Re: Difficult language ideas

li [Jim Henry] mi tulis la

> On 9/19/06, Leigh Richards <palomaverde@...> wrote:
>
> > 1. As unambiguous as possible, especially in full
> sentences; it's easy to
> > clarify any ambiguities.
>
> I would suggest studying some existing conlangs intended
> to be unambiguous, such as Lojban.
>From what little I've learned of Lojban so far, it appears it can be as
ambiguous or as precise as the speaker wants it to be.  Maybe someone
who knows more than I do can clarify this better.

> > 2. Hard to learn, and easy to say the wrong thing. Small
> and subtle changes
> > have a large impact on the meaning, and it's unpredictable
> in that guessing
> > something new from what you already know will rarely work.
>
> Some things to try:
>
> - a high lexicalization density, so that most of the possible
> words with a given phonological shape are actually instantiated
>
> - a large phoneme inventory, with many actual words distinguished
> by only one distinctive feature
>
> - a large number of verb and noun paradigms, with the appropriate
> one for a given word not predictable from the form or semantics of
> the root
>
> - a large number of grammatical categories to be marked
> mandatorily in certain circumstances, the exact categories
> required/allowed differing for different paradigms.
>
> - some categories marked in multiple ways; e.g. distant past
> tense might be marked with a prefix, recent past tense
> with an initial consonant mutation, present tense with a
> suffix and future tense with a modal auxiliary.  Direct experience
> evidentiality might be marked with a suffix, hearsay
> evidentiality with an adverb, & inferential evidentiality gets
> null/default marking.
I've actually start working on what could be classified as a
philosophical language.  Various features are going to be used to mark
things lexically and grammatically.  So far things like voiced/unvoiced,
aspirated/unaspirated consonants can indicate small differences.  Tonal
qualities are added and so far are designated to indicate tense and
maybe noun inflection.  This obviously is just an experiment and would
not be intended for use by humans due to the many difficulties involved.