First I'd like to say a little more about the manuscript I'm encoding,
since I'm afraid I left out details of importance to the discussion in my
first email.
The manuscript is a single leaf with, on the recto side, cast list -
including requests for which actors Ibsen prefers - and setting for the
drama "The league of youth". The verso side contains notes for a planned
(but never completed) preface for "Emperor and Galilean" (we encode the
verso side as a separate XML file).
The cast list + setting probably is a draft for an attachment to a letter,
known but lost, from Ibsen to the theatre, concerning the performance of
the drama.

Then to Lou Burnards suggestion to use <front> for the cast list + setting
and put <gap/> in <body>. I agree with Paul Schaffner, it doesn't seem
alright using <gap/> here, not in the encoding of this particular
manuscript anyway. There has most likely never excisted anything more (of
the drama) but this cast list + setting in this particular manuscript, cf.
the description above. If there never was anymore than this in the
manuscript, <gap/> shouldn't, in my opinion, be used.

And on second thought, I was perhaps wrong to call it a fragment as well?
The text in the manuscript is probably copied (by Ibsen) from another
manuscript (containing the complete drama) to be used in the letter.
Depending on how one considers it, it could either be seen as a fragment
of the drama or as a draft for an attachment to the letter. As fragment of
a drama Lou Burnard's suggestion for encoding would make a good solution,
but viewed as a draft for something else, and in this case a draft for an
attachment to a letter, the text (i.e. cast list + setting) in the
manuscript surely could be seen as the body of the doument where the
doument = the attachment? Or, maybe even as the back matter for the
letter, since an attachment to a letter probably would be considered as
back matter.

I still think I'll continue to regard the contents of this manuscript as
the body of the document.

Bu putting the <castList> in <body> leaves the question of what to do with
the setting unanswered.
According to the P4 guidelines, one should use <stage>: "This element
should not be used outside the front matter; for similar contextual
descriptions within the body of the text, use the <stage> element" (cf.
"Note" in description of <set> in the "Elements" chapter)
According to the guidelines then, it would be possible to use <stage
type="setting"> here, but I find that unsatisfying as the text in question
is a setting - it belongs with the cast list and it "contains a
description of the setting, time, locale, appearance" (quoted from the
definition of <set> in the "Elements" chapter) - and it's not a "stage
direction within a dramatic text or fragment" (quoted from the definition
of <stage> in the "Elements" chapter). Rather than

 <stage type="setting"/>

I think I prefer


Or even allow <set> in <body>. Is there a reason why <castList> is allowed
in <body>, while <set> is not?

On Wed, January 31, 2007 00:18, Paul F. Schaffner wrote:
> A very common analagous situation arises when a *piece* of a work is
quoted in such a way as to reproduce its original structure, e.g. for
purposes of commentary or rebuttal or abuse. At least, I think it is
analagous, since unless someone can offer a better option, a quoted (say)
preface, or perhaps only the head or opener of the preface, often seems
best treated with <q><text><front>..., since only this really represents
its role, but the rest of the text is not there
> is
> perhaps only quoted in a later chapter). Since half the books in the
seventeenth century reproduce chunks of the other half in order to hold
them up to derision, this sort of thing is depressingly
> commonplace:
>    <p>Mr. Smith cannot seem to tell the truth even in his epistle of
>      <q><text><front><div1 type="dedication">
>      <head>A Humble Submission.</head>
>      <byline>By John Jones, Esq;</byline>
>      <opener><salute>To the Lord Bishops of England, greeting.</salute>
>      <p>I have ever loved you, my lords</p>
>      </div1>
>      </front>
>      </text>
>      </q>
>    since his name is not Jones, he is anything but humble, and "love" is
a strange word for his implacable enmity, as of darkness for the sun. And
then he heaps deceit upon mendacity by signing himself,
>      <q><text><front><div1 type="dedication">
>      <closer>
>      <dateline>Brasenose College, Oxon., <date>31 June 1645 novo
>      </dateline>
>      <signed>Your ever loyal servant, <name>J. Jones</name></signed>
>      </div></front></text></q>
>    since everyone knows he is a Cambridge man! (and June has only thirty
> We allow some things that canonical TEI doesn't, but even so, in such
circumstances, I have found myself driven to 'fake' things that aren't
there, e.g. an empty <p> to follow a quoted <head>, or an empty
> <body> to follow a fragmentary <front>. In this case, <gap> doesn't seem
quite right. Or does it? What do you do?
> pfs
> --------------------------------------------------------------------
Paul Schaffner | [log in to unmask] | 316-C
Hatcher Library N, Univ. of Michigan, Ann Arbor MI 48109-1205
> On Tue, 30 Jan 2007, Lou Burnard wrote:
>> Seems to me that the only pure and honest way of dealing with this
>> is to do something like
>> <front>fragmentary front elements</front>
>> <body><gap><desc>Only the front matter survives</desc></gap>
>> </body>
>> If part of what you want to encode about a fragmentary title page is
that it
>> is in some sense "front matter" then putting it into the <body> simply
>> on. As Michael SpMcQ used to say "if there is a choice between an
>> lie and a difficult truth, always prefer truth".  If it's "front
>> then there has to be something for it to be in front of -- and in a
fragmentary manuscript, what it's in front of is a gap!
>> Lou
>> On Tue, 30 Jan 2007, Hilde Bøe wrote:
>>> Thanks Brett,
>>> this is clearly similar to our problem.
>>> I think it would have helped (I would certainly have appreciated it)
>>> e.g. either the chapter on "Default Text Structure" or the chapter on
"Transcription of Primary Sources", could include a section on
>>> fragments, where the discussion included examples on fragments
>>> of only front and back matter. At present the TEI Guidelines does not
allow documents consisting of only <front> or <back> (and that's
>>> with me), but moving front and back matter into the <body> isn't an
>>> solution, as far as I can see.
>>> All the best, Hilde
>>> On Tue, January 30, 2007 18:54, Brett Barney wrote:
>>>> Hilde,
>>>> On the Whitman project, we encountered a similar problem a couple of
>>>> back: We were encoding a manuscript title page with a list of
>>>> titles. Unlike <castList>, <titlePage> can occur only within <back>
>>>> <front>, though. We ended up going down the first of the paths you
>>>> encoding the variant titles within <body> (in a <div> with a type
attribute). Even though this was a few years back, I'm still not
>>>> at ease with the issue: the first choice seems intuitively right from
>>>> a commonsense and a TEI-theory purist perspective, but the second
>>>> has a pretty strong pragmatic appeal--it's easier to work with
>>>> components across a collection when they are marked up the same.
Unfortunately, none of this does much to enlighten your particular
conundrum, I'm afraid. But maybe it's useful as a related problem of
manuscript encoding?
>>>> Best regards,
>>>> Brett Barney
>>>> --------------------
>>>> Walt Whitman Archive
>>>> (402) 472-4547
>>>> Hilde Bøe <[log in to unmask]>
>>>> Sent by: "TEI (Text Encoding Initiative) public discussion list"
<[log in to unmask]>
>>>> 01/30/2007 09:11 AM
>>>> Please respond to
>>>> [log in to unmask]
>>>> To
>>>> [log in to unmask]
>>>> cc
>>>> Subject
>>>> <castList> and <set> in <front> and <body>
>>>> In the TEI P4 guidelines ("Default Text Structure") the overall
>>>> of a unitary text is defined as including front, body and back. We've
had a discussion on wether a document can consist of <front> only
>>>> (although the encoding isn't valid unless one adds <body> inside
>>>> The background for the discussion is a manuscript containing only a
>>>> list.
>>>> Some of us argue that the cast list ought to be encoded in <body>
>>>> it is the body of the text it represents:
>>>> <body>
>>>>  <castList>
>>>>   <castItem/>
>>>>   <castItem/>
>>>>   <castItem/>
>>>>   <castItem/>
>>>>  </castList>
>>>> </body>
>>>> Others argue that since the cast list is usually interpreted as front
matter, it should always be encoded within <front>. Encoding it
>>>> <front> makes it necessary to add <body>/<div>/<p> to get valid
>>>> <front>
>>>>  <castList>
>>>>   <castItem/>
>>>>   <castItem/>
>>>>   <castItem/>
>>>>   <castItem/>
>>>>  </castList>
>>>> </front>
>>>> <body>
>>>>  <div>
>>>>   <p/>
>>>>  </div>
>>>> </body>
>>>> We would like to hear from others on the list who might have had
>>>> problems. What have you done?
>>>> Henrik Ibsen's cast lists nearly always include settings, and so does
>>>> one in the manuscript mentioned above. Alongside the choice between
encoding <castList> within <front> or <body>, we have encountered a
problem concerning the setting. Encoding a cast list in <front> lets the
>>>> encoder use <set> for the setting (see 10.1.1 The Set Element in 10
>>>> Tag Set for Drama):
>>>> <front>
>>>>  <castList>
>>>>   <castItem/>
>>>>  </castList>
>>>>  <set>
>>>>   <p/>
>>>>  </set>
>>>> </front>
>>>> <castList> can go nearly everywhere: <front>, <body>, all kinds of
>>>> <back> and so on. <set> on the other side is only allowed in <front>
>>>> <back>. Thus we can't use <set> when encoding the <castList> in
>>>> <performance> is defined as "a section of front or back matter
>>>> how a dramatic piece is to be performed in general", and could thus
include the setting, had it not been that <performance> also is only
allowed in <front> and <back>.
>>>> We're therefore, it seems, left with the choice between <stage> and
>>>> but neither of them seem to be more than an emergency solution since
>>>> can't use <set> (preferably) or <performance>.
>>>> Our question is thus: what's the reason for this difference? If you
>>>> have a cast list in the body of a text (as some of us believe we have
>>>> one of Ibsen's manuscripts), why can't you have a setting as well?
>>>> we
>>>> forced to encode the setting in <back>?
>>>> <body>
>>>>  <castList>
>>>>   <castItem/>
>>>>   <castItem/>
>>>>   <castItem/>
>>>>   <castItem/>
>>>>  </castList>
>>>> </body>
>>>> <back>
>>>>  <set>
>>>>   <p/>
>>>>  </set>
>>>> </back>
>>>> We are using P4 and are therefore not familiar with P5. Is the
encoding of
>>>> cast lists and settings changed in P5?
>>>> On behalf of Henrik Ibsen's Writings, Hilde Bøe
>>>> --
>>>> Hilde Bøe
>>>>      Ass. Editor
>>>> [log in to unmask]
>>>> Henrik Ibsens skrifter, Postboks 1166 Blindern, 0316 OSLO
>>>> Tlf. +47 228 591 52/fax +47 228 59 169
>>>> URL: