There has never been a more exciting time to be working on Shakespeare’s language. If you have an internet connection, fire it up.

When Shakespeare and his contemporaries thought about language, they thought of speech: breathy, ephemeral sounds cast into the air. This very ethereality was taken as a proof of one of the key Renaissance ideas about language: inasmuch as it allowed humans to make evident their ability to reason, it was a divine gift, distinguishing humanity from, and elevating it above, the rest of creation. The gift of language could raise the monstrous to the level of the human, as it does Caliban, and the voluntary abandonment of language suggested a descent: as when the arrogant Ajax, swollen with pride at the prospect of single combat with Hector, loses his ability to distinguish social rank, along with his humanity, to a mumbling silence:

Language is social in the Renaissance in a formal, public sense we no longer quite appreciate: there is a clear link here between Ajax’s loss of language and his mistaking the lowly Thersites for the highest of the Greek generals. His loss of language is not complete: Ajax can still manage a dismissive ‘Thanks, Agamemnon’, and he can proudly declare that speaking is for beggars – but he has lost his reason, and therefore his membership of society. He is instead its laughing-stock.

SINCE THE late nineteenth century, critics have tried to group some or all of the plays Shakespeare wrote late in his career – Coriolanus, Cymbeline, The Winter's Tale, The Tempest, Pericles (co-authored with George Wilkins), Henry VIII and The Two Noble Kinsmen (both co-authored with John Fletcher) – into a single critical category, usually on the basis of thematic, dramaturgical, or linguistic similarities among members of the group. While there is no consensus on which category is most appropriate for such a grouping (designations such as ‘late plays’, ‘romances’, and ‘tragicomedies’ have been proposed), there is nevertheless a persistent feeling among Shakespeare's readers that something distinguishes several of these plays from the others and that this ‘something’ ought to be the object of critical analysis. At times the perception of such a similarity has been so strong that it has led to categorical declarations of the sort made by Gerard Eades Bentley in the mid twentieth century, who asserted that ‘no competent critic who has read carefully through the Shakespeare canon has failed to notice that there is something different [about these plays]’.

TONGUE

In ‘Shakespeare's talking animals’, Terence Hawkes makes a fundamental claim about language and Shakespeare's work. The plays, he says, contain ‘ideas about language’ which we neglect ‘because we are anaethetized to them by our own literacy’ (Hawkes: p. 69, this volume). Nothing could be more important in seeking to understand Early Modern ideas about language and use of language than becoming aware of our own narcotic unawareness of them. We are used to historicizing Shakespeare in every respect except his language, and, as Hawkes implies, our ignorance is matched only by our ignorance of our ignorance. But I would go further than Hawkes: as I will try to show in this introduction, there are not only ideas about language we miss; there are usages of language we misinterpret because we mistake the nature of language in the Early Modern period.

From the point of view of linguistics, and taken as a product of human cognition, language can be assumed to be the same thing in all cultures, and at all times in attested human history. However, taken as a cultural entity, within literary or cultural criticism, language changes radically between the Early Modern period and our own – as radically as other cultural entities such as government, religion, and duty change. In the first part of this introduction, I will try to make language strange, to give an idea of its different cultural status in the Early Modern period; in the second, I will examine the curious reality our culture has bestowed on ‘wordes’, and what this does to our readings of Shakespeare.

Introduction

Linguistic historians of English like to claim that they have the nature and origin of Standard English nailed. The standard, as any fule kno, is a non-regional, multifunctional, written variety, historically based on the educated English used within a triangle drawn with its apexes at London, Cambridge and Oxford. Even more specifically, the propagation of this ‘incipient’ standard can be linked to a particular branch of the late medieval bureaucracy: the court of Chancery.

At least, that is the standard account of the rise of Standard English in most classrooms and textbooks. In a series of articles on business texts, however, Laura Wright has now challenged the second part of this account, pointing out that the central governmental bureaucracy is not the only place where the necessary conditions for standardisation obtained; and other chapters in this book offer further evidence of a growing unease with the status of ‘Chancery Standard’ as the simple and sole source of Standard English. In this chapter, I want to question the first part of the standard account – and in particular the general theoretical basis of the hypothesis, which I take to be the evolutionary, family-tree model of language change. I claim that linguists have tended to accept what I will call the ‘single ancestor-dialect’ hypothesis (the SAD hypothesis), not because the linguistic data supports it (in fact it does the opposite), but because the family-tree metaphor demands it.

Authorship studies in the early Modern period

There is a great number of Elizabethan and Jacobean plays which are either anonymous, wrongly ascribed, or thought to be the work of more than one writer. This situation is a result of the particular context in which early Modern plays were written, acted, and published – and an understanding of this context explains many of the most common authorship problems which arise (Bentley 1971, and Wells et al. 1987:1–68 provide excellent accounts of these issues in more detail than can be given here). Such an understanding can also help in explaining the shortcomings and limitations of much previous authorship work.

The single most important factor in this context is that early Modern plays were only very rarely regarded as ‘literature’ in a sense recognisable today. They are better regarded as raw material fuelling the profitable entertainment industry of early Modern London, much as film scripts are the raw material of today's film industry. Like film scripts, they were bought from writers by acting companies and, just as today, once a script was sold, the writer lost control over it.

The demand of the theatres for new material, especially in the later sixteenth and early seventeenth centuries, was insatiable, and, again as often in the film industry, this favoured collaboration as a means of producing material of the required standard in as short a time as possible. Writers might specialise in certain types of writing – opening scenes, closing scenes, comic scenes, love scenes; and unsatisfactory scripts might be touted round various authors who would add to and cut the script in the hope of producing something playable.

In this book, I introduce a new method for determining the authorship of renaissance plays. This method relies on what I term socio-historical linguistic evidence. Simply put, the use of socio-historical linguistic evidence involves the determination and comparison of linguistic usages of renaissance dramatic authors: at this time the English language is changing so rapidly that it is possible to distinguish between the grammatical usages of certain writers, even though they are writing in the same place (London), and at the same time (c. 1590–1625).

The methodology I use to determine and compare linguistic usages is based on theories of language variation and change developed in the fields of socio-linguistics and socio-historical linguistics by William Labov and Suzanne Romaine (see Labov 1972, Romaine 1982). As I argue in the Introduction, the use of socio-historical linguistic evidence offers a more reliable means of resolving authorship debates surrounding Elizabethan and Jacobean plays than has previously been available.

This book is divided into three parts. The first part concerns the methodology of socio-historical linguistic evidence, and the grammatical features investigated in this particular study. It should be of particular interest to historical linguists and authorship scholars who want to understand the precise basis of the method. One of the strengths of socio-historical linguistic evidence as opposed to other current approaches to authorship, is that its basis should be readily appreciable by anyone familiar with the language or literature of the period: no statistical background is required.

The use and limitations of socio-historical linguistic evidence

This book has shown that it is possible to use the varying rates of standardisation in the work of early Modern dramatists as evidence in authorship studies. Socio-historical linguistic evidence can be used to its fullest effect in cases where the investigator wishes to divide a play between two named candidates, for each of whom there exists a group of non-controversial plays suitable for use as a comparison sample. This is the case in the Shakespeare–Fletcher collaborations, where marked differences in linguistic usage between Shakespeare and Fletcher allow detailed comments on the likely authorship of individual scenes to be made. In such cases, knowledge of the minimum significant sample size, gained from tests on plays of known authorship, allows an assessment of the relative certainty of any ascription to be made. In some cases however, the linguistic usages of collaborators may be so similar as to preclude the use of this type of evidence in authorship work.

Even where specific candidates for authorship have not been agreed upon, or where there are not comparison samples available for all candidates, it is still possible for this type of evidence to make limited contributions to the authorship debates surrounding texts. Where suitable comparison samples exist for only one suggested collaborator in a play, for example, it may be possible to rule that writer out of certain portions of the play, or indicate the relative likelihood of their having written other sections.