tlhIngan-Hol Archive: Mon Jul 05 02:39:22 1993

Back to archive top level

To this year's listing

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

Re: House languages

[email protected] wrote on Mon 5 Jul 93 14:47:39 +0800 (Subject: House
languages):-
  > ... a house language for *my* house ... It would have to have the
following attributes: Very easy grammar. Easy to generate enough vocabulary to
talk about pretty much anything that we want to. Phonetically easy, but
"foreign sounding" (That knocks Khoisa [= the language of the South African /
Namibian Bushmen] right out:-) ).
  - [Tolkien's] Quenya ... no less complex linguistically than other (?)
natural languages. It'd be hard to learn  -- I can't imagine getting to the
point where I could converse with my 10 year old in it.
  - Klingon. To some extent, the same problems as with Quenya ...
  - We could make something up ...

  Real languages are complicated because life is complicated. Limited
vocabulary, limited range of clause subordination constructions, limited range
of verb tenses and moods, etc, cause lack of precision, as is well known with
trying to understand people who call every object "that" or "the thing" and
call every process "doing", instead of using the correct noun or verb. Look at
a grammar of any real language for the variety of situations that separate
grammatical forms are needed for. As regards trying to disambiguate by other
methods: tone can't be expressed on paper without diacritical tone marks;
context and circumstance are unreliable. E.g. having only one "if X, then Y"
construction is not enough: Greek felt a need to distinguish: ordinary
conditions; habitual conditions; past and present unreal conditions; future
less likely conditions; and oddsorts including where there is a second implied
condition. E.g. also Tolkien Quenya 'vanya' = "goes, vanishes" (verb) does not
seem to me to distinguish "go, presumably to another place" from "vanish,
perhaps by ceasing to exist": in English and Greek these two meanings are
distinct: 'ienai / erkhesthai' and 'aphanizesthai'. And the usual quota of
ambiguities of what a pronoun refers to, and what word belongs with what, etc.

  > Hmmm. Has anybody written a language generator program?

  That would probably need a (genuine or simulated) neural net computer
capable of understanding real world life and its temporal and spatial etc
situations non-linguistically, so that the language that it invents is its own
ideas and not largely the ideas of whoever programmed the language generator.
I don't think that AI (= artificial intelligence) has got that far quite yet.

Back to archive top level