18.713 exam question

From: Humanist Discussion Group (by way of Willard McCarty willard.mccarty_at_kcl.ac.uk>
Date: Fri, 15 Apr 2005 08:11:43 +0100

               Humanist Discussion Group, Vol. 18, No. 713.
       Centre for Computing in the Humanities, King's College London
                     Submit to: humanist_at_princeton.edu

         Date: Fri, 15 Apr 2005 08:02:39 +0100
         From: Martin Holmes <mholmes_at_uvic.ca>
         Subject: Re: Fwd: 18.710 exam question

Hi there,

At 08:52 AM 14/04/2005, you wrote:
>>"Tagging is a subspecies of critical reading". Discuss.
>>This is precisely what I tried to show (not in an exam, but in a PhD
>>thesis), segmenting and tagging a (100.000 words long, Latin, and
>>religious) prose text with a set of stylistic values (TEI "seg" element
>>with "ana" fields); the set evolved during reading, and reflects how I
>>"learned the style" of the text.
>>The counter-arguments of my mentor were:
>>1) what if somebody else tagged the same text completely differently?
>>2) could the "level of detail" I decided on for a prose text (how deep
>>markup should be) be applied to poetry?

This is something we've argued about at length during workshops on markup
for faculty members. Markup constitutes a theory of the text. Even my
surface-level markup of the ACH/ALLC abstracts makes various claims that
are open to question -- for example, I've tagged the names of software
programs as title level="m", making them equivalent to other publications
at the monograph level. (I'm primarily a software developer...). Among
faculty members new to the concept of tagging, we frequently encounter the
belief that it is some kind of mechanical operation best done by "hired
help", and is somehow beneath them. Nothing could be further from the
truth, especially in the case of deep or detailed tagging. If you are, for
instance, tagging the metrical structure of a poem (perhaps dividing it
into feet, and classifying the feet according to some set of conventions --
iambic, trochaic, etc.), then even with the most metrically regular poetry,
there is vast room for disagreement over any given line. But that's the
point -- tagging forces us to be specific about our theoretical claims,
opening them up to discussion and rebuttal; this is the very heart of
humanities teaching and research. It often seems, though, that some fear
this explicitness, because exposing their claims in such a rigorous and (as
Neven suggests) falsifiable manner lays them open to more criticism than
would be the case with a discursive prose argument, where difficult areas
can be fudged and bets hedged.

>>Re 1, I believe that precisely the differences of two "critical readings",
>>when they are made much more explicit --- and, shall we say, falsifiable
>>--- by markup, or computer modelling, enable us to discuss what style,
>>and the perception of it, really is. Why did you mark _this_, and _here_,
>>when I marked _that_, and _there_?

Here's another little example emerging from the ACH/ALLC abstracts, which
will be a familiar problem for anyone using TEI regularly:

Many of the papers came in to us as word-processor documents. full of style
attributes like italics and quotes. For the sake of editorial consistency
across the document, we have tried to convert styles to descriptive tags,
in particular using the tags "term" (for technical terms, "soCalled" (for
scare quotes and similar types of authorial "distancing", "emph" (emphasis)
and "mentioned" (for words or phrases mentioned rather than used, as for
example in a linguistic context). Sometimes it can be quite difficult to
categorize specific instances, but usually it's fairly straightforward.
Now, we are currently producing PDF output with the "soCalled" tag rendered
with single quotes, and the others in italics (double quotes are reserved
for quotations or article titles). This means that some authors submitted
something in double quotes, but they see it styled in single quotes because
(according to me) it's a "soCalled". This "feels wrong" to them; they
"meant" double quotes, and single quotes is somehow different. If pushed,
they'll usually agree that their intention was an instance of "soCalled";
they'll also agree that it's OK for a house style to render such things in
single quotes; nevertheless, they feel somehow misrepresented. The tagging
forces us to make claims about authorial intention explicit, and (in many
cases) reveals new levels of complexity in the text, and interesting
discrepancies between an author's perceived intention and the reader's

Now back to the tagging grind...


Martin Holmes
University of Victoria Humanities Computing and Media Centre
Received on Fri Apr 15 2005 - 03:19:20 EDT

This archive was generated by hypermail 2.2.0 : Fri Apr 15 2005 - 03:19:21 EDT