response to nick on dikyjvo

[Logical Language Group <cbmvax!uunet!pucc.princeton.edu!lojbab>]

Nick proposes:
>y should be allowed within le'avla. without that, it is impractical to allow
>any lujvo to be a le'avla prefix.

Can't happen.  Morphology breaks down.  Indeed, we are having problems
with the proposed le'avla-lujvo algorithm using ",iy," between units.
Nora just drove a couple of trucks through the morphology algorithm that
incorporated this proposal.  We can probably salvage most of this idea,
but there will likely be some complicated restrictions.

The breath of permitted forms for le'avla cannot cross into lujvo space,
and there is no algorithmic way we know of to prevent this if 'y' is
permitted.  Indeed, the "iy" le'avla lujvo hyphen proposal came about
because 'y' can't even appear in a lujvo also containing a le'avla without
it sometimes breaking down into rafsi.  (e.g. protonynukli)

If there is anyone manic enough to want to work on the morphology algorithm,
volunteers are welcome.  You may recall that I posted the proposal in early
July.  Alas, it still doesn;t work.

Three choices:
1) don't use lujvo as prefixes - I certainly don't intend to - the words are
   long enough as it is.
2) use a vocalic consonant in place of a lujvo 'y'
3) jam the prefix components together in some other way - since the
   prefixing rules are not as regimented as the lujvo-making rules, but
   more like names.

I have no comments on Nick's dikyjvo proposals, other than that there is
nothing wrong with them.  I don't expect to support any rules for
dikyjvo prior to the 5 year usage period, but I also won't stop others
who choose to use them (as if I could), provided that they recognize
that not everyone will do so.

lojbab


P.S. I may discuss the morphology problem in a couple of days - we already
think we have a solution, even since I wrote this message.

lojbab@grebyn.com