You could set up a system of classifiers by which the nouns are given
categories based on the verbs they are associated with.  Example
(exclusive of tones):=20

=09ma-ringo-nara gua-dang-busho-er ma-dang-ring-luo; a ma-er-nara
=09gua-wo-dakdak-er ma-dang-busho-luo a.

=09The cue ball struck the object ball; which went into the
=09pocket.

=09dang-ring =3D cue ball (ball of driving)
=09dang-busho =3D object ball (ball of objectivity)
=09wo-dakdak =3D pocket, basket (place of nuts)
=09ringo =3D drive

=09ma- =3D wheel classifier: round objects, objects moving
=09cyclically
=09gua- =3D rock classifier: any object at rest

=09-luo =3D primary object relational
=09-er =3D secondary object relational
=09-nara =3D atemporal time (time unspecified)

Note that dang-busho changes from gua- to ma- (static to motive) when
it is now the subject of a verb of (circular) motion.=20

The blow by blow action:

[round class action]-action of striking-[atemporal] [at rest class
object]-objectball-[secondary relational] [round class
object]-cueball-[primary relational]; [open dependant clause] [round
class action]-action of going-[atemporal] [at rest class
object]-pocket-[secondary relational] [round class object]-objectball-
[primary relational] [close dependant clause]

Padraic.

On Wed, 10 Nov 1999, FFlores wrote:

>I have a question: how do nouns agree with verbs, besides
>the usual stuff (person and number)? Do you know of langs
>that use different pronouns for different kinds of verbs?
>Because I've discovered, to my dismay, that Nolt Lethris,
>the Old Tongue (ancestor of Drasel=E9q) seems to have this
>kind of agreement; pronouns (maybe clitic ones, but indeed
>free at some point) that include a consonant, as an affix or
>infix, which varies according to the verb. These pronouns
>are also marked for tense, I guess like Teonaht does, and
>in later history their position becomes fixed (postverbal)
>and even later they merge with the verb. For example:
>
>        ndar=E1s tuyerhe        (>> modern _drast=FCer_)
>        cross  2p.PST
>        'you crossed'
>
>The past tense is here marked by the preffix <-uy->; _t-erhe_
>is the pronoun root. But if the verb is of another declension:
>
>        tese buyerhe    (>> modern _tesb=FCer_)
>        lie  2p.PST
>        'you lied'
>
>Here the pronoun root is _b-erhe_. It's not just a matter of
>leaving the consonant on the verb, since the phonotactics doesn't
>allow it most times, and the positions could be reversed (tensed
>pronoun first, and then the verb). This characteristic consonant
>doesn't appear (at this stage) in finite forms of the verb (though
>it does appear in the infinitive and participles). It doesn't seem
>to have any semantic meaning (though it could have had one, in
>earlier times). Plus (forgot to mention) it appears not only as
>a prefix to the 'real' pronominal root (here, _erhe_), but also
>as an *infix* sometimes!
>
>What can I do? Help please!
>
>
>--Pablo Flores
>  http://draseleq.conlang.org/pablo-david/
>