9.338 text-encoding

Humanist (mccarty@phoenix.Princeton.EDU)
Fri, 1 Dec 1995 00:33:21 -0500 (EST)

Humanist Discussion Group, Vol. 9, No. 338.
Center for Electronic Texts in the Humanities (Princeton/Rutgers)

[1] From: Patrick Durusau <pdurusau@emoryu1.cc.emory.edu> (38)
Subject: Re: 9.332 text-encoding

Ian Lancashire's post on text-encoding with SGML appears to confuse the
use of SGML with the adoption of specific editorial principles.

> SGMLers remind me of missionaries passionately trying to convert
> heathens, in this case scholars whose editorial practices don't follow
> ISO standards. What those on a mission may forget is the specific
> purposes that these scholars have in mind in making the e-texts.

Editorial practices are not dictated by any ISO standard in general or
SGML in particular. SGML provides a mechanism for documenting both the
editorial practices used and their application to particular texts.

SGML is not synonymous with various specific guidelines for the use of
SGML such as CALS or TEI. It there is some specific editorial practice
that cannot be used with the TEI guidelines for example, perhaps it will
be posted to this list.

> These purposes -- not any promise to conform to SGML -- get them awarded
> research grants. These purposes also drive editorial principles.

Conformance to SGML such as represented by the TEI Guidelines will make
encoded materials useful for more researchers over a longer period of
time and therefore represents better use of scarce funding resources. One
would hope that the use of SGML will become a condition of funding for
text encoding projects in the near future.

> SGML by its very nature imposes
> interpretations on texts. If I am going to impose an interpretation on
> French character sets, views of people like Russon Wooldridge will
> always have to weigh heavily ...

I am not quite sure how anyone could reach the conclusion that SGML
"imposes interpretations on texts." A particular encoding of a text
could certainly inpose an interpretation, but that is the responsibility
of the editor and not the method used for the encoding.

I would be very interested in discussing specific examples of problems
with using SGML, particularly under the TEI Guidelines, on or off list.
I am less interested in discussions concerning the preservation of present
practices for reasons unrelated to the voiced goals of scholarship.

Patrick Durusau

Patrick Durusau
Information Technology
Scholars Press