Classification and language history

Typology and language history

We have so far dealt with two kinds of language classification, namely genealogical classification and areal classification. The first of these groups languages together into language families on the basis of shared features which have been retained during a process of divergence from a common ancestor, the second groups them into linguistic areas on the basis of shared features which have been acquired through a process of convergence resulting from spatial proximity. It will thus be seen that both the genealogical and the areal systems of language classification depend upon the interpretation of shared isoglosses as resulting in one way or another from the past history of the languages concerned. It is this diachronic aspect of both genealogical and areal classification which opposes them jointly to a third purely synchronic method of classifying languages. This is typological classification, which groups languages together into language types on the basis of isomorphism of structure without any regard to either their historical origin or their present or past geographical distribution (Greenberg 1957: 66–74; Robins 1973). Various structural characteristics have been proposed as a basis for typological classification but at the grammatical level there are essentially two systems, one based on morphological and the other on syntactic criteria.

The first of these, which is also the older, classes a language as being of isolating, agglutinative or inflectional (fusional) type according to the morphological structure of the word (Robins 1973: 13–17; Bazell 1958).