Re: THEORY: Tenses (was: Re: THEORY: ... Auxiliaries...)

In a message dated 7/10/2005 12:49:42 PM Eastern Daylight Time,
ray.brown@FREEUK.COM [quoting me] writes:

>> He [Comrie] says the record for distinct past tenses is 6 or 7 (his source
is
>> lamentably unclear) in Kiksht (a Chinookan language of the northwestern
>> USA).
>Does he?
Yes he does.

>Is he using 'tense' in the strict meaning 'correlating directly
>with distinctions of time', that is no aspect (or mood) is involved?
Yes, he is.

>If  so, I am a little skeptical of these figures. Do we have any details?
His description is essentially the same as the one that butsuri@myrealbox.com
has kindly supplied.  There are 4 tense prefixes covering various degrees of
remoteness in the past, and either 2 or 3 of them can be further broken down
into earlier & later segments using the prefixes u- and t-, giving a total of
either 6 or 7 past tenses.  Comrie lists the forms & meaning as follows:

ga(l) . . . u- = remote past
ga(l) . . . t- = from one to ten years ago
ni(g) . . . u- = from a week to a year ago
ni(g) . . . t- = last week
na(l) -        = yesterday or the preceding couple of days
i(g)-          = earlier today

The last may or may not be divisible into

i(g) . . . u- = earlier on today, but not just now
i(g) . . . t- = just now

which would bring the total to 7 past tenses.

That about exhausts my knowledge of the subject.  I have no independent
confirmation of what he says.

Doug