[YG Conlang Archives] > [jboske group] > messages [Date Index] [Thread Index] >
At 12:19 AM 1/7/03 +0000, And Rosta wrote:
> >A good logical language is one that is not only logically precise but > >also concise > > Why? Because a notation that is insufficiently concise is a deterrent to using it. If it takes ages to say what you want to say, then you are more likely to choose to say something shorter than what you had wanted to say. Creating a logical language is a pretty trivial undertaking, because logic notations have already been invented. The challenge is to create a more concise (and hence more usable) notation.
If logicians haven't done it, what makes you think we could?
> Formal predicate notation is concise only in using single symbols for > concepts. I never made it far in logic class, but it seemed to take very > little to end up with quite complex "sentences", and there was no sense > that there needed to be abbreviations presented. That is because in logic class there is no call for concision, but there is a call for a notation that is homomorphous with the logical formula, so that the logical structure is the more apparent.
Most people think "spoken predicate logic" means precisely a language which matches the forms that one would see in a logic class. Unfortunately, beginning logic classes when I attended school did not mention lambda or tense logic or Montague semantics, so most of this stuff is way over their head.
> The concept of using > discursives as abbreviations for logical constructs was largely my own, > though JCB had discursives that were logically unanalyzed. (I think jimc > also had the idea of analyzing discursives logically, and he and pc may > have had some go-rounds on this issue before I became involved). JCB's > examples of "spoken symbolic logic" in Loglan 1 were generally quite > long-winded I have always presumed that was because of a failure on JCB's part to understand that lexical structure needn't be homomorphous with logical structure, and -- due to inexperience -- a failure to apprehend the importance of concision.
He recognized the importance, but not the importance of having concise forms being defined precisely as concisions of longer forms. In other words, he presumed (I don't think he explicitly did so, but it was there) that concision would lose information, and it was merely a matter of figuring what stuff would be acceptable. He did not accept enough to leave a language that was speakable as well as logical enough. Not enough abbreviations and those that he added were even vaguer than ours.
I remembered the contributions from jimc and the final assignment by John, a UI meaning "without parallel". I have nothing against that UI, and it may save us from so often wishing to express 'logical only', but it does not itself express 'logical only', does it.
I dunno. What's "logical only" if not that?
> > > I am not sure what > > > happens to that meaning when within a negation. I am quite sure that> > > people won't consider what happens to that meaning, until someone does the> > > analysis and calls them on it > > > >It is easy to say what happens to logical 'only' within negation > > What is "logical 'only'"? If we could have agreed as to what it was, it > would have probably been explicitly adopted "only x is broda" = "every broda = x".
How about an example where this differs?
> >We can't say what happens to {po'o}, because that has no logical
> >definition, and nor is there any general account of how UI that
> >might encode logical meaning interact with other logical elements
> >in the bridi
>
> There is *in general* very little account of how cmavo interact with each
> other.
In cases where CLL is explicit enough about the individual cmavo, we
can deduce how they interact.
But we often did not consider how they would interact in making them explicit, and I suspect that most of the problems that have cropped up have been precisely those sorts of interactions, where people are making too much of wording that was adopted without considering what people would do in interactive cases. (And in many cases, Cowan was perfectly content to NOT consider what the interactions would lead to).
> Logical semantic precision would be equally infinite as non-logical > semantic precision You seem to me to be making confident pronouncements about a subject you don't understand, but I often have that impression and sometimes it's wrong. Maybe you could explain to me why you think logical semantic precision would be infinite? Take some logical formula: in what respects is it imprecise? Perhaps you mean that nonlogical semantic precision, in definition of nonlogical semantic predicates must perforce involve a logical element? If so, then that's not what I meant by logical precision.
Since I don't know much about logical semantics, and I have the impression that it is not in the least bit simpler than nonlogical semantics, I cannot imagine how it would result in simpler expressions. But we may simply be talking at cross purposes as to the nature of "complete semantic precision" both logical and non-logical.
> Jorge indeed raised this very question, and Cowan answered it in the way I > would expect (though he typoed si'e) > http://balance.wiw.org/~jkominek/lojban/9412/msg00005.html You're doing an admirable job of navigating archives I have always found unnavigable!
google searches do wonders when you learn what keywords work.
There is no escaping the fact that the ma'oste says: "x1 is an (n)th portion of mass/totality x2" which I don't find ambiguous. Examples elsewhere show that pimusi'e is also legitimate, but that definition shows that resi'e is legitimate.
Only if you read the definition in a certain way never supported by any examples or textual explication. Thus the wording is poor - mea culpa. It is telegraphic writing and hence has some amount of slop, probably too much, and I'll agree that your reading is a legit one. But it is not the intended one. That is what byfy is for: to fix this sort of thing and make non-telegraphic definitions, preferably with examples (CLL originally started as a project to come up with defining examples of each cmavo in each grammatical construct wherein it could be used, now we need something more like that, and we have the people-power to do it)
As I say, it is something that the BF can rule on, though.
Yep.
> >resi'e = x1 is a 2th portion of mass x2 > > The definition was poorly worded. Trying to define things baseline-ably in > 100 characters with no examples was something we never intended, and no one > checked my wordings before baselining (which I never wanted to do before > the dictionary). But a "2th" portion is not English-grammatical anyway > (nor is a 1/2th, I'll admit) We all agree that the vlaste are rife with poor wordings. Sometimes that wordings are unclear, sometimes they give rise to contradiction, sometimes they fail to express the designers' intention. Si'e is an instance of the last of these. But these are the baselined materials, and they define current SL.
Thus we bear the fruit of people insisting on baselining a document that wasn't written with the intent of being a baseline document (none of the wordlists were - they were LogFlash file inputs).
Yes, current SL is a total mess, and that's why there have been calls for BF and AL. > >lo resi'e = at least one thing that is a countable half of something > > Only if you assume that a 2th is a half I do. I would consider anybody who didn't to be an imbecile, in this particular context, just as I would consider imbecilic anybody who thought ordinal remoi couldn't mean "second" because the definition of moi uses "(n)th". Indeed, I am pretty disgusted at you for even raising the notion, because it lowers the calibre of discussion so much. You often give me the impression that when you are trying to argue for your views you throw up any and every old argument without heed to its foolishness, which is why I sometimes dismiss your arguments overhastily.
Sorry. I guess I can't understand people taking the wordings so literally. Probably because I reject the literalism that is apparently necessary for the logical precision you want. I KNOW that English isn't unambiguous and precise enough to specify things precisely, and thus didn't try.
> >You don't have to assume it is elliptized. You could take the veiw that > >the absence of an overt quantifier means the absence of a covert one > > We did not make that assumption, which never would have occurred to me Which assumption didn't you make? That the quantifier is elliptized?
That a missing quantifier where one is permitted could be me'ozi'o (which I presume you mean by "absence of a covert one").
I know you didn't assume that the absence of an overt quantifier means the absence of a covert one, but it would have been sensible to do so had it occurred to you.
Since quantifiers are grammatically allowed (and the TLI language did not have either tu'o or zi'o) it would not have occurred to us. I don't know what we would have done "if".
> zi'o was not around at the time, and I still don't really accept zi'o even > if I can vaguely understand it when used tu'o is mo'e zi'o, according to CLL, though according to ma'oste it is mo'ezo'e (it's pretty clear that CLL is right here).
tu'o wasn't around either; it arose when Cowan redesigned MEX, and was not considered for any usage outside of pure MEX where it solved a specific problem.
I think that tu'o is ambiguous between mo'ezo'e and mo'ezi'o in its definition. In one grammatical context, that of a dummy argument in PN or RPN, it seems clearly to be mo'ezi'o. When used as a digit variable in a digit string such as retu'o for twenty-something, it is clearly mo'ezo'e.
Lojban has unquantified sumti -- "ko'a ce ko'e", for example. So it surprises me that you cannot imagine "lo'i broda" -- or an expression meaning "the set of all broda" -- as unquantified.
I define whether it is or is not quantified by whether the syntax allows it. I am a pragmatist.
> >This one is not a very good example, because we can say "su'o broda";
> >{lo} is a comparatively useless gadri
>
> Until late 1994, su'o broda was distinct from su'o lo broda, and I never
> agreed to conflate the two, and I still don't like it.
But what do you see as the difference between them? (But see below.)
You'd have to reread whatever I said in 11/94.
> I believe that su'o broda was
> su'o da poi broda and lo broda was not defined in terms of da poi broda so
> su'o lo broda was not. I don't remember and likely did not understand the
> justification for merging the two, and at one time HAD understoof the
> reason for treating them separately
Ah. I see what you mean. Yes, I have in the last day or two shown
that {lo broda} is not equivalent to {da poi broda}, so if {su'o
broda} were short for {su'o da poi broda} then yes, the conflation
would be annulled and {su'o lo broda} would not be equivalent to
{su'o broda}. It makes only a slight difference to the meaning of
{su'o broda}, and I'm not sure which I prefer.
Interesting.
%^) -- lojbab lojbab@hidden.email Bob LeChevalier, President, The Logical Language Group, Inc. 2904 Beau Lane, Fairfax VA 22031-1303 USA 703-385-0273 Artificial language Loglan/Lojban: http://www.lojban.org