[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

expanding rafsi



>Date:         Tue, 21 Jan 1992 16:22:35 GMT
>From: CJ FINE <C.J.Fine%BRADFORD.AC.UK@cuvmb.cc.columbia.edu>
>X-To:         Lojban list <lojban@cuvmb.cc.columbia.edu>

>A little speculation that occurred to me while walking home the other
>day:

>How do you use a brivla when you can only remember the rafsi?

>I practice my lojban as I walk, by saying things - often little sentences
>of the form "ta broda". The other day I saw some grass.
>Now I recall from recently reading The Open Window that "lawn" was
>"saurfoi", and that "sau" was glossed as "grass", but I've never looked
>to see what the full form is. So what do I do?

Interesting problem!

>I started with
>   "ta se rafsi zo rau"
>Unfortunately the thing I was mentally pointing at was some growing
>vegetable matter, not a linguistic object. In order to improve that you
>seem to need
>   "ta mela'elo se rafsi be zo rau"
>which I think does the job, but is rather long-winded.
>[I guess you could be even more precise with something like "terbridi"
>rather than "me"].

Yeah, "ta se rafsi zo rau" would imply that "ta du zo srasu", i.e. "ta" is
the *word* srasu.  Personally, I expect to see a lot of the idiom "mela'e",
especially in poetic language.  I can see someone generating an
emotion-selbri from a UI, e.g. "mi mela'ezo .ui" might mean "I am happy",
or perhaps the pace structures would be different...

You could also do "ta se cmene lo se rafsi zo rau", or, more Englishly, "lo
cmene be ta cu se rafsi zo rau" (assuming that "cmene" is broad enough in
meaning; I suspect it is).  This leads me to another idiom I'd like to use
but haven't had the opportunity.  Just like "tu'a" can be used nicely to
mean "an action/event somehow involving X", I think we'll be wanting a word
meaning just "something unspecified related to X", allowing for easy
metonymy (which ties in well with my recent post on synecdoche in sci.lang,
but I digress).  Fortunately, we don't need another cmavo for that, the
compound "zo'epe" seems to do the job well, and my Zipf-er doesn't rebel at
its polysyllabicity.  I wouldn't have thought that it makes sense to thow
"pe"'s and "poi"'s on "zo'e", but here it really does.  Thus, "zo'epe ta cu
rafsi zo rau" -- "something-about this is-rafsied-by 'rau'".  I think it
works nicely.  I like conceiving of "zo'epe" as a compound, and thus sort
of single, like "tu'a".

>Note that the "se rafsi" is needed - otherwise you would be (assumed to
>be?) talking about the WORD "rau" not the RAFSI - and I'm not entirely
>happy about how this distinction creeps in. Is it purely pragmatic, so
>that "lo se rafsi zo rau" is actually ambiguous but rules out one
>meaning ("the brivla whose rafsi is the CMAVO rau") simply because it is
>nonsense?

The quoting removes the tags on "rau".  Unless you carefully put them back
(as in my "la'ezo .ui", using la'e), I feel that a quoted word is no longer
a cmavo or cmene or brivla or anything; it's juts a word, and not
necessarily even that.  It's some collection of sounds that Lojban
recognizes as not falling apart any more.  So it can be a rafsi or
whatever.  You need the "se rafsi", though, otherwise you wouldn't know how
"ta" relates to "rau".

>What I want is a non-specific rafsi (it has to be a rafsi, not a cmavo,
>so I can't use do'e [I think that's the non-specific pro-brivla isn't it?]),
>so that I can turn any rafsi into a lujvo with meaning identical to its
>corresponding gismu. I guess I could use "saurbroda", but unless we
>establish this usage, that is likely to confuse.

The non-spec pro-brivla is "co'e", and you *can* use it.  More accurately,
you can use its *rafsi*.  "co'e" has a rafsi, which is just  like the rafsi
of a gismu.  It happens to be identical with the word, but that's just
engineered coincidence.  So "saurco'e" is a perfectly good lujvo, meaning
just what you want ("grassily is/does something-unspecified").  "saurbroda"
would depend on the assignment of "broda", and would be confusing if it's
unassigned.

~mark
      o o     o   o             o o   o     o        o   o     o o
              o                       o o   o      o       o     o
  o   o     o     o         o     o     o   o      o     o     o     N2KOT
Mark E. Shoulson:  shoulson@ctr.columbia.edu