The computer-generated concordance is probably the earliestdeveloped, least problematic, and most useful application of electronic data processing to literary texts. Not only do computers prepare concordances quickly and accurately, but they allow great flexibility of format and provide more information in their output than the traditional manual concordance. Freedom in formatting can be found in a "combinatorial" concordance program that produces contexts where words with identical roots cluster in groups of two or three.1 The possibility of increased output information is exemplified by the fact that in a concordance of dramatic works one can include the speaker of each line or word as well as the usual location.2 Such concordances can serve more than the customary needs of literary scholars. With the aid of a conventional concordance, one can trace the peculiar meaning given by an author to certain words; classify his vocabulary as nominal or verbal, concrete or abstract; or study his method of associating words in sets to create image patterns and themes. The words that yield this kind of information are usually nouns and verbs, adjectives and adverbs. The compiler of the concordance calls them "significant" words and lists every instance of them in an appropriate context; other words, the "most common" ones, are altogether omitted or listed partially. But these common words, which linguists call "structure" or "function" words (as opposed to "content" or "lexical" words), constitute the sole headings of a grammatical concordance. They consist of prepositions, conjunctions, determiners, quantifiers, whwords, the various kinds of pronouns, and the modal and tense auxiliary verbs.3 They mark syntactic relations and give formal
Read full abstract