Root lists.

[Koontz John E]

On Mon, 14 Oct 2002, Linda A Cumberland wrote:
> Is that list still available somewhere? It sure would be a lot easier than
> sifting through Boas and Deloria! I'm finding a lot of differences between
> Dakota and Assiniboine in the list I have so far - roughly 300 roots.
> Seems to warrant continuing the search.

It might be, and it's easy enough to repeat.  The algorithm is just:

1) find headword that matches a regular expression that reprents stuff
that can precede instrumentals followed by an instrumental (in the
orthography of the file)

2) print unmatched coda followed by comma followed by matched header
followed by first n characters of the definition followed by the page
reference

3) repeat

I wrote the code in AWK and processed (I think) a version of the Siouan
Archives Buechel that had been recoded into the Siouan Dictionary
character set we had developed for DOS.

That's possibly the main problem with resuscitating the files if they
still exist - being able to view the character set.

> I embarked on this project to get a handle on reduplication, since
> deletion and coda nasalization (the latter being a common effect at
> borders in Assiniboine) makes some reduplicated forms look related when
> they're not.  But the list is proving useful in a lot of ways.  Since
> there seems to be interest in it, I'll probably include it in my
> dissertation as an appendix - with the caveat that glosses are
> provisional!

That sounds like a good idea if space permits!

JEK