Theiling Online    Sitemap    Conlang Mailing List HQ   

Re: Logical?

From:And Rosta <a-rosta@...>
Date:Wednesday, June 12, 2002, 1:00
Mathias:
> And Rosta <a-rosta@...> wrote: > >>> > Self-segreting morphology is not necessarily a requirement; > nonselfsegregating morphology will not necessarily result > in lexical ambuity, and nonselfsegregating words result > only in holistic ambiguity, not in specifically syntactic > ambiguity. > <<< > i had quite another experience learning to speak a language with a almost > all-CV pattern, namely japanese.
[...]
> when you learn a language, you don't know all the vocabulary, so that you > can't guess where words start or end within a cluster. i could have heard > "aratani" ("anew") as "Ara-Tani" ("Rough Valley") and "akumademo" > ("forever") as "aku ma demo" ("even the opening space"), "juuichi-kyou" ("a > bit more than eleven") as "eleven kyou" and wondered "what's a "kyou"?",
[...]
> a good example was "otoko no ko to" "with the boy" from another thread which > could also mean "otoko no koto" "regarding the boy" (as in "otoko no koto > suki da"). > > therefore, practical experience makes me believe that a spoken language > whose grammatical tags or shape of words don't make words clearly > identifiable from each other is difficult to understand for a student > listener. i also think that very few students would be able to parse a > sentence according to only a prosody that is foreign to them. although i > must concede that the stress on the penultimate syllable of words helped me > a great deal understanding some other spoken natlangs.
As you concede, natural languages may have phonologies that make word boundaries easier to determine (by means of stress placement, final obstruent devoicing, phonologically unique suffixes, etc.), and so drastically reduce the amount of ambiguity that causes communication problems. But afaik (& I may be wrong) no natlang has completely self-segregating words. In the case of Livagian, it is trivially easy to determine roughly where word boundaries are, but more complex rules are need to determine exactly where they are. If Livagian were all artlang and no engelang, I would scrap the more complex rules, for the needs of efficient quotidian conversation does not require them. --And.