WebAIM - Web Accessibility In Mind

E-mail List Archives

Re: abbreviations


From: John Foliot
Date: May 26, 2010 2:57PM

> Surely that would also be best practice for screen reader users?

Surely web accessibility is for more than just developing for screen
readers? <grin>

Denis Boudreau wrote:
> Geof, were you implying that Jaws will read the content of the <abbr>
> or <acronym> element if it's defined by the @title? What would Jaws
> read if it stumbled across:
> <acronym title="North Atlantic Treaty Organisation ">NATO</acronym>
> <abbr title="Monday">Mon</abbr>

See my comment above...

> My understanding has always been that by default, Jaws would not read
> the @title attribute on anything but <img>, <area>, <input> or <frame>.
> Can it read it on <acronym> or <abbr> too?

You are correct in stating that by default, JAWs (as one of many
screen-readers on the market) does not read out the title attribute values
on any element that has a title attribute. However, this alone is not, nor
should it be, a reason or non-reason to use any HTML mark-up or attribute
when appropriate. There is a very old axiom in some accessibility circles
which suggests "author proposes, user disposes", and in this case it is
quite apropos:

WCAG2, Guideline 3.1 Make text content readable and understandable

Guideline 3.1.4 - Abbreviations: (Ensure) A mechanism for finding the
expanded form or meaning of abbreviations is available.

Success Technique H28: Providing definitions for abbreviations by using
the abbr and acronym elements states:

"It is always appropriate to use the abbr element for any
abbreviation, including acronyms and initialisms. When using HTML 4.01,
XHTML 1.0 or XHTML 1.1, initialisms and acronyms may be marked up using
the acronym element. XHTML 2.0(*) proposes eliminating the acronym element
in favor of the more general abbr element."

See also:

(* Note to self - contact the WAI WG on referencing XHTML 2 in this
document. FWIW, HTML5 is also obsolescing the acronym element in favor of

> Besides that, Ben Meadowcroft had an interesting article on the subject
> a couple of years ago that sums up really well how I feel about using
> <abbr> and <acronym>:
> <http://www.benmeadowcroft.com/webdev/articles/abbr-vs-acronym.shtml>;.

Times change, so should opinions...

The 'difference' between acronym and abbreviation is one of those "how
many angels on the head of a pin" discussions/debates that just goes
'round and 'round. For the sake of providing appropriate fallback
information to our users, it's a moot point: providing the expansion is
what is important, not what we call it.

> Part of my dislike for those tags also goes back to a few years ago,
> one or the other wasn't properly supported in MSIE (can't remember which

> one if was or what the support problem was).

You are correct: early versions of IE did not support the <abbr> element.
Since IE 7 this has not been the case however, and if we want IE 6 to die,
we need to just stop supporting it - please(?).

Geof Collis wrote:
> It would appear that by default my version of JAWS 10.0 has the
> abbreviation and acronym turned off and that's just the way I like it.

Again, this is one of those 'author proposes, user disposes' situations.
Geoff might not want to hear acronyms and abbreviations expanded, however
he is not the only user out there <grin> and I personally take the stance
that a belt and suspenders approach to content mark-up is always the
better approach - we can never assume anything of our users: some users
might *want* to have this information. I have also seen (but cannot find
the URL today) a neat example of mixing a bit of JS and Print CSS to
"write out" in a printed version of an HTML document the <abbr> value - a
value-add proposition for those with cognitive issues (for example).

Dan Conley wrote:
> The issue as I understood it (and I admit that my accessibility training

> is now a few years old, but the content of the site hasn't changed much,

> so I assume/hope it's all still relevant) is that the nature of the web
> allows for users to jump into a page without starting at the top, and so

> it's possible they would have missed that first expansion (each article
> has a table of contents with links to each h2, so if they wanted to they

> could skip the introduction).

Yep, that's a fair synopsis. Again, for power-users of screen readers,
they quite often do not have the title attribute values read aloud for
them, as is their choice. However, in many modern user-agents today,
<abbr> and <acronym> also render on screen traditionally with a dotted
line under the abbreviation, and mousing over them traditionally provides
a tool-tip: again, this could benefit those with cognitive issues related
to difficulty in reading skills, or second-language issues. Sometimes it
can be helpful to the main-stream population as well. (An old story from
my early Government of Canada consulting days: the main branch responsible
for Canada's governmental policy tracking and enforcement is called the
Treasury Board Secretariat, or as it is often referred to in Ottawa - TBS.
Yet in the United States, TBS is the Turner Broadcasting System...)

> What I didn't realize then, and what my issue really is now, is that
> there are some acronyms ('Portable Document Format') that actually may
> be MORE confusing spelled out than in abbreviated form. Out of context I

> think AIDS could go either way (we're a rehab site, so communication
> aids, etc), but in an article about HIV and AIDS it seems unnecessary.
> (Originally my concerns were also for users with intellectual
> disabilities, but while we do provide less technical versions of complex

> articles on the whole these are still aimed at rehab professionals,

So this is actually the trickiest part of the whole discussion - when is
it appropriate to *not* use an acronym/abbreviation expansion? Like many
things related to accessibility, this becomes a judgment call we must make
(coupled with consistent implementation/usage). The most oft quoted
example is Radar (which was actually RADAR - RAdio Detection And Ranging),
but numerous other examples exist, including the 3 Dan mentions here: PDF,
HIV and AIDS, where in fact those abbreviations have entered our
collective vocabulary as bona-fide 'words'. In this case, using some
common judgment (based on the premise that you/we have a general idea of
our target audience) can be applied: a rehab site targeted at medical
professionals likely needs not expand on AIDS.

(One test I use is to do a Wikipedia and www.merriam-webster.com query on
the acronym or abbreviation - if it's there then I go with the assumption
that is has become a mainstream 'word' in its own right. For example, AIDS
at Wikipedia has its own page, but my other example of TBS leads to a
Wikipedia disambiguation entry, thus AIDS likely does not need the
expansion, TBS does)