Re: [lojban] Gerunds, infinitives and other technicalities

[John E Clifford <kali9putra@yahoo.com>]

Now that is tempting, slashing three (or more) abstractors to one.  I seem to 
recall, however,  that there were a number of gismu that suggested more than one 
type.  If this is an ambiguity in the definition it should be repaired (or is 
the different type meant to do the disambiguization?).  'ni' and 'jei', as far 
as I understand them, aren't really abstractors in the sense that they create 
references to intensional entities.  But the types of 'nu' remain a problem, 
since so many words have (in English at least and apparently in Lojban) several 
readings, one at least, in each of two or more of the subtypes.  Maybe we need 
either to disambiguate these words or provide some other clues when the nature 
is unclear (there are obvious relations between these types and aspects, for 
example). If we were to get down to one type, I would vote for 'ka' on 
metaphysical grounds: evert abstraction can be reduced to a property (if you 
allow "properties" with no gaps -- see earlier). (Of course, it works the other 
way too, if you allow propositions or events with gaps.) 'li'i' was designed for 
a paraplegic. who is, I think, no longer with us, but it might have value for 
phenomenologists, to give the residue of bracketing.  And it may be part of The 
Quest discussed elsewhere.
The main problem that I see is in formal semantics, where we have all thee 
critters to deal with and I am not sure just what some of them might turn out to 
be.  For a simple predicate, we have the extension (the set of all the things 
that have the property), the intension (the property itself -- maybe some 
function on possible worlds) the meaning (its place in a matrix of related 
terms  -- a Platonic definition, if you will, except rarely binary), and its 
truth function (maybe the same as the characteristic function of the set -- 
another item -- or, in non-binary logics, maybe not).  Moving to sentences 
merely increases this plethora.
How are 'ji' and 'nei' used for indirect questions.  What could be the truth 
value of a question or the amount?  Or are they just used in place of 'xu kau' 
and 'ma kau' (with appropriate restrictions)?  back to questions about UI, I 
suppose.


----- Original Message ----
From: Jorge Llambías <jjllambias@gmail.com>
To: lojban@googlegroups.com
Sent: Sat, July 30, 2011 9:53:42 AM
Subject: Re: [lojban] Gerunds, infinitives and other technicalities

On Sat, Jul 30, 2011 at 11:06 AM, John E Clifford <kali9putra@yahoo.com> wrote:
> Well, I'm not quite sure what a generic abstractor  or generic bridi
> subordinator would do.  It seems that each case where subordination of the 
sort
> involved here is concerned, only a limited number of  types (by the way how do
> we fill in x2 of 'su'u'?) are possible and each gives a slightly different
> meaning.  What is left for the generic case?  I suppose, following xorlo by
> analogy, that it would, in each case, be the indeterminate middle of all the
> possible inserts, assuming that makes any sense at all (what is the middle of
> propositions, events, and properties?).  It seems to add obfuscation just 
where
> clarity us sought.

It seems to me that whenever a subordinate clause is needed, only one
type (of the three main ones) makes sense and the others don't just
give a slightly different meaning, the others give nonsense.

For example, the subordinate bridi that goes in x1 of frili must be a
nu. It makes no sense to say of a proposition or of a property that it
is frili. The subordinate bridi that goes in x2 of jinvi must be a
du'u, it makes no sense to jinvi an event or a property. The
subordinate bridi that goes in the x2 of mutce must be a ka, it makes
no sense to say that something is mutce in an event or a proposition.
As far as I can tell there are hardly any cases where we have a
choice, and then what's the point of duplicating the information that
is already there in the meaning of the predicate? And in the cases
where we hesitate which one is "right" (usually between nu and du'u),
it's only because the predicate is not well defined, not because there
are two separate meanings that the same predicate would distinguish.

That's as far as the "big three" are concerned: du'u, nu and ka. In
the case of the four types of nu (pu'u, za'i, zu'o and mu'e) the
determination comes from the inner bridi rather than from the slot
where the bridi is inserted, but again there is little or no choice,
the event described by the bridi practically determines the subtype of
nu, so there is no point in duplicating that information with
different NUs.

ni and jei are special cases because they do encode additional
information, although they have the problem that they are badly
defined, so that each of them has two separate usages. Their main
usage is in encoding an indirect question. jei can be just avoided and
replaced by "du'u xu kau". ni (in the indirect question sense) is
usually something like "ka se la'u ma kau".

I can't say much about li'i and su'u because I don't really understand
them, but they are hardly used anyway.

In summary, I tend to agree that a single subordinator would have made
things simpler without really losing anything important.

mu'o mi'e xorxes

-- 
You received this message because you are subscribed to the Google Groups 
"lojban" group.
To post to this group, send email to lojban@googlegroups.com.
To unsubscribe from this group, send email to 
lojban+unsubscribe@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/lojban?hl=en.

-- 
You received this message because you are subscribed to the Google Groups "lojban" group.
To post to this group, send email to lojban@googlegroups.com.
To unsubscribe from this group, send email to lojban+unsubscribe@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/lojban?hl=en.