Re: Simple alternation

On 18/10/16 21:13, C. M. Sperberg-McQueen wrote:
> On Oct 18, 2016, at 1:25 PM, Lou Burnard wrote:
>
>> On 18/10/16 20:15, C. M. Sperberg-McQueen wrote
>>> Er, why tag it as unclear?  It looks perfectly legible to me
>>> in the image.
>> Because (I assume) it's not clear which is the right reading, specifically because of lack of clarity as to the initial letter.
> The 'unclear' element has a documented meaning; it marks
> spoken material which was made hard to hear by ambient noise
> or other problems, or written material which is hard to read
> because the handwriting is difficult or the page is damaged or
> whatever else can cause writing to be hard to read.  The
> 'unclear' element does not mean "we are not sure what to do here" or
> mark any other form of lack of clarity.

Fair point. This is probably why both Peter B. and I want to specify 
that the <unclear> should wrap just the letter because we think it's 
just the letter which is unclear. Remember that we're looking at a 
printed version of something else in which there are question marks 
around both the letters proposed. So we interpret that text as saying 
"there's a letter here which might be one thing or it might be another 
and we can't determine which". This doesn't seem so different from 
"there are four characters here which I can't read".

>
>>> Like Magdalena Turska, I think this looks like app and rdg.
>>> Both readings are from the same witness, but that’s not
>>> unknown.
>> True, but it is rare (I would suggest nonexistent) to have *only* readings from a single witness.
> Rare, yes.  Non-existent?  In all the wide range of textual
> transmission, do you really believe that there are no marginal
> alternatives or inline corrections in ANY of the thousands
> of manuscripts which are the only witnesses to the text they
> contain?
Why no. That's crazy talk. I'd just not recommend using app to record them.
>
> Wittgenstein's manuscripts are full of alternate readings; I have
> heard distinguished professors tell me that authorial variants
> are not AT ALL like the manuscript variants of medieval and
> ancient texts, structurally or in any other way, but for the most
> part it was immediately clear that they didn't know what they
> were talking about.

You know, I think I may have met some of those very same critters myself.

You (in particular) should be pleased to know that they or their 
representatives pop up and complain about the TEI's deplorable laxity in 
this matter at regular intervals,  roughly every five years or so. 
Depending on the constitution of the TEI Council at the time, further 
weasel wording subsequently sneaks into (or is expunged from) the 
Guidelines to nudge users in favour of one or other position.

And thank you for the suggestion of a way to handle acrostics using 
<choice>!

L
> Any text for which multiple versions exist
> or can be constructed from the textual tradition can be
> represented by app and rdg; it's what they are good for.
>
>
>>> The documentation for ‘choice’ says
>>>
>>>      Because the children of a choice element all represent alternative
>>>      ways of encoding the same sequence, it is natural to think of them
>>>      as mutually exclusive. However, there may be cases where a full
>>>      representation of a text requires the alternative encodings to be
>>>      considered as parallel.
>>>
>>> It would be a more obvious choice if the Guidelines specified one
>>> interpretation as a default and provided an attribute for overriding
>>> that interpretation.  But I read the paragraph quoted as saying that
>>> you as an encoder are at liberty to take either interpretation; I’d put
>>> a remark in the header to inform later readers of the document.
>> I'd forgotten about that weaselly worded paragraph.  In all the years we've been using this element I have yet to see any case where "the alternative encodings" are also somehow "to be considered as parallel".  Maybe because I'm having trouble working out what that might mean. If they're alternatives, how can they be parallel? Can you suggest an example?
> Why are you looking at me?  Not my weaseling.
>
> But since alternative ways of encoding the same thing might be
> alternatives each of which capture part but not all of the truth,
> the paragraph quoted does seem to be trying to leave the door open
> to an imaginable use case.  (At least, I can imagine that such a case
> might arise.  Can I imagine such a case for real?  Hmm.)
>
> Well, I don't know if this is a good use of choice, but since you
> asked so nicely ... Suppose we have a poem whose first letters
> spell out an acrostic message which does not otherwise appear
> in the poem.  We want the acrostic message to be easily findable
> in a simple search of our document, so we want to transcribe
> those characters twice.  Could one say that one has a choice
> between reading them  horizontally (the 'normal' text) and
> vertically (the acrostic message), and encoding the normal text
> as one child of a 'choice' element, and the acrostic as a second?
> The two transcriptions do not, in this case, exclude each other
> in the same way that the w and m do in the example this thread
> started with.
>
> Michael
>