On Thu, Jan 3, 2013 at 5:33 PM, Ian Johnson <blindbravado@gmail.com> wrote:

On Thu, Jan 3, 2013 at 7:08 PM= , Jonathan Jones <eyeonus@gmail.com> wrote:

On Thu, Jan 3, 2013 at 4:47 PM, Ian Johnson= <blindbravado@gmail.com> wrote:

No. I think it should mean exactly three. The differe= nce is that I think "exactly three" should only be considered wit= hin context. If we're talking about the house next door, and I say {ci = prenu} =3D {ci da poi prenu}, it should mean that there are exactly three p= eople in the context of the house next door, not that there are exactly thr= ee people, in the whole of time and space, real and imagined, etc..., whate= ver the scope of "universe of discourse" is.

This answer isn't really {na go'= ;i} so much as {na'i}. I was saying to choose a fixed universe of disco= urse, then decide what an outer quantifier means. In other words I wanted t= o simplify the issue of outer quantifiers by decoupling it from the issue o= f determining the universe of discourse. You're saying that the univers= e of discourse shrinks and expands rapidly from sentence to sentence, and t= hat the idea of holding one constant isn't reasonable. I think this is = troubling in general, partly because everything done in logic depends on a = background universe of discourse. In other words the first thing you do is = say what you're talking about precisely (for example, the real number s= ystem), and then you start saying things about it. Rejecting this idea is s= a'u illogical, but not necessarily irrational nor necessarily damaging.=

I would say situation to situation,= not sentence to sentence, because context isn't determined by what is = said, but by the environment it is said in. Strictly speaking "environ= ment" isn't really the right word, but you can't define contex= t as "context".

The way you described it, it sounded li= ke it would be perfectly [...] fine, if there were 5 people in a house all = of whom have already been=20 mentioned, to say {ci prenu cu zvati lo zdani}, since they had become=20 the topic of discussion as of this sentence.

<= br>Yeah, no. Sorry if that's the impression I gave. I suppose I should&= #39;ve emphasized the "context" bit more than I did.
=A0

And as far as logic wrt Lojban is concerned, Lojban is logical only in = the sense that its grammar is built from predicate logic. Beyond that, it i= s no more nor less logical than any other language.

No, it's not true that just its grammar is built from predic= ate logic. If that were literally true, its semantics could be not at all t= ied to predicate logic, whereas the definitions of things like {da} clearly= reflect semantic ties to predicate logic. People use the term "gramma= r" in a terrible way, grammar alone is just structure, it doesn't = mean anything. Instead they use "grammar" to refer to grammar in = combination with "structural semantics", which are essentially in= variants, aspects of semantics that persist within a given grammatical stru= cture no matter what components you put in place. {lo broda cu broda} is on= e.

At any rate, I reject the premise that the grammar of Lojban is built f= rom predicate logic only for it to completely throw out the semantic ideas = of predicate logic at random. I think it's built this way so that it wi= ll be good at expressing the things that predicate logic is good at express= ing. Quantifiers are one such thing, but only if you are reasonably careful= about the universe of discourse.

I'm not going to quibble about this. = I'm apparently being much less capable of expressing myself than I'= d like. I am in agreement with you wrt this specific issue.
=A0

My problem with making the "verbosity and precision correlate tightly&= quot; doctrine into law is that precise statements all vanish for being too= verbose. It's like pedantic English, you can be very careful in Englis= h if you try really hard, but we don't speak that way, and those that t= ry are criticized for sounding awkward.

I'm not sure that's true, and I'm not certain th= at's really much of a problem if it is. I highly doubt that requiring m= ore words to be more precise- which, by the way, is already true in the vas= t majority of Lojban as it is- automatically makes higher precision stateme= nts cease to be made.

I phrased that the way I phrased it for a rea= son. Insisting that verbosity and precision correlate tightly makes = it so every precise statement is necessarily verbose. I think part of the p= oint of Lojban is that logic cleanly and succinctly expresses certain ideas= (quantifiers, for example, and in particular ro, su'o, and to a lesser= extent pa), and by reworking ordinary language carefully we can bring ever= yday ideas into that structure.

I don't like the phrase "e= very precise statement is necessarily verbose". I would agree with &qu= ot;more precise statements are necessarily more verbose", however. A t= ight correlation means that the desired amount of precision would require a= higher degree of verbosity, but not that every precise statement would be = too verbose to actually be used, which is what it seems you are asserting.<= br> =A0

= I also don't think your comparison with English is very on the ball, bu= t that's neither here nor there.

It's an exaggeration, but I think t= he basic point is valid. That is, once you get to a certain level of verbos= ity, your listener tunes out. The threshold is probably higher in Lojban be= cause the grammar is both simpler and less obtrusive, but it's still th= ere.

Well, certainly. I just don't t= hink it's as big a problem as you seem to. Also, what I'm suggestin= g isn't a radical change from the way things are: the vast majority of = Lojban does have a strong correlation between verbosity and precision, I= 9;m just saying this correlation should be "all", not "most&= quot;.