Re: conlang classification and productivity

Irina:
> On Sat, 10 Feb 2001, And Rosta wrote:
>
> > (a) a notional absolute reckoning of the elaborateness of each of
> >     P, L, G, T, C (in their completed forms).
>
> How?
Note that my suggestions were meant to be for what would be informative
but not necessarily for what would be practicable. For example, Dan
posted that he had about 20 conlangs, and David replied admiringly
that he, David, has only one, which happens to be about twice as old
as Dan. But we know of course that the products reflect David's
preference for pouring his great energies into a single project, and
Dan's for spreading his great energies across many.

At any rate, the absolute reckoning could be relative to certain
standard reference points, such as Laadan, say.

> > (b) a percentage for the extent to which each of P, L, G, T, C are
> >     borrowed/derived from, modelled on, prior sources.
>
> Well, yes.
>
> > (c) a percentage for how much each of P, L, G, T, C has been created.
> > (d) a percentage for how much each of P, L, G, T, C has been documented.
>
> The problem with that is that I don't know how much what I've got is
> of what there is; I can't see the whole language, only the parts of
> it that I have and just a little bit beyond. How can one *ever* know?
> It's not as if I'm making something with the intent of *finishing*
> it.
I see. Had I the resources of time, I would wish to finish P and G for
Livagian, and at least get L and C to stages where they're not plainly
incomplete (e.g. enough of L complete to cover the semantic range covered
by a normal-sized dictionary, and enough of C as could be described in
a 30,000 words encyclopedia article). The problem with not reckoning
(c) as a percentage is that small areas of the language could be worked
out in great and lengthy detail, but with much huger chunks left entirely
unworked-out; the result is as if you have pages 1735-1935 of a _Grammar
of X_, but no other pages, and this is qualitatively very different from
having the entire 50 pages of a _Grammar of Y_.

If (d) is a problem as an absolute measure it could be expressed as a
percentage of (c).

--And.