Syd,
I think this is problematic:
At 04:07 PM 7/11/2003, you wrote:
>It is my belief that there are many projects out there that take a
>flattened TEI DTD file (i.e., TEI Lite) and make modifications to it
>directly. The TEI Council recently discussed this issue, and came to
>the conclusion that while we have no particular objection to people
>doing that (I'd personally say no more objection than to people using
>MS Word :-), the TEI does object to the end result being referred to
>as a TEI DTD, and the documents that conform to it as being in TEI
>format, because they're not.
Take my Scholia demonstration project, for example. (Those who've never
seen it can look at http://www.piez.org/wendell/Amsel/Amsel.html.) For
local use, I created a flat DTD containing just the parts of TEI I needed,
plus my extensions. Just as you describe, I started with TEI-Lite. In order
to be able to claim conformance in good faith, however, I also created a
canonical model in the prescribed way, by calling in modules (generated by
the Pizza Chef), declaring my extensions in an extension module,
documenting them and the rest. So far so good.
The documents (well, actually there is only one -- it's a demonstration!)
conform to both DTDs.
Does the TEI really object to my local processing DTD being described as a
TEI DTD? If so, how do I get across that apart from my four new tags, it's
TEI? (I do call it an "TEI with local extensions", and I presume that's
okay.) But more to the point: if the document currently has a DOCTYPE
declaration that points to the local processing DTD, does that mean the
document is "not in TEI format", even though with a single tweak of the
declaration to point to the other DTD, my TEI-pizza-with-extra-toppings, it is?
Or maybe, since my modifications are in an external module, and not the
document's internal subset, it isn't TEI in any case?
I submit that by using TEI in my project, I am getting all the advantages
promised therewith. As you know, for me this is a choice, not a necessity
-- frequently, the overhead of analyzing and modifying TEI to special
requirements is actually greater for me than it would be to create a
lean-and-mean DTD from scratch, that would be tuned to my requirements *in
utero*, as 'twere -- which is what I have done for some other projects. (In
this, I am confessedly an exception to the rule that adopting TEI saves
work: unlike many operations getting started with this stuff, I am already
tooled up to create DTDs and schemas and can do it pretty quickly and
easily.) But in this case, the promise of interchange and interoperability
that TEI gives me is too great to justify going it alone: the project is
not far off from a "classic" TEI-type project, and while I could certainly
write a transform from a bespoke tag set if I ever wanted TEI, in this case
few extensions are called for, and they're really just syntax sugar, so
what the heck. (In fact, in my case, since the actual requirement for
interchange is entirely hypothetical and long-term, the advantages here are
more *social* and *political* than they are technical: it's now easier for
me to bring in interested colleagues, leveraging their TEI knowledge, than
it would be if I had my own schema with my own tags -- and I don't have to
defend it to Lou! ;-> But that's okay: interchange is a social and
political problem, as well as a technical one.)
Last year at the Members' Meeting in Chicago I argued that it was mistaken
and dangerous to identify TEI with XML, and indeed with any particular DTD,
including P4. Since we already recognize that "TEI" is more than DTD
conformance in any case (the DTD doesn't describe or qualify tag semantics;
for that we need the Guidelines), should we really be splitting hairs and
saying that conformance to a particular, normative DTD (or schema) is of
the essence, and faithful application of the tag set in adherence to the
Guidelines is subordinate to this?
There are many areas in life where we don't have machines to give us thumbs
up or down -- should this be one where we do? I know this complicates the
OP's question (which I am glad he raised, since it's a critical one) -- but
I think it's a misconceived effort to try to boil the question of
"conformance" down to a rigid prescription of (a) validation regimen and
(b) system architecture (e.g., whether your extensions are in an internal
or external subset: whee!). This is not just a syntax problem; it's
inevitably about tagging semantics as well. Wisely, the Guidelines
recognize this. Interchange is about *communication* -- hence its social
and political aspects -- and I'm afraid if the TEI places too much emphasis
on conformance to the letter of the DTD in preference to the less
formulaic, but no less important aspects of document tagging -- the real
advantages of using a widely known tag set in a familiar way -- many babies
will be thrown out with the bathwater, and many who might have looked to
TEI to *enable* them will see it as simply an encumbrance. (Besides babies
and bathwater, other homely cliches about horses, carts, dogs and their
tails come to mind.)
Indeed, it's often been remarked how it's possible to have a document that
conforms to TEI as far as the letters in the machine are concerned, but
which strays very far from it in spirit.
<!DOCTYPE TEI.2 SYSTEM "/home/wherever/mylocalteistuff/tei2.dtd"
<!ENTITY mobydick SYSTEM
"http://www.ibiblio.org/gutenberg/etext01/moby10b.txt" >
<TEI.2>
<teiHeader>
<fileDesc>
<titleStmt>
<title></title>
</titleStmt>
<publicationStmt>
<p></p>
</publicationStmt>
<sourceDesc>
<p></p>
</sourceDesc>
</fileDesc>
</teiHeader>
<text>
<body>
<p>&mobydick;</p>
</body>
</text>
</TEI.2>
Okay okay, if that text file contains any '<' or '&' it won't parse -- but
you get the idea.
Mightn't we say that "conforms to TEI" is in itself underspecified, and
that you need to say which TEI DTD you have in mind to make a claim that's
very meaningful? The more successful TEI is, the more true this will be
over time....
Cheers,
Wendell
======================================================================
Wendell Piez mailto:[log in to unmask]
Mulberry Technologies, Inc. http://www.mulberrytech.com
17 West Jefferson Street Direct Phone: 301/315-9635
Suite 207 Phone: 301/315-9631
Rockville, MD 20850 Fax: 301/315-8285
----------------------------------------------------------------------
Mulberry Technologies: A Consultancy Specializing in SGML and XML
======================================================================
|