[tei-council] genetic draft -- from Brett, pt. 2

[Brett Barney]

Lou,

Thanks for the encouraging responses to my first set of notes. I'll try to
follow up on issues you and Sebastian raised in a next message. For now,
here are thoughts about the two questions I skipped before and a handful of
others based on a reading of the next page or so of the "Preliminary
updates" document.

> 6. discussion of spanTo needs expansion. Some more real examples would
> help.

I don't know that I can help with expansion of the prose, but I'll spend
some time looking for other examples. Is there a particular sort that would
be especially welcome?

> 7. Should <metaMark> possibly have a <desc> child to describe it rather
> than relying on brief characterisation as attribute values?

Yes. I know, we should try to generate one, then.

--------------

> The element patch is useful in cases where the written surfaces
> constituting a document are not homogeneous.

I don't think this description of the cases in which patch is useful is
quite right. Many manuscripts comprise multiple leaves that are not all
"homogeneous" (e.g., one torn from a notebook, others written on the backs
of envelopes, etc.), but I don't think those manuscripts are the sort
meant. Perhaps something along the lines of "The element patch is useful in
cases where an inscribed leaf subsumes one or more physically distinct
scraps."

> Most writing is linear, in the sense that it is composed of discrete
> tokens organized physically into groups, typically organized in a
> sequence corresponding with the way they are intended to be read.

This sentence seems to me to waffle unnecessarily--e.g., I don't see a
reason for both qualifiers "mostly" and "typically." More important,
though, I think that the last two-thirds ("organized . . . to be read."),
is more opaque than "linear," for which it's presumably a gloss. Most
important, though: The observation that most writing is linear doesn't by
itself constitute to my mind an argument that we need an "element to hold a
complete group of such tokens." I'd be much more convinced by the sort of
reasoning, common elsewhere in the Guidelines, that having the contents
marked up in this way allows something that people want to do.

> Where, however, the lineation is not considered useful or significant,

Small point: I'd prefer just "significant" to "useful or significant." Or
maybe "purposeful or significant."

Maybe @flipping/able="no" should be added to the example encoding of the
Whitman draft?

> The encoder may choose to combine graphics with the transcription at
> whatever level is considered effective, or not at all.

The word "combine" seems not quite right, since the graphics themselves
aren't combined with the transcription but referred to, right?

> Equally, the encoder may choose to provide only graphics without
> transcription, or with a structured (non-embedded) transcription, or any
> combination of the three.

As this section is headed "Embedded transcriptions," this strikes me as an
odd place for this information. Maybe the solution is just to point
explicitly to the sections where these other approaches are outlined. Also,
"any combination of the three" might be made more specific; this is
basically getting at the idea that a TEI document might be made up of any
combination of <text>, <facsimile>, and <sourceDoc [or whatever]>, right?

> when dealing with authorial manuscripts,

Is this supposed to mean (and should it therefore read) "literary
manuscripts"?

> To encode such transcriptions, we propose a simple model in which writing
> traces are marked within a transcription, using an appropriate tag such
> as mod, del, etc. to indicate their function in the document, as
> described in the remainder of this chapter.

The voice/tone here seems out of step with the rest of the Guidelines,
which don't typically "propose" things (IIRC) and don't use the 1st person
plural. Also, I supressed a smile at "simple model." What about something
like "The remainedr of this chapter describes a model for encoding such
transcriptions using tags that mark the functions of various writing traces
within the document.

> termed a layer

STOP THE PRESSES! AAAAAGGHH! BURMA! Based on the listserv discussion back
in early June, I was assuming that "layer" was off the table. When Lou
started that thread on 6/6 I immediately started formulating a detailed
response, complete with multiple exhibits to try to persuade the group that
layer is egregious--far, far worse than diploma, which Sebastian
characterized as "a disaster" because the word "has such precise, other,
semantics in normal conversation." That observation applies far more
strongly in the case of layer, I'd contend. I am running out of time for
today, but I'll make it my first order of business tomorrow morning to put
together the full case that I had thought was unnecessary back in June,
based I suppose on the lack of objections to Martin's post that followed
almost immediately after Lou's. (I'm referring here to this bit: "  -
<changeSet> instead of <layer> (and friends). I think this would be
uncontroversial, since some of us voted for it already."


Thanks,
Brett

------------------
Brett Barney
Research Associate Professor
Center for Digital Research in the Humanities
University of Nebraska-Lincoln
bbarney2 at unl.edu
http://cdrh.unl.edu