What is accessibility?

As soon as the word accessibility is mentioned very strong feelings and opinions come into motion and the discussions go on all night. Therefore, I felt the need to take a shot at explaining my view on accessibility.

To me, it is all about making web sites accessible to people with disabilities and at the same time to people using different operating systems, web browsers and devices. I’m sure that the general notion when the term accessibility initially was coined that it was to focus on, and cater to, people with special needs that don’t have all the prerequisites as everyone else. A very noble initiative and a corner stone if we ever want the web to be taken seriously.

But when making a web site accessible to people with disabilities, why wouldn’t we at the same time make it accessible to people who aren’t using Windows and Internet Explorer? It’s a mindset and an attitude that go hand-in-hand for me. Surely, everyone wants to reach an audience as wide as possible, right?

A thing that bothers me, though, is when accessibility advocates proclaim that we have to stay away from using JavaScript, Flash et al, all in the name of making it accessibility. Accessibility and using JavaScript, for example, aren’t mutually exclusive. It’s all about progressive enhancement. Build a common ground and then implement enriching features in an unobtrusive way that doesn’t rule out accessibility.

So, let’s stop bickering about what we read into the word accessibility, and instead start focusing on reaching as many people as possible with this wonderful medium called the Internet!


  • Tommy Olsson says:

    Good luck!

    I've been trying to advocate this for years, only to be told that I'm wrong.

    People with some types of disabilities are treated 'specially' in the physical world. In some cases that is unavoidable. Someone who needs a wheelchair cannot be expected to climb stairs, so building a ramp is a necessary 'special' treatment. (It does, however, benefit some groups of people who don't use wheelchairs at all.)

    On the Internet, we're building a brave new world. I would like that world to be as free from discrimination and 'special' treatment as possible.

    Most of the practices involved in 'web accessibility for people with disabilities' are beneficial for far larger groups. Using <code>ALT</code> texts on images, for instance, benefit me as a dial-up user (images off) as well as search engines.

    So why claim that this is only for assisting people with severe visual impairments, who need assistive technology (e.g., a screen reader) to browse the Web?

    I get a feeling that some of the proponents of the accessibility-is-only-about-disabilities concept just want to feel good about themselves. 'Look at me, I'm nice enough to help those poor disabled people.'

    I'm not disabled, so I cannot purport to know what it's like. I believe, however, that if I were disabled in some way, I'd prefer web pages to be written in such a way that they work well for everybody, rather than being pointed to a 'special' version created exclusively for people with my particular disability.

    A web designer/developer should be aware of things that put up barriers for others. Whether that's because someone has a disability or because they use some fringe (but standards compliant) technology to access the Web is not the main issue.

    Making the Web accessible for people with disabilities is very important. By making the Web accessible to everyone (as far as possible) we don't exclude people just because they happen to have a disability. And we don't patronise or condescend by referring them to a 'special' section of the site.

  • Allan Haggett says:

    @Robert: Well said.

    @Tommy Olsson: Well said.

  • You can definately add my name to the list of 'accessibility for all' advocates, and it's refreshing to hear someone else who believes it's fine to include javascript and flash enhancements in sites that also claim to be accessible. Implemented with thought, I really don't see why the two need to be mutually exclusive either!

    @Tommy – you're not wrong, they are.

  • Olly says:

    Hear Hear!

  • Put my name on the "I fully agree" list.

  • Jules says:

    For me, Flash and JavaScript are accessibility barriers. What I mean is that I have little experience with either and although I understand the principles surrounding what may be required to make these technologies accessible, I have neither the experience, nor the skills, to make or ensure that they are accessible. Therefore, I avoid them.

  • Tommy says:

    There is also a political/commersial benifit of talking about the obvious advantage you will have if everyone should be able to access the site, as compared to talking about it as extra cost and trouble to support a small group with accessibility problems.

    As I understand it, the accessability community is getting more tolerant about javascript. For example, the preliminary version of WCAG 2 is much more allowing in this aspect than WCAG 1.

    This is an effect of that a lot of accessiblity techniques, for example screen readers, no longer has problems with javascript (except for the parts of javascritp that assumes that the user has a mouse).

  • Tommy you are wrong, totally wrong. Now, if you remember correctly I showed you proof that others in high-places also thought you were talking nonsense (no names mentioned). 😉

    I have heard of people with disabilities before but they are rare anyway (it's hard to find them ever posting on places like this) so there's not much need to consider them as user's so much as troublesome minority people.

    I mean come on, how many people does the average person know that uses specialised "Assistive Devices" to help them surf the internet?

    Advocatus Diaboli

  • Robert de Mildt says:

    not only do we share the same name, we share the same opinion on this matter! 🙂 I do wholeheartedly agree!

  • Robert Nyman says:

    Thanks for your comments, and I’m happy to hear that you agree! 🙂


    Humbly, I would say that the technologies aren’t to blame for that, but it is, as you say, a matter of experience. I know many people that know how to implement, hence I want to spread the word that it can be done.

    Regarding JavaScript, Unobtrusive JavaScript is a good read.


    Point is taken, and definitely an interesting one. I’ve been thinking of elaborating on this in a post in the future, but basically making something accessible doesn’t mean that everyone will automatically be able to take advantage of all the features available.

    I think there should be a common ground where it’s accessible for as many as possible, but naturally, people with certain prerequisites will be able to do something more.

  • Agreed.

    There is one issue however that has never been clear to me (and has also troubled me), to help explain I will recall an example I read somewhere:

    A bank has worked fine for many years with no accessibility issue. Then they decide to place an ATM (automatic teller machine) outside the bank. Transactions are now quicker for those who use the ATM outiside. A person confined to a wheelchair attempts to use the ATM but cannot reach it. They claim the bank is discriminating against them by offering services the person cannot access (assuming the service using the ATM is dramatically faster). Who is at fault? (also assuming the height issue cannot easily be fixed). Is the bank ethically and/or legally responsible to provide equal access here? If you think not, then what if the ATM was accessible out of hours, but not the tellers inside?

    My immediate reaction is that there was no problem without the ATM previously, and so the bank is not at fault, but at the same time, a website is a sort of "progressive enhancement" for many businesses, and it's not ok to make those inaccessible…

    This applies because if we progressively enhance a website with javascript, at what point are we excluding disabled users from new features, even if it is just ease of use?

    P.S. I'm not taking any side here, just merely something to consider.

  • Jens Meiert says:

    A definition search shows the diversity of meanings. And it must be said that (web) accessibility is rather a theoretical concept, since universal accessibility can never (!?) be achieved – just think of "natural" barriers created by language. (I wanted to keep this short.)

  • Robert Nyman says:


    Universal accessibility is a hard word. To me, the general concept is, to some degree, make things accessible to as many as possible.

    Accessibility means many different hings to many different people, so I just wanted to explain how I see it, express how I'm tired of people fighting over theory than making it happen, and about a mindset I think everyone working with the web should have.

  • trovster says:

    Bingo Robert, bingo. What frustrates me is people citing accessibility saying "can't do this/that/whatever" when they're clearly wrong… anyway, that's a long long argument.

  • Robert Nyman says:


    Thanks! 🙂

  • […] de ook nog onder Accessibility? Waarom trek ik dit nu zo in vraag? Wel naar aanleiding van Robert Nyman zijn artikel over “Accessibility”. “To me, it is all […]

  • Jewel says:

    Currently, I am like Jules in my understanding of javascript and Flash. 🙁 However, I have recently obtained copies of Unobtrusive Javascript, and DOM Scripting, so hopefully enlightenment is not too far away. 😉

  • Robert Nyman says:


    Most likely important steps to enlightenment! Good luck! 🙂

  • Felix Miata says:

    Practice what you preach. Strip lines 10 & 15 from http://www.robertnyman.com/wp-content/themes/east…, and advocate others do likewise. You reduce accessibility by arbitrarily setting fonts smaller than users have chosen in their browser settings.

  • […] bin nicht der Einzige. Robert Nyman, zum Beispiel, denkt über das Thema in seinem Artikel “What is Accessibility” nach. Wenn ich Zugänglichkeit in diesem Artikel […]

  • Siegfried says:


    i strongly do agree! The only drawback in making accessible sites would be that it needs more to build them. You need better trained and better motivated web designers and more time. So it would cost more. So many companies feel they do not need such an investment. Ironically accessible sites tend to not only be more accessible, but also to be much much better maintainable. Maintaining such a site would then cost much less than maintaining a traditional code monster. Any investment into general site quality would quite soon pay off.

  • […] too narrow, and I’m not the only one. Robert Nyman, for instance, ponders the subject in What is Accessibility?. When I mention accessibility in this article (and elsewhere) […]

  • Robert Nyman says:


    With setting the font with the <code>em</code> unit, it will adapt to the users text size setting in their web browser. As you can see in my CSS file, the basic setting of the font is 100% of the default font setting (normally about 19 pixels), and then making the general font respectively smaller since I don't want a minimum of 16 pixels font on all pages (something I think you agree on).


    To me it's about where you draw the line with accessibility. If you from the start write correct semantic code, separate CSS and JavaScript from the HTML-content it doesn't take longer to develop such a web site.

    When it comes to web designers and companies, I think it's their responsibility that they know how to do the job correctly in the first place.

    Where some factors of accessibility can take longer time is in the test phase, if one will want use tools as screen readers etc.

  • Robert Nyman says:


    First, I definitely don’t declare myself as an accessibility expert; I plainly express my personal opinions from my experience.

    And while I get your point, there are several factors I want to mention:

    1) First and foremost, not being able to get larger text than it’s possible in Internet Explorer is a shortcoming in IE. Then if it’s dad’s responsibility to use a web browser that gives him more possibilities or not, I’m not sure.

    2) This is a classic example of clash between classic accessibility and everything else one wants to accomplish with a web design. If one wants the default font size to be equivalent of around 12 pixels with the normal text size setting (that a majority of users have) it needs to be scaled down from web browsers’ default 16 pixels, but then naturally with the possibility to resize it.

    3) Take whatever major web site that is released: there will be design guidelines how large the default font should be for a majority of the visitors. It’s not about my or any other person’s personal preference, it’s about what will look the best for the target audience. Of course I know this is a compromise and not the best option for IE users that need a very large text setting, but it is a deliberate design decision. At least the font here in my web site is resizable at all in IE, as opposed to most web sites out there.

    Conclusively, IE is the only web browser that doesn’t offer pretty much unlimited text resizing, and that will change with the zoom feature in IE 7. So while I acknowledge the problem, I’m happy to see that it’s one that will go away.

  • Felix Miata says:

    I'm constantly amazed how many professed accessibility experts fail to understand the most elementary component of accessibility: text legibility. You cannot be making people struggle to read tiny text and call it accessible. Quite simply, text to be legible needs enough contrast, and enough size. Everything else is secondary.

    Here's why your 75% font-size is too small, even with your 1.5 line-height and white on black P text:

    Dad has typical senior eyes. He can read, but not as well as he used to. As much as possible he reads only books in large print editions. He struggles to read the small print in most newspapers and magazines. He buys a 17" flat panel to replace his 15" CRT, thinking that 3" actual difference going to make everything bigger and easier on the eyes. He gets it home, and finds its instructions recommend using its native resolution for best results. That's 1280×1024, and he's astonished to find unstyled web page text is actually smaller before.

    What happened? It turns out that he needs 14pt type to read web pages comfortably. Due to the error of applying 96 DPI on a 14" CRT screen, which in fact is only 91 DPI on the median 1024×768 resolution he was using, his IE 12pt default was actually only slightly less than 14pt. Now with a smaller error applying 96 DPI to a larger screen that is in fact 94 DPI at its native resolution, the default 12pt is smaller instead of larger. The quality is better, but that doesn't make up the difference, so he manages to find the IE text resizer, and finds that the larger setting gives him the approximately 14pt size that works for him, and many other senior adults. How should text be presented within a website?

    Now he comes to your 75% page and finds his 14pt text turned into 10.5pt text. So, he goes and finds that IE text resizer widget again, and changes it to the largest available choice, and guess what, he gets maybe 12pt, and that's too small for him to read comfortably on his fancy new larger display. In the top IE window of this screenshot is what he sees. He doesn't know every browser he's never heard of can give him what you should have given him in the first place, shown in the firefox window underneath.

    So, it doesn't matter that you size in ems or keywords or % if you're usurping your visitor's decision what size works for him with an arbitrary reduction from whatever that choice happens to be. It's rude and unwarranted. If you think your browser default is too big, then change your own browser default, and don't mess with dad's. He's probably where you'll be in 20-30 years, and you won't like it any more than he does.

  • Felix Miata says:

    Your points 2 & 3 are a statement that design trumps accessibility. Users on average want 12pt text, but designers on average want 3/4 size 9pt text. That may OK for print, but it's not OK for the web. It means the designers who want 12px either don't really know what they're doing, or don't care about the impact on accessibility, or think design is more important than accessibility. Whichever reason actually applies, their rude imposition on visitors is the same. Users shouldn't need to adjust for it on a per site basis. That's what they rightly expect their computers to do for them. Sites should be designed to accomodate rather than overrule whatever size a user finds makes his best default automatically. And, they can be so designed, if only designers would so choose. Truly astute web leaders like the W3C and WebAIM already do it.

  • Robert Nyman says:


    Correct me if I'm wrong here, but are you saying that no designer should have the possibility to decide what text size is best suitable and consistent with the rest of their design (e.g. a smaller font in a sidebar listing than the default in the web browser)?

    In that case, it rather sounds like accessibility trumps and holds back design. I don't think we will ever have a web where every web site per default has got a consistently large (and in many peoples' opinion oversized) font no matter what design is being used, just to have it a 100% according to web browsers' default font size.

    Personally I think it's ok to have a font that would, for example, be equivalent to 10 pixels if it suits the design, but then that it's naturally resizable as well. That way we get the compromise of the two worlds, without holding design or accessibility back (except for the mentioned IE problem above, of course).

  • That “How should text be presented within a website?” document is flawed and isn’t a very good reference example.

    Likewise it would be folly to assume size and contrast were the main barriers to reading text. Though I agree it does need to be legible. For example blue body text containing blue hyperlinks isn’t always legible.

    Like Robert said theoretically: User Agent Accessibility Guidelines (UAAG) is there to that help lower barriers to Web accessibility obviously the M$ browser seem to disregard some of them.

    We know Video Systems know no concept of DPI at all and only work upon pixels. As you will notice that the word "dpi" simply does not appear in any user manual for any monitor or for any video board 96 dpi setting in the Windows video settings is called “Font Size”.

    Though I agree you have to be careful to make sure you don’t accidentally add too many barriers to a website.

  • Felix Miata says:

    Wellock, the only flaw with arguable significance in that web site is that, as pt and px sizing is problematic for screen media, it fails to reconcile the 12pt and 14pt sizes that most readers want into the sizes every competent web design best practices source recommends using. That doesn't actually matter, because by choosing medium as a base, or not explicitly setting any base size, designers are actually giving users 12pt or 14pt or whatever other size it is users want.

    Size, brightness, and contrast are basics. If you can't get them right, there's little more you can do for most people, people who have no problem reading as long as text is at least the size they require. Poor legibility is the #1 web user complaint. Fully accessible pages do not produce that complaint.

    Nyman, I'm not at all saying no designer should ever size text. I'm saying designers need to know the consequences of what they're doing, and understand why and for whom they're doing what they're doing. There's no good reason they can't style to suit every personal whim if they're only designing to please themselves and/or other web designers, all of whom are not normal web users. Designers certainly have the power to produce pixel perfection for the environment they're using to do their designing.

    However, the environment of the visitor is an infinitely variable unknown. The web is an inherently fluid and adaptable medium. The designer doesn't know the display size, resolution, viewport size, visual actuity, or other user local factors that determine what works acceptably for each user.

    A designer designing for the majority of the planet's users needs to embrace the fluidity that is the essence of the web. This requires building adaptable pages, pages that can and do work for the visitor regardless of his local variables. Adaptability is accessibility.

    Small text is for legalese. Big type is good business. Browser makers give users adjustable defaults precisely so users can determine in advance of reaching your page how big is big and how small is small.

    You as designer can't actually know what size any particular visitor's default is, so you must make sure your design doesn't depend on some particular size, instead, working well whether the default happens to be 11px, 3cm, medium, largest, smaller, 24pt or anything else that fits reasonably in the viewport of whatever reasonably sized visual device the user happens to be using.

    The user's actual default size shouldn't matter to anyone except the user, so arbitrarily setting a size for normal P content at 75% of whatever it happens to be serves no purposes other than barring some users, impeding other users, and keeping people who have sampled the web and found it too difficult from trying again. It limits/reduces accessibility.

    As in print, contextual text sizing for the web is a valid means of communicating. But text sizing by the designer should be restrained by the designer himself to applying user default based contextual (relative) sizing only to things that need it, like headings, sidebars, breadcrumbs, footers and such, leaving content P text, the basic or dominant size, at the user default. In the overwhelming majority of cases, if the designer thinks his design requires a smaller base size text, then either his own browser default is wrong, or he's not designing for most of the web's users.

  • Robert Nyman says:


    As I expressed above, I don't entirely agree. Like I said, I find it to be a compromise of different goals, and trying to make the best out of it.

    Given that a mjority of the web users have the medium text setting and also that they don't even know how to change this, to get a text size that looks good designwise, you'll want to decrease the font size.

    Then I think it also comes down to the tools, i.e. the web browser manufacturers, with features like zooming.

  • I agree points (pt) units are mostly irrelevant to screen so any web reference that recommends points for the screen is talking; should we say nonsense though I know many state 16pt or greater.

    The style properties of the font will affect perception though it would be too narrow to assume that size is the major factor with readability, it depends what you want to add colour and spacing of letters, lines and the wetware factor, etc.

    One could produce a black and white website without any CSS and it would be pretty much inaccessible to a large proportion of users.

    I myself have disability and don’t find the ‘Useit Alertbox’ legible; admittedly it does use CSS and focuses on Web Usability. Also many people wrongly think “Style Switchers” help accessibility, whereas in most cases they actually confuse the user.

    There were one or two contradictions you made with the last post but we’d be swaying off-topic too much. It also wouldn't matter if the text size was 200px I’d still would have issues reading it.

    Anyway it was fun discussing the topic. 😀

  • Felix Miata says:

    Whether they know how to adjust their defaults and whether and/or how many do is simply irrelevant. You can't know that changing it can improve anything except in your own viewing environment.

    That the defaults are too big is an argument with a genesis in the mid '90's, when the default was 640×480, 800×600 was frequent, and 1024×768 was very rare. In those conditions, when the windoze 96 DPI default was conceived and applied to screens of under 50 dpi, making the default 12pt 50% bigger than 12pt print, the defaults were indeed usually too big, and quite ugly due to the pixelation and jaggies of low resolution. This is no longer the case over 10 years later. 1024×768 is the new standard, including a large proportion of 13"-14" laptops no larger than the CRTs of 640×480 days; 640×480 is history; 800×600 is fast becoming history; and high resolution of 1280 wide or higher, including for 16" or smaller laptops, is now and the future.

    What the absolute size default is that a user prefers is simply irrelevant in a good accessible web page design. It just happens to be that 12pt is what users most commonly prefer, and on average what they'll get from web pages with no text styles in today's environments.

    Doze still defaults to 96 DPI, and often that is actually close to accurate now (not infrequently erring in the other direction), making the IE 12pt default pretty much what most users prefer on average. This table shows 96 can now be pretty close, or even go under. Laptops are now often configured reasonably at 120 DPI by their vendors. This means the IE 12pt default is more likely close enough to what users want than what any arbitrary CSS adjustment you apply can hope to achieve.

  • […] 义的范围太窄,并且不仅仅是我一个人这么认为。比如,Robert Nyman 在《 What is Accessibility? 》中也探讨过这个问 […]

  • […] s of cool stuff. So, what am I bitching about? Why is JavaScript dependency bad My posts What is accessibility and AJAX, JavaScript and accessibility explain my stance and h […]

  • ReaderX says:

    What is the practical application of this approach? I mean to ask, should everything be done in em, relative to the default size?

    Can you provide one or two helpful links on where to get started understanding this issue in detail?

    More and more Windows machines are set to '120 DPI' and that increases font sizes, which breaks website designs, and also increases images sizes, for which there seems to be no fix at all!


  • Robert Nyman says:


    Regarding units to use and design: Yes, DPI is definitely a problem when it comes to designing web sites. More information can be found in: Resolution vs. browser size vs. fixed or adaptive width.

  • ReaderX says:

    Thanks very much for the reply. I will go read the article.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.