[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [lojban] Re: A (rather long) discussion of {all}
On 6/4/06, Jorge Llambías <jjllambias@gmail.com> wrote:
On 6/3/06, Maxim Katcharov <maxim.katcharov@gmail.com> wrote:
> So each predicate is marked for distributivity/non-distributivity
> (bunch-indiv/bunch-together)?
Each of a predicate's argument places _can_ be marked for it. It is
not always marked, in the same way that tense is not always marked.
It wouldn't get marked in the same way as tense. Tense marks the
entire bridi, or each sumti, and not the predicate place. It also uses
very specific words for it, as opposed to the trick that you want to
use with the outer quantifier (which, by the way, disallows "only 10
of the group surrounded the building and wore hats").
Regardless, this isn't the explanation that I've been repeatedly asking for:
Bunch, individually: We are not treating Alice this way, so this does
not apply. (If this were the hat example: Alice, herself, wears one+
of the bunch of hats implied by the blank inner of {lo mapku})
Bunch, together (but not in the sense of mass or group): Alice's
relationship to the surrounding of the building is ???
What is the difference between the latter two relationships?
For example, if it is indeed similar to tense, I'll offer some rough
example explanations that show the differences between pu/ca/ba when
they are used within the sumti:
{pu}: an entity that at some point in the past was running, but may or
may not be running now or in the future. {ca}: an entity that is
running now. {ba}: an entity that at some point in the future will
run, but my or may not be running now or have run in the past. The
difference between the three is that {pu} will not refer to any entity
that was /not/ running in the past, ca will not refer to an entity
that is not runnig now, and ba will not refer to an entity that is not
running in the future. This means that in {lo ca tadni cu sruri lo
dinju}, Alice is definitely currently a student, if she is to be a
referent of {lo ca tadni}.
> What is {lo tadni cu bebna} marked for?
It isn't marked, but it could be marked for distributivity.
Ok, what is the difference between the two markings of distributivity,
especially from the perspective of Alice's relationship to the
foolishness?
> How about {lo tadni cu sruri
> lo skori}? What if they're playing tug-of-war? What if they're
> standing around a rope looking up at the person climbing it? What if
> several paths surround the building, and we're talking of them?
Indeed there are many situations that could be described by that
sentence. What if it's happening now? What if it happened two weeks
ago? etc. etc.
Right.
> No, "surrounded the building" is not inherently collective, and
> neither is any other selbri or sumti-slot.
Right. And "sumti-slot" is the right thing to say, because it is the
slot that should logically get the mark, not what fills the slot.
No other part of language works this way, and with good reason. But
this is putting the cart before the horse.
{dasni} is not inherently distributive either. For example in "the students
wore the hats", we would not normally want full distributivity for both
arguments, which would give "each of the students wears each of the
hats", we only want "respective" distributivity: "the students wore their
respective hats", "each student wore their own hat".
It's interesting to note that while Lojban has gadri corresponding to the
{joi}-connective, it has no gadri corresponding to the {fa'u}-connective, so
to get the "respectively" reading fully explicited you have to duplicate
the sentence:
ro le tadni cu dasni pa le mapku ije ro re mapku cu se dasni pa le tadni
ro ri mapku, yes.
Of course {le tadni cu dasni le mapku} will be clear enough in most contexts,
because we know that hats are normally worn by one person only (at
a given time) and that a person normally wears just one hat, but that comes
from our general knowledge of how the world is, not from what the sentence
states.
> > There is no difference in the referring expression. The difference is in
> > the predicate.
>
> Predicates don't have default distributivity/non-distributivity.
I agree.
I am, of course, referring to your"predicates", and not
my"predicates". My predicates don't need to be marked, because they
simply don't behave in this (strange) way.