3.374 old spellings and coding (138)

Mon, 21 Aug 89 20:23:48 EDT

Humanist Discussion Group, Vol. 3, No. 374. Monday, 21 Aug 1989.

(1) Date: Mon, 21 Aug 89 11:57:04 CDT (36 lines)
From: Steven J. DeRose <D106GFS@UTARLVM1>
Subject: Old spellings, clarification

(2) Date: Sat, 19 Aug 89 15:41:12 CST (83 lines)
From: "Robin C. Cover" <ZRCC1001@SMUVM1>

(1) --------------------------------------------------------------------
Date: Mon, 21 Aug 89 11:57:04 CDT
From: Steven J. DeRose <D106GFS@UTARLVM1>
Subject: Old spellings, clarification

>From Bob Kraft's response, I gather I didn't clearly express what I
meant in my last posting -- for which I plead the unoriginal excuse
of hurried composition.

I have no objection to encoding the details of particular manuscripts,
even down to the slant of each letter, the positions of coffee stains,
and whatever else. The point I was making was that for *most purposes*
this is not the desired information. Certainly anyone working in
papyrology, or working in other related areas, both wants and needs
that level of detail. But as I meant to make clear, that is a
question of audience -- "different encodings for different folks", to

I'll also take this opportunity to add a point which I meant to make in
the previous posting, but omitted: that having things both ways is
what hypertext is for. I should be able to view the "original" spellings
if I want to, or (as I have been reminded) even a detailed facsimile of
the original; yet I should be able to see a modernized or otherwise
"enhanced" version as well.

In a hypertext system, all the marginalia should be instantly available.
For example, a student should have glosses and alternate spellings
available, not just a pop-up inquiries but also as replacements in the
text (say, choosing old or new spellings via a menu option). Similarly,
a scholar should be able to view any variant manuscript at will, or
several side-by-side with differences highlighted.

In that world, all of the encodings we've been discussing are
available, and each user can choose his or her own view(s) from
the widest possible range of possibilities.

(2) --------------------------------------------------------------86----
Date: Sat, 19 Aug 89 15:41:12 CST
From: "Robin C. Cover" <ZRCC1001@SMUVM1>

The discussion on retaining or modernizing "old-spellings" seems to have
diverged in several directions since the original query. With this
contribution, it will wander yet further...

(1) Leslie Morgan rightly points out that most publication traditions do
not force an editor to make an absolute choice whether to retain the
archaism or to modernize it: marginal notes or footnotes may supply
alternate readings if the editor wishes to modernize orthography. With
hypertext, we may conceal at the immediate textual locus MANY such
readings, as well as annotations of various kinds.

(2) The "textual apparatus" mentioned by Bob Kraft moves the discussion in
a different direction. Readings in a textual apparatus traditionally
preserve witnesses that compete for antiquity, or authenticity or
otherwise contribute to a modern understanding of the ancient evolution of
the text. But usually the "readings" are designated by the editor as
belonging to antiquity, not to modern creativity or demands for
perspicuity/readability. Of course, modern editorial comments and
emendations are also introduced into the app crit, but
they are distinguished as such.

(3) To introduce the notion of "transcription" is yet another issue, for
it includes paleography and a host of more complex factors relative to the
original document. Cuneiform texts (tablets, but also lapidary
inscriptions and many other media) are customarily published in several
formats because each format has its own scholarly purpose: (a)
photographs, so that scholars may personally collate difficult readings,
examine the scribal ductus, and so forth; (b) hand copies using line
drawings and shading, where tablet features are simplified, but still
represented in 3-D perspective; (c) "transcriptions" having the barest
subjectivity: sign-by-sign algebraic representation of the graphic units,
establishing only character-boundaries and identity of wedge-cluster; (d)
normalization, involving various levels of interpretation, including
resolution of word-boundaries, logographic, determinative and other non-
phonological usages of the signs, etc. OK -- I won't inflict more of
this on HUMANISTS who may still be reading. The point is this: even if we
are interested in the *simplest interpretation* of an ancient text, dozens
of factors may be involved in its textual representation. Shall we call
all this "encoding?"

(4) Now suppose we move to "coffee stains" and other features of original
documents that are essential in the joining of textual fragments or of
text restoration (papyri, clay tablets, leather). To speak of "encoding"
is to press the term too far, it seems to me. Scholars make fragment
joins by noting the fingerprint of rare trace-elements as determined by
neutron bombardment of baked clay -- shall we record color and chemical
composition of stone, leather, parchment as part of the "textual
encoding?" All information about the physical medium of writing is
important, of course, but should this be called (textual) "encoding?"

(5) I finish with a instance which may find more sympathy among advocates
of SGML: when "encoding" an ancient text (ephemeral or official), shall we
account for ("encode") the two-dimensional position, stance, style and
size of characters on the written text? This is *not* just a variation on
the paleographic theme mentioned above (where paleography can assist in
textual reconstruction), but an issue of document structure and content.
A text's appearance on the original document *may* provide a clue to the
author's understanding (or perhaps's the scribe's) of the document
content. Are poetic lines shorter than prose lines? What does
indentation mean at various loci? Why are some characters in larger size,
or in alternative (archaizing) style? What do the different regions of
"white space" mean? Why regions of cursive script in the middle, or end,
of lapidary text?

The discussion of "old-spellings" is useful, I think, in helping isolate
the different roles that "markup" or "encoding" play in rendering various
kinds of documents. SGML in a modern structure-editor or other authoring
system makes sense because the author presumably knows why s/he wants to
create such-and-such a textual object, and can apply a name to the object.
Markup of recently-published texts is also relatively straight-forward,
since we usually understand editorial and authorial conventions of our
contemporaries. Markup ("encoding") of ancient texts presents numerous
problems, not just tactically, but because we are imposing interpretations
upon texts which may be inadequate or entirely incorrect.

Robin Cover

(no flames, please...."the Texas heat made me do it")