2 ROBERT D. VAN VALIN, JR.
information management and discourse patterns; it is not a system in the
sense that has been assumed in twentieth-century linguistics.
RRG falls between these two extremes, differing markedly from each.
In contrast to the Chomsky an view, RRG takes language to be a system of
communicative social action, and accordingly, analyzing the communicative
functions of grammatical structures plays a vital role in grammatical
description and theory from this perspective. It is in this sense that RRG is
functionalist, but it is not radical functionalist like the emergent grammar
view.^3 Language is a system, and grammar is a system in the traditional
structuralist sense; what distinguishes the RRG conception from the stan
dard formalist one is the conviction that grammatical structure can only be
understood and explained with reference to its semantic and communica
tive functions. Syntax is not autonomous. In terms of the abstract paradig
matic and syntagmatic relations that define a structural system, RRG is
concerned not only with relations of cooccurrence and combination in
strictly formal terms but also with semantic and pragmatic cooccurrence
and combinatory relations. Hence RRG may be accurately characterized as
a structural-functionalist theory, rather than purely formalist or purely
functionalist.^4 It also differs from these other two views with respect to cog
nitive questions. Hopper (1987) rejects any cognitive interpretation of lan
guage, while Chomsky's claims about innate, autonomous formal linguistic
structures are well known. The RRG approach to language acquisition,
presented in Van Valin (1986, 1991a) rejects the position that grammar is
radically arbitrary and hence unlearnable, and maintains that it is relatively
motivated (in Saussure's sense) semantically and pragmatically. Accord
ingly, there is sufficient information available to the child in the speech to
which it is exposed to enable it to construct a grammar, and therefore the
kinds of autonomous linguistic structures posited by Chomsky are unneces
sary.
RRG differs from other theories of syntax in terms of its technical fea
tures. The way in which it most strikingly sets itself apart from Govern
ment-Binding theory [GB] and Relational Grammar [RelG] is that it posits
only one level of syntactic representation, and from this it follows that there
are no syntactic rules akin to traditional transformations, Move a, or the
relation-changing rules of RelG. The posited syntactic level corresponds to
the actual structural form of utterances, and it is linked directly to a seman
tic representation. This parallels, in general terms, the major features of
Lexical-Functional Grammar [LFG], which also does not posit any kind of
