Introducing HUM19

Abstract

This report introduces the HUM19 corpus of 19th century British and Irish novels and describes its construction. We characterise the HUM19 corpus as a specialised reference corpus that can be used both as a comparator against which other corpora or texts may be compared and as a study corpus, searchable for the linguistic features and patterns associated with 19th century British and Irish fiction. We explain how these envisioned uses informed the construction process and show that the construction of the HUM19 corpus required principled compromise with regard to sampling but that this was achieved without necessarily forfeiting corpus quality.