Jespersen's cycle revisited: formal properties of grammaticalization

In this article I present a description and analysis of the rise and fall of multiple sentential negation in English. The cyclic development identified by Jespersen (1917), in essence a case of grammaticalization, can be traced in detail, with the proviso that in English multiple sentential negation is attested rather earlier than has been generally assumed. I analyse the development from a theoretical perspective, making crucial use, for reasons explicated below, of the current generative practice of projecting grammatical categories such as Tense, Mood, Negation as syntactic constituents, each according to the standard phrase structure format. With respect to negation, this entails postulating the NegP introduced by Pollock (1989). This perspective allows a very precise and insightful account of the historical development, which in turn shows the relevance of structure. This is important when we consider the historical development from the point of view of a typology of change: Jespersen’s negative cycle could count as a schoolbook case of grammaticalization: the semantic and morphosyntactic weakening of an erstwhile independent constituent and its subsequent entrenchment in a system of grammatical marking, as in Meillet’s original definition (1912). Grammaticalization theorists such as Hopper and Traugott (1993) generally regard this type of change as a longterm, diachronic and semantically motivated process. The account here, on the other hand, implies that grammaticalization is primarily a morphosyntactic change, and shows that the longterm development is necessarily punctuated by synchronic shifts. We will see that the history of English negation is shaped by a delicate interplay between various negation positions and strategies.