Dot's example and suggestion raises an interesting question for the P5
Guidelines. Dot wrote:
> I'm working on a collection of carolingian laws, and we're using <div>
> for each law. But these laws are divided into books, and each book is
> its own <text>. I'd say it depends on how much of a collection your
> collection is. If there is an internal organization, better to use
> <div>; if it's more random, use <text>
The Guidelines are currently mute on <text> element below the level of
<group> or <body>. The Guidelines (Section 7.0) do provide examples of
composite and unitary texts [i.e. objects to be marked up]. They also
devote space to the possible uses of the <group> element (Section 7.3).
There is a brief mention of using text in the manner described by Dot.
"This element may also be used to represent embedded texts, as when one
narrative is contained by another."
Embedded texts are currently mentioned in the chapter on multiple
hierarchies and the chapter on infrastructure.
Would it be useful in some future version of the Guidelines to have an
index with such entries as "text, composite", "text, embedded" and "text,
unitary"? Presently surfing the Guidelines is convenient is one is
following an element name (e.g. <text>). Checking for concepts, less so.
Aside: <div> for each law and <text> for each book --- is there some
textual matter between <div> and <text> in this use case? or has the
schema/dtd been modified to allow for <text> to nest directly within