ERCIM News No.28 - January 1997

Web Style Sheets for the Visually Impaired

by Håkon Lie and Steven Pemberton

Work is currently going on to extend web style sheets to include facilities for the visually impaired as part of the HTML work within the World Wide Web Consortium (W3C) based at the Massachusetts Institute of Technology in the USA and INRIA.

HTML was originally intended as a structural markup language (defining, for instance what the top-level heading is for a document, but not how it should be displayed), but later additions by, amongst others Netscape and Microsoft, added new markup to HTML that was only there to influence the presentation, for example <BLINK>.

The Style Sheets work brings HTML back to the original aim, while still allowing authors to influence the presentation. The first specification of Style Sheets for HTML, CSS1, is currently a working draft within the W3C process, and it allows authors in a fairly simple manner to influence aspects such as font, colour, margins and so on.

One advantage of HTML being a structure language over a presentational language, is that it is consequently relatively easy to make an Aural Browser using a speech synthesiser, that just reads a document to you. The advantage, for instance, of using <EM> to specify emphasis, rather than <I> to create italic text, is that such an aural browser can read the emphasised text with more emphasis. This is clearly of advantage to all people who are functionally blind. While this of course includes visually impaired people, it can also include, for instance, people driving, or people involved in other eye-busy tasks.

Work is currently going on to extend CSS to include speech and other aural properties. For instance it might be possible to specify that headings come from the left of the audio field, and body text from the centre, or alternatively, that headings get read in a deeper, or louder voice, or both. This work was initiated by T.V. Raman of Adobe Systems, who is himself blind, and whose PhD. thesis was on the subject of voice synthesis and automatic reading of LaTeX documents. He surfs the web using the Emacs W3 web browser hooked up to a speech-synthesiser that implements an early version of the specification that is currently being worked on. T.V. Raman's web page with demos:

For more information:

Please contact:
Håkon Lie ­ W3C, INRIA
Tel: +33 4 9365 7771

Steven Pemberton ­ CWI
Tel: +31 20 5924138

return to the contents page