Loading ...
Sorry, an error occurred while loading the content.

200827Re: replies on loglangs I: pertaining to Lojban, Xorban, Liva

Expand Messages
  • And Rosta
    Feb 2, 2014
    • 0 Attachment
      F, On 29/01/2014 10:49:
      > And Rosta said:
      >> * Some bits of Lojban aren't human-usable.
      > Would you mind listing those bits? I have seen selma'o SE being
      > brought up before, but what are the other parts that you would say
      > aren't human-usable?

      SE is the clearest example. I strongly think the official formal grammar too is not human-usable.

      I gave an off-top-of-head list of (inter alia) unnatural features in <http://listserv.brown.edu/archives/cgi-bin/wa?A3=ind1104c&L=CONLANG&E=8bit&P=158204&B=--&T=text%2Fplain;%20charset=ISO-8859-1&header=1>, but I can't think of futher exx of human-unusability.

      >> * Altho Lojban is perhaps the best-described of all conlangs, its
      >> description still covers only a small part of what even a minimally
      >> complete description of an actual language would require. In particular,
      >> the rules specifying what sentences mean are very incomplete.
      > How would you go about specifying such rules? Is there a good way to
      > do it systematically? When is the job complete?

      I think a loglang grammar should specify precisely what predicate--argument structure (which I take to subsume quantifier--variable structure) a sentence encodes. But I don't place any very stringent requirements on it specifying the semantics of the predicates; I see semantics as essentially extralinguistic, and the job of describing it belongs to the writers of encyclopedias rather than the writers of lexicogrammars. So when I said "the rules specifying what sentences mean are very incomplete", I really meant "the rules specifying what PAS the sentences encode are very incomplete".

      >> Xorban gets much closer to unambiguous lossless encoding of
      >> predicate--argument structure in a way that is concise and ergonomic.
      >> Xorban does so much better than the others that only Xorban deserves to
      >> be considered state-of-the-art.
      > Xorban is the purest loglang I know. It has no extra baggage, but
      > that's also what makes it extremely hard to speak in real-time. The
      > mental stack can't seem to handle all the variables, so I can only
      > really see it being used as a written language as of now.

      I tend to agree, tho I doubt that giving it extra baggage, while maintaining the same functionality, would make it any easier to speak. The problem seems to me to be not the baggagelessness but rather the having to hold in short-term memory the temporary register of variable names.

      MorphemeAddict, On 29/01/2014 17:30:
      > What remains to be done with Xorban grammar? I was under the impression
      > that there were still issues that had to be worked out, and when that was
      > done, a (speakable) phonology would be hung on it.

      IIRC, the basic grammar could handle most things, tho doubtless with minor gaps (e.g. for mathematical expressions, as was mentioned recently). There was virtually no lexicon at all, not by design but because that had been left till later. There was not consensus on morphology and phonology, and that had deliberately been left till later too. There was disagreement about the nature of the grammatical rules that define the language: I think those rules had not yet been written out explicitly. IIRC, we were hotly debating that when for me the demands of the academic year 2012--13, which had inconventiently begun shortly after the birth of Xorban, finally kyboshed my further ability to participate.

    • Show all 57 messages in this topic