Linguistic typology and theory construction: Common challenges ahead

The goal of this article is twofold: to present a brief overview of differences and similarities between formal grammars (regardless of particular mode of implementation) and typology, and to outline the common challenge faced by both. To anticipate our conclusion, typological approaches to language and theory construction have much to offer each other, but need to engage in a real dialogue, rather than merely talk about the need for such a dialogue. Otherwise both orientations run the risk of becoming (increasingly) irrelevant. The idea that language structure is shaped by communicative needs of speakers has constituted a major thread in linguistic discourse over the last century – this idea is prominently represented in the Prague Linguistics Circle, in Roman Jakobson’s work, in Simon Dik’s Functional Grammar, and in much of linguistic typological work, which rests on the foundation of such earlier researchers. We agree with Nichols (2007) that the functional orientation of typology does not entail rejection of structure; rather, functionalism in typology manifests in the recognition of communicative, social, or processing needs as the reasons grammars are the way they are. To borrow a contrast from a different linguistic dialect, typology needs E-language to explain I-language (Chomsky 1986). This inherently functional premise of typology is sometimes misunderstood, both by non-typologists as well as by typologists themselves. Among typologists, some are eager to use functions as operational criteria for the definition of a grammatical phenomenon, which leads to statements like “[i]n a prototypical active voice, the agent [. . . ] is the most topical participant of the clause” (Givón 1990: 566), or “the structural complexity of a form tends to reflect conceptual or experiential complexity” (Newman 1996: 17), or “discourse factors – such as the saliency of the first argument – favor interpretation of the first argument as subject” (Chelliah 1997: 121). Pronouncements of this type provide