> Seems odd to me that we have
> elements defined for things like epigraphs and notes, not to mention
> paragraphs, but don't have elements for the more "gross" structure of a
> book. There aren't that many, and they're well defined in such works as
> the Chicago Manual of Style.
This is a design decision that's been with us since the days of P1. The
rationale for it is simple: although the gross structure of some kinds
of work may be well defined by such manuals as the one you cite, not
everyone follows the rules. Chapters are not particularly problematic,
but in some works groups of chapters are called "sections", and in
others they are called "books"; in yet other works, "sections" are
subdivisions of "chapters", and "chapters" are grouped into "parts"...
and so on. And that's just talking about ordinary print books! When you
start thinking of other kinds of written material such as newspapers,
manuscripts, pamphlets.... and then there is the historical dimension too.
The decision was taken therefore to represent the basic principle that
texts are hierarchically structured, but to regard the labelling which
may be attached to a particular hierarchic level as something unsuitable
for standardization. Over the last decade I think the wisdom of this
decision has been born out by the way that people have managed usefully
to apply the TEI to a very wide range of material.
As Sebastian points out, of course, if you wish to provide "syntactic
sugar" for your div types, then you should. For the reasons given, this
would simplify the task of data entry -- though it complicates slightly
the job of data interchange.
I say nothing (this time) on why we have both numbered and vanilla divs!
That topic has already been addressed quite extensively on this list and
I lost the argument...