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

Re: Is there an inverse relationship between lexical richness and grammatical complexity?

Expand Messages
  • Demian Terentev
    Hypothesis fails on Sanskrit and Ancient Greek that feature both lexical richness and grammatical complexity. As for conlangs, Toki Pona is an example of
    Message 1 of 56 , Mar 19, 2013
    • 0 Attachment
      Hypothesis fails on Sanskrit and Ancient Greek that feature both lexical
      richness and grammatical complexity.

      As for conlangs, Toki Pona is an example of grammatically and lexically
      sparse language.

      I believe, it is harder to create a lexically rich language than a
      grammatically complex one, so, most conlangs tend to be grammatically
      complex. Although, it would be interesting to develop an isolating conlang
      with lots of absolute sinonyms for example, I can hardly see anyone
      investing that much effort in a conlang.


      2013/3/20 Daniel Bowman <danny.c.bowman@...>

      > Hi All,
      >
      > Alex Fink and I had a very interesting conversation today where we
      > considered how lexical richness may (or may not) have an inverse
      > relationship with grammatical complexity. I am interested to hear what
      > others on the list think of this concept, and I'm particularly excited to
      > know if anyone's considered this while designing their conlangs.
      >
      > This is how it works:
      >
      > If a language has a large lexicon, it may be able to use words to describe
      > situations that other languages grammaticalize. For example, English does
      > not grammaticalize formality (unlike Korean and Japanese). Therefore,
      > English speakers have to use words to describe a situation that a Korean
      > speaker would mark using a certain formality inflection. English is richer
      > in vocabulary for formality, whereas Korean is richer in grammar. If
      > Korean has less words for formal situations than English, this would lend
      > support to the inverse relationship hypothesis.
      >
      > Another example showing the inverse: my conlang Angosey has evidentiality
      > markers. One of these markers indicates that the speaker considers the
      > source of information doubtful. In this case, I have obviated the need for
      > the word "doubt" since I have a grammatical construction for it. I can
      > likely do away with "dubious, unsubstantiated, unlikely" etc, or at least
      > greatly reduce my usage of these terms.
      >
      > I think the absolute inverse relationship is unlikely to hold - I am sure
      > there's a situation where I would need a word for "doubt" in Angosey and be
      > unable to replace it with my evidentiality marker. However, such markers
      > may push certain words - such as "doubt" below the "common use" threshold
      > we recently discussed in the English word count thread. In other words,
      > the word "doubt" will exist, but it will be used quite seldom since the
      > evidentiality marker replaced most of its occurrences.
      >
      > Irrespective of whether or not the lexical richness vs grammatical
      > complexity holds for natlangs, it poses an interesting puzzle for
      > conlangers. Is it possible to design a very lexically rich, grammatically
      > minimal conlang? Is it easier to do this than to make (and use) one that
      > is both grammatically and lexically sparse?
      >
      > Conversely, is it possible, or do we have examples of, languages with a
      > very minimal lexicon with a correspondingly rich grammar? Perhaps Ithkuil
      > is an example of this?
      >
      > Danny
      >
    • R A Brown
      ... [snip - all points noted] ... plus was simply taken from Trask, and being used for convenience (i.e. not having to think of something else - the same
      Message 56 of 56 , Mar 28, 2013
      • 0 Attachment
        On 28/03/2013 13:30, And Rosta wrote:
        > R A Brown, On 27/03/2013 15:32:
        [snip - all points noted]

        > I agree with the basic idea, but "plus" needs to be
        > tightened up, in ways too complicated to fit in an
        > off-topic email discussion, but in simple terms "{CAR} +
        > [plural]" means "{CAR} when it is the phonological shape
        > corresponding to a plural noun node in syntax.

        'plus' was simply taken from Trask, and being used for
        convenience (i.e. not having to think of something else -
        the same applies to shape of brackets). Trask, of course,
        was merely giving a fairly simple dictionary entry, rather
        than elaborating any particular theoretic viewpoint.

        > In {CHILD} + [plural], the shape of {CHILD} is the stem
        > //tS.I.l.d// + //rn// (roughly), but I don't see any
        > grounds for saying that [Plural] is instantiated as
        > //rn//.

        I don't think we're many miles apart - probably coming at
        things from different angles.

        But I'm not intending at the moment to work out any
        hard-and-fast system - I haven't got time for one thing.

        It seems that it is only you and I now exchanging emails on
        this off-topic discussion, and it has certainly helped clear
        some of my thinking - not enough, perhaps, but it can wait.

        We are, I think, both agreed that morphemes, whatever they
        are, are not identical to "units of meaning", which is what
        sparked off this thread. As I say, I don't think we're
        miles apart.

        --
        Ray
        ==================================
        http://www.carolandray.plus.com
        ==================================
        "language … began with half-musical unanalysed expressions
        for individual beings and events."
        [Otto Jespersen, Progress in Language, 1895]
      Your message has been successfully submitted and would be delivered to recipients shortly.