Suggestions for the forthcoming/ongoing (?) TEI revision
- Hilde Bøe, Ellen Nessheim and Stine Brenna Taugbøl for Henrik Ibsen's
Writings, 04.12.2001.
Our encoding principal, Tone Merete Bruvik, has encouraged us to post this
overview of changes and additions that Henrik Ibsen's Writings has made in
its dtds, on the TEI discussion list as suggestions for revision of the TEI
Guidelines.
1 Suggestions for corrections
1.1 <app>, <lem> and <rdg>, <add> and <del>
At HIW we have discussed how to encode complex manuscript substitutions,
additions and deletions on a number of occasions, and we have decided that
our main task in the encoding is to give a photographic impression of the
changes rather than try to reconstruct the order of the changes, e.g. to
let the encoding show the genetic order of the substitutions, additions and
deletions.
Our need for the first change arises from our basic interpretation of how
to encode manuscript corrections. When i.e. an item in the cast list has
been deleted from the list, we want to record the whole element as deleted,
not just the content of the element, cf. simple examples of encoding below.
We interpret the deletion as covering not just the text of the item, but
the item as a whole.
<castItem><del>CATILINA</del></castItem>
<del><castItem>CATILINA</castItem></del>
This interpretation principle has made it necessary to allow <lem>, <rdg>,
<add> and <del> to contain complete structural elements (i.e. <castItem>,
<sp> etc.), cf. element definition below.
<!ELEMENT del - -
(#PCDATA | abbr | address | date | dateRange | expan | measure
| name | num | rs | time | timeRange | add | app | corr
| damage | del | orig | reg | restore | sic | space | supplied
| unclear | distinct | emph | foreign | gloss | hi | mentioned
| soCalled | term | title | ptr | ref | xptr | xref | caesura
| seg | formula | fw | handShift | bibl | biblFull | biblStruct
| cit | q | quote | label | list | listBibl | note | witDetail
| camera | caption | move | sound | tech | view | castList
| figure | stage | table | text | ab | l | lg | p | sp |
witList | appSpan | clarification
| lgSpan | shy | witEnd | witStart |
alt | altGrp | index | join | joinGrp | link | linkGrp |
timeline | cb | lb | milestone | pb | addSpan | delSpan
| gap | lacunaEnd | lacunaStart | anchor
| epilogue | performance | prologue | set |
divGen | titlePage | byline | docAuthor | docDate | docEdition
| docImprint | docTitle | epigraph | imprimatur
| titlePart | head | div1 | div2 | spOpener | speaker
| roleDesc | role | actor | castItem | stageRole)* >
<!ATTLIST del
id ID #IMPLIED
lang IDREF #IMPLIED
n CDATA #IMPLIED
cert CDATA #IMPLIED
resp IDREF #IMPLIED
hand IDREF #IMPLIED
status CDATA "unremarkable"
rend CDATA #IMPLIED
type CDATA #IMPLIED
TEIform CDATA "del" >
Because of this change it is thus only when corrections cross structural
boundaries that we use the span elements. When corrections, deletions or
additions span several whole elements we prefer <app> with <lem> and <rdg>,
or <add> or <del>. These elements make the processing of the encoded texts
easier, and because of this we try to avoid span elements when possible.
This change has made it necessary to include <app>, <add> and <del>
globally. We also thought this to be a good solution because in
transcribing the Ibsen material we have found that substitutions, additions
and deletions appear nearly everywhere in the text.
1.2 Global inclusion
We suggest that these changes are considered for inclusion in the TEI
Guidelines.
1.2.1 <fw>
We use the <fw> element to record pagination, foliation of leafs, sheet
signatures, column numbers, headers and footers. These appear everywhere,
and we have found it useful to include this element globally.
Our guidelines on the use of this element are as follows:
"<fw> should where possible be put outside structural elements
(acts/scenes/speeches/stage directions) [...] <fw> should be placed inside
elements that span page breaks etc., eg. within speeches beginning on one
page and continuing on the next." (cited from HIW's guidelines on text
encoding)
1.2.2 <figure>
We use the <figure> element to encode illustrations, logos, separating
lines and ornamental bars. These also tend to appear everywhere in the
text, and we have thus included the element globally.
Our guidelines on the encoding of separating lines and ornamental bars are
as follows:
"We consider separating lines and ornamental bars as bars that conclude
preceding textual elements. On title pages <figure> is placed between the
other elements. The same goes for headings; <figure> is placed between or
after headings. When occuring at the end of an act or a scene, <figure> is
placed within the element it concludes. A separating line at the end of the
last act of a play that concludes the play as a whole, should be placed in
<body> outside the <div1> element. Separating lines appearing in connection
with songs and poems in plays, should be placed outside the <lg> elements."
(cited from HIW's guidelines on text encoding)
1.2.3 <note>
It would be useful if it was possible to use <note> everywhere. At HIW
notes are used i.e. to record information between the encoders and thus we
need to be able to use them everywhere. We have therefore added <note> to
our global inclusion in the extension files and we suggest that this change
is included in the TEI Guidelines.
--------------------------------------------------------
2 Suggestions for additions
We suggest that these elements are considered for addition to the TEI
Guidelines.
2.1 New element: <printer>
HIW has constructed a new element, <printer>, for the recording of
information on printers for use in <docImprint> in <titlePage> and
<imprint> in <bibl>. We find this element useful for information presently
not covered by any of the excisting TEI elements.
Here is the element definition:
<!ELEMENT printer - -
(#PCDATA | app | hi | corr)* >
<!ATTLIST printer
id ID #IMPLIED
lang IDREF #IMPLIED
rend CDATA #IMPLIED
n CDATA #IMPLIED >
Here is an example of the use of the element in <imprint>:
<bibl id="Br66a"><monogr>
<author>Henrik Ibsen</author>
<title type="main">Brand</title>
<title type="desc">Et dramatisk digt</title>
<edition>2. opplag</edition>
<imprint>
<pubPlace>Kjøbenhavn</pubPlace>
<publisher>Den Gyldendalske Boghandel (F. Hegel)</publisher>
<printer>J. H. Schultz</printer>
<date>1866</date>
</imprint>
</monogr></bibl>
2.2 <omit>
In the TEI Guidelines there are a few remarks about an element called
<omit> (cf. remarks on <del> and <supplied> in chapter 35 "Elements"). This
element is not implemented in the Guidelines, but we think it should be.
In the Ibsen material we have some text witnesses that were printed as
serials. In these we sometimes have to omit text or illustrations that are
not part of the text we are transcribing, but that for some reason appear
in the middle of our text. For these instances we have used a gap element;
<gap reason="irrelevant"/>, but we feel that an <omit> element would have
been a better solution. This should then be defined as one of the tags for
editorial changes along with <corr> etc., and it should at least include
the reason and resp attributes.
--------------------------------------------------------
3 Suggestions for changes and additions for chapter 18 "Transcription of
Primary Sources"
If there are plans for revising chapter 18 "Transcription of Primary
Sources", and we hope that there are, we have a couple of proposals for
additions and changes now, and we would like to contribute with further
suggestions if given the opportunity on a later occasion.
3.1 New element: <clarification hand="" place="">
We have constructed an element for a manuscript phenomenon we call
clarification, that is, when (parts of) words or letters are overwritten
for the sake of legibility. We record the hand in a hand attribute if this
is another than the original hand. We also record where the clarifying is
done with a place attribute, because we have seen that Ibsen on some (rare)
occasions clarify a word by repeating it somewhere offline (in the margins,
above the line etc.).
Here is the element definition:
<!ELEMENT clarification - -
(#PCDATA | abbr | address | date | dateRange | expan | measure
| name | num | rs | time | timeRange | add | app | corr
| damage | del | orig | reg | restore | sic | space | supplied
| unclear | distinct | emph | foreign | gloss | hi | mentioned
| soCalled | term | title | ptr | ref | xptr | xref | caesura
| seg | formula | fw | handShift | bibl | biblFull | biblStruct
| cit | q | quote | label | list | listBibl | note | witDetail
| camera | caption | move | sound | tech | view | castList
| figure | stage | table | text | appSpan |
clarification | lgSpan | shy | spOpener
| witEnd | witStart | alt | altGrp | index | join
| joinGrp | link | linkGrp | timeline | cb | lb | milestone
| pb | addSpan | delSpan | gap | lacunaEnd | lacunaStart
| anchor)* >
<!ATTLIST clarification
copyOf IDREF #IMPLIED
corresp IDREFS #IMPLIED
sameAs IDREF #IMPLIED
select IDREFS #IMPLIED
synch IDREFS #IMPLIED
next IDREF #IMPLIED
exclude IDREFS #IMPLIED
prev IDREF #IMPLIED
id ID #IMPLIED
lang IDREF #IMPLIED
rend CDATA #IMPLIED
n CDATA #IMPLIED
hand IDREF #IMPLIED
place CDATA #IMPLIED>
Here is an example of the use of <clarification>:
<clarification hand="h2" place="inline">illegible</clarification>
We believe that clarification is a rather common phenomenon in manuscripts,
and, since it is not covered by the existing TEI Guidelines, that it should
be included. Others must have handled this problem earlier, and we are
curious of how they have dealt with this phenomenon.
3.2 The hand attribute
We have extended the elements <fw>, <hi> og <emph> with the hand attribute
for use in the transcription of manuscripts because we have learned that
the phenomena recorded with these elements quite often are added by other
writers than Ibsen or that they are added with other pens than the first
pen (and for us this qualifies as another hand). Others working with the
transcription of manuscripts must have encountered the same as us, and we
believe that adding a hand attribute to these elements, would be of
interest to others working with transcription of manuscripts.
--------------------------------------------------------
4. Suggestions for additions for chapter 9 "Base Tag Set for Verse"
We add notation of verse structures on all of Ibsen's verse texts, e.g.
verse dramas and poems. Main structure, substructures, variations and
deviations are described for both rhyme and metrical pattern. In doing this
we have had to expand TEI's tag set for verse.
4.1 New attributes for <l> and <lg>
4.1.1 Anacruses: <lg an="">
The pattern for anacruses (unstressed syllables before the first stressed
syllable in a verse line) may be varying even though the remaining metrical
pattern is regular. And the presence of anacruses infer on the counting of
syllables in the verse lines. Therefore we have found it reasonable to give
information about anacruses in a separate attribute, called an.
4.1.2 Deviating anacruses, metrics and rhymes: <lg realAn="" realMet=""
realRhyme="">
In the TEI Guidelines there is a real attribute intended for notation of
actual realization or deviation from the main formula described in the met
attribute. There is no such possibility for the rhyme attribute, although
this is often necessary. Therefore we have given each of the verse
attributes such possibilities. We have renamed the real attribute realMet
and supplied with the similar realRhyme and realAn.
Here is the element definition of <lg> (the definition of <l> has the same
new attributes as <lg>):
<!ELEMENT lg - -
((argument | byline | dateline | docAuthor | docDate | epigraph
| head | opener | salute | signed)*, (lb | fw | pb | l | lg | stage
| appSpan | addSpan | delSpan | anchor | app | add | del
| clarification | note | figure)+, (byline | closer | dateline
| epigraph | salute | signed | trailer)*) >
<!ATTLIST lg
copyOf IDREF #IMPLIED
corresp IDREFS #IMPLIED
sameAs IDREF #IMPLIED
select IDREFS #IMPLIED
synch IDREFS #IMPLIED
next IDREF #IMPLIED
exclude IDREFS #IMPLIED
prev IDREF #IMPLIED
id ID #IMPLIED
lang IDREF #IMPLIED
rend CDATA #IMPLIED
n CDATA #IMPLIED
sample (initial | medial | final | unknown | complete) "complete"
part (Y | N | I | M | F) "N"
type CDATA #IMPLIED
org (composite | uniform) "uniform"
real CDATA #IMPLIED
met CDATA #IMPLIED
rhyme CDATA #IMPLIED
realMet CDATA #IMPLIED
realRhyme CDATA #IMPLIED
an CDATA #IMPLIED
realAn CDATA #IMPLIED
TEIform CDATA "lg" >
4.1.3 Example
Here is an example of the new attributes in use:
<lg1 type="stanzaic Heinestrofe" an="single" met="bi+ free (7 6)2"
rhyme="(A b)2">
<lg2 type="stanza">
<l met="v-v-v-v" rhyme="A">Nu er det Dagens Trolde,</l>
<l met="v-v-v-" rhyme="b">Nu er det Livets Larm,</l>
<l met="v-v-vv-v" rhyme="A">Som drysser alle de kolde</l>
<l met="-v-v-" realAn="N" rhyme="b">Rædsler i min Barm.</l>
</lg2>...</lg1>
As shown in the example we have also made a system for verse notation of
(germanic) verse. It would be very useful to have more specific guidelines
of the detailed verse notation (e.g. the attribute values) in the TEI
Guidelines. But this is perhaps an issue for TEI P5?
4.2 New elements: <tune>, <composer>
Beneath the title of poems a line with information of the melody for the
text is often printed/written. There is no appropriate element for such
information in the TEI Guidelines. We are planning to add a new element
<tune> for this information. <tune> may appear inside <opener> or <closer>
and contain the elements <title> and <composer> (yes, we want to add the
last element too).
--------------------------------------------------------
Hilde Bøe
--------------------------------------------------------
Research Assistant, Cand.philol.
Henrik Ibsen's Writings, P.O.Box 1166 Blindern, N-0316 OSLO
Phone +47 2285 9152 - Fax +47 2285 9169
URL: http://ibsentexts.hit.uib.no/
Email: [log in to unmask]
|