[MEI-L] Colla parte

Roland, Perry (pdr4h) pdr4h at eservices.virginia.edu
Sat Sep 3 00:58:15 CEST 2011


Hi Joachim,

Let me re-state the issue to see if I understand it correctly --

1. Weber (is there any other composer worth talking about?) draws some version of a slash-symbol-thing to indicate that the current measure should be filled in with material from another instrument.

2. You want to record the original version (with the slash symbol) and give its "expansion" (as it were)?

First, are you sure this problem can't be resolved using <orig> and <supplied> (or perhaps <reg>)?  For example, each measure of the flute part might be encoded --

<measure>
  <staff n="1">
    <layer>
      <choice>
        <orig> <!-- some repetition sign(s) here --> </orig>
        <supplied> <!-- a copy of the content of the vln. part --> </supplied>
      </choice>
    </layer>
  </staff>
<dir staff="1" tstamp="1">colla parte ...</dir>
<!-- instead of the words there might be a <line> -->
</measure>

This seems like a good solution to me because it is explict.

I'm using <supplied> because the actual substitution doesn't take place in the score, there's only an indication that it "ought" to happen. Actually making it happen is up to the editor -- the human being, not the software.

Of course, what's missing is a link between the <supplied> element and the words "colla ..." (or a "wiggly" line), which I presume is in the flute part). At present, the editorial intervention can only be explained in the editorialDecl.  But, if we put our heads together, we might be able to think of another method of linking these.  (Perhaps with an <annotation>, which has pointers to the "participants"; that is, the <supplied> and <dir> elements, and a type attribute value of "collaparte"?)

I don't know if you remember, but MEI originally had a different definition for <mRpt>.  It didn't necessarily indicate the repetition of the preceding measure, but rather the repetition *of any other measure*, "repetition" being defined somewhat loosely, of course.  This definition went the way of the dodo bird when the editorial elements (add, del, orig, reg, etc.) were added since the attribute (don't remember what it was called off the top of my head) pointing to the source measure duplicated these editorial elements' function.

However, I recognize that there's currently no good way of implementing a reference to content given elsewhere in the document.  So, perhaps we should consider adding such a specialized element.  In the example above, the content of <supplied> would be the <colla> (or some such name) element.

<measure n="1">
  <staff>
    <layer>
      <choice>
        <orig> <!-- some repetition sign(s) here --> </orig>
        <supplied> <colla/> </supplied>
      </choice>
    </layer>
  </staff>
</measure>

Of course, it would need an attribute pointing to the content of the vln. part.  In a way, this element would be similar to <ref>, although  <ref> is intended for navigation, while <colla> indicates "go get the content and put it here."  This is also not unlike internal parsed entities. This element would also need to be related to the <dir> or <line> elements. We could think about using its generic corresp attribute or give it a new, specialized attribute.

(I can already hear some noises in the back of the room about some attribute or attributes for describing whatever transformation must be applied to the source, such as "transposed up a 5th", "inverted and retrograde", etc.  I'm not ready to go there yet!  There be dragons!)

It might also need a more generic name than "colla".  Then it could be used for other situations where content given in one location needs to be referenced in some other.  Can't think of a good name off the top of my head, too late in the day.

Just in case someone asks, I don't think XInclude can point to a location in the current document.

Enough to think about for the weekend?  :)

--
p.

__________________________
Perry Roland
Music Library
University of Virginia
P. O. Box 400175
Charlottesville, VA 22904
434-982-2702 (w)
pdr4h (at) virginia (dot) edu
________________________________________
From: mei-l-bounces at lists.uni-paderborn.de [mei-l-bounces at lists.uni-paderborn.de] on behalf of Joachim Veit [veit at weber-gesamtausgabe.de]
Sent: Friday, September 02, 2011 3:43 PM
To: mei-l at lists.uni-paderborn.de
Subject: [MEI-L] Colla parte

Dear MEI-L-Readers,

when encoding manuscripts MEI has the possibility to encode
measure-repeats or half-measure-repeats which are found in the original
manuscripts in a form similar to: // or .//.  (two strokes through the
barlines or something similar) with <mrpt> or <halfmrpt>. This has
always a clear reference to the bar (or half-bar) before.

The same symbols are used in manuscripts in combination with the "col
Basso" or other "colla-parte"-instructions.
Again in Weber's Freischuetz-Overture (we never do something other...)
we have a contrabbasso with <staff n="16"> and above this the
violoncello with <staff n="15>. Weber notates the contrabbasso in a
normal way and in the cello-staff we find only "c. B." (= bar 1) and
afterwards //   //    // etc.
For a modern edition we could label the layer of the contrabbasso with
an xml:id="2011" and put in the layer of staff 15: <layer n="1"
copyof="2011"/>.

But in this case we want to encode the "original" situation! So we first
have to define in the scoredef:
<scoredef>....
<symboltable>
<symboldef xml:id="symCollaParte"/> <!-- here we describe the symbol(s)
Weber and others use in this case -->
</symboltable>
</scoredef>

and within our staff 15:
<staff n="15">
<layer n="1">
<symbol ref="symCollaParte"/>
</layer>
</staff>
In this case the reference is not always clear: it might go to the staff
below (as in this case) or the staff above (if the cello is written out
and the cb-staff is pointing to the cello) or even from a flute to the
first violin 5 systems below. So there should be some mechanism to make
clear where the model is situated. And second: We should have the
possibility to use "c. B." (or something similar) as a symbol-phrase
(which at the same time "defines" the model) and for the following bars
simply use "//" (or something similar) (maybe even one symbol for 2-3
bars together??).
Is this case already considered in the future guidelines?

Best greetings and a happy weekend,
Joachim


More information about the mei-l mailing list