Sent from my iPad
On Sep 11, 2012, at 9:06 AM, And Rosta <and.rosta@hidden.email> wrote:
Jorge Llambías, On 09/09/2012 02:57:
> On Sat, Sep 8, 2012 at 9:36 PM, And Rosta<and.rosta@hidden.email>
> wrote:
>>
>> As I wrote in the other thread:
>>
>>> There's a slightly different quantifier that I'd been meaning to
>>> mention, but have kept on forgetting. It's an existential
>>> quantifer with scope over the whole discourse:
>>>
>>> "Once upon a time, there was a poor woodcutter. He lived in a
>>> hut with his dutiful daughter."
>>>
>>> In Livagian I treat it as a separate quantifier in its own
>>> right. (Livagian doesn't have Xorban's lV.)
>>
>> How to do this in Xorban? Just "la [poor]a [woodcutter]a"? -- No,
>> tthat's "The poor are woodcutters"? Maybe "sa
>> [hereby-newly-introduced-into-discourse]a je [poor]a
>> [woodcutter]a"?
>
> There seem to be two issues here: how to introduce it, and how to
> carry it over to other sentences. For the first part, I don't see
> any problem with plain s-:
>
> la je drna cdra se je pnde mdrktne cbneka Distant era-A, some-E
> poor(E)& woodcutter(E): at(E,A)
I agree this does the job, in the same way that "It is not the case that every woodcutter was not poor" would, so you're claiming the world contains at least one poor woodcutter, and then subsequent definites can be presumed to be picking out a certain one of those poor woordcutters.
Discussion with John has brought me round to the view that every quantifier can project into the UoD some sort of 'personified' individualized thingo whose interpretation would be that of the variable bound by the quantifier if the quantifier had scope over the rest of the discourse.
Dear Lord, what could I have said to suggest that. I just said that an initial quantifier picks out a thing ( or bunch of things) that remain the same through it's scope -- which is what particular quantifiers do. No projection, just selection, no personified (whatever that is), just the same thing referred to over and over.
I haven't yet thought through exactly what's going on with this, tho. It seems to me that "Two out of every five people will live to 100" can project into the UoD (i) the one who lives to 100; (ii) the lucky pair who out of every fivesome are the ones who live to 100"; (iii) the 20% of people that will live to 100.
This is the worst sort of statistical fallacy. That _expression_ is not a quantifier one and is self contradictory if taken as one.
> Now as long as you keep everything within the same sentence, you can
> keep using "e":
>
> la je drna cdra se je pnde mdrktne je cbneka je li je dtfli txnike
> lo gnmokeki lu cmlzdnu xbjoku Distant era-A: some-E poor(E)&
> woodcutter(E): at(E,A)& (dutiful(I)& daughter(I,E))-I:
> they(O,E,I)-O: hut-U: lived-in(O,U)
>
> But this can get painful soon. We don't have a way to quantify over
> multiple unconnected sentences.
Mike has now proposed one, a unary operator ("no", say) whose complement is a formula and extends the binding on the variable indefinitely (until rebound, I guess).
It is probably easier to take (ala Kamp) that quantifiers have indefinite scope to begin with and sort out the logic later (it's what Lojban does except for sorting out the logic). If you go sentence by sentence you inevitably lose continuity, since each s is a whole new selection officially.
> If you use the implicit binding rule you could keep using "e" with
> "le je pnde mdrktne" restriction and recover it in a new sentence:
>
> la je drna cdra se je pnde mdrktne cbneka Distant era-A, some-E
> poor(E)& woodcutter(E): at(E,A)
>
> li je dtfli txnike lo gnmokeki lu cmlzdnu xbjoku (dutiful(I)&
> daughter(I,E))-I: they(O,E,I)-O: hut-U: lived-in(O,U)
>
> where the binding of each E in the second sentence is an implicit
> "le je pnde mdrktne".
If the first sentence were "si pnde mdrktne" then the implicit restriction rule wouldn't work, but Mike's version would.
The simplest solutions would be (i) to adopt Mike's "no", or (ii) to designate a series of vocalic morphemes to act as variables that preserve their values until rebound, or (iii) to make this the interpretation of all unbound morphemes.
i see the following problems with these:
(i) There's no mnemonic quality to the variable names. E.g. the form "ua" is no help in remembering that this is the poor woodcutter.
(ii) It feels to me as though these variables projected as individuals into the UoD become ontologically the same as other individuals. We treat other individuals as predicates, e.g. "mV xrxe".
(iii) Option (iii) suffers from the problem of wasteful use of expensive memory.
A solution to take account of these problems might be to be to have a counterpart of mV that assigns names -- maybe an illocutionary applied to mV -- so "la za bbi" or "la co ma bbi" is "that which I hereby nonce-name _bbi_", possibly supplemented by a version of mV (e.g. "da ma bbi") that indicates that the name is a nonce name assigned in the discourse rather than a name from the permanent onomasticon.
I'm thinking aloud here, not making any definite proposals.
--And.
The memory problem is a real one for a language and no solution is ideal. You seem to have boxed yourselves in a bit with you morphology at the beginning. Of course, that is in aid of decomposability, but perhaps that can be obtained in other ways one more of the structure is in place. |