At first sight it might seem that corpus linguistics and humor studies make odd bedfellows. Corpus linguistics seeks regularities across large bodies of text; humor often lies in the peculiarity of a single instance. The former has a reputation for concerning itself with decontextualized chunks of language, while humor often resides in the subtle exploitation of interactional cues in the context of situation. Corpus linguistics is able to search for lexical items through very large bodies of language data but, with certain exceptions, such as canned jokes, there is no special vocabulary of humor to locate. And, indeed, until recently there has been little collaboration between the two fields.