In the absence of a global "resp" attribute, I was wondering if Ingo's
example of disambiguating or making explicit the speaker implied in the
stage direction could be marked up using some of the attributes available
when working with the simple analysis tag set.

<speaker ana="encoder_supplied">Name of the speaker as read by
<stage>Stage direction as present in the witness</stage>
the speech...
<note id="encoder_supplied">Disambiguated reference...</note>

The "ana" attribute links the element <speaker> to an element <note>
through an IDREF/ID mechanism.

The "corresp" attribute can also be used to link for example <speaker> and
<stage>. Both "corresp" and "ana" attributes can take multiple values.

<speaker id="abc1" corresp="abc2 encoder_supplied">...</speaker>
<stage id="abc2" corresp="abc1">...</stage>

An XSLT stylesheet can then be applied.

Francois Lachance, Scholar-at-large
per Interactivity ad Virtuality via Textuality