`already'

[ucleaar <ucleaar@UCL.AC.UK>]

ivAn:
> I vote against a specific ZAhO for `already'.  It is not an aspectual
> operator; it is orthogonal to the ZAhO scale ({pu'o} appears to be the
> only ZAhO whose meaning is incompatible with the idea of alreadiness).

They are quasi-aspectual. What we wish to express is something like

 * starting sooner than expected
 * starting later than expected
 * continuing longer than expected

where for the last one Jorge has been using {zao}. I don't think that
that is quite what superfectivity is supposed to be though. So at
present I know of no decent way to render any of these three meanings.

> There's a lot of linguistic work on `already' and its cousins `still'
> and `finally' (I could provide references; there was a detailed study
> by J van den Auwera in _Linguistics and Philosophy_ about 2 years ago),
> and there's a lot of controversy, but people seem to agree that they
> characterise the event or state as taking place respectively sooner,
> longer and later than it might (than anticipated, than on some other
> occasion etc.) (ignoring a host of derived meanings).  I think we
> should be able to express such things by means of attitudinals.

[We mean "discursive", not "attitudinal" - attitudinals are things
like "Wow!", while discursives are things like "frankly" and
"unfortunately" (unless one of those is what indicators are).]

(a) Which UIs?
(b) UIs are the right solution iff they don't affect the truth
conditions of the bridi. If {koa krici kuau koe xao broda} means
  "koa believes koe brodas for longer than koa would expect" or
  "koa believes koe brodas for longer than I would expect", or
  "koa believes koe brodas for longer than one would expect"
then a UI is not the right thing. I.e. if koa does believe that koe
brodas, but koa doesn't believe koe brodas for longer than
expected, should {koa krici kuau koe xao broda} be false? If yes,
then {xao} should not be in UI. If no, then {xao} should be in UI.

Very tentatively, I conclude from this that they shouldn't be in UI,
reasoning woolily that if {sohi} and {sou} are "subjective" but do
affect truth conditions (as I *presume* they do), then {xao} & co.
shd behave likewise.

coo, mie And