Re: semantic roles

(back from 2 weeks off. i can't read all the mails and picked that one=20
randomly.)=20

Dans un courrier dat=E9 du 16/06/99 07:14:29, Nik Taylor a =E9crit :

> I'd just like to share some verb types from Watakassi', tendencies in
>  case assignment:
> =20
>  "Normal"
>  Agent: Ergative
>  Patient/subject[of intransitive verb]: Absolutive
> =20
>  Experiencer-subject (sensory/emotion verbs)
>    Experiencer: Ergative
>    Object: Absolutive
>    Indicates control by experiencer
>  Experiencer-Object
>    Experiencer: Dative
>    Object: Absolutive
>    Experiencer has little or no control
>  These can both be used with the same verb, for instance:
>    Flassi'tas naDtya'nal pikatu' (Saw John-erg stone =3D John looked at
>  the     stone)
>    Flassi'tas naDtya'naz pikatu' (Saw John-dat stone =3D John saw the
>  stone)
>  Experiencer-Equal
>    Experiencer: Absolutive
>    Object: Commitative
>    Most common with "state" verbs, like "love"
>  Mental Action I
>    Person: Absolutive
>    Object: Dative
>    Common with certain emotions
>  Mental Action II
>    Person: Ergative
>    Object: Absolutive
>    Most common with rational mental actions
>  Mental Action III
>    Person: Dative
>    Object: Absolutive
>    Suggest that the person is affected by the mental action
>  Learn and study are indicated by the same verb, _pyasu'n_, in the sense
>  "learn", MA-II is used, while MA-I is used for "study"
>  Believe and have faith are also the same verb, _tya'iya_, MA-II is used
>  of "believe", while MA-III is used for "have faith in"
i was very glad to read this because i have pondered the same issue for a=20
long time. i can see we solved it in different ways but still i know i'm not=20
the only one :-)

mathias