At 08:45 AM 5/6/2008, you wrote:
>I find it oddly difficult to deal with something so simple as the
>following opening sentence from a chapter in Bleak House:
>The name of MR. VHOLES, preceded by the legend GROUND FLOOR, is
>inscribed upon a door-post in Symond's Inn, Chancery Lane.
>If you want to be very fussy about it, the sentence presupposes the
>existence of a distinct document that is rendered in it its entirety,
>but not in the order of its actual existence:
>GROUND FLOOR MR. VHOLES
>I have seen:
>The name of <hi rend="smallcap">MR. VHOLES</hi>, preceded by the
>legend <hi rend="smallcap">GROUND FLOOR</hi>, is inscribed upon a
>door- post in Symond's Inn, Chancery Lane.
>Small caps is what actually appears in the text. I could imagine, with
>or without a rend attribute:
>The name of <quote type="inscription">MR. VHOLES</hi>, preceded by the
>legend <quote type="inscription">GROUND FLOOR</hi>, is inscribed upon
>a door-post in Symond's Inn, Chancery Lane.
>One could imagine ways of expressing the relationship between the two
>parts of what is a single inscription or advertisement. But at that
>point I begin to wonder whether life is too short and whether the
>necessary overhead justifies the meagre payback.
>Are there best practices about encoding such mini-documents in fiction
>(they are quite common once you start looking for them). If you were
>in the business of encoding epigraphic documents you would necessarily
>pay attention to all the semantic and palaeographical minutiae. But
>when the context is sprawling nineteenth century fiction, should the
>appropriate maxim be "de minimis non curat lex"?
Surely the question of how to achieve an adequate representation in
the abstract is impossible to answer, short of a definition of
"adequate". And then one must ask, "adequate to what".
I am not saying life is too short. It is, but I am still happy to
indulge what time I have in questions like this one. But if one is to
make progress with the question, surely it helps for one to have some
notion in mind of what one is progressing toward.
The observation that there are texts latent in texts is both keen and
commonplace. Yet those texts, implicit or interpolated, are also
never fully there. (For example, Dickens' narrator does not tell us
if anything else is inscribed on the door post. I submit this is true
even when the text-in-the-text appears to be fully rendered, like
the epitaph that appears at the end of Sartor Resartus chap IV, or
any of multitudes of similar such cases.) Immediately this poses a
problem for the markup practitioner, since it is at least an implicit
claim of our markup (which, after all, we go to great pains to make
machine readable) that the markup is both comprehensive and complete
in its description, however we define the boundaries of what we seek
It may be that a markup practitioner chooses not to attempt such
comprehensiveness (loving angle brackets for their own sake?), or
does so only by narrowly circumscribing the "intent" (the presumed
functional requirements) of the markup to the immediate case, giving
this poem (or this speech or this letter) specialized "bespoke"
markup to reflect that it is, after all, special. But to the extent
we do this, we also give up on the pretense that our markup should be
generic, in the sense that it promises all those advantages of
application independence, interoperability and the rest. A world of
special cases is a noisy world, in which it can be difficult to tell
what is special about the special.
Accordingly, I suggest you think of your markup of such a case as a
rhetorical act, considering not only what do you want to say, but who
are you saying it to? There may be any of a number of machines
(presently or in the future) "listening" to see what they can make of
your markup. Many of them might not understand your language, at
least not without instruction.
Are you content for your markup to be noise to them? If not, then you
have to consider whether you prefer to say something in a language
they can understand, or stay silent, since no one is there to hear.
Or maybe you can build machines of your own to recognize these tags
and do things with them that further audiences might consider meaningful.
Certainly, I acknowledge that many would think it wise not to be
consumed with such trifles. On the other hand, depending on what you
intended to do -- and especially, what your stylesheets or
applications knew how to do -- I can imagine that systematic and
broad treatment of such inter- and subtextual phenomena in
19th-century fiction (and elsewhere) could be both fun and the basis
of a very intriguing study.
But codified best practice? I doubt it. My guess is that when you do
try to attempt this systematically, you will soon get taken up into
hard questions of modeling features and phenomena in your texts that
are overlapping, non-contiguous, and out of their "natural" order.
(Notice, for example, that the two pieces of the door post
inscription are given to us in reverse order.) Individual projects
here and there may have wrestled this bull to the ground, at least
long enough to tag its ear. But we don't yet have good generalized
tools to deal with this sort of thing. And until we have clearly
defined tasks for those tools to do, we're not likely to.
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