The Necessity of Syntax

[Östen Dahl]

Exactly what does "syntax is motivated by communicative needs" mean?
Does it mean that individual syntactic rules are motivated by
communicative needs? This is what Dan Everett's formulation "(most of)
syntax is
motivated by communicative needs" implies. But if one language puts
adjectives before nouns, another puts nouns first, and a third language
allows for both possibilities, are all these choices motivated by
communicative needs? Or is it the existence of syntax as such that is
motivated by communicative needs? That is what Steve Long seems to mean
when he says "any human language -- prewired or invented by humans --
will have syntax if it is to function as well as human languages do".
But that will hardly help us in comparing different languages as to
complexity, since presumably all languages have some kind of syntax. And
the postings seem to be talking past each other.

- Östen Dahl