Re: Encoding the speaker of a set of <lg> or <p> elements

If,  as I assume, you're dealing here with Sanskrit literary materials, 
I think the "XXX said" lines are formulaic labels which behave just like 
the things you would tag with <speaker> inside a <sp> in a 
conventionally presented dramatic script, so it certainly seems logical 
to wrap the whole thing up in an <sp> with a <speaker> followed by <p> 
or <l> as need be.

The trouble with <speaker> is that it's a bit of a misnomer: I certainly 
think of it as containing the *whole* of any such indication, so it may 
contain a name as well as other words.  I think tagging "said" as a 
<stage> is a stretch too far, though again I can see why you might want 
to use it for things like "Lou then uttered the following words, rolling 
his eyes in disbelief". But do such formulations occur in the case in 
hand? For me <stage>  really belongs to drama, not to this kind of 
narrative verse.

In fact, <stage> is permitted both between and within ungrouped <l>s 
except when the <l>s are grouped into <lg>s, in which case it's only 
allowed within an <l>, or between <lg>s. This doesn't seem entirely 
wrong to me, though I can see why it might be considered a bit odd.

  On 01/09/11 15:18, Paul F. Schaffner wrote:
> It's not clear to me whether the phrases in question mark the
> beginning of a speech, or simply a reminder of who is speaking;
> nor whether (if they *do* mark the beginning), there is any
> clear spot at which you can say, 'here this speech ends.' The
> answers to these questions would determine how we would tag
> this, with<sp>  or not.
>
> If the text is fully dramatic in form, i.e. is broken into
> distinguishable speeches with beginnings and ends and identifiable
> speakers (plural), then yes, it would be feasible to tag the
> speeches with<sp>; if not, not.
>
> As for the phrase "Speaker said...," I think that we would
> probably put it in<stage>. We allow<stage>  within<lg>;
> P5, I think (rather surprisingly, since<stage>  is a kind of
> <note>), allows it only within<l>. In either case, wrapped
> in<l>  or between<l>s, you'd get something like this:
>
> <l>....</l>
> <l>....</l>
> <stage>Speaker said:</stage>
> <l>....</l>
>
> If we wished to draw out the Speaker into a<speaker>  tag,
> we would combine<speaker>  and<stage>
>
> <sp>
> <speaker>Speaker</speaker><stage>said:</stage>
> <lg>
> <l>....</l>
>
> I believe that<stage>  fits very well the 'openerComplement' role
> described below. At least, there are examples in the Guidelines that
> seem to use it more or less in that role.
>
> If you do go the<sp><speaker>  route, and the speech begins in
> the middle of a stanza, then of course you'll have to break the
> <lg>  in two and distribute the halves between the<sp>s.
>
> pfs
>
>
> On
> Thu,
> 1
> Sep
> 2011, Bertrand Gaiffe wrote:
>
>> You probably are right. In such a case, you could as well have an<ab>  (read
>> garbage element) that would do :
>>
>> <sp who="#s1"><speaker>the speaker</speaker><ab
>> type="openerComplement>said:</ab>
>>
>> <lg>
>>   <l>Lorem ipsum dolor sit amet</l>
>>   <l>consectetur adipisicing elit</l>
>> </lg>
>> </sp>
>>
>>
>> I understood that interoperability is an issue (reading previous mails on the
>> list). There are a lot possible choices here !
>>
>> the definition of<speaker>  reads :
>>
>> 1) A specialized form of heading or label,
>> 2) giving the name of one or more speakers in a dramatic text or fragment.
>>
>> If you emphasize on (1), you could end up with such things as :
>>
>> <speaker><name>The speaker</name>  said :</speaker>
>>
>> If you emphasize on (2)
>>
>> you would probably get something like either
>>
>> <sp who="#s1"><speaker>the speaker</speaker><ab
>> type="openerComplement>said:</ab>
>>
>> or the solution you propose...
>>
>> It seems that both parts of the definition together are challenging in a lot
>> of cases !
>>
>> 	bertrand
>>
>>
>>
>> Le 01/09/2011 09:37, Espen S. Ore a écrit :
>>> Would you really say that 'said' is part of the<speaker>  element? In
>>> the Ibsen-project we have found that we would like to keep the name (or
>>> other person-references such as 'the speaker') in the speaker element
>>> but other stuff outside - including typical trailing dots. For that
>>> reason we have constructed a new element: spOpener, so that a typical
>>> Ibsen-opening would be something like:
>>>
>>> <sp><HIS:spOpener><speaker who="PEER">Peer Gynt</speaker>  <stage>(uden
>>> at standse)</stage>.</HIS:spOpener>
>>> <l>...</l>
>>> </sp>
>>>
>>> Espen Ore
>>> University of Oslo
>>>
>>>
>>>
>>> Den 01. sep. 2011 09:09, skrev Bertrand Gaiffe:
>>>> Perhaps stupid answer...
>>>>
>>>> Why not :
>>>>
>>>> <sp who="#s1"><speaker>the speaker said:</speaker>
>>>> <lg>
>>>> <l>Lorem ipsum dolor sit amet</l>
>>>> <l>consectetur adipisicing elit</l>
>>>> </lg>
>>>> </sp>
>>>>
>>>> whith of course<l>  at the first level in case<lg>  may contain more
>>>> than a speaker.
>>>>
>>>>
>>>> bertrand
>>>>
>>>>
>>>>
>>>> Le 01/09/2011 08:15, Arun Prasad a écrit :
>>>>> Hello all,
>>>>>
>>>>> I have a fairly simple question to ask, but I'm stuck on the answer to
>>>>> it. I hope one of you will be able to help me out.
>>>>>
>>>>> First, some background. I'm working with a narrative poem that consists
>>>>> of a few hundred verses (each wrapped in<lg>  tags). These verses are
>>>>> almost all regular units, but every tenth verse or so will be prefixed
>>>>> with "so-and-so said:" . That is, if this is a standard verse:
>>>>>
>>>>> Lorem ipsum dolor sit amet
>>>>> consectetur adipisicing elit
>>>>>
>>>>> Then occasionally a verse will appear as follows:
>>>>>
>>>>> [speaker] said:
>>>>> Lorem ipsum dolor sit amet
>>>>> consectetur adipisicing elit
>>>
>>
>>
>> --
>>
>> B. Gaiffe
>> ATILF, 44 Bd de la libération, B.P. 30687, F 54063 NANCY CEDEX
>> Nouveau numéro de téléphone : 03 54 50 53 41
>>
>>
>>
>
> --------------------------------------------------------------------
> Paul Schaffner | PFSchaffner <at> umich.edu | http://www.umich.edu/~pfs/
> 316-C Hatcher Library N, Univ. of Michigan, Ann Arbor MI 48109-1190
> --------------------------------------------------------------------