Subject: Re: Lojban ML: Syllogism and sophism

[Logical Language Group <lojbab@ACCESS.DIGEX.NET>]

>>>It does make sense if {le ni} is a number, as Cowan also says.
>>
>>But what does it mean to say that it is a "number" semantically?
>
>Something like {li nopire} for example.

This is a conclusion you reach, but it is not the only possible answer,
which is why no one can be sure what Cowan meant (possibly including
Cowan, though he is rather lee likely than most to write something
without being aware of its ramifications).  In the Lojbanic mindset, I
consider everything that is a quantifier under Mex to be a number, and
thus must be covered by the semantics of number.  Thus a dimensioned
number under mo'e, or even something totally nonsensical like "mo'e la
xorxes." is "semantically a number".

>> Is "five inches" a number?  It is - a dimensioned number, but not a pure
>>number.
>
>Right. But {le ni}, under the number definition,  would have to be a
>pure number, like {le se grake} or {le se klani}, not a dimensioned
>number like "five inches". Lojban doesn't make much use of
>dimensioned numbers in any case. Most place structures that take
>numbers take them dimensionless.

Since Lojban expresses dimensioned numbers either as two sumti, or as a
single sumti marker by mo'e which then has the grammar of a Mex
quantifier, pretty much it is impossible that a Lojban place structure
could make much use of dimensioned numbers directly.  However, through
the use of mo'e, and then converting back to a sumti (li vei [operand])
every dimensioned number can be used any place where a simple number can
be used.  In Lojban, all parts of speech are pretty much fully
interchangeable in this way, with the limitations primarily being
imposed by context.

But I suspect that most place structures that take a "number" indeed do
not take them dimensionless. in th sense that the number is almost
always carried in one place and a ddimension is carried in a different
place.

>>but I don't thin kit is
>>going to stand up to the kind of word-twisting semantic0-logical
>>analysis that you seem to apply to every word of the text.
>
>Why word twisting? I find example 5.3 to show very clearly {le ni}
>used as a number and example 5.5 shows a clear case of it being
>used as a xokau-property. It is you who's twisting words to get
>5.3 to mean something different than what it obviously means.

It doesn't mean anything!  Cowan doesn't say what it means.  It is not a
sentence.  It has no context.  It is merely a sumti with no obvious
referent.  You can give a literal trasnlation as Cowan does, but he
gives no indication how such a translated sumti might be used.  I can
think of no English sentence where I could use (with meaning) "1-B,
where B is the blueness of the picture", without in some way introducing
some units to measure blueness at the very least (and even then it would
be a most unusual sentence indeed).

>Also, you can look at the next section about {jei} where exactly
>the same issue resurfaces with examples 6.1 and 6.3.
>6.1 shows the use of {jei} as a truth value and 6.3 shows its
>use as a xukau indirect question.

I think we discussed this one rather recently.

>The way I would say it is:
>
>    le nu ko'a sutra le nu broda cu cenba [le ka ce'u klani li xokau]
>    Koha's speed at brodaing varies [in how much it is].

That first sumti is discussing the event of ko'a being fast, and not the
measureable property of that event, which is the speed of le nu broda
taking place.  All you have said is that this event of ko'a being fast
has varied in some specific property of being associated with a number
that has some value.  What numerical property of the event that is
varying is not in any way made clear and indeed it could be any
numerical property.

Now I am the last person to claim to undersatnd ce'u, but I would have
thought that ce'u in that sentence could in some way be correlated with
x1, based on how I would presume this example of yours would map into a
zmadu predicate comparing two speeds.  Which suggests to me that you
think that the x1 of klani is a lenu clause.

>Of course, if broda was more specific we could be more
>specific about le se cenba. For example if broda was
>travelling by car, le se cenba could be kilometers per hour.


Yes, but the event of John being fast at travelling by car can vary in
the number of times John stops for gas.  Another number unrelated to
speed that is a variation in the event of John being fast.  Isee no way
that this interpretation is excluded by your ce'u klani format.  Now if
you put a scale of miles per hour in the x3 of klani, in le se cenba, I
might have to work harder to come up with another number about the event
of John travelling by car that is measured on that scale which is NOT
speed, but the point is that nothing in your formulation says that it is
SPEED that is varying in that event, merely that it is a number having
unstated propeties.

>And of course, we also have the simpler forms:
>
>    ko'a cenba le ka ce'u sutra le nu broda kei sela'u li xokau
>    Koha varies in the extent to which it is fast at brodaing.
>
>    le nu broda cu cenba le ka ko'a sutra ce'u kei sela'u xokau
>    The brodaing varies in the extent to which koha is fast at it.

And these equally in no way require that the quantity be the speed, not
to meantion being sloppier semantically in general.  I get for the first
one

ko'a varies in the property (NOT the "extent") of being fast at
broda-ing, and this varaince is associated with some number which is the
focus of our comment.

and the second one

Some specific event of broda-ing varies in the property of ko'a being
fast at it, and this variance is associated with some number.

I dunno - the latter in particular seems to suggest acceleration to me
more that speed.  Perhaps so does the former though I cannot make much
sense of it at all.  And acceleration is just as much a property of a
chance in how fast something is as is speed.  But acceleration is not
itself a ni sutra while posibly being part of ka sutra for a specific
event.

>Let's see if we agree on how {klani} works first. Are these acceptable
>uses to you:
>
>(1)        le birka be la djan cu klani li pa le ka mitre
>             John's arm is 1 in meters.

No.  John's arm is not a quantity, I'm happy to say %^).  The LENGTH of
his arm could be a quantity, as could be the width/diameter of his arm.
But there is no disambiguation present to indicate what about John's arm
is 1 in meterness even if this were an acceptable use of klani.

>Other examples:
>
>(2)    lei va plise cu klani li mu le ka mi kancu
>         Those apples there are 5 by my reckoning.

Those apples are a quantity?

>(3)    le gugde cu klani li ciciki'oki'o le ka namcu pe lei xabju
>         The country is 33,000,000 in number of inhabitants.

The country is a quantity, measured as 33 million in a numberness
associated with the mass of inhabitants I have in mind.

There are a lot of numbers associated with said inhabitants, the count
of which is only one - see the Statistical Abstract ofthe US for 1000
pages of numbers measuiring the inhabitants of the US in a variety of
ways.

> >What then SHOULD the place structure of the underlying ni abstraction be
>>such that this works - that le ni broda fits in the x1 of klani, and if
>>a le se ni broda exists, that it be wwll-defined.
>
>I don't know. I didn't invent {ni}, nor do I have much use for it.

Well you are trying to pretend it isn't needed, but when you are not
careful, no one knows what you are measuring.

>Of course, there are lots and lots of possible variations, especially
>for the last one:
>
>    i mi do prami sela'u li piso'i
>    I love you in quantity a lot.

Vague as the relationship of the quantity with the predicate.

>    i mi klani li piso'i le ka prami do
>    I amount to a lot in loving you.
>
>    i do klani li piso'i le ka mi prami
>    You amount to a lot in being loved by me.


As with the earlier examples, you and I have somehow become quantities.
I always thought we were people.  And the scale leka prami on which the
measurement piso'i is not only vague, but is also not open-ended since
the "pi" implies a maximum value of 1. Sounds like a truth-value to me.

>>Or do you simply maintain that the time relationship in the two cases
>>that leads us in English to use the word "old" is simply too distinct
>>and that there needs to be two separate brivla?
>
>That's my point, yes. At least as distinct as the difference between
>cnino and citno. If the difference is irrelevant in the case of  old1/old2
>why is it relevant in the case of new/young?

Well, I have to admiot that the fact that English uses two different
words was likely a major influence here.  Therefore we presume some
significant difference without worrying about what it is.

>>I believed that the concepts were compatible and attempted to find a
>>wording that worked.  Obviously I failed.  But I don't want my failing
>>to limit the usage of the word if it is merely a failing of words and
>>not an irrationality of idea. can youi do better?
>
>I already proposed my solution: use tolci'o for "old" in the sense
>of many-yeared and keep slabu for the sense of familiar/well known.
>If you want slabu for both meanings then glossing it as "old"
>suffices. But then why keep two words for the equally related
>concepts of young/new? (Let me guess: JCB had two words.)

Absolutely.  (I do think that there is a difference between old/familiar
and young/new, but I cannot characterize it,and would not be sure that
my characterization isn't as malglico as the original decision.
Possibly it is how we measure them. slabu as measured is a time which
increases with the degree of slabu-ness.  I am not sure the cnino is the
exact opposite in that higher degree of newness does not always imply
less time, but could also imply greater unexpectedness. citno is more
valid as an opposite, but we simply do not measure youngness, but rather
age so it is hard to be sure.  So the best I cna perhaps offer you that
is definitely not malglico is that the measurement of age, being
openended in one direction only, is not symmetrical, If negative ages
were meaningful and associated with newness and/or youngness this
argument, however weak, would probably collapse indeed.)

>>If you don't like it, how would YOU propsoe to fill in the scale places
>>of these various predicates, and in the generic sense, ckilu and gradu.
>>If you don't think a "si'o" fits, what would you prefer instead.
>
>{ka}, as I showed above. A scale is a two place relationship
>between the thing being measured and the resulting measurement.

But your examples except possibly for the one involving kancu, do not
say what about the thing being measured, is being measured.  There is no
indication of whatthe units might be.

>>You seem to
>>love analyze every decision I/we made to the point of breaking it,
>
>That's not my goal. I just want to learn how to use the language.
>I just don't see how an idea can be a scale. If you have a good
>argument why it can be, I want to hear it so that I can make sense
>of it when I see it used.

I am not sure that an idea can be a scale.  What I am sure of is that a
sclae is a dimensioned number line with certain units, and that scale is
a mathematical ideal - an abstraction that can never be pointed to.  So
in that sense a scale is an idea of a particular kind, notnecessarily a
syummetric relationshiop.  IT seemed to me that mathematical ideals, and
indeed ideals in general, are what we mean by si'o abstractions, and
that is therefore what Cowan and I chose for those place structures (I
think he was in on that decision which was relatively late).

>My decision for {ni}, as far as I now understand it, would be
>to ditch it, since it is never really needed.

Your examples in this post have not shown this.

lojbab