Tolowa reflexives

[Gary Holton]

Tom,

Though I am unfamiliar Victor's original comment, the facts of Tolowa
agree well with those in Northern Athabascan (and Apachean).  For
example, Tanacross

n-Ek-'e~h
THM-1sg/L-see
'I see him/her/it'

'ede-n-Eg-'e~h
RFLX-THM-1sg/L,D-see
'I see myself'

or Dena'ina

tgh-esh-'ih
FUT-1sg-see
'I will see him/her/it'

hu-tgh-esh-t-'ih   (t + 'ih --> [t'ih])
RFLX-FUT-1sg-D-see
'I will see myself'

Hupa is the odd one out here in not employing the D classifier change:

'adi-w-cis
RFLX-1sg-see
'I see myself'

but even here the nominative subject is retained. I'd be interested to
hear of Athabascan examples where this is not the case.

Gary Holton




On Mar 8, 2005, at 12:21 PM, Tom Givon wrote:

>
> Dear FUNK people,
>
>
> A while back Vic Golla commented on my suggestion, in a previous note,
> that in Tolowa Athabascan reflexive clause, the agent retains the
> subject/nominative grammatical role. At the time I was away from my
> Tolowa files, so I could not respond properly, tho in private I
> conceded
> to Vic that I may have been in error.
>
> I have now had a chance to review my files on the subject, and to my
> great surprise I was right (it very seldom happens to me in an argument
> with Vic). Here are some examples that suggest the nominative status of
> the agent in the reflexive clause. This may, of course, be another
> grammatical innovation in Tolowa grammar, where the reflexive marker
> /dU/ is used in addition to the old de-transitive ('classifier') /D/.
> Tolowa may thus not reflect the pan-Athabaskan situation (of which Vic
> is infinitrely more knowledgeable than I am). My analysis is based on
> the one (most) reliable subjecthood criterion in Athabaskan verb
> morphology -- subject pronominal agreement. The data is taken from Dave
> Watters' paper on reflexives & reciprocals.
>
> For third persons, there is no real subject agreement, so the criterion
> cannot be used. The old yU-/bU- Athabaskan contrast has been
> re-analyzed, whereby yU- is just a transitivity marker. Thus  (U =
> schwa; lh = voiceless l; i~ = nasalized i):
>
>              yU-lh-ts'a's
>              TR-L-whip
>               's/he is whipping it/him/her'
>
>              dU-d-lh-ts'a's
>              REFL-D-L-whip
>              's/he is whipping him-/herself'
>
> For 1st or 2nd person subjects, however, the subject-agreement
> criterion
> is available:
>
>              'U-sh-k'Usr
>              THM-1s/SUBJ-shave
>              'I am shaving him/it'
>
>              dU-sh-d-k'Usr
>              REFL-1s/SUBJ-D-shave
>              'I am shaving myself'
>
>              na-sh-tlh-mi~sh
>              ADV-1s/SUBJ-L-hang
>              'I am hanging it/him/her'
>
>              naa-dU-sh-d-lh-mi~sh
>              ADV-REFL-1s/SUBJ-D-L-hang
>              'I am hanging myself'
>
>              ghee-s-ii-'i~'
>              THM-PFV-1s/SUBJ-see
>              'I saw it/him/her'
>
>              dU-ghee-sU-s-d-'i~'
>              REFL-THM-PFV-1s/SUBJ-D-see
>              'I saw myself'
>
>              naa-'ii~-tlh-te
>              ADV-2s/SUBJ-L-care
>              'You care for it/him/her'
>
>              naa-d-ii~-d-lh-te
>              ADV-REFL-2s/SUBJ-D-L-care
>              'You care for yourself'
>
> However uncharacteristic Tolowa may be of the rest of Athabascan (we
> know it has done much re-structuring in other areas of the grammar),
> the
> agent in its reflexive clause certainly abides by the much more common
> cross-linguistic pattern, whereby it remains the nominative/subject.
>
> Cheers,  TG
>
>