applicative (again) (was: Re: Voices)

Steven Williams wrote:
> Another idea, quite-thoroughly discussed about a month
> ago, is the applicative voice. Basically, it works
> like this:
>
> Active voice:
> "The cat [ACT] slept on the mat [OBL]."
>
>..... To make it applicative, you'd
> do something like this:
>
> Applicative voice:
> "The cat [ACT] (on)slept the mat [PAT]."
>
> Where the cat is still the actor, but the mat is
> raised to the status of the patient, and if you like,
> you can indicate the exact role the former oblique
> phrase took as a verbal adfix (which is why I put 'on'
> in parentheses there).
>
> English has some constructions that look suspiciously
> applicative,
Yes, things like: to water (the plants), to cage (a bird), to pocket (a
coin) etc. basically "put X in/on NOUN"

>... but I can't think of any
> natlangs offhand that have the applicative, though
> I've heard that it's fairly common at any rate.
Indonesian does this with its verbal suffix -i:
air 'water' > meng/air/i 'to irrigate s.t.'
minyak 'oil' > me/minyak/i 'to oil s.t.'
kurang 'less' > mengurang/i 'to reduce s.t.' (however, there's
near-synonymous mengurang/kan 'to reduce s.t.'-- The exs. in the dictionary
don't make clear what the difference is, and I'm no longer sure either...if
there is a difference.

Note that in Indonesian, just as in Engl., "skin ~to skin" (kulit
~menguliti) means "_remove_ the skin from...".  Peculiar.

The meanings sometime change rather radically: datang 'come' but
men/datang-i 'to attack'.

Our sleeping cat ex. works nicely in Indo. too; the (a) exs. are SV+PrepP,
the (b)s are SVO--
1a. kucing tidur di kursi 'the cat slept in/on the chair'
1b. kucing meniduri kursi 'the cat slept-in/on the chair'  OR

2a. Ali tidur di tikar 'Ali slept on a mat'
2b. Ali meniduri tikar 'Ali slept-on a mat'

Problem: my non-native Sprachgefühl says that 1a simply states a fact, while
1b suggests that the cat shouldn't have been in the chair (though 1a could
also imply that). But conversely, 2a suggests either habitual or one-time
occurrence, but 2b suggests only _habitual_.  I'm not at all sure of this,
however-- but it could be an interesting distinction to make in a conlang.

The main use of the -i forms (since they're transitive) is in relative
clauses where a passive may be required:

Kursi mana yang ditiduri kucing? 'Which chair (is it) that is/was-slept-in
(by) the cat?
Tikar yang ditiduri Ali, kurang bersih 'The mat that is/was-slept-on (by)
Ali is/was not very clean.'

In these cases, only context will tell whether it's habitual/one-time,
permitted/non-permitted action. (Of course an auxiliary verb would clear
things up-- suka tidur di../meniduri... 'likes to sleep-on...', tak mesti
tidur di.../meniduri... 'mustn't sleep on...')