[tei-council] Proposal for choice, model.pPartEdit, and model.choicePart

Daniel O'Donnell daniel.odonnell at uleth.ca
Mon Jul 23 15:03:08 EDT 2007


I think we need to be a little careful with Matthew's proposals: I've
long wanted some kind of emend element (i.e. something similar to edInt,
but I'm not convinced that edInt is actually different from a relatively
clear existing distinction in some of the use cases proposed:


> 1. Scribal and editorial emendations
> The elements <add> and <del> should be retained and used as they are
> currently, i.e for additions and deletions physically present in the
> manuscript, whether the actions of the scribe at the time of writing
> or of a later hand. A <subst> element should be added for indicating
> cases where <add> and <del> constitute a substitution:
> 
> <p>my brain <subst hand="#scribeA"><del
> rend="overstrike">burns</del><add
> place="margin">hurts</add></subst>.</p> 

I like this, but I think we should see it for what it is: a
specialisation of choice (in this case, choice at type="substitution").
I'll deal with the claim that it is not choice below. The reason I like
this as a proposal is that semantically substitution is a recognised
example of (let's call it authorial for the moment) alterity in the
proposal. In terms of my proposal, it helps address the issue of the
janus-y pair del : add, since it provides a specific "authorial" axis
for a specialised form of choice to work along. This may also help
clarify restore, which can also be seen as a semantically specific form
of choice--albeit one that for reasons of parsimony we don't have the
"other" state explicitly mentioned.

So yup. I'd add "subst" to model.pPartEdit alongside choice and restore,
and model.choicePart. I wouldn't change model.choicePart, but you could
either just add add and del to the content model of "subst" or we could
create a class model.pPartSubstitution parallel to model.pPartRestore.

> 
> The <supplied> element should be retained for text now illegible or
> lost through damage but assumed originally to have been in the
> manuscript (what used to be supplied reason="illegible|damage"), and
> for the expansion of abbreviations -- these are both arguably things
> which are present in the manuscript, but where a degree of
> interpretation is required on the part of the editor.

I like this. It is a little bit more narrow than the current use of
substitution (which can also be used with gap, where the point is that
the text might NOT have been in the original), but the new semantics are
easy to explain and very clear ("This element is used to supply text
that arguably once existed in the physical witness"). 

It does raise the question (which also came up in Berlin), if we should
not consider whether gap is doing to much, when it represents both
physically missing material (i.e. due to a hole in the MS) and
intellectually missing material, due to editorial sampling decisions,
for example, or saute meme a meme. We could end up in a situation where
the constraint on whether supplied is used with gap is purely semantic:
if you are using gap because the text that was there is now physically
missing, use supplied to restore the text; if you are using gap to
describe text missing due to an error or a sampling decision, do not
restore with supplied.

How about if we split these two functions: <missing> for material that
once was in the text but is now lost (for which <supplied> would be an
appropriate indicated of the alternate state) and <gap> for material
that was never present in the physical witness or has been omitted from
the transcription for sampling reasons. In this case, "supplied" would
be inappropriate for indicating the alternate state and "edInt" should
be used instead.

> 
> Add an <edInt> element, for "editorial intervention", where an editor,
> believing the manuscript to be incorrect or corrupt, supplies,
> suppresses or otherwise alters the text (note the specific means of
> indicating each of these three cases in traditional printed
> editions). 

I like this (though not the name) if we consider this as basically an
"editorial" parallel to "supplied." What we are talking about here is
basically emendation. Rather than edInt what if we called it emend or
edit?

Now come the problems:
> In the first case, the editor supplies text assumed to have been
> inadvertently omitted by the scribe:
> 
> <p>my brain hur<edInt type="supply" reason="omitted" resp="#MJD"
> cert="high">t</edInt>s.</p> 

Sure. But since we also think it is an error (hurs is clearly understood
here as sic for hurts), what is wrong with <sic> and <corr>?

> 
> In the second, the editor suppresses superfluous text in the
> manuscript, for example arising through a dittography:
> 
> <p>my brain hur<edInt type="suppress" reason="dittography" resp="#MJD"
> cert="high">ur</edInt>ts.</p> 

I suppose: but again, what about sic and corr? Again, hururts is also a
misspelling of hurts.

> 
> In the third the editor corrects a misspelt word, substitutes one word
> for another, or rearranges the order of the words in the text of the
> manuscript. In this case the existing <sic> and <corr> should be used:
> 
> <p>my <edInt type="emmend" resp="#MJD"
> cert="high"><sic>brian</sic><corr>brain</corr></edInt> hurts.</p> 

This I really disagree with. First of all, apparent we are privileging
dittography and simple omission of letters over alteration in order. In
my view all three are spelling mistakes, something we already have an
editorial tag set for describing, sic and corr. Secondly, this specific
use of edInt is quite different from the previous two: here edInt is
being used as a special kind of choice, parallel to subst (pace to
below); in the other two it is a special kind/alternate editorial form
of supplied. I think this distinction needs to be cleaned up.
> 
> The resp and cert attributes may be used to indicate the person
> responsible for the addition, suppression or emendation and the degree
> of certainty. Where the reading of another witness supports the
> reconstruction the source attribute may also be used, providing for
> example the sigil of the other witness.
> 
> Note: I don't think this makes <subst> or <edInt> sub-types of
> <choice>, since one would frequently want both the added and deleted
> or original and emended forms to appear, whereas <choice> implies that
> only one of the two options can be used at any one time.) 

Only the name "choice" suggests this. Both actual usage and the
description (narrative and the slightly erroneous version in the
Appendix indicate that choice describes alterity. A use case showing
that both options of choice can be showing is choice/abbr|expan. We
commonly print both options the first time a name that will subsequently
be abbreviated is used; after that we tend to use either abbr or expan
(without choice). So something you might very easily encode thus


> This proposal is intended for the
> <choice>
> <expan>Text Encoding Initiative</expan>
> <abbr>TEI</abbr>
> </choice>

Will very commonly be rendered


> This proposal is intended for the Text Encoding Initiative (TEI).


I would strongly argue that choice is to subst (and this use of edInt)
as ab is to p, and seg to any phrase element: a generic case that is
semantically and syntactically narrowed by the other options in order to
cover specific cases.

Secondly, I would strongly argue that the basic pattern of all
choiceLike elements is choice/option|option where option can be either a
child of choice or used on its own when alterity is not required--but
that options should not also be able to substitute for choice. So we
should allow this (option is a currently non-existent generic element
intended to stand for elements like abbr and expan):

<choice>
 <option>
 <option>
</choice>

or

<choice>
 <option>
  <choice>
   <option>
   <option>
  </choice>
 <option>
</choice>

But neither

<option>
 <option>
 <option>
</option>

nor 

<choice>
 <optiin the case of nested elements, on>
  <option>
  <option>
 </option>
 <option>
 </option>
</choice>

In other words, we should consider very carefully whether an element is
*either* the choiceLike container, or the optionLike option; but the
same element should not be capable of both expressing an alternate state
and directly containing options expressing alternate states.
> 


On Mon, 2007-07-23 at 11:08 -0400, Dot Porter wrote:
> On 7/23/07, Arianna Ciula <arianna.ciula at kcl.ac.uk> wrote:
> 
> > I think lots of people (especially modernists) will be happy to see a
> > proposed differentiation between authorial and editorial interventions.
> >
> > Arianna
> 
> I agree that differentiating between authorial and editorial
> interventions would be a valuable addition to the Guidelines.
> 
> Dot
> 
> > >
> > > Comments?
> > >
> > >
> > >> Hi all,
> > >>
> > >> Here is the first of two proposals relating to my concerns around choice
> > >> and app/rdg:
> > >> http://people.uleth.ca/~daniel.odonnell/Research/a-proposal-for-revisions-to-choice-app-and-modelppartedit
> > >>
> > >>
> > >> I'
> > > _______________________________________________
> > > tei-council mailing list
> > > tei-council at lists.village.Virginia.EDU
> > > http://lists.village.Virginia.EDU/mailman/listinfo/tei-council
> >
> > --
> > Dr Arianna Ciula
> > Research Associate
> > Centre for Computing in the Humanities
> > King's College London
> > Strand
> > London WC2R 2LS (UK)
> > Tel: +44 (0)20 78481945
> > http://www.kcl.ac.uk/cch
> > _______________________________________________
> > tei-council mailing list
> > tei-council at lists.village.Virginia.EDU
> > http://lists.village.Virginia.EDU/mailman/listinfo/tei-council
> >
> 
> 
-- 
Daniel Paul O'Donnell, PhD
Department Chair and Associate Professor of English
Director, Digital Medievalist Project http://www.digitalmedievalist.org/
Chair, Text Encoding Initiative http://www.tei-c.org/

Department of English
University of Lethbridge
Lethbridge AB T1K 3M4
Vox +1 403 329-2377
Fax +1 403 382-7191
Email: daniel.odonnell at uleth.ca
WWW: http://people.uleth.ca/~daniel.odonnell/




More information about the tei-council mailing list