<<
>These seem to me to be paradigm cases (well, not quite, since none of them
>has {lo'e ...} as first argument) of talking about lo'e ..., which is an
>intension (in some sense or other -- I am not at all such which), isn't it?
>You say you like it or that that is like it or that is a picture of it (a
>notion I have a lot of trouble with -- abstract expressionism?)
You know perfectly well that is not what I mean.
I mean "I like chocolate", "that is like a sofa" and "that is a
picture of a boa". They don't mean "there ia some chocolate such
that I like it", "there is some sofa such that that is like it"
or "there is some boa such that that is a picture of it". To get
those latter meanings I would have to use {lo} instead of {lo'e}.
>>
As I had said repeatedly then, I no longer knew *what* you meant; the various explanations did not jell into any one thing -- except that it should work as you say and that it involved types, which two did not jell in the that terminology.
<<
I'm not sure why paradigm cases need to be in x1, but here are
some: {lo'e cinfo cu xabju le friko} "Lions live in Africa",
which is different from "some lions live in africa" (lo),
"all lions live in Africa" (ro), "most lions live in Africa" (so'e).
{lo'e mlatu cu kavbu lo'e smacu", "Cats catch mice", which is
different from saying that "some cats catch some mice", etc.
>>
That is just the way the paradigms run; there is obviously nothing special about it except presentation. I would have seen the examples as simply false sentences types. And, even from my new point of view I have trouble with them -- now if I am to separate your {lo'e} from Lojban's, since these seem perfect cases of the latter and imperfect ones (if I now understand it) of the former.
<<
> Does {lo'e sfofa} refer to the proximate type of
>sofas? Apparently not. What then does it refer to?
{lo'e sfofa} does not refer. It is like {zi'o} to that extent.
Even in the most restricted sense of "the typical" it has to
be like that to make any sense.
>>
I now understand and agree with the starter here (properly understood -- this may not be the best way to put it "its reference is oblique'). But the reference to {zi'o} is totally misleading, since {zi'o} is a semantic plug, functionally like {se}, for creating new predicates from old. Pointing to "the typical" is better, but that does not work like {zi'o} in any way I can think of and yet makes perfectly good sense in various contexts (different kinds of sense, to be sure, in different contexts)
<<
>It is obviously not a
>meaningless expression (or you would not fight so hard about it).
Obviously it has the meaning of {sfofa}. It certainly maintains
the intension.
>>
No, it doesn't (if I finally have it right. No, in any case). The sense of {sfofa} is a property, the reference of {sfofa} is a set and neither of these work for {lo'e sfofa}, which (as you just noted) has no reference and, in fact, no sense neither.
<<
> So, it has
>a sense, that would pick out something in the world, if there is the
>appropriate sort of thing in the world.
No, it doesn't pick anything in the world. It just puts to use
the sense of {sfofa}. It does not get anywhere near the extension.
>>
I guess it is the locution "puts to use the sense of {sfofa}" that makes this seem a muddle. If {lo'e sfofa} works off the essential properties of {sfofa} as the Lojban version does off the typical ones, then I suppose this is a way of putting it -- though then the examples above are all bad, since they do not deal with essential properties. If only typical properties are involved, then how is it different from the Lojban version?
<<
>Otherwise it fails to refer, perhaps
>accidentally, because the world is shy this sort of object.
It intrinsically does not refer, like {zi'o}. But unlike {zi'o}
it adds some sense to the predicate from which it removes a place.
So {simsa lo'e sfofa} behaves just like the predicate "x1 is like
a sofa in property x2". (I suppose {simsa zi'o} would behave like
"x1 has property x2" maybe.)
>>
As noted before, the first part here is just what any real sumti does; there is nothing special about {lo'e}. {x simsa la djan y} behaves like the predicate "x is like John in property y." What is special about (lo'e}? It seems to refer somehow else how does it mean (there is a good answer to this, but I haven't seen it in your stuff yet).
<<
> Since you seem
>to think that {ta simsa lo'e sfofa} is true, the approriate sort of thing
>must be in the world
No, there is nothing in the world that is a referent of {lo'e sfofa},
neither in my usage nor in the more restricted definition
as "the typical".
>>
?! Your {lo'e} is more general that Lojban's?! Now, even if I have it finally in the right category, I've lost the specifics. I can think of at least four members of this category: the typical (about properties common across the set), the ideal (about properties the members of the set should have), the stereotypical (about properties we think are common across the set), and the average (about averages in whatever sense somebody wants to work them out, so possibly open to several subtypes). But, yes, the references made by {lo'e sfofa} are oblique and {lo'e sfofa} by itself has no referent -- nor, strictly speaking, sense neither.
<<
>(we have disallowed some weeks ago the possibility that
>some places are inherently opaque [what I used to call intensional before
>there go to be too many things getting called by that name]
We agree there. The x2 of {simsa} is not inherently opaque.
It is perfectly possible to say {ta simsa lo sfofa}: "there is
at least one sofa such that that is like it". Which does not
make exactly the same claim as {ta simsa lo'e sfofa}.
>>
Yes, there may not be anything that has just the properties that {lo'e sfofa} puts into the mix. I suspect that that means that the expansions of sentnces containing {lo'e sfofa} turn out to be intensional from top to bottom or at least hypothetical (and that probably amounts to the same thing). Does {mi nelci lo'e sfofa} means something like (we can prise out the details later) "I would like anything that had the properties delimited in {lo'e sfofa}"? [I'm skipping over the question of what to do with liking or being like something that does not exist, which I now see as a separate issue.]
<<
Unfortunately we don't have the la-version of lo'e:
lo le la
lo'e le'e ??
But we can use {lo'e me la santas}.
>>
Cute! But doesn't that require that there be an instance of Santa? I think the text is just unclear on this point (maybe never even brings it usp -- a wise move).
<<
>What the fatal fandango is it? How (in
>addition) does taking {lo'e sfofa} to refer to the proximate type of lo'i
>sfofa, take the type (which one?) as a token of types --
Tokens of the class "types" are the things we talk about in
this meta-discussion. Abstract entities like sets and numbers.
Not things we want to talk about in ordinary discourse.
>>
Well, sets and numbers aren't types, but I don't suppose that is what you mean. The quotes here are confusing as is the talking about tokens of a class (types have tokens, classes only have members). I still can't unpack this with any confidence. I guess you mean that we ordinarily talk about ordinary things, even if obliquely, and only in metathery do we talk about abstract things directly. But I am not sure how this comes out of what you actually say (and no amount of adding or removing quotes helps in this one).
<<
>and what does that
>mean? The proximate type of all the sfofa is, of course, a token of the
>type
>type, since it is a type (this gets hairy in practice, but has a variety of
>solutions).
I talk about sofas, (not about _some_ sofas, not about _each_ sofa,
not about all the sofas that exist or could exist taken en masse,
also not about the property of being a sofa, but just about sofas).
>>
Yes, this fits. Except, of course, that you do talk about all sofas, just not directly -- and ditto the proerty of being a sofa (at least in some cases: I assume that {lo'e sfofa cu sfofa} is trivally true).
<<
>So, as noted earlier, what would be an example of talking about a type?
This whole conversation has been mostly about types and such, not
about sofas.
>>
Yeah, but you keep saying that the other is not about sofas, either, though it also is not about types.
<<
>Putting {lo'e broda} in first place? Obviously not. Using {li lo'e broda
>li'u}? Hopefully not. What then?
We don't have a special article for talking about tokens of the
class "type", of course. It would make no sense to have one.
It is bad enough that we have a special article to talk about
tokens of the class "set", something we rarely want to do in
ordinary conversation.
>>
Well, but it would be -- like sets (which we have made reference to quite often in this discussion, not) -- useful for discussion like this. As it turns out, we would not have used the notion of type in this discussion, since the target point was not about tokens and types at all, but about oblique reference (I just made that fairlya ccurate term up on the fly to sort out this case from others that were also intensional in very different ways).
<<
To talk about types we need a word that means "x1 is a type of
property x2" or "x1 is a type of set x2" or some such. Maybe {cnano}
is one such predicate? (Probably it won't always be used in that
sense.) But then we can talk about le cnano be le ka sfofa, the
type of class "sofa". I certainly do not want to claim
{ta simsa le cnano be le ka sfofa} in that sense of {cnano}!
>>
{cnano} won't work for "type" in any clear way, but it is a good start for I take it what is going on with {lo'e} and kin ("the average" version, anyhow). It gives values not types and it works off the properties too directly. I am -- depite my conviction that Lojban could use a good way to talk about types and tokens -- presently fervently hoping we don't get involved in it for a long time.
| Yahoo! Groups Sponsor | |
|
|
To unsubscribe, send mail to lojban-unsubscribe@onelist.com
Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.