dorsey film conversion questions and estimate

[Koontz John E]

On Fri, 9 Jan 2004, Rory M Larson wrote:
> No, the 20,000 figure is rough estimate of the total number of words,
> and it's probably high.

The estimate antedates microfilming.  I think they probably took an inch
of cards, counted the cards in it and multiplied by the number of inches
in the file boxes.  At least that's what I would have done. I doubt they
counted the actual cards, and I don't think Dorsey did, either.

> Each frame includes typically 4, but often fewer, index card images.

I'd forgotten the four cards per frame reduction.  It's been a while since
I've been able to consult the microfilm.  The CU Library microfilm
facilities (if not changed lately) must be much worse than those in
Minnesota.

> I suspect Dorsey originally wrote his notes on cards by hand, then typed
> them on another card, but all these cards were kept and sorted
> alphabetically in the same deck.

Is there substantial duplication?  I don't remember that.  I'd expect him
to remove handwritten cards as they were typed.  Elsewhere he speaks of
notes and slips as needing to be typed.  What I do remember is that
somewhere he has a manuscript on how one might go about generating the set
of all OP words of certain lengths - basically an approach based on
knowledge of the canons of OP words qualified by their morphosyntax.  It
looked to me like some slips were more or less blank, except for having a
potential form on them, and I think these may have been "potential form"
cards that he kept so he would be reminded to check whether such a form
actually existed and, if so, what it might mean.  This was only one of his
discovery procedures, of course. There were a certain number of such
cards, but I don't think they were anything like half the number.  I
couldn't say what percentage they might have been.

I once did a count of unique word forms in the texts and I believe I came
up with about 5-6K, though it has been a while since I thought about that
and I don't have the file handy.  I remember immediately concluding that
the slip file had material from sources other than the texts, because that
count certainly includes different inflected forms of a single stem.  The
slipfile itself, however, lists inflected forms of a single stem on the
same slip.  Different slips may be different derivations from the same
underlying stem.