Background data for: Some obstacles to replication in corpus linguistics

This dataset contains tabular files recording occurrences and frequencies of modal verbs in the Brown family corpora; nine modal verbs (can, could, may, might, must, shall, should, will, would) and six corpora are considered (Brown, LOB, Frown, FLOB, BE06, AmE06). Tokens were retrieved using the CQPweb interface provided by the University of Lancaster, and the tables include information on several text-level variables (text length, broad genre, text category, corpus, time period, variety). The data are provided in two formats: (i) in case form, where each token (77,872 in total) is listed separately, including information on the context of occurrence (10 words to the left and 10 to the right); and (ii) in frequency form, which aggregates occurrences by providing information on how often each modal verb appears in every text, thus including one row per text-modal combination (27,000 in total: 6 corpora x 500 texts x 9 modals).