A Preliminary Report on the TEI Technical Review Meeting,
Eynsham Hall, Oxon., 14-16 May 1993
The Text Encoding Initiative held a Technical Review meeting last week
to review the current draft of the TEI Guidelines (TEI P2) and resolve
outstanding technical issues preparatory to the meeting of the TEI
Advisory Board, to be held in Chicago next month.
Like the two previous Technical Review meetings in Myrdal and Chicago,
this one brought together a broad cross-section of TEI participants and
resulted in a number of new ideas, refinements of old ideas, and general
Among the topics of general interest discussed at the meeting were
* Information about who is responsible for given aspects of the markup,
and the certainty which attaches to it, may be given informally using
the standard <NOTE> element; this meeting further developed a proposal
for a specialized <CERTAINTY> element (originally made at the Myrdal
meeting) to define such aspects more formally, including the ability to
express alternations and conditional confidence levels.
* The syntax used to declare and locate "canonical references" is to be
refined, and revised to exploit the extended pointer syntax defined in
chapter SA (Segmentation and Alignment)
* Specific elements are to be provided for simple segmentation of a text
into clauses, phrases, words, and morphemes
* The current <PARTICDESC> (participant description) element supports
demographic description of the participants in a language interaction;
its description (and possibly the names of its components) will be
revised to make more explicit the fact that it is intended also for
demographic description of the authors of written texts.
* Several refinements are to be made to the definition of TEI conformance
In some respects, however, the most important news from Eynsham Hall is
that most of the technical discussion concerned the resolution of
relatively localized questions with limited knock-on effects throughout
the rest of the guidelines. Many of the decisions resulted in
clarification of the proper usage of, or minor repairs to, existing
constructs, rather than their wholesale replacement by others.
We read this as an encouraging sign that the present draft of the
Guidelines is converging to a stable core, and reflects (we think) their
increasing maturity and robustness.
A revised version of the current state of TEI P2, incorporating the
simpler or more urgent results of this meeting, will be mailed to the
TEI advisory board about the beginning of June. We expect to publish a
further revision, incorporating the more time-consuming results of the
meeting, sometime this fall. This will be the first full edition
of the Guidelines, and will have the document number TEI P3.
-C. M. Sperberg-McQueen