Prolegomena to pictorial concordances

One of the most revered tools of literary scholarship is the concordance. This deceptively simple device lists the locations in a corpus where a given word (the 'target word') occurs. Each reference accompanies the word in its context. The popularity of Key Word In Context (KWIC) concordances in computer-assisted literary study attests their continuing usefulness. Such a list of references serves us well if we are lexicographers seeking different meanings for a word, or grammarians investigating its inflection in different contexts. That is, as long as our focus of attention is on the individual word or phrase, the classical concordance meets our needs. Sometimes, though, we wish to look at the text as a whole, and identify which words and themes predominate in its various parts. It has been, from time to time, popular to invoke such data in an effort to detect differences of authorship within a document. More recently, students of literature discuss such concentrations of linguistic phenomena in seeking to explain the present structure of a text. 1 These studies of literary architecture rest implicitly on three hypotheses concerning a work of literature.