Dear List,
I just wanted to point out that there is one place in the Guidelines
where this particular issue (of grouping and labeling list items with
brackets) is explicitely addressed: in the context of <CastList>. In
order to keep TEI encoding more consistent ("orthogonal"), maybe <list>
and <castlist> should try to learn from each other?
In section 10.1.4, "Cast Lists", the Guidelines recognize the existence
of bracketted lists for grouping and collectively naming characters of a
play; towards the end of the section, a coding example is given:
<castlist>
<castgroup>
<head rend=braced>Mendicants</head>
<castitem><role>Aafaa</> <actor>Femi Johnson</></>
<castitem><role>Blindman</> <actor>Femo Osofisan</></>
...
</castgroup>
<castitem>
...
I've used this structure to encode a cast list within Joyce's Ulysses
that in print looks like:
MEDICAL DICK }
and } (two birds with one stone)
MEDICAL DAVY }
MOTHER GROGAN (a watercarrier)
FRESH NELLY
and
ROSALIE (the coalquay whore)
(The last grouping is different from the first, because FRESH NELLY is
*not* connected with the role description "(the coalquay whore)".)
I find this solution (having a group of <item>s under one <head
rend=braced>) quite workable for cast lists; maybe it carries over to
the other lists of the world?
The case of only some items braced (as in the <castlist> above) would
call for an <itemgroup> allowed within <list>, and advocate the use of
<head rend=braced> instead of <label> in most of the posted examples.
Alternatively, one could treat braced groups within a list as
recursively nested sublists. Within the given TEI DTD, we'd have to see
such groups as nested sublists and stuff them into an <item> of the
outer level list.
The problem already posed by Lou's original riddles was the various uses
the brace can be put to. "My" solution applies well to things like
Lou's fruit list or Christina Powell's:
John Smith }
Squires
James Jones }
which becomes
<list>
<head rend=braced>Squires</head>
<item>John Smith</>
<item>James Jones</>
</list>
But it feels semantically off the mark in Lou's example of
The causes }
The extent } of human misery
The remedies }
This is, in my opinion, a typical problem of the informality of lists
and similar structures noted down in fluent prose; similar problems
appear in bibliographical pointers, like
"...as it is noted somewhere in the TEI Guidelines, section 6.7, as
far as I remember, but section 6.6 might also apply (the former can
be found on page 177, by the way)."
As much as I appreciate this human way of speaking, it is a nightmare
for structurally tight SGML encoding. In an abstract way, Lou's example
*is* a headed list, like:
Human Misery
============
- its causes
- its extent
- its remedies
If <head> seems to disturb the sequence of the original text too much,
maybe the following goes down smoother? :-)
<list>
<item>the causes</>
<item>the extent</>
<item>the remedies</>
<tail rend=braced>of human misery</>
</list>
But if I focus more on the look and feel of the prose than the
abstraction of the structure, what I see is a form of two-dimensional
prose. And I am really at a loss to make a nice, meaningful, and
TEI-compatible construct out of that.
((Is there a subcomittee for an auxiliary tagset 'TEI.crossword' ?))
((Has anyone found a meaningful encoding for e.e.cummings "Grasshopper" ?))
With excuses for verbosity --
Tobias
APPENDIX: (by e.e.cummings)
============================
r-p-o-p-h-e-s-s-a-g-r
who
a)s w(e loo)k
upnowgath
PPEGORHRASS
eingint(o-
aThe): l
eA
!p:
S a
(r
rIvInG .gRrEaPsPhOs)
to
rea(be)rran(com)gi(e)ngly
,grasshopper;
--
.............................................
(_) Tobias Rischer
"===' [log in to unmask]
" "
...still.loving.gnu..........................
|