On Wed, 28 Sep 2011 15:01:12 -0700, Padraic Brown <[log in to unmask]> wrote:

>On Wed, Sep 28, 2011 at 11:16 AM, R A Brown <[log in to unmask]> wrote:
>> I think the present thread suggests that things are simpler
>> if time is shown by the verb (and nouns are considered
>> present in relation to the time of the verb.
>
>Quite -- but I'd argue that this is mostly because we are used to
>having time shown by the verb rather than the noun. If we were native
>Tupi speaking conlangers, we might be arguing the exact opposite.

Certainly many would. But statistics is on our Anglophone side, on the side
of marking tense on the verb; and I don't think there's any mystery to that
fact. I agree with Ray that it's not at all an obvious move to dis-identify
the same agent at different points in time (e.g. we don't name them
differently; you're clept Padraic whatever time it is; similarly probably
even in Tupi). So actors in a event have no particular relevant temporal
boundedness, they continue existing for days or years to either side. But
the event itself (typically) has clear temporal boundedness. So it's
natural, I think, that the thing in the clause we mark for what time it
occupies is the thing for which that question has a contained answer.

Alex