statistic mentioned the percentage of blind users, which is quite small. Then vision impaired, also small enough for businesses to safely ignore. Then, he went on to point out that 25% of adults are functionally illiterate. Of course, there’s also the aging process. All of us computer geeks who read blogs all night are sure to have impaired vision quicker than most. So when talking about designing accessibility, we’re designing for ourselves.

    The final point Mark made was this: accessible design is good design. Make something accessible and you’re making it easier for everyone, not just “disabled” people.

    Looking at WCAG 2.0, they also allude to this principle of design for all. Take the intent of guideline 2.1 (Make all functionality operable via a keyboard interface) in Understanding WCAG 2.0, where they mention “Some devices do not have native keyboards — for example, a PDA or cell phone”. So what camp are the W3C WAI in, if it matters? Seems clear to me


  10. Being a Software Engineer at heart, I follow a methodology when developing any web application. I use a hybrid DSDM/WISDM approach, with Web Accessibility a mandatory non-functional requirement in every project.

    Clients still ask me what Web Accessibility means, and I give them a combination of the statements made by “camp 1” and camp 2”: “It makes your site as accessible as possible, on the widest range of devices, and will allow those with disabilites to view it also. It avoids unnecessarily shutting out part of your possible audience”.

    If the term “Web Accessibility” satisfies both “camp 1” and “camp 2” critereon, then why do we argue about it?

  11. No it isn’t, and we’re not doing that - we simply used that to illustrate the mindset of many businesses.

    But by following their framing, aren’t you simply reinforcing that mindset? Aren’t you reinforcing the notion that making a site accessible to the blind is on par with ensuring a float is cleared?

    That’s what needs to be changed. We need to differentiate the importance of access to the blind (or any other handicap) as being not the same as browser compatibility. That’s what’s being lost.

  12. Thank you for a sensible (and religion-free) look at this topic.

    Just finished reading a transcript of the last Geek in the Park talk by Patrick Lauke and Bruce Lawson: “Where the rubber meets the road: web accessibility and pragmatism”. It’s an interesting (and less formal) take on the topic, and makes good reading/listening back-to-back with this article.

  13. October 25, 2006 by Tommy Olsson

    Drawing from the progressive enhancement technique for the addition of javascript or flash to a site, do you first ensure accessibility for those with disabilities, then move on to tackle dialup users and minority browsers/mobile devices etc? Or some other approach?

    Steve, I can only tell you how I work, but I think it’s very likely that Roger’s methods are quite similar.

    I focus first and foremost on the content: the text that conveys information. I’ll make sure that it is properly marked up as far as the limited semantics of HTML will allow. I’ll mark up words and phrases in foreign languages; abbreviations and acronyms; quotations; lists; paragraphs, etc. Then I’ll make sure the document validates, of course. Using valid, semantic markup offers the best possible foundation for all user agents and assistive technologies.

    This part will guarantee that the content is accessible (as in ‘possible to access’) for any user agent supporting HTTP and HTML. Anything from Lynx to Opera, search engine ‘bots, and even IE.

    In order to improve accessibility for several groups of visitors, I’ll also try to write the text in a certain way. Putting the most important information first; near the top of the page, at the beginning of each paragraph, etc. This helps users with visual impairments (even those who need a screen reader) who prefer to skim, because reading (or listening) is an arduous task. It also helps people with some forms of dyslexia, for similar reasons. And it will even be beneficial to users who have no disability at all, because we often skim as well.

    I now have valid, semantic markup where the content appears in a logical source order and the language is as clear as I deem necessary. This will be accessible and usable for people with certain types of disabilities, for dial-up users and mobile users (and almost everyone else).

    Accessibility is relative to the target audience. If I expected to have users with certain types of cognitive disabilities, I’d try to provide non-textual equivalents for the most important parts of the text. This is really beyond my skill set so, if possible, I’d ask for help from an expert.

    Then I start adding progressive enhancement. Some of that will be aesthetic, but some of it will also improve accessibility and usability for some user groups. CSS, for instance, and JavaScript. The important part is that the document will ‘work’ without those embellishments. The add-ons merely provide improvements for those who support them.

  14. I agree with all this, and the majority of the comments. “Camp 2” accessibility has always felt to me to be a really limiting approach. It feels like a subset of Camp 1, unless there are subtleties I’m missing.

    I strive to make the best possible sites on the planet - in every way - Functionality, usability, graphic design, accessibility. I probably only get 50% there, 100% of the time, but if I was only aiming for 50% (the ‘only for disabilities’ subset), then I’d be more likely to hit 25%….

  15. Great article!

    I agree with Ben that a big part of the discussion between Camp 1 and Camp 2 is semantic.

    As you say on the article, building sites is about the people that will use it, and, in my particular point of view, accessibility means that the site will be accessible for the audience is meant for.

    That may include visually disabled people and may include people with poor computational resources as well. Not to mention the visually impaired people that, additionally has poor computational resources.

    In another hand, a site whose purpose is merely visual (say, a gallery of optical illusions) may not count visually impaired people among the people that may be interested on the site, and accessibility takes a completely different shape.

  16. Outstanding. Good job Roger and Tommy. It’s good to see this thinking echoed more and more. IMO Everybody includes the disabled! One can do no worng with this logic.

    To add to this I’d like to offer two, perhaps very relevant, Accessites articles: Ten Reasons [to use for selling the concepts of accessibility to clients] and Breaking Barriers .


  17. Tommy, thanks for a very interesting and detailed reply, much appreciated!

  18. I’m too interested in making sites “Accessible”, but when I make these sites, I wonder how many of my visitors are really going to be “Physically challenged ones”?, When I think that… I say leave accessibility aside, But again it comes to me a Search Engine bot is the worst considering its disabilities compared to a Human. and then again I work with accessibility. :)

  19. I am learning about accessability and usability and have found the contents very interesting. I think there is more education needed for this stuff as people are not really aware of these standars.

    Atleast for me, right now, if I say my manager that we need to fix our website for these, he will say we will look into that later, you work on your current priorities…

  20. You have good point and very informative reading here.

Comments are disabled for this post (read why), but if you have spotted an error or have additional info that you think should be in this post, feel free to contact me.