#14 <mentioned>


Please add an optional "type" attribute to <mentioned>,
specifying the ontological or linguistic type of the
mentioned entity.

Sample value could include:

* term (typically rendered in quotes)
* symbol (typically rendered as is or in quotes, too)
* graphs (typically rendered in italics)
* graphemes (typically rendered as "<...>")
* phones (typically rendered as "[...]")
* phonemes (typically rendered as "/.../")


  • Lou Burnard

    Lou Burnard - 2004-08-22

    Logged In: YES

    If you simply want this facility in order to provide different
    rendering styles, then the rend attribute should be used.

    The purpose of <mentioned> is to distinguish cases
    of "mention" from use. If you additionally want to say
    something about the ontological status of the mentioned
    entity, the way to do it is by embedding a more ontologically-
    specific tag (e.g. <term>, <ident>) within the <mentioned>
    tag, I think. (And <ident> *does* have a type attribute)

  • Andreas Nolda

    Andreas Nolda - 2004-08-22

    Logged In: YES

    The value of the proposed "type" attribute should *not*
    specify the rendition. (The examples in my proposal above
    only gave *typical* renditions.) In fact, the rendition of,
    say, a <mentioned type="graphs"> element can vary
    according to style or context (e.g. block vs. inline).

    Of course, you can always specify the ontological type of
    the mentioned entity by some subelement. Personally, I'd
    prefer <seg> to <ident> because <seg> also allows for
    non-PCDATA content.

    However, I'd like to argue that an (optional) type distinction
    on <mentioned> makes sense. A mention always involves
    mentioning something; and mentioning a word, say, qua
    graphic entity or qua graphematic entity results in different
    mentions. Distinct rendition conventions only reflect these
    type distinctions.

  • Lou Burnard

    Lou Burnard - 2004-09-01
    • labels: --> TEI: New or Changed Element
  • Syd Bauman

    Syd Bauman - 2004-09-19

    Logged In: YES

    Maybe I'm being dense, but when is a grapheme or a phoneme
    or a phone ever actually used as opposed to mentioned? That
    is, isn't some element (<seg type="phoneme"> jumps to mind
    for interchange, although one might prefer to have a
    <phoneme> element for local usage) that indicates its
    content is a phoneme sufficient without being inside a
    Perhaps some examples will help straighten me out.

  • Lou Burnard

    Lou Burnard - 2004-09-22
    • status: open --> closed
  • Andreas Nolda

    Andreas Nolda - 2004-09-27

    Logged In: YES

    So far three different approaches have been suggested on
    this page:

    1. <mentioned type="graphemes">Karl</mentioned>
    2. <mentioned><seg type="graphemes">Karl</seg></mentioned>
    3. <seg type="graphemes">Karl</seg> (or, "for local usage"

    In my view, (3) does not express that "Karl" is mentioned as
    opposed to being used. <seg> just attaches some type to its
    content and is neutral with respect to the mention/use
    distinction. (Cf. a corpus encoder tagging some lemma by
    <seg> with part of speech-information: he does not imply
    that the lemma is mentioned, instead of being used, in its

    As to (2), that method is more verbose than (1). What is
    more, (2) is also less restrictive than (1). As far as I can
    see, a string is always mentioned *either* as a sequence of
    graphemes *or* as a sequence of phonemes *or* as whatever;
    there are no 'mixed' real-world examples like:

    <seg type="phonemes">Karl</seg>
    <seg type="graphemes">kommt</seg>

    In sum, I'd opt for method (1).


Get latest updates about Open Source Projects, Conferences and News.

Sign up for the SourceForge newsletter:

No, thanks