On Thu, 12 May 1994 09:16:32 CDT <[log in to unmask]> said:
>I wish that we could get away from the idea of coding an apparatus of
>variants. This is a hangover from the constraints of print technology
>where it simply was not feasible to publish all versions of a manuscript.
> ...
>We don't have to worry about that now. ...
 
Prof. Faulhaber is, of course, quite right that the computer is changing
the economics of transcription and publication of texts with variations.
If nevertheless I cling to the term 'apparatus', and indeed to some
aspects of the thing itself, it is for several reasons:
 
- it provides a familiar term and a useful metaphor for a collection of
information on the variations in a text; the use of the term 'apparatus'
in TEI P3 should not be taken to mean that the encoding methods
described there are slavishly modeled on the form of printed apparatus,
but only that, like the printed apparatus, they allow one to record
salient information about textual variation
 
- printed apparatus identifies points of particular interest in the
study of the textual tradition: the variations.
 
- it is more economical to store, just as it is more economical to
print, a single reference version of a text, with notes of differences
in other versions. Such economy may not always be necessary, but in
general if we cut the size of our data representation by 50%, we cut our
storage costs by 50%. (This motivates the design of revision control
systems like RCS and SCCS; it is not limited to print.)
 
- storage of a single reference version with an 'apparatus' of
differences allows us to be as exhaustive or selective as we like in
recording the differences.
 
I don't insist that an apparatus, or an apparatus-like data structure,
is the ONLY way to deal with textual variation. But it does seem to me
to make clear the nature of the textual variation and that of the
textual object itself in useful ways.
 
C. M. Sperberg-McQueen