non-clausal ke'a
tijlan
[.i] mi ke'a citka
I suppose it's syntactically valid. But what relativised sumti could
it refer to? Would it make more sense if it followed another sentence:
mi viska lo plise .i mi ke'a citka
Could we say that this {mi ke'a citka} is a sentential expansion of a
clause that could describe {lo plise}, as in this:
mi viska lo plise poi mi ke'a citka
I'm asking this as I've been thinking about the difference between
"le/la/les" and "en" in French:
Je vois une pomme. Je la mange.
(I see an apple. I eat it.)
mi viska pa plise .i mi ri citka
Je vois des pommes. Je les mange.
(I see apples. I eat them.)
mi viska su'o plise .i mi ri citka
Je vois des pommes. J'en mange.
(I see apples. I eat "of them".)
mi viska su'o plise .i mi ___ citka
http://en.wiktionary.org/wiki/en#French
http://en.wikipedia.org/wiki/French_personal_pronouns#The_pronoun_en
Would {ri/ra/ru} alone be an accurate translation of "en", which
differs from "le/la/les" in the referent's quantity (i.e. whereas "Je
la mange." means that the whole of the object is eaten, "J'en mange."
doesn't imply such an entirety)? My guess is that {su'o ri}, that is
{su'o su'o plise}, would be more accurate than a bare {ri}. But I also
vaguely feel that "en" might have something to do more with
relativising "ke'a" than with back-counting "ri/ra/ru" (or the lerfu
solution, for that matter -- "py" in this case).
Jorge Llambías
> What could this mean:
>
> [.i] mi ke'a citka
>
> I suppose it's syntactically valid. But what relativised sumti could
> it refer to? Would it make more sense if it followed another sentence:
>
> mi viska lo plise .i mi ke'a citka
>
> Could we say that this {mi ke'a citka} is a sentential expansion of a
> clause that could describe {lo plise}, as in this:
>
> mi viska lo plise poi mi ke'a citka
I don't like overloading "ke'a" like that, because the two
interpretations of "ke'a" could come into conflict:
mi pensi da poi ge ke'a cpana lo jubme gi mi ke'a ba citka
Does that mean that I will be eating the table, or thinking about
what's on the table and will be eating that?
> I'm asking this as I've been thinking about the difference between
> "le/la/les" and "en" in French:
>
> Je vois une pomme. Je la mange.
> (I see an apple. I eat it.)
> mi viska pa plise .i mi ri citka
>
> Je vois des pommes. Je les mange.
> (I see apples. I eat them.)
> mi viska su'o plise .i mi ri citka
>
> Je vois des pommes. J'en mange.
> (I see apples. I eat "of them".)
> mi viska su'o plise .i mi ___ citka
>
> http://en.wiktionary.org/wiki/en#French
> http://en.wikipedia.org/wiki/French_personal_pronouns#The_pronoun_en
>
> Would {ri/ra/ru} alone be an accurate translation of "en", which
> differs from "le/la/les" in the referent's quantity (i.e. whereas "Je
> la mange." means that the whole of the object is eaten, "J'en mange."
> doesn't imply such an entirety)? My guess is that {su'o ri}, that is
> {su'o su'o plise}, would be more accurate than a bare {ri}. But I also
> vaguely feel that "en" might have something to do more with
> relativising "ke'a" than with back-counting "ri/ra/ru" (or the lerfu
> solution, for that matter -- "py" in this case).
This is the usual problem of mixing pro-sumti with quantifiers. A
quantifier has a scope, and when you use a pro-sumti to pick something
bound by a quantifier from outside the scope of the quantifier,
trouble follows. This is called a "donkey pronoun", and to find out
why you can start by reading:
http://en.wikipedia.org/wiki/Donkey_pronoun
To see this more clearly, expand the "su'o plise" to a properly formed
logical expression:
[su'o da poi plise zo'u (mi viska da)] .i mi citka ri
"ri" is outside the scope of "su'o", and yet it points to the bound
variable "da". What can that possibly mean? From a strictly logical
point of view, it is nonsense, a bound variable doesn't have a
referent that "ri" could pick from it. The bound variable is just a
place holder for all the things in the domain of quantification, in
this case all the apples.
In this particular case, you may say that it's obvious that "ri"
should pick just those members of the domain of quantificationj that
make "mi viska da" true, but if the context is slightly more complex,
this won't always work.
My own opinion is that in "mi viska su'o lo plise .i mi citka ri",
"ri" points to "lo plise", not to "su'o lo plise" which is not
logically well defined.
mu'o mi'e xorxes
John E Clifford
discourse analysis) would be that 'ri' picks up 'le su'o lo plise' or some such
construction, which does not actually occur but refers to the right things,
i.e., the apples on the table in this case.
--
You received this message because you are subscribed to the Google Groups
"lojban" group.
To post to this group, send email to loj...@googlegroups.com.
To unsubscribe from this group, send email to
lojban+un...@googlegroups.com.
For more options, visit this group at
http://groups.google.com/group/lojban?hl=en.
Jorge Llambías
> The other possibility, which also has some support in logic (or at least
> discourse analysis) would be that 'ri' picks up 'le su'o lo plise' or some such
> construction, which does not actually occur but refers to the right things,
> i.e., the apples on the table in this case.
Yes, that's what I meant by:
> In this particular case, you may say that it's obvious that "ri"
> should pick just those members of the domain of quantificationj that
> make "mi viska da" true, but if the context is slightly more complex,
> this won't always work.
The most obvious case where it doesn't work is when the quantifier is
"no" instead of "su'o", since "ri" can hardly be expected to pick up
"lo no lo plise". In:
mi viska no lo plise .i ki'u bo ri se mipri
"I saw none of the apples, because they were hidden."
I expect "ri" to pick "the apples", not "the none of the apples that I saw".
But that's not the only problematic case. If "su'o" is in turn within
the scope of some other quantifier, there can be trouble too.
I guess pragmatically it is inevitable that people will use "ri" and
such to point to "(lo) su'o..." even when there is no actual "lo" in
the antecedent, but I prefer the rule to be that "ri" looks for an
actual sumti, and any pragmatic deviation from that is just a
pragmatic deviation.
John E Clifford
approach would be just to extent the scope of the 'da'. but again, the 'no' case
rather cuts into that, and, indeed, reference back to a no case seems always to
be peculiar -- or to point to the more general suggestion you make, which seems
to cover all the possibilities (until we consider other quantifiers with
overlapping scopes).
----- Original Message ----
From: Jorge Llambías <jjlla...@gmail.com>
To: loj...@googlegroups.com
Sent: Sat, November 20, 2010 12:01:20 PM
Subject: Re: [lojban] non-clausal ke'a
--
Jorge Llambías
> I was going to note that, in the expanded form, though not the original, another
> approach would be just to extent the scope of the 'da'. but again, the 'no' case
> rather cuts into that, and, indeed, reference back to a no case seems always to
> be peculiar -- or to point to the more general suggestion you make, which seems
> to cover all the possibilities (until we consider other quantifiers with
> overlapping scopes).
When overlapping quantifiers are involved, the simple rule of picking
just the sumti also makes the most sense:
ro lo prenu cu citka su'o lo plise .i ri kukte
"Everyone ate at least one of the apples. They were delicious."
If "ri" just points to "lo plise", there's not much of a problem.
The alternative here is that it points to "lo ro lo plise poi se citka
su'o lo prenu", but to get to that you have to apply some major
transformations to the sentence as given. For each combination of
quantifiers you would need to come up with a specific rule.
Pierre Abbat
> What could this mean:
>
> [.i] mi ke'a citka
It's syntactically valid, but it's nonsense, just like "du ra'o", "te.u",
and "pe naku". It's like saying "Which I eat.", except that "which"
corresponds to "poi ... ke'a" together. (Hebrew and Modern Greek have a word
like "poi", but no special pronoun like "ke'a".)
> I'm asking this as I've been thinking about the difference between
> "le/la/les" and "en" in French:
>
> Je vois une pomme. Je la mange.
> (I see an apple. I eat it.)
> mi viska pa plise .i mi ri citka
>
> Je vois des pommes. Je les mange.
> (I see apples. I eat them.)
> mi viska su'o plise .i mi ri citka
>
> Je vois des pommes. J'en mange.
> (I see apples. I eat "of them".)
> mi viska su'o plise .i mi ___ citka
>
> http://en.wiktionary.org/wiki/en#French
> http://en.wikipedia.org/wiki/French_personal_pronouns#The_pronoun_en
>
> Would {ri/ra/ru} alone be an accurate translation of "en", which
> differs from "le/la/les" in the referent's quantity (i.e. whereas "Je
> la mange." means that the whole of the object is eaten, "J'en mange."
> doesn't imply such an entirety)? My guess is that {su'o ri}, that is
> {su'o su'o plise}, would be more accurate than a bare {ri}. But I also
> vaguely feel that "en" might have something to do more with
> relativising "ke'a" than with back-counting "ri/ra/ru" (or the lerfu
> solution, for that matter -- "py" in this case).
I'd say "mi su'o ri citka", but this pronoun is peculiar to French, Occitan (I
think), and Catalan and is bound to cause some trouble in translating.
Pierre
--
When a barnacle settles down, its brain disintegrates.
Já não percebe nada, já não percebe nada.