Re: Aelya question

On Sat, 22 Jan 2000, Aidan Grey wrote:

> [snip interesting stuff on preverbs]
>
>     I don't want to keep the preverbs in most cases (otherwise I'd end up
> with the problem of Mn. Irish, where a zillion verbs begin with t-  because
> of the old do- preverb), keeping them only when they're meaningful
> (differentiate several different verbs with the same root). In this case,
> I'm thinking to use the meanings of those preverbs to form derivational
> prefixes, or to use phrasal verbs (jump up to, e.g.). But, I do want to keep
> the "infixing" of objects. I thought I'd use a generic preverb (like the
> ever handy a-) for cases like this (instead of OI's ro-, too many uses of
> that one preverb!).
Lakhota has person/number marking which is infixing in a
particular class of verbs. Boas and Deloria (1941) speculate
that the material which preceded the p/n marker came from a set
of locative prefixes. Not exactly like what your proposing, but
similar enough to warrant a "go-for-it".

> I also like the idea of preverbs as aspect marking, and
> I really would love to incorporate the zillion markers various people on the
> list have used, and that are common in many Native American langs. BUT, I
> don't want a zillion marlkers, cause I'd never use them all - too complex.
I believe that the origin of the West Germanic ge- prefix for
past participles lies in the set of preverbs which Germanic
inherited from PIE. In fact, for a large class of verbs in
German, the past participle is formed without ge- because there
is already a prefix present on the stem. This is shown in the
following pairs of verbs in which the first member appears
without a prefix and takes the default ge- in the past
participle, and the second member takes a prefix, which
displaces the default ge- in the past participle.

1.      lehren 'to teach (someone)'
        gelehrt 'taught'

        belehren 'to teach (something)'
        belehrt 'taught'

2.      arbeiten 'to work'
        gearbeitet 'worked'

        verarbeiten 'to process'
        verarbeitet 'processed'

>     So, I guess what I'm asking is this/these:
>
>      Anyone else faced this problem? What did you do?
>     Any idea on how to incorporate preverbs but on a smaller scale, but
> still as useful?
>     Is my generic preverb idea any good? Anyone have better ideas?
I think the generic preverb is a nifty idea. As I showed above,
German has something similar with the past participle marker
ge-, presumably derived from an inherited IE preverb. If Aelya
is derived from Old Irish (also IE) a parallel development
doesn't seem far-fetched.

>     Second query:
>
>     I have another problem - I love morphology. So, I keep trying to add
> ooodles (3 o's is on purpose!) of neat stuff, like cases like Finnish,
> morphosyntactic qualities like Navaho, a gajillion (it's a technical term)
> deictics, and so on. On top of that, I'm never happy with the affixes or
> umlauts I assign. How do you folks deal with this - I can't be the only
> person here (and I know I don't have time to start ANOTHER lang - I'd like
> to finish this one!) with this problem?
Well, we certainly have different esthetics here. I like
morphology well enough, but my approach has been a bit more
modest with respect to the number of morphological markers I
include. I've found that after a while the morphology settles
down a bit into something I can live with all on its own. The
Shemspreg verbal paradigm I gave in an earlier post (what, you
didn't immdediately save and print out that message?!! :-) is
all you get in the active form of a verb. It's pretty lean
compared to the other PIE projects posted by Padraic and Andrew:
no person and number inflection, only past/non-past
distinctions, and aspect only in the past tense. Earlier drafts
had a much more traditional PIE feel to them with person/number
inflections and several different moods, etc. It just didn't
feel right to have all of that morphological baggage, so I
chucked it. No *logical* reason to do so, but it just didn't
fit. I also worked with several T/A marking schemes before
settling on the one I have now. That's how I solved that problem
for Shemspreg; probably not very helpful to you, though.

Dirk

--
Dirk Elzinga
dirk.elzinga@m.cc.utah.edu