W3C home > Mailing lists > Public > w3c-wai-gl@w3.org > October to December 2005

Language-related success criteria (was Re: appology)

From: Christophe Strobbe <christophe.strobbe@esat.kuleuven.be>
Date: Fri, 04 Nov 2005 11:22:20 +0100
Message-Id: <>
To: <w3c-wai-gl@w3.org>

Hi David and All,

At 00:11 4/11/2005, David MacDonald wrote:
In our conference call I believe I made an unfair comment. About the 
Acronym issue, I said "it appears that industry is lining up against 
disability consultants." I think every one on this committee is concerned 
about making a document that will provide the maximum accessibility to 
people with disabilities in a way that is realistic and sustainable. And I 
trust our group conscience and I trust our process. For that reason I 
apologize for the comments. I have a great respect for everyone on the 
committee and hope you will accept my sincere apology.

I believe the context for this was a comment that most success criteria 
that are meant to benefit people with cognitive disabilities are at level 
3, whereas text alternatives are mostly at level 1 (GL 1.1). Many of these 
success criteria have to do with language (e.g. GL  3.1 L3) and there is a 
lot of resistance to push these up to a higher level. I think this has more 
to do with the limited way in which current information technology can 
handle linguistic syntax and semantics, than with an unwillingness or 
reluctance to tackle linguistic issues per se:
- Ontologies, thesauri and other mechanisms for disambiguation meaning 
(e.g. concept coding framework) that can actually be used on the Web are 
still in their infancy (they cover too few words or meanings, cover too few 
languages etc)/
- How to make text readable and understandable is something you can't 
describe in a language-independent way, unless you concentrate on a very 
limited number of issues (see GL 3.1 L3).
- Some linguistic terms that seem intuitive to laymen can't be used because 
there's no definition that spans all languages, e.g. the term "word".
- Some linguistic criteria that seem intuitive to laymen don't have an 
objective or testable basis in linguistic data or research: e.g. what are 
"words used in an unusual or restricted way"? This is a concept that 
requires research based on concordances [1] (which many dictionary 
publishers use nowadays) and the definition of a threshold level between 
"normal" and "unusual" or "restricted" (which comes close to a judgement call).
- What you can mark up or describe depends on the technology you use: 
abbreviations, definitions, definition lists, links to glossaries or 
dictionaries, ambiguity, etc cannot be marked up in all technologies, so 
techniques sometimes come close to hacks and this can cause resistance to 
certain success criteria.

That's just my view on the situation.

[1] http://en.wikipedia.org/wiki/Concordance_%28publishing%29


Christophe Strobbe

Christophe Strobbe
K.U.Leuven - Departement of Electrical Engineering - Research Group on 
Document Architectures
Kasteelpark Arenberg 10 - 3001 Leuven-Heverlee - BELGIUM
tel: +32 16 32 85 51

Disclaimer: http://www.kuleuven.be/cwis/email_disclaimer.htm
Received on Friday, 4 November 2005 10:23:53 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 16 January 2018 15:33:57 UTC