W3C home > Mailing lists > Public > w3c-wai-ig@w3.org > January to March 2000

RE: Validation as test for basic accessibility

From: Charles F. Munat <charles@munat.com>
Date: Thu, 20 Jan 2000 12:09:03 -0800
To: "Steven McCaffrey" <smccaffr@MAIL.NYSED.GOV>, <charles@munat.com>
Cc: <w3c-wai-ig@w3.org>

I appreciate your response. I did take a look at Bruce's original challenge, but I still don't see it your way.

Bruce said:
"If a page validates, odds are that it is accessible!"

This is not the same as "if a page validates, it is accessible."

I think that the real problem here is a fundamental miscommunication. I read Bruce's original comments as being relative, not absolute. And he was clearly asking for anecdotal evidence to support his assumption. Viewed in this light, I think he makes a good point, and that it is worth pursuing as an intellectual exercise at least. 

For example, it raises questions such as these in my mind:

1. Are there other ways to approach "selling" or teaching accessibility that might be more effective?

2. Should different groups be approached differently?

3. If so, how can we improve our current methods?

4. How many different methods are required for maximum effectiveness? What is the minimum number of methods required for reasonable effectiveness? (And at what point do we reach "reasonable effectiveness"?)

5. Are some methods more effective than others and if so, why?

6. What role does valid code play in accessibility?

It seems to me that some of us have gotten perhaps a little too comfortable with the status quo. When Bruce first raised this issue, there were several reactions that seemed a bit knee-jerk to me. Instead of viewing Bruce's idea as an opportunity to ask questions, probe other possibilities, and perhaps improve our methods, some seemed to view his words as a threat. How else to explain the argumentativeness of some replies, yours included?

I read somewhere that people think and speak in a variety of ways, and that these can be seen as a continuum. At one end are people like Bruce (and me, to be honest), who use the group as an extension of our own minds. We will often have only a vague idea, sometimes a crazy idea, but we will immediately bring it to the attention of the group to see how it plays out. We have no real attachment to it, it's just a wild idea.

At the other end of the continuum are those who remain silent until they've thoroughly evaluated each new idea. By the time this type of thinker opens his or her mouth, every argument is in place and he or she is fairly convinced of the value of the idea.

The problem occurs when people from both ends of the continuum are in the same group. When person A mentions his wild idea, person B thinks that he means this as a serious suggestion and responds accordingly. Likewise, when person B first mentions an idea, person A might not take it seriously enough, thinking that person B is only making a suggestion.

I think that's what occurred here. Bruce can correct me if I'm wrong, but I think he meant to stimulate discussion (he succeeded admirably) and to play devil's advocate. Not that he doesn't believe it's a good idea, just that taking it to the group is not an announcement that he's proved his theory, it's the process by which he hopes to prove his theory. It's a challenge, if you will. So to ask him for a "correlation analysis" misses the whole point.

Unfortunately, instead of considering the idea with an open mind, several respondents immediately took up arms to insist that Bruce is wrong. That pretty much defeats the whole purpose of Bruce's post (IMHO).

I suggest that instead of insisting that valid sites are automatically accessible or vice versa, we look at Bruce's idea as an invitation to explore other methods for promoting accessibility. After all, have any of us done scientific studies to prove that our current methods are the best? Kin, you reacted quite strongly and defended your own teaching methods, but aren't those based on your own experiences rather than a formal and rigorous scientific study?

Steve, you mentioned shifting the burden of proof, but I think it is you who did so. When you demanded a correlation analysis from Bruce, did you offer a correlation analysis to support your own theory? As Bruce did not intend his comment as a statement of fact (if my interpretation is correct), he had no burden of proof. By interpreting his post as a statement of fact and demanding proof, you have yourself incurred the burden of proof. After all, you are making the claim that Bruce's statement IS false, whereas he is only ASKING if it is true.

I suggest we move beyond the arguments and instead use Bruce's idea as a starting point from which to explore new methods for teaching and/or promoting accessibility. Bruce, what say you?

Charles Munat,
Munat, Inc.
Seattle, Washington

P.S. Steve: I didn't mean my comments about sarcasm as a reprimand or a put down, nor was I offended by them. I meant them literally: I don't think that the sarcasm helped your argument. That's all.


-----Original Message-----
From: w3c-wai-ig-request@w3.org [mailto:w3c-wai-ig-request@w3.org]On
Behalf Of Steven McCaffrey
Sent: Thursday, January 20, 2000 8:17 AM
To: charles@munat.com
Cc: w3c-wai-ig@w3.org
Subject: RE: Validation as test for basic accessibility
Please look at Bruce's original challenge to the group and then look up the invalid argument form known as 
"shifting the Burden of Proof" and maybe my response will be more clear.
I do concur with you however, I was too sarcastic.  I stand corrected.  You are correct.
I was wrong to use sarcasm.  I hereby publicly apologize to Bruce and any who also was offended in any way by that.
>From now on, I will endeavor to site the point I think is invalid and just say what invalid form I think is being used.
Thank you for correcting me.
-Sincere regrets to all,
I still stand by my substance though :) that
even if it were the case that it just so happens that the "state of the web" happened to have say 90% of the sites pass validity checks and are also accessible, does not proove,, even by the odds, that just because a page checks out to be valid it will b accessible.  So, it should not be taught 
"If you make your page valid, it will be accessible".  This is simply false.

Let me use some notation.
What does everyone agree with?
1. For all W, [if V then A]
2. For all W, [if A then V]
3. For all W, [V equivalent to A,
that is 1 & 2 is true.
W = Web page, V = Valid, A = Accessible.
These are quantified implications, the last going both ways.
If you want to show one, you have to use a valid form of reasoning.


>>> "Charles F. Munat" <charles@munat.com> 01/19/00 12:56PM >>>
Bruce wrote:
"My own personal experience is that producing valid code is an invaluable
step towards producing accessible code."

Steve replied:
"Precisely, Bruce, it is a step, not all steps.  Glad you now agree with me and others."

Bruce also wrote:
"It is my impression that authors who care about
validity care about content and much less about presentation, and therefore
they don't bother with multimedia nor fancy CSS -- so it happens that their
pages ARE accessible, even if they did not work to achieve this."

And Steve replied:
"It's my impression that descent, good-natured, honest, law-abiding people who follow the law follow the law even though they did not look up every law on the books and consciously decide to follow them.  wow, isn't that amazing?  I bet we really don't need laws, what about you Bruce? Think we're on to something big?"

My response to Steve:

Putting aside your obvious sarcasm (which, frankly, does not help your argument), you're barking up the wrong tree. I've been following this thread with some interest, and I hope Bruce will permit me an attempt to restate his argument.

First, I don't recall Bruce saying that validation = accessibility. In fact, he clearly acknowledged the difference. So to point out that it is "a step, not all steps," is a bit disingenuous. You are wasting time arguing where there is no disagreement.

Ditto for your second point, which implies that Bruce thinks the WCAG unnecessary. He said nothing of the sort, neither does he believe that or he wouldn't be on this list. Again, you're being disingenuous.

What Bruce DID say, as I understand it, includes the following:

1. HTML validation is an important part of ensuring accessibility. Not the whole of it, but an important part.

2. Selling people on validation might be easier than selling them on accessibility.

3. Once sold on validation, the jump to accessible code is much smaller (thus less frightening).

4. While we continue to push for accessibility, perhaps we should further augment our arsenal by pushing for validity as well. A sort of one-two punch combination.

As I see it, the real question here is whether pushing validity will help us to sell accessibility, or dilute our message. So far, I haven't seen anyone address this issue.

I currently build/consult on web sites and I advertise myself as specializing in accessible web site design. Furthermore, I refuse to take assignments if the client won't agree to meeting accessibility standards, at least to single A (with most AA and AAA checkpoints as well). I have been using the validity argument for some time, and, properly done, I believe it does help.

That said, I would like to hear what others think on this issue. Instead of arguing about whether validity = accessibility (which was never Bruce's point anyway), let's address the real issue that Bruce has raised.

Finally, speaking as someone who is often too busy to participate in this forum, I think we could significantly reduce the amount of wasted ink (pixels?) if we read carefully the arguments of others and responded only after we were certain we a) understand the argument and b) have something of value to contribute. Then, perhaps, instead of receiving twenty messages on a topic, fifteen of which are off the point or redundant, we might receive only five. But those five would clearly and concisely argue the issue (sans sarcasm, please).

For what it's worth, that's my two cents.

Charles F. Munat,
Munat, Inc.
Seattle, Washington
Received on Thursday, 20 January 2000 15:08:57 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 13 October 2015 16:21:07 UTC