CSD (privately)

[Koontz John E]

On Sun, 5 Oct 2003, R. Rankin wrote:
> Many of the cognate sets were assembled during the Summer of 1984 with
> participation by the above four, plus John Koontz, Paul Voorhis, Pat
> Shaw, Willem DeReuse and I'm sure another one or two that I'm
> forgetting.

Randy Graczyk, and briefly Allan Taylor, Josie White Eagle, and Ray (?)
Gordon that I can immediately recall.

> The CSD is in a DOS file, but it uses a special DOS font.  It is the
> font that needs to be converted from ASCII to Windows (ANSI).

This is really the main issue.  A few things were done with diacritics,
when we couldn't squeeze everything into one font.  David Rood has, I
think, a set of fonts for this in Windows, and those are easy enough to
create, but I don't know that any recoding software was ever created.

> I think John, who has done most of the computer work with the file, has
> it in a Unix format for formatting purposes.

No.  It's formatted with a nod to AskSAM's textbase format, but I always
converted to the SIL SFM format for practical use.  This involves the
major notational change of

fieldname[ field contents

to

\fieldname field contents

I then would convert the various asterisks, daggers, and quotation marks
used in the fields to what was in the last versions |subfieldcode{...}.
The SIL tool that I used to convert the file to an MS Word DOC file needed
that to encode formats.  The main issue here was that subfields were often
missing, or inconsistantly coded, or interlarded with arbitrary
commentary, so I would have to deduce the correct coding heuristically.
This worked 9 times out of 10, but that's a lot of not-working cases.

I ultimately tried to get the editors to work with the data in the form

\lg phonemic form | source form | gloss | source code
\rem-lg ....

but I didn't come up with this version until I had exhausted their
patience with earlier cruder versions.  Using those |subfieldcode{...}
constructs directly, for example.

> I don't think anyone minds giving people access to information from the
> files at this point, but the file itself isn't very "portable" because
> of the font problems and the fact that we use a special Program Editor
> to read and search it.  Wes Jones put the PE together for us as I
> recall.

I don't mind myself, but I think the actual, senior participants - i.e.,
not me - have to make that decision formally.

> What we need to do now, I think, is to convert the database so that it
> can be "sucked up" into Doug's IDD and edited further there.

I could be wrong here, but I wonder if the IDD has the right set of
fields and/or subfields?  The general technology would work, of course.

> That should ultimately produce publishable copy.

There are various ways of getting from a database format to a publishable
format.  In a way that's a less strenuous problem than just getting the
database edited into the regular form that feeds whatever is used.  Many
of the changes needed there are trivial, but it's often the
non-automatable trivialities that are most laborious.

> The remaining tasks include a more thorough work-up of the kinship
> terminology that Dick Carter promised to do, some pronominal
> reconstruction that needs to be refined and maybe one or two other
> things.

I thimk there was one other "domain" issue (like kinship terms), but I
can't think what it was.  Also, I discovered that many of the missing
Omaha-Ponca forms are readily available in Dorsey's texts.

> The dictionary could be expanded enormously if words found only in the
> Mississippi Valley Siouan subgroup were allowed to constitute sets.

That seems almost certain, though in the long run I'd like to see it
actually done, of course.  The CSD shouldn't be delayed for that, however.

> That's about where things stand today.  David and John have a published
> paper in which they talk about the logistics of the project.  They can
> provide the source for you.

Rood, David S., and Koontz, John E.  2002. The Comparative Siouan
Dictionary Project.  pp. 259-281 in Making Dicitonaries:  Preserving
Indigenous Languages of the Americas, ed. by William Frawley, Kenneth C.
Hill, and Pamela Munro.  U of California Press, Berkeley, CA.