Resumen de Two quantitative methods of studying phraseology in English
Word frequency lists are a standard resource for many theoretical, descriptive and applied questions. However, due to severe problems of definition, there are no equivalent lists which give the frequency of phrases. This paper proposes two independent methods of studying the frequent phraseology of English. First, using a data-base of the most frequent collocations between word-forms in a 200-million word corpus, the strength of attraction between pairs of content words is discussed. Second, using a corpus of 2.5 million words, some of the most frequent phrases, in the sense of strings of uninterrupted word-forms, are identified, and their lexical, grammatical and semantic features are discussed.
