- From: Charles McCathieNevile <charles@w3.org>
- Date: Thu, 22 Mar 2001 05:32:26 -0500 (EST)
- To: Wendy A Chisholm <wendy@w3.org>
- cc: Anne Pemberton <apembert@erols.com>, <w3c-wai-gl@w3.org>
For Japanese there are published vocabularies... (I don't know about grammar checking in Japanese - mine is never good enough to understand the checking). I have built a spell-checker for a language that uses inflection to identify most grammatical function - the knowledge more or less had to be built-in to make it work, so it would be feasible to build a checker based on structures used. (Finnish works like this). As a non-expert it took me a few days. In general, vocabulary testing is not complex, and we can expect it to be understood by people in each language. One of the important tests is whether to use dialectical forms - there are a number of ways that people say something in english. Some of them mean one thing in en-US and something else, or nothing at all, in other variants of english, such as en-gb. A good reference such as the Oxford English Dictionary tells you this information, and there are trained linguists in most languages. We should be able to get help in this from the i18n group, who could point us to groups like the summer institute of linguistics - http://www.sil.org I think - who in turn can point us to relevant reference work in different languages. I agree that the general process is an important part. I would invite people to look at the latter part of the 19 march draft of Techniques for authoring tool accessibility (or later drafts as they are available) which includes some material on how to test for conformance. http://www.w3.org/WAI/AU/ATAG10-TECHS chaals On Wed, 21 Mar 2001, Wendy A Chisholm wrote: what about languages other than English? Do we know of tests to run on written text in Japanese? Chinese? German? Finnish? Spanish? Hebrew? Arabic? How do the tests differ between dialects? Addressing the process used to determine if a site is accessible is very important. When I presented to a group at the U. of Wisconsin a couple weeks ago, they said one of the biggest lessons they learned was about the process that I use to determine if a site is accessible. This includes which tools I use and how I use them and most importantly, what questions do I ask myself as I go through a site. This process has not been documented for WCAG 1.0 and should definitely be part of making the requirements clear for WCAG 2.0. This gets back to a testing checkpoint that Len suggested. Perhaps we list tests that an author or evaluator must run: 1. spelling 2. reading level or grammar check 3. validate syntax 4. lists of questions to ask:... I'm not sure where this fits in. Is it a checkpoint? Is it a technique? Regardless, something like this would be valuable and something I get requests for all the time. Thoughts? --wendy At 05:01 PM 3/14/01 , you wrote: >Kynn, > > I've exerpted from your note to Al: > >At 07:35 AM 3/14/01 -0800, Kynn Bartlett wrote: > >At risk of beating a drum, I really need to emphasize that the above > >description -- which has been used by a number of people to illustrate > >the problem -- is incomplete because it is really only along one > >axis. There are -numerous- reasons to try to communicate, and applying > >a fog index (e.g.) to -every- form of written or verbal communication > >is improper. E.g., editorials, advertisements, parodies, humor, > >fiction, and a vast number of other content types are even -harder- > >to apply such a standard to. > >First of all, Kynn, a reading level index is NOT a fog index. It will not >point out stupid word choices, or unsupported claims. Reading levels are >used in all areas of education, and routinely include parodies, humor, and >fiction. Even advertisements (if they have enough words in them to >measure), have been tailored for young readers as well as their parents and >grandparents... Even the ubitquitous Weekly Reader mags include editorials >and think pieces aimed at children just learning to read. The Reader's >Digest, which is aimed at the average reading levels, has pages of jokes >that can be read without having a sheepskin on the wall! > >Yes, there is a mindset in all this, a mindset that, for me at least >boggles at the resistance to setting a checkpoint that ensures that general >content will be readable by a general audience. Not all disabilities are >aided by the slick "devices" that make the web possible by varying the >presentations possibilities. Sometimes it is the mundane, everyday stuff, >such as reading level and illustrations that carry the day. Never >underestimate the value of getting some mud under your nails ... > > .... From warm and sunny Virginia where the bulbs are > blooming, the temp >is in the sixties and seventies, and the itch to get into the garden is >rampant.... > > Anne > > Anne > > > > > >Anne Pemberton >apembert@erols.com > >http://www.erols.com/stevepem >http://www.geocities.com/apembert45 -- wendy a chisholm world wide web consortium web accessibility initiative madison, wi usa tel: +1 608 663 6346 /-- -- Charles McCathieNevile http://www.w3.org/People/Charles phone: +61 409 134 136 W3C Web Accessibility Initiative http://www.w3.org/WAI fax: +1 617 258 5999 Location: 21 Mitchell street FOOTSCRAY Vic 3011, Australia (or W3C INRIA, Route des Lucioles, BP 93, 06902 Sophia Antipolis Cedex, France)
Received on Thursday, 22 March 2001 05:32:27 UTC