Abstract

The English particle verb alternation exemplified in “John put on the kettle” vs. “John put the kettle on” is one of the best – and best-researched – examples of a referentially-free word order alternation in English. This talk situates the variable historically, sociolinguistically, and grammatically, and goes on to discuss two acceptability judgment experiments, work conducted in collaboration with Bill Haddican. The experiments show that the known effects of object weight and information structure can be observed in perception as well as production data, and their independence can be seen more clearly. The judgment method also allows us to see how these factors affect the two word orders separately. Instead of imposing a binary-choice framework on the alternation, we argue that in some cases one emerges from the data. Speakers’ judgments give evidence for “grammar competition”. The between-speaker correlations in the experimental results take us into more uncharted territory. We explore the possibility that the quantitative co-variation of two effects can be taken as evidence for their grammatical similarity. Interaction between effects, again under some theoretical assumptions, means they must operate at the same level. But despite these tools, we are far from an elegant understanding of the particle verb ‘alternation’. For if both phonology and discourse affect word-order variation, through a processing-based competition, we are led to question the role of syntax in variation, if indeed any remains.