> > Thats an odd assertion. Why do you think the TEI mandates "Make
> > explicit the implicit" ?

> Because I've read it (I think in the late eigthies) in the "Gentle
> Introduction".

Indeed, the sentence you are refering to appears in the "Gentle
Introduction to SGML" chapter of P2:

  Encoding a text for computer processing is in principle, like
  transcribing a manuscript from <foreign lang=LA>scriptio
  continua</foreign>, a process of making explicit what is
  conjectural or implicit, a process of directing the user as to how
  the content of the text should be interpreted.

(Note the lack of quotation marks around the attribute value -- P2
was written in SGML, not XML.) And has since become the following in
P4 (and has yet to be changed for P5, either):

  Encoding a text for computer processing is in principle, like
  transcribing a manuscript from <foreign lang="LA">scriptio
  continua</foreign>,<note place="foot">In the <q>continuous
  writing</q> characteristic of manuscripts from the early classical
  period, words are written continuously with no intervening spaces
  or punctuation.</note> a process of making explicit what is
  conjectural or implicit, a process of directing the user as to how
  the content of the text should be (or has been) interpreted.</p>

But I hardly think that defining markup in this way lays TEI open to
a charge of "mandating" anything. Even in those cases where TEI does
recommend particular features be explicitly marked up, that's all it
is -- a recommendation, a guideline, a suggestion, some advice.

And furthermore, either of the two encodings Peter describes (the
laborious manual encoding of some repetative detail, or the simple
specification of the detail once in the header) is making explicit
the same feature of the text; the difference being that the former is
more direct while the latter is more efficient.