Re: [lojban] fancu

[And Rosta <arosta@uclan.ac.uk>]

pc:
#jjllambias@hotmail.com writes:
#> Your assumption is that to refer to a function we must use something
#> that looks like one of its values. Is there a justification for that?
# Not my assumption, just the usual way of doing it;  how would you like to do 
#it?  A set of ordered pairs (but you hate sets) with the condition that for 
#each first member there is only one second member?  That is not what you have 
#presented.  You have presented a proposition or a function to propositions; 
#that is what {du'u} does, le du'u bridi is by definition something we want to 
#call the propsotition that bridi.  Putting two variables in makes it a 
#property, i.e., a function from (in this case) an ordered pair to a 
#proposition. That is a perfectly good way to talk about taht kind of 
#property/function, but that is not the kind of function we have here.  I 
#don't have a good third idea at the moment, except of course to use names and 
#fill in the last place of {fancu} -- la mamfanc fancu lo'i danlu lo'i fetsi 
#le nu roda mapti le mamta da.

If it is in the nature of functionhood that for every x there is at most one f of x
(where f = a function), then {mamta} seems inappropriate as part of a locution
that expresses the mother-of function (e.g. {le mamta be ce'u}) because 
there is nothing intrinsic to the sense of {mamta} that says that something
can have only one mother. {mamta zei fancu} would be a better selbri,
or conceivably {pa zei mamta}.

I would not be saying this, if Lojban had a way to use {mamta} as an applied 
function rather than only as a predicate. E.g. if *{mamta la djan} functioned
as a sumti that referred to the mother of John. That seems to be how you
conceive of {le mamta be la djan}, but really that means "x is such that
it is nonveridically said to be the case that x mamta la djan", where x is
not bound by a quantifier.

#> In my view {makau} stands for the value that the relationship gives
#> when the ce'u place is filled. {makau} will take a value from x3
#> for each value taken from x2 and placed in {ce'u}.
#Ahah!  I have accused you of that view several times and you have almost as 
#often denied it, swearing that you believed that the answer to a question was 
#a proposition not a thing.  Now, to make a point you will go back to your 
#true view.  OK.  

I'd be steaming if you'd written that to me! 
Jorge does believe, contrary to your accusations, that the answer to a question 
is a proposition not a thing. He does not say anything in the quoted passage 
that contradicts this. He says that (loosely) {ma kau} stands for a thing.

#But notice that will make {la djan djuno le du'u makau mamta 
#la bil) into perfect nonsense (of a highly forbidden kind: we can't use 
#{djuno} for people).

According to Jorge, "du'u ma kau" is a category of propositions -- a 
category of answers that replace {ma kau} with a value that makes the
proposition true. So {la djan djuno le du'u makau mamta la bil) is not
made into perfect nonsense.

It's utterly pardonable that you fail to understand Jorge, for for all of us there
are things that we fail to understand, but it must try the patience if you fling
around these accusations ("Now, to make a point you will go back to your
true view"), even when flung at someone so imperturbably equable as
Jorge. (Yes, yes, I know that Jorge will say "That's the nature of pc; you take
the rough with the smooth", but one can still hope for a slight smoothing of
the rough!)

#Ah, but maybe what you mean is that somehow it is built into the operation of 
#indirect questions that they generate the proposition with the right critter 
#in for the {kau}.  But then, of course, it is impossible to get the answer 
#wrong, which, alas, goes against our experience: {mi jinvi le du'u maku mamta 
#la bil}  guarantees I get it right (so only essay questions from now on). 

A good objection, which, it seems to me, applies to any variety of the set of
answers analysis.

I don't know what Jorge will say, but I'd suggest that maybe {du'u ma kau}
gives the set of all answers (including false ones), but that the semantics
of {djuno} means that any answer that is se djuno is perforce true. I'm not
sure how that fits with {mi jinvi le du'u ma kau pendo la bil}, but then I'm
not clear about exactly what that is supposed to mean.

#And's view -- if I have it somewhat right -- at least misses that problem and 
#only runs into all the intensionality or interchange problems -- as well as 
#missing several good answers.  

That's right. [I won't say more, because we have agreed to postpone
discussion to another fresh thread.]

#> Why would its values be more representative of a function than the
#> relationship that gives rise to it?
#
#"Is mother of,"  {le ka/du'u ce'u mamta ce'u}, is a relation and, indeed, a 
#function, as a set of ordered pairs --though the order is reversed here, so 
#{le du'u ce'u se mamta ce'u} .  There are many functions for which it is 
#somewhat unnatural to think of the corresponding relation (sum, product, and 
#the like, for example) and, indeed, the relations can usually be expressed 
#only by an equation between the function with an argument and its value for 
#that argument (though one way of doing Logic does take this notion as basic, 
#to simplify some kinds of metatheoretical proofs). 

I think it would be very helpful to use Sum rather than Mamta as an example.

--And.