Suppletion in personal pronouns: Theory versus practice, and the place of reproducibility in typology

Abstract

Reference grammars often set out the personal pronouns in paradigms similar to those of nouns. When the pronoun stems are phonologically different, this approach implies that we have instances of suppletion, though this is typically not discussed explicitly. In contrast, those linguists who have raised the issue directly tend to suggest that we are not dealing with suppletion. They claim rather that personal pronouns are not opposed for number (thus we would not be the plural of I ). The issue is particularly important if we wish to develop a typology of suppletion, since personal pronouns are potentially one of the commonest instances. If we separate out the two parts of suppletion (irregularity of form and regularity of meaning), the question becomes clearer. As far as form is concerned, while personal pronouns are frequently irregular, there are languages with morphologically regular pronouns (which makes the suppletion analysis for irregulars more plausible). And in terms of meaning, we find different types of number at different points on the Animacy Hierarchy; it is not a major issue that the plural of personal pronouns frequently involves associativity, since this may be found also for other items high on the hierarchy. These two main points lead us to conclude that there are languages where the personal pronouns show suppletion for number. Two subsidiary arguments, based on data from Latvian and Bengali, support this conclusion. Thus a typology of suppletion should include personal pronouns as an important part. This interesting but specific problem is set in the broader context of the need for reproducibility in typology.