The collocations presented in this book were mainly selected from those identified as significant by the Cambridge International Corpus of written and spoken English (now known as the Cambridge English Corpus) and also the CANCODE corpus of spoken English, developed at the University of Nottingham in association with Cambridge University Press. The Cambridge English Corpus is a vast database of real English taken from a range of sources, such as books, newspapers, advertising, letters and emails, websites, conversations and speeches, radio and television. By studying this corpus we obtained a representative picture of how English is really used and which words naturally and frequently go together. We also made much use of the Cambridge Learner Corpus, a corpus of learner English made up of exam scripts from students taking Cambridge English examinations all over the world. This particular corpus showed us what kind of collocation errors learners tend to make.
