Search

The purpose of this paper is to consider how grammaticalisation, which is generally considered to be a diachronic process, can be fitted into a theory of language change that is based on the idea that change is brought about by the speaker, and hence is essentially a synchronic matter. First the question of the relation between explanation, the theory of grammar and the theory of change will be discussed, on the basis of which a number of guidelines will be suggested which should direct empirical research in the area of language change. In the second part of the paper, one particular case of grammaticalisation will be investigated, namely the development of have to in English from a possessive, full verb to a modal semi-auxiliary. It will be shown that this case contains both diachronic and synchronic aspects, which need to be kept apart. In keeping them apart, this particular case of grammaticalisation can be seen to accord with the principles of language change argued for in the first part.

Personal pronouns have distributional properties which are in important respects different from those of full NPs. Pronominal objects of a preposition, personal pronoun subjects and objects have therefore been analysed as clitics. A clitic analysis has also been proposed for adverbs. In this paper I concentrate on the question of how we can recognise clitics. Some proposed criteria are tested on the Old English material, leading to the conclusion that there is some support for a clitic analysis, but that it cannot be concluded that personal pronouns are always clitics. It is likely that they behave sometimes like ordinary NPs. Adverbs are unlikely to be clitics.