Re: [lojban] Re: tersmu 0.2

I am still puzzled by what this stream is all about, but I can comment on a few peripheral (I think) topics.

I apologize for "veridical". I expect it is mine, since I can't find it in Brown and it has the smell of exposure to Philosophy about it. It is, then, just another case of Logjam taking over a technical term and misusing it to death (cf. pseudo sciences generally). It does indeed originally apply to propositions and is here used for descriptions in place of "accurate" or some such. Of course, descriptions (in Logic, at least) are meant to be accurate and so there is no point in mentioning that fact. But Logjam thought it had a description with {le} which could function even if not accurate, so we needed to distinguish the normal description somehow. Of course, it turns out {le} phrases aren't logically descriptions, so we can drop the whole thing.

{zo'e noi broda cu brodu} = {zo'e ge broda gi brodu} is almost correct logically though dubious linguistically, which would prefer {zo'e ge brodu gi broda} since {zo'e} is defined on the head predicate, not the peripheral. It is, in fact, the final stage in a process of talking about what to put in for places on a predicate that one doesn't want to fill. The first solution was just to leave them blank and cover the logical problem with that by understanding a short-scope particular quantifier, Griceanly, though, this implicated that either you did not know or did not care what filled that slot. While this was often true, it also was often not, though the reasons for omitting a reference were varied: it was obvious from context or impolitic to mention, for example. So, Griceanly again, we stuck in a marker that said we did know and care but weren't saying. But that seemed paradoxical, so we dropped that clause and came down to what we have now “refers to an object that makes the bridi true” But, in fact, it is usually not the case that just any thing will do, but rather that some salient one is required or, at least, intended, so {zo'e} as bearer of this role, has this feature as well (and this is crucial, though ill-used, in a further feature to be discussed, even though not mentioned). In short, logically, {zo'e} stands for a generalized Skolem function dependent the predicate place and all the other arguments to that predicate and yields the salient object(s) that fit in that place in the given context. Briefly, {zo'e broda} = {lo broda cu broda} (where {broda} includes the relevant information).

This means that the “definition” floating around for {lo broda} is, at best, circular. But, in fact it is not that good. In the first place, {zo'e} is only defined when filling a place and in the definition it is used separate from any predicate (shades of the independent use of Russell's description and the messes that causes). So, insert the definition into its proper place: {lo broda cu brode} = {ze'a noi broda cu brode}. This latter is a tautology, since {ze'a} always refers to a thing which makes the predication true. But the first is not a tautology (except when the head predicate is {broda} again). The definiens has the matter bass ackwards, making the essential property of the referent an incidental feature and the incidental the essential (the use of {noi} rather than {poi} for the essential feature emphasizes this inversion). As a result, the meaning of {lo broda} so defined changes with each use, emphasizing that it is definition in terms of the incidental not the essential. (I pass over the question of scope here, since the problems there – and the variety of solutions – are too well known and only arise if this definition is accepted, which I suppose it should not be,) This is a wormrunner theory, ignotum per ignotius: the only ground for using {zo'e} is salience, which is there implicitly from history but not there in fact and what is in its place is a mystery.

Finally, a practical note. Whatever the point of all this is, dealing with MEX now is bad idea. Given that the creators of Lojban don't know how MEX works and, to a great extent, don't even know how it should work, trying to explain it now is a thankless task foreddomed to failure. Eventually, when you have a successful theory (of whatever sort you are working on) for the core of Lojban, you will be in a position to apply that theory to MEX to explain how parts of it work and to criticize and guide the development of other parts. Hopefully. But not now and not yet as a part of e\developing that theory.

On Wednesday, October 8, 2014 2:40 AM, And Rosta <and.rosta@gmail.com> wrote:

Jorge Llambías, On 07/10/2014 21:59:
> On Tue, Oct 7, 2014 at 11:00 AM, And Rosta <and.rosta@gmail.com <mailto:and.rosta@gmail.com>> wrote:
> Jorge Llambías, On 06/10/2014 23:10:
> On Mon, Oct 6, 2014 at 9:55 AM, And Rosta <and.rosta@gmail.com <mailto:and.rosta@gmail.com> <mailto:and.rosta@gmail.com <mailto:and.rosta@gmail.com>>> wrote:
>
> So {lo broda cu brodu} is not equivalent to {zo'e ge broda gi brodu}?
>
> I would say they are not equivalent because that it brodas is in one
> case presupposed and in the other case asserted.
>
> Okay. That answers my question. What's the rationale for your answer (i.e. for holding that the lo description is presupposed)?
>
> I take "lo broda" to be a referring _expression_, not a claim, and
> therefore the veridicality of its description can only be
> presupposed. I'm not sure what other kind of rationale there might
> be.

If {lo broda cu brodu} were equivalent to {zo'e ge broda gi brodu}, you could still say {lo broda} was a referring _expression_ by virtue of the {zo'e} it is equivalent to. Referentiality needn't entail presupposition of the description. As rationales, I thought you might be arguing that it's better to have different ways to express different meanings rather than merely different ways to express the same meaning; or maybe you had discovered logical pitfalls with a nonpresuppositional version of {lo}.

> Since "lo" is marked neither as definite/indefinite nor as
> specific/generic, it is useful for identification of its referents
> that it is at least veridical, This also allows maintaining the
> original definition of "lo", prior to CLL: "veridical descriptor: the
> one(s) that really is(are) ..."

What do you mean by 'identification'? The speaker knows what the referent is, and doesn't need to identify it to themself; and the referents aren't necessarily identifiable to the addressee.

It's not clear to me -- without knowing any literature on the matter -- that presupposition can rightly be considered veridical, but I can see why you'd think presupposition is at least tantamount to veridicality, and so is appropriate for the meaning of CLL-defined {lo}.

> I have seen attempts to define {lo} periphrastically using {zo'e}. Since afaik Lojban has no words for marking presupposition, any periphrasis (without the requisite neologistic presupposition-markers) is doomed to fail.
>
> Yes, the usual paraphrase for "lo broda" is "zo'e noi ke'a broda",
> which changes the presupposition into a side-claim, which is as close
> as we could make it.

Far better to introduce an experimental cmavo for presupposition than to put about an incorrect paraphrase.

>But even with that paraphase "lo broda cu brodu"
> is not equivalent to "zo'e ge broda gi brodu", because "naku lo broda
> cu brodu" = "naku zo'e noi broda cu brodu" = "zo'e na broda .i ta'o
> ri brodu" is not equivalent to "naku zo'e ge broda gi brodu" = "zo'e
> ga na broda gi na brodu".

It might be -- coherently -- that "lo broda cu brodu" = "zo'e ge broda gi brodu", while "na ku lo broda cu brodu" = "zo'e ge broda gi na ku brodu", which is what I'd had in mind. That is, the "lo broda cu brodu" = "zo'e ge broda gi brodu" equivalence is not a rule for exchanging word-strings but rather for deriving more basic logical forms from less basic ones.

--And.

--
You received this message because you are subscribed to the Google Groups "lojban" group.
To unsubscribe from this group and stop receiving emails from it, send an email to lojban+unsubscribe@googlegroups.com.
To post to this group, send email to lojban@googlegroups.com.
Visit this group at http://groups.google.com/group/lojban.
For more options, visit https://groups.google.com/d/optout.