> Daniel O'Donnell wrote:
>> Lou-at-home wrote:
> While it is possible to see del > add, or sic > corr (or orig > reg or
> even rdg > lem) as a sequence, they are also parallel states (i.e.
> when I pick up a corrected manuscript, both states are present).
> I think this is true of corr/sic and orig/reg but I don't think it's
> true of the add/del pair you cited in exactly the same way.
> Let me try harder to explain why. If I transcribed only the deletion
> (or only the addition for that matter) you would reasonably say that I
> have left out something important.
Depends on what you are doing. It is quite common for non-diplomatic
print transcriptions/editions (e.g. semi-diplomatic) to record only the
corrected state, sometimes indicating the fact that there is another
underlying form somewhere by use of italics, but sometimes not. Think,
to give an extreme example, of some of the various forms Gabler's
Ulysses takes: from synoptic diachronic transcription to clear-text
emended and corrected edition. Whether the sequence is essential depends
on the level of interpretation, the amount known, and the purpose of the
encoding and/or reading.
> So they are not alternatives: they are both necessary. On the other
> hand if I gave you only one of the corr/sic pair or the orig/reg pair
> you might say it wasn't appropriate and you;d rather have the other or
> both, but you wouldn't (I think) feel you'd been seriously
> shortchanged. So these *are* alternatives: you can do without one or
> the other.
Again, I've seen lots of editions that do the post-correction text only;
some that do the pre-correction. I don't think we should shut down the
idea of simultaneity in addition/deletion, sic/corr, orig/emend pairs
since what is sequential at one point can be simulataneous at another.
> I might want
>> to see the sequence, I might want to see a final state (if that can
>> be determined or is applicable), I might want to see both states at
>> the same time. The whole point of the element/class (as I understand
>> it) is to mark a point in a text in which more than one alternative
>> state exists. Since very few things actually (from a processing point
>> of view) actually enter a text simultaneously, most of these will be
>> a sequence as well (from an actualy diachronic production point of
>> view); but from an idealised (and synchronic) textual point of view
>> they are parallel. To have a coherent and sensible reading (in the
>> absolute strictest sense--e.g. for use in text summation or
>> translation software), one needs to choose one form in each of these
>> cases from the alternatives presented (the only reason for keeping
>> choice as a name). The whole point of the element, however, is to
>> indicate that there is a synchronic multiplicity for whatever reason.
> I agree with this, I think, and apologise for confusing the issue by
> bring time into it!
>> This finally does raise my point about emendation and correction. I
>> think the 600 pound gorilla we are not discussing directly here is
>> the scope of the markup in question. That is to say what is being
>> encoded. In any transcription there are two texts in play: a physical
>> witness and an abstract text. Structural markup like TEI on the whole
>> assumes that textual features have specific meanings--i.e. that
>> italics means foreign, or emphasis, or highlighting even. Diplomatic
>> transcription, on the other hand tries to avoid interpreting the
>> significance of features (though obviously it makes a selection based
>> on interpretation). The problem we are running into is coming at
>> places where we are jumping from a descriptive to an interpretative
>> markup. E.g. the problem with
>> <sic><add>foo</add></sic><reg>phoo</reg> (however the elements are
>> ordered) is that the sic and add pairs and the add (and in my view
>> implicit delete) pairs are describing different orders of
>> information. The fact that the foo has been added is an
>> (interpretative) description of what has actually happened in a text:
>> there wasn't foo and then there was. The fact that sic is a mistake
>> for phoo is an (interpretative) description of what the idealised
>> text should read: although as witness says foo, it should actually be
>> spelled phoo for whatever reason (because foo is wrong, a dialect, or
>> a form I don't expect my audience to understand).
> I'm with you 100% up to here!
> Both <choice><add>foo</add><del>fat</del></choice> and
>> <choice><sic>foo</sic><corr>phoo</corr></choice> are perfectly clear
>> because in each case the scope of the phenomenon being discussed is
>> likewise clear.
> But here you lose me. Are you saying that if the children of <choice>
> are <add> and <del> then I should interpret their relationship in a
> different way from the way I would if they are <sic> and <corr>? I
> don't think that's a good idea. It's easy enough with the cases cited,
> but what's going to happen when people start throwing arbitrary <seg>s
> <choice> <seg type="a">xxxx</seg> <seg type="b">yyyy</seg> <seg
> Does this mean that the A and B are both about the physical state, and
> the c is an interpretative judgement of some type? Or does it mean
> that the A is the physical state, and the B and C are alternative
> interpretive judgments?
I don't think we disagree here. My point was that we have no trouble
seeing what the choice is when we are describing things of the same
scope (i.e. choosing between physical text and physical text or
editorial text and editorial text). The problem that I and I think
Sebastian when he's not sleeping with the neighbours had with your
original encoding of
<add><choice><orig>foo</orig><reg>phoo</reg></choice></add> (or was orig
actually sic?) is that the regularisation is not being added to the
text, it is correcting an editorial reading that is based on a witness
reading that was added. That's given me a headache, but I hope you can
see the scope difference.
> The problem comes when you try to describe a physical
>> feature and interpret its abstract significance in one choice.
> No reason why one choice shouldn't be "I'm just going to talk about
> the physcial features" is there? But then it must be ONE of the
> chilfdren of <zzzz> not several of them.
>> BTW, although we have voted for choice, I really think we should be
>> paying attention to the difficulty we are having with this: some of
>> it may well be semantic.
> I agree with that, certainly! I wonder whether everyone else on this
> list is asleep, watching television, or down the pub? Oh no, it's
> Sunday. They must all be in Church.
> This SF.Net email is sponsored by: YOU BE THE JUDGE. Be one of 170
> Project Admins to receive an Apple iPod Mini FREE for your judgement on
> who ports your project to Linux PPC the best. Sponsored by IBM.
> Deadline: Sept. 13. Go here: http://sf.net/ppc_contest.php
> Tei-choice mailing list
Daniel Paul O'Donnell, PhD
Associate Professor of English
University of Lethbridge
Lethbridge AB T1K 3M4
Tel. (403) 329-2377
Fax. (403) 382-7191
Home Page <http://people.uleth.ca/~daniel.odonnell/>