Abstract

The paper discusses current assumptions about constraints on language diversity in formalist models and the functional-typological approach, with particular regard to the explanations proposed for these constraints. Whereas formalist models posit qualitative constraints inbuilt in a speaker's mental grammar, linguists working in the functional-typological framework postulate general language preferences motivated in terms extragrammatical factors rooted in human cognition, social interaction, and information processing. These differences are traditionally described in terms of a contrast between language internal vs. language external, or formal vs. functional explanations. Another aspect of these explanations, however, has received comparatively less attention. In spite of the differences between formalist and functional-typological explanations, both types of explanations are result-oriented, in the sense that particular structural configurations are assumed to arise because of factors specifically pertaining to the properties of the configuration. In many cases, however, evidence about the actual historical processes that give rise to individual configurations suggests that these processes are motivated in terms of the properties of particular source constructions, rather than properties of the resulting configurations. This points to the need to adopt a source-oriented approach, one in which constraints on language diversity are actually constraints on the distribution of particular source constructions or the occurrence of particular developmental processes, rather than constrains pertaining to the resulting configurations in themselves.