Research into the automatic recognition of anaphoric features in English is illustrated by an
examination of the pronoun “it”. Various non-referential usages of “it” are described, and
simple surface rules are proposed for identifying them. A computer program applies these
rules to a text and classifies all cases of “it” according to mode of usage. Runs were
performed on the Technical Section of the Lancaster-Oslo/Bergen Corpus of printed English;
this was divided into a development section, which was used for refining the rules, and a test …