On Pawley’s conjecture

Abstract

Abstract This paper shows that Pawley’s conjecture that the frequency of lexical items in text corpora is positively correlated with the number of phrasal lexical items which have those lexical items as heads of phrase is confirmed. Data for testing Pawley’s conjecture are taken from two sources: Kilgarriff’s lemmatized frequency lists from the BNC of the 6,318 words which appear more than 800 times (http://www.kilgarriff.co.uk) and the around 14,000 PLIs in the Syntactically Annotated Idiom Dictionary (Kuiper et al., 2003). Why this statistical fact should be the case is a matter for further research.