A final pre-publication version of: Culpeper, Jonathan (2011) A new kind of dictionary for Shakespeare's plays: An immodest proposal. In: Ravassat, Mireille and Jonathan Culpeper (eds.) Stylistics and Shakespeare: Transdisciplinary Approaches. London: Continuum, 58-83 [A revised (and hopefully improved!) version of Culpeper 2007]. It may contain minor errors and infelicities. Chapter Three A new kind of dictionary for Shakespeare’s plays: an immodest proposal 1 JONATHAN CULPEPER 1. Introduction The best-known classic Shakespearean ‘dictionary’ is probably Charles T. Onions’s Glossary (1986 [1911]), written in the philological tradition that characterized the Oxford English Dictionary (OED), and providing pithy definitions and illustrative quotations. 2 The proposed dictionary of the language of Shakespeare’s plays is analogous to more recent developments in dictionaries of general English, and, more specifically, the departure from the philological tradition brought about by the Collins Cobuild Dictionary of the English Language (Sinclair 1987). The Collins Cobuild is a corpus-based dictionary. This implies both a particular methodology for revealing meanings and a particular theoretical approach to meaning, as we shall see in this chapter. In particular, there is a strong empirical emphasis. There is less reliance on the vagaries and biases of editors, and a greater focus on the evidence of usage. The question of ‘what does X mean?’ is pursued through another question: ‘how is X used?’ To answer the ‘how’ question, corpus approaches deploy the whole gamut of computational techniques, in order to reveal patterns of usage in context. This inevitably involves matters of frequency. Frequency is not in fact as alien as it might seem to the literary critical ear. Any textual analysis that identifies a pattern implicitly involves frequency, as a pattern is the (full or partial) repetition of elements. In fact, the proposed dictionary goes beyond what one might find in the Collins Cobuild in a number of ways. Crucially, an additional feature proposed for the dictionary that makes it like no other is that it aims to be comparative. 3 Saying that X word occurs Y times in Shakespeare’s plays and that it has W and Z senses is less informative than contrasting those facts with those of his contemporaries (and not just writers of literary texts but writers of various text-types, including records of spoken interaction). In this way, we can reveal not just the denotative or conceptual meanings of words but also their stylistic, discoursal and pragmatic values in the general language of the period. Similarly, the plan for the dictionary is that it should also conduct internal comparisons, taking account of the distribution of items over internal genres (e.g. comedy, tragedy, history, particular characters, particular plays) and social categories (e.g. gender, role). Of course, what is revealed through these internal comparisons can be further pursued through external comparisons. For example, having identified that X is typical of women in Shakespeare, one could examine whether X is typical of women in plays by other contemporary playwrights, in ‘real life’ trial proceedings, and so on. In this chapter, I will deploy a number of case studies to show how techniques developed in corpus linguistics can be used to produce this new kind of dictionary based on usage and frequency. The case studies below are chosen to illustrate particular issues relating to the dictionary; each case study is not complete in itself.