Re: Lexical determination of word order
From: | Gary Shannon <fiziwig@...> |
Date: | Wednesday, May 31, 2006, 19:31 |
For what it's worth, some years ago I worked out a
system where each word in the language would include,
in its dictionary definition, argument placeholders,
like the dummy variables in a C++ function definition.
The arguments would be attached where indicated, and
that could vary from word to word. For example, some
verbs might be defined as <S>+verb+<O> and others
might be defined <O>+verb+<S>. Likewise some nouns
might be defined as having their adjectives come
before (<adj>+noun) and some might be defined the
opposite way (noun+<adj>). There were no word order
rules for the language in general since word order
would be entirely determined by the linkageage
specified in the dictionary for each word.
--gary
--- Peter Bleackley <Peter.Bleackley@...>
wrote:
> Khangaþyagon has a single, invariable word order,
> VSO. Its descendent,
> Mágikimnaz, had a variable word order, determined by
> definiteness (definite
> NPs occur before the verb, indefinites afterwards).
> Most of Mágikimnaz's
> descendents have found other ways of marking
> definiteness, and adopted a
> fixed word order - SOV, SVO, or VSO (OVS and rarely
> OSV can occur as marked
> word orders in some languages, but not as basic word
> orders, and no
> language ever uses VOS). Languages with SOV word
> order generally leave
> definite arguments unmarked and mark indefiniteness,
> those with VSO leave
> indefinite arguments unmarked and mark definiteness,
> while those with SVO
> either mark both or mark for definiteness on objects
> and indefiniteness on
> subjects.
<snip>