Re: [xml-dev] xml + style.

[David <dlee@calldei.com>]

I suggest the core problem is simplier then this.    Ignoring the unique 
charactoristics of legislation (as the blog was mainly about).
I work with Clinical authors and documents.  But I don't think that is 
unique.  I suggest the problem is the difficulty of humans to map 
tangible representations to more abstract ones.  As well as their 
feelings that the visual representation is the *preferred* 
representation.    There was a great quote at Balisage a few years back 
(pardon me for forgetting the speaker or the exact words) but it went 
like this.
Its about the advent of desktop publishing now the push towards semantic 
markup.

"For 20 years we've been empowering people to create their own 
documents.  With desktop PC's and programs like Word or Pagemaker or 
other document editing we've been teaching people that they can *own* 
their own document presentation.   Put headers here, bold there, italics 
here ...
Now we come and say 'Forget everything you've been taught ! Don't put in 
style markup.  Put in semantic Markup and "trust us" it will look right".

No wonder we have problems with getting people to use XML and when we do 
they still want to do presentation.

Its deeper then that though.


I have a difficult time with my authors having them using a tag like 
<cite> instead of <italics> ... because they want (really badly) to 
control how something *looks*.  Even when explained that on different 
devices it will look different, and some wont even *have* italic fonts,  
and that a majority use case for this data is *data* not presentation.   
That just doesn't sink in.   People intrinsically seem to think in a 
visual representation for writing.

I suggest this is 'human nature' or at least a few thousand years of 
training.   Document authoring for the lifetime of the human race has 
been mostly about presentation.  I suggest this crosses all languages, 
cultures, and generations (some exceptions but I think very few).
It may be because people are *taught* to think this way or maybe its 
*ingrained*.   Much of the early written language is hyrogliphic in nature.

Even today, with intelligent, well trained people, it seems 'against the 
grain' to get document authors to think in terms of meaning instead of 
presentation.
They want Paragraph, Bold, Italics, Indentation, Bullets.  This visual 
display is how we think of documents.
Its very very difficult to think in terms of "Section" , "Citation", 
"Emphasis" , "list".

And this is when your starting from scratch.  I've had the opportunity 
to author clinical document schemas and its literally pulling teeth to 
tell the authors "No you cant specify Italics, you have to use a 
<medication> tag instead".    And if you give them a visual Preview 
function ... (as they really want) then they end up using <medication> 
tags for anything that that should 'look that way in the preview' even 
if its not a medication.
And these are highly intelligent educated people.   I've had to get over 
thinking "they just dont get it" and start thinking "this is how humans 
are wired to think".

And that's when creating a schema and document type from scratch and 
paying people to author explicitly for it.

I really do feel sorry for the author of the blog posting about 
Legislation.  In that case your dealing with centuries old practices of 
a particular format of written (presentation) documents and workflows.  
And the participants don't want that changed.    Someone else from 
outside is coming in and saying "We need electronic versions of these 
documents".   But the benefit is all for someone else, the authors and 
participants don't benefit so not only is it difficult technically but 
its difficult socially.

The same thing happens in the Medical field.   Big Organizations are 
coming in and saying "You need electronic health records".  But the 
doctors and nurses dont want them - at least for their everyday 
practices.   They don't benefit personally from having to use a computer 
to enter or view health records.
There are generations of practice of using written records and they work 
very well for the individual practitioners.  This is in some ways worse 
then the legal field, because documents are *hand written* not printed.  
Its not an issue so much of "Header" and "Bold" but more of margin 
writing, tick marks, shorthand notation  ... things that are 'out of 
band' of even presentation markup based schemas.


-------------------------
David A. Lee
dlee@calldei.com
http://www.calldei.com
http://www.xmlsh.org


On 6/20/2010 9:33 AM, Greg Hunt wrote:
>
> What sometimes seems to happen is that people try to map multiple sets 
> of equal-valued but not really overlapping meaning about something in 
> the world (like perhaps legislation) onto a model with a "real" 
> meaning and some other stuff that must be presentation or otherwise 
> not quite as important (I suspect that that is the problem that the 
> blog author has).