* Saturday, 2014-10-18 at 13:12 -0300 - Jorge Llambías <jjllambias@gmail.com>:
> On Fri, Oct 17, 2014 at 9:45 PM, Martin Bays <mbays@sdf.org> wrote:
> > I don't see it that way at all. It's the same rule. It's just that the
> > rule is applying to the implicit sub-bridi of the description clause; in
> > general it applies to the immediately enclosing bridi (just like any
> > other bridi operator, e.g. tenses).
> >
> > (I'm not particularly happy with this "bridi" and "sub-bridi"
> > terminology, btw, though I hope you can see what I mean by it; feel free
> > to suggest alternatives)
>
> Well, we know for sure how the operator "ko'a .e ko'e" operates when it
> occupies the place of an argument of an atomic predicate: It outputs the
> conjuntion of the two atomic formulas created by using "ko'a" and "ko'e"
> respectively as the argument of the atomic predicate.
(Since we were talking about terminology, let me be picky: it isn't
just a matter of atomic formulas, since e.g. in {ko'a .e ko'e ro da
broda} ~= {ge ro da zo'u ko'a da broda gi ro da zo'u ko'e da broda}, the
formulae being conjoined aren't atomic.)
> When "ko'a .e ko'e" occupies some other place (e.g. the argument of a
> function LAhE) we cannot apply the above rule directly. We need to either
> say how "ko'a .e ko'e" operates when it occupies the argument of an atomic
> function (which to me implies a new rule, even if the rule is very similar
> to the rule for when it occupies the place of an atomic predicate
While, if you'll excuse me reiterating, I would say that given the
profusion of logical connectives and other bridi operators throughout
the language, it's better to give abstract rules which cover many cases
uniformly rather than think in terms of specific rules for each
situation; so in this case I would explain the situation with two rules:
(i) a bridi operator operates on the bridi immediately under
construction, i.e. the lowest in the tree;
(ii) a logical connective acts as a bridi operator, connecting the
two formulae obtained by substituting each connectand in turn
for the connected element.
> ) or we can say that there are no atomic functions, all functions are
> in fact built out of predicates. In this case, we just need to express
> the function in terms of a predicate, and then apply the rule we
> already have for "ko'a .e ko'e" as a pseudo-argument of an atomic
> predicate.
If I understand correctly, (i) and (ii) also explain these semantics;
the only difference is in whether LAhE introduces a bridi of its own.
> > So it does... I guess to get the "lo (re lo mlatu .e ci lo gerku)", you
> > have to use
> > lo tu'o boi re lo mlatu .e ci lo gerku
> > ?
>
> I suppose. I guess with ".a" it would make more sense: "lo po'i ga re mei
> gi'e me lo mlatu gi ci mei gi'e me lo gerku".
Hmm... hold on, in {lo tu'o boi re lo mlatu}, we have to parse {re lo
mlatu} first as {re da poi me lo mlatu}. So I don't think any remei are
really involved.
If we assume that {lo [quantifier]} does introduce a sub-bridi, I guess
I'd expect
lo tu'o boi re lo mlatu .a ci lo gerku
-> lo poi'i me re lo mlatu .a ci lo gerku
-> lo poi'i ga me re lo mlatu gi me ci lo gerku
-> lo poi'i ga re da poi me lo mlatu zo'u ke'a me da gi ci da poi me lo
gerku zo'u ke'a me da
(which is a rather bizarre thing to refer to!).
> > > I have no idea what the rules for mekso are in detail, so I don't know
> > > whether for example "li na'u sinso mo'e ko'a .e ko'e" is supposed to be
> > > equivalent to "lo sinso be ko'a .e ko'e" or to "li na'u sinso mo'e ko'a
> > > lo'o .e li na'u sinso mo'e ko'e".
> >
> > What is the operand {mo'e ko'a .e ko'e} such that sine of it is
> > something which is sine of both ko'a and ko'e?
>
> Is "mo'e ko'a .e ko'e" an operand, or just a pseudo-operand
> (morphologically an operand, but logically not an operand, like "ko'a .e
> ko'e" is morphologically a sumti, but logically not a sumti)?
>
> We cannot claim that the rule for operand-3 always returns true logical
> operands, in the way that we could claim that sumti-6 (almost) always
> returns true logical sumti. "gek operand gik operand-3" is an operand-3,
> for example.
I would say that 'sumti' always returns a term, and so does 'operand',
where "term" is meant in the sense of logic, and where "returning" is
something that only makes sense in the context of the full parsing
algorithm, but is such that e.g. {ko'a .e ko'e} returns what ko'a is
bound to to one branch, while returning what ko'e is bound to to the
other branch - where these "branches" are what are created by the
semantics of logical connectives as in (ii) above. Similarly, {ro da}
returns the variable quantified over by the quantifier it introduces.
> So, I would not want to insist that "mo'e ko'a .e ko'e" is a logical
> operand, and indeed in my example I was not taking it as one. Since
> sumti and operands are logically the same thing, I was just ignoring
> "mo'e" and "li", and the question was whether "ko'a .e ko'e" operates
> on the predicate "sinso" or on the predicate containing the li-clause
> as its argument.
OK; so in other words you're taking {mo'e} to be transparent, i.e. not
to introduce a sub-bridi, but for the production 'mex' to be opaque and
to introduce a sub-bridi corresponding to the predicate "is the result
of applying [operator] to [operands]"?
So this handling of {mo'e ko'a .e ko'e} involves a "new rule"?
> > > > > Do we even know what "na ku zo'u broda .i bo brode" means?
> > > > I believe it's the same as "na ku ge broda gi brode".
> > > I suppose that's one choice, although it's not a necessity that
> > > juxtaposition be equated with conjunction, or that the negation of two
> > > separate propositions has to be the negation of their conjunction.
> > Is there a plausible alternative?
>
> Well, I'm not advocating this but we know that "na ku zo'u broda" expresses
> the negation of the proposition expressed by "broda", so it would not be
> out of the question to say that "na ku zo'u tu'e broda .i brode tu'u"
> expresses the negations of both propositions, rather than just the negation
> of their conjunction. The operator "na ku zo'u" is only well defined when
> applied to a single proposition. When applied to multiple propositions at
> once, who knows how we want it to act. Maybe it should be distributive.
I guess it could! Yet another thing to be decided. Do you think it might
be a bad idea to decide that it's equivalent to "na ku go broda gi
brode"?
Attachment:
signature.asc
Description: Digital signature