Re: why are selbri relations so arbitrary?

["Bob LeChevalier (lojbab)" <lojbab@lojban.org>]

At 09:09 PM 10/23/99 -0600, reciproc@freenet.edmonton.ab.ca wrote:
>From: reciproc@freenet.edmonton.ab.ca
>la jocua. pu ciska di'e
> > ex. vencu    x1 (seller) sells x2 (goods) to x3 (buyer) for x4 (price)
> >
> > Four completely arbitrary arguments which are somewhat specific to time 
> and culture, and if nothing else, requre simple memorization. What's 
> wrong with the prepositions this replaces?
> >
> > I would envision the language being as consistant and simple as 
> possible, requiring as little memorization as possible with the exception 
> of the rules of grammar. Why so arbitrary with the selbri?
>
>The argument for doing selbri the way they are is that prepositions like
>"to" and "for" have no real meaning outside the particular verb. "To" in
>English can be a receiver, a destination, or a result, for example. Even
>if one were to have various different preposition-like things to
>specifically, the point is that using them would require memorization
>of even more words, and the prepositions chosen would have a tendency be
>biased towards English.

Actually it is plausible that we could have come up with a preposition 
system not biased towards English.  But it is not plausible that we could 
come up with a preposition system that WORKS, with the unambiguity that 
Lojban demands.

A preposition system with no biases, is effectively a "case grammar" - the 
prepositions are serving as tags to indicate certain cases or roles in the 
relationships of a sentence.  Linguists have been working on 'case grammar' 
approaches to analyzing languages for a couple of decades now, and while 
they can explain some of the semantic structure of languages, none of the 
theories seems to work comprehensively.  In particular, I believe we found 
that it is fairly easy under any reasonable set of semantically distinct 
and unambiguous prepositions to come up with a predicate/sentence which 
requires the same preposition twice.  At which point you are back where you 
started, having to memorize some place structure paradigms to disambiguate 
the multiple prepositions.

Meanwhile a proper preposition system would have to include prepositions on 
all places including x1, so even the shortest 1-place sentences are 
lengthened by at least one word.

And of course none of this would really prevent the kind of preposition 
semantic drift that leads to languages having prepositions with multiple 
meanings like "for" and "of" and "to" and "in".

As a response to Lojban, some people tried at one point several years ago 
to design a Lojban-like conlang with prepositions/case tags for all 
places.  The language proposal was called "Voksigid", and it never got 
close to completion; you might be able to find traces of it on the Web and 
especially in the Conlang mailing list archives.

>In any case, the syntax of any language's predicates is more or less
>arbitrary, and most of the places are ordered the same throughout similar
>gismu--I can't think of any gismu where the place for the sender follows
>the receiver, for example.

And indeed, in learning any natural language with prepositions, you have to 
learn what prepositions are typically used with each verb/predicate and 
what they specifically mean when attached to a prepositional phrase in a 
sentence.  Easy example from colloquial English: "I am going in the store", 
which can mean that "I am going INTO the store (from outside)", or that "I 
am moving within the store (presumably to a new unspecified location)".

The Lojban preposition system ends up being no harder than that of any 
other language.

And you can indeed use prepositions - the case tags of selma'o BAI are a 
larger set than any natural language has, and will cover most needs, with 
the possibility of creating new ones using fa'o if you manage to come up 
with something that fits none of the words in BAI.  There is a plausible 
dialect of Lojban that uses BAI tags as prepositions on every sumti of 
every predicate; the important thing to remember is that if you use BAI 
words as prepositions on standard place structure places, that all 
succeeding places in the sentence after the first such usage need a 
preposition as well (exceptions exist, but this is a first approximation to 
the rule).

lojbab
----
lojbab                     ***NOTE NEW ADDRESS***           lojbab@lojban.org
Bob LeChevalier, President, The Logical Language Group, Inc.
2904 Beau Lane, Fairfax VA 22031-1303 USA               703-385-0273
Artificial language Loglan/Lojban:
   see Lojban WWW Server: href=" http://xiron.pc.helsinki.fi/lojban/ "
   Order _The Complete Lojban Language_ - see our Web pages or ask me.