Re: minimalistic grammar
From: | Garrett Jones <alkaline@...> |
Date: | Tuesday, April 9, 2002, 4:17 |
Andrew Strader wrote:
>>>
Hi, I searched the archive for this topic, but with no luck. I wonder if
anyone has experience with designing a language with a minimal grammar, that
is, with the fewest number of rules, but with which it is still possible to
construct all correct sentences.
<<<
it is simply not possible to specify a grammar for a speakable human
language with only a few rules. Any rulesets that specify only a few rules
generally assume the basic ruleset of eurolangs, and thus to provide a
*full* description of the grammar, that would require the description of all
the general european grammar rules. A person fluent only in cherokee or
japanese would render some pretty funky sentences if given only a couple
rules for some random conlang, because they only have the rules of their
native language to fall back on. It all depends on what you mean by a
minimalist grammar: minimalist for europeans, or minimalist for any language
background?
There are loads of events that have to be taken into account:
-word order of heads vs dependents and the sentence in general
-marking of semantic meaning (word order, constituent marking, etc)
-number of elements each verb can take (transitive vs intransitive)
-formation of new words by compounding or derivation (morphology)
-construction of subordinating clauses
-forming questions (yes/no questions, wh-questions)
-negation (this is actually very complicated: see the lojban grammar)
a grammar isn't complete until these and many other issues are taken care
of.
--
Garrett Jones
http://www.alkaline.org