W3C home > Mailing lists > Public > w3c-wai-ig@w3.org > July to September 1999

Re: How to describe Flowcharts, Schematics, etc

From: Steven McCaffrey <smccaffr@MAIL.NYSED.GOV>
Date: Fri, 27 Aug 1999 14:10:53 -0400
Message-Id: <s7c69c85.007@MAIL.NYSED.GOV>
To: karl.hebenstreit@gsa.gov, jay@peepo.com, w3c-wai-ig@w3.org
   

Hi Jonathan:
You make thought provoking points here.  Let me first say that Len Kasday's responses ( Len's excellent question started this thread) to my posts on this thread have already went a long way, and in many respects, even further, in getting at what would be required to provide "equivalent" access to a chart/diagram for the blind.
See the archives:
http://lists.w3.org/Archives/Public/w3c-wai-ig/1999JulSep/0285.html  

I am trying in this thread and other related threads to state precisely what "equivalent" means.  I feel I am only qualified to attempt to clarify this from the perspective of a blind person since I am blind myself now, and have had sight in the past and so can contrast what I used to gather with my eye (one eye was blind since about the age of 3) to what I can now gain only with my ears (well, and fingers with Braille).  

What is the nature of information gained by the eyes when viewing a chart
diagram?

">If you are seeking one simple answer that works for all problems, probably
>this is a little tenuous to be a thread."

I am not seeking one simple single answer, (although that would be nice), but a single question that would be at the root of a tree of questions a web page author should ask when trying to describe the information contained in graphical representations to the blind. 
I am thinking of the WCAG guideline,
"        1. Provide equivalent alternatives to auditory and visual content."

As more complex information comes onto the web, it will be less and less obvious the what "equivalent" means and thus the greater the need for the kind of understanding I am asking for.  
I realize the problem is difficult, but worthwhile and necessary to think about.  Why not try?
Len's food chain example (see archive reference above) is an excellent start.  What we need to do is build up a set of examples grouped into classes.  Each class would begin with one like Len's.  The other members of a class would be derived by systematically and gradually generalizing to higher levels of abstraction.  Then, to give guidance to web authors, we go in the reverse direction: Create a set of questions that would lead a web page author down the tree of example classes. That is, from higher levels of abstraction to lower levels all the way down until the example is sufficiently similar to the flowchart/diagram the page author  wants on her/his  page.  This is one way, there may be others.  As has been discussed on this thread, An automated tool/program could be designed to help with this, like a high level bobby, but interactive.

     All information has an underlying structure: how it is rendered causes problems in communication.  If you do not know what information you are trying to convey in a perticular medium how can you "provide an equivalent alternative"?
 
">However if you at first accept that a subjective view, possibly within a
>context has a value then, we no longer need to worry too deeply for the
>present (about the semantics).

Yes, exactly.  Although there may not be a way to state the general rule, it should be possible to break the question into cases.  How to do this?  Your statement helps to point the way: context.  So, to restate my goal:
If you are trying to "provide an equivalent alternative" ask yourself:
1. What is the context?
2. What is the subjective information you want the readers/viewers of your web page to understand.  
Universal can only be universal if you can translate the subjective information in a given context into other mediums (e.g. graphics to text, text to graphics, video file to audio file, audio to video, 
information without color to  information conveyed by color, etc.)

">In a recent copy of wired (september?) I was stunned that all the graph
projections for internet stock growth were displayed as continouos growth,
whereas the history was saw tooth.

">Describing the nature of the saw tooth is evidently of benefit to the blind.
">However, this is bound to be subjective, and always will be, forecasting is a dodgy area.

I'm not sure I understand.  What is subjective?  
a) The description of the saw tooth?
b) Some say the growth is continuous and others say it is represented by a saw tooth pattern? or c) both?
  
A description  of the nature of the saw tooth is what I am after.
One characteristic to describe you have already mentioned: A saw tooth represents discontinuous growth.  This is already a big step forward.  It's not so hard after all.  The next level is to ask
What can a person answer by lloking at the saw tooth?  Well, how about:
1. What is the year in which the   biggest jump in stock growth is forcast?
2. Is there an average size jump?  If so, what is it?
and other questions.
If the saw tooth is more complex, showing superimposed saw tooths (teeth?) perhaps for individual stocks, or sectors of stocks, I might want to know
What stock, sector, is projected to have the largest jump?

">The first step is to notice the valid description, and this may not always
">be available.

I'm not sure  I agree, or at least, I think one should try a little longer and harder in any given case before one decides that it is not possible to provide a valid description.  How long and deeply should someone think before they give up on trying to provide an equivelent alternative.  I believe that if people reflect even say 5% longer and deeper, they will find it is really not so hard after all.  We all just need to Practice cross-modal communication.  People who are readers who  have a friend who is a non-reader should practice.  Take a non-trivial text description and try to convert it into a picture.  If you give up after only 5 minutes and say to yourself "I can't do it" or even worse "It's impossible", I am simply saying try a little longer and harder.  Are you a sighted person who has a blind friend/colleague?  Take a non-trivial chart/diagram and try to describe it. If the chart is really very complex,  can you both  design a series of questions that the blind person could ask you to get the mot from the chart/diagram.

Something is impossible until it isn't any more.

">I saw a childs smiley face drawing, however in reality it was a house upside
">down.

If the child were the one putting his/her drawing on the web and wanted to communicate something to a blind person, what I would like is a text description that corresponds to what the child wants to convey with his/her drawing.  So the Alt= tag should be "Smiley Face" (or "Picture of someone happy"...)and not "House upside down".  If the child is putting the information on the web the child's subjective view should be communicated with his her alt= description.  If you were putting it on the web, you would use your description.  In both cases,  what you are trying to convey is the information to be translated from a graphic to a description if you want to provide an "equivalent alternative" for a blind person.  

">The artist Bridget Riley in the forward to her recent book positively
">stated, and it is a commonplace amongst many artists that an attempt to
">describe what she is attempting to create saps the energy from the >creative process.

Off topic of thread comment: 
That's sad.  Some people say that knowing the physics behind a rainbow takes away from their  enjoyment of the beauty of the rainbow.  The late physicist Richard P. Feynman said he actually has an increased sense of beauty because he could appreciate two aspects of beauty: 
1. the visual appearance of the rainbow which he shares with other people who find it beautiful and 
2. he also has the knowledge of the beautiful laws that underly the rainbow.

">This approach will not suit everyone"
If you mean the approach of asking a web page author to ask deeply herself/himself  what she/he is  attempting to convey with a picture, text description, audio file etc. in order to provide an equivalent in another format?  If this approach is not used, either manually or with the help of an automated tool, then, in my view many people with disabilities may have a limited alternative, but it will be far from "equivalent".
  
Will you be satisfied with a page author who says to you:
"I'm sorry, but there is no way to provide a picture equivalent of this text for non-readers.  I've tried for 5 minutes and could not think of a way, therefore I know it is impossible"?
We can do better than that.  Try for the broadest sense of the word "equivelent".  Try again, and again.  As a community, we will get better if we think it is possible.  It is possible, I believe.  
Yes, in *all* cases, to some degree.  What I am asking for is a greater and greater degree.  The time to really pin the meaning of "equivalent" down is now before really complex information is everywhere on the web, if it isn't already.

Thanks for these thought provoking points Jonathan.  I had to think hard before replying.  I had to write, edit, refine, re-write...  I had to express better what I am asking for.
-Steve


jay@peepo.com 

a www for those learning to read.

Please send us links to your favourite websites.
Our site www.peepo.com is a drive thru.
When you see a link of interest, click on it.
Move the mouse to slow down.
It is a graphical aid to browsing the www.
We value your comments.


  





------
Steven McCaffrey
Information Technology Services
NYSED
(518)-473-3453
Received on Friday, 27 August 1999 14:14:57 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 5 February 2014 07:13:33 UTC