Stylistics and Shakespeare's Language: Transdisciplinary Approaches

Stylistics and Shakespeare’s Language Advances in Stylistics Series Editor: Dan McIntyre, University of Huddersfield,...

Author: Jonathan Culpeper | Mireille Ravassat

134 downloads 1230 Views 8MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

Stylistics and Shakespeare’s Language

Advances in Stylistics Series Editor: Dan McIntyre, University of Huddersfield, UK Editorial Board: Beatrix Busse, University of Berne, Switzerland Szilvia Csábi, Eötvös Loránd University, Hungary Monika Fludernik, University of Freiburg, Germany Lesley Jeffries, University of Huddersfield, UK Jean Boase-Beier, University of East Anglia, UK Geoffrey Leech, Lancaster University, UK Larry Stewart, College of Wooster, USA Manuel Jobert, Jean Moulin University, Lyon 3, France. Titles in the series: Corpus Stylistics in Principles and Practice Yufang Ho D. H. Lawrence and Narrative Viewpoint Violeta Sotirova The Discourse of Italian Cinema and Beyond Roberta Piazza I. A. Richards and the Rise of Cognitive Stylistics David West Opposition in Discourse Lesley Jeffries Style in the Renaissance Patricia Canning The Stylistics of Chick Lit Rocío Montoro

Stylistics and Shakespeare’s Language Transdisciplinary Approaches

Editors

Mireille Ravassat and Jonathan Culpeper

Continuum International Publishing Group The Tower Building 80 Maiden Lane 11 York Road Suite 704 London New York SE1 7NX NY 10038 www.continuumbooks.com © Mireille Ravassat and Jonathan Culpeper and contributors 2011 All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage or retrieval system, without prior permission in writing from the publishers. British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library. EISBN: 978-1-4411-6425-4

Library of Congress Cataloguing-in-Publication Data A catalog record for this book is available from the Library of Congress.

Typeset by Newgen Imaging Systems Pvt Ltd, Chennai, India Printed and bound in India

Contents

List of Illustrations

vii

Notes on Contributors

ix

Introduction Mireille Ravassat and Jonathan Culpeper Chapter 1: ‘Strange deliveries’: Contextualizing Shakespeare’s First Citations in the OED Giles Goodland Chapter 2: Shakespeare’s Vocabulary: Did it Dwarf All Others? Ward E. Y. Elliott and Robert J. Valenza

1

8

34

Chapter 3: A New Kind of Dictionary for Shakespeare’s Plays: an Immodest Proposal Jonathan Culpeper

58

Chapter 4: ‘If I break time’: Shakespearean Line Endings on the Page and the Stage Peter Kanelos

84

Chapter 5: Subject-Verb Inversion and Iambic Rhythm in Shakespeare’s Dramatic Verse Richard Ingham and Michael Ingham

98

Chapter 6: Shakespeare’s ‘Short’ Pentameters and the Rhythms of Dramatic Verse Peter Groves

119

Chapter 7: Wholes and Holes in the Study of Shakespeare’s Wordplay Dirk Delabastita

139

vi

Contents

Chapter 8: ‘a thing inseparate/Divides more wider than the sky and earth’ – of Oxymoron in Shakespeare’s Sonnets Mireille Ravassat Chapter 9:

‘Rue with a difference’: a Computational Stylistic Analysis of the Rhetoric of Suicide in Hamlet Thomas Anderson and Scott Crossley

165

192

Chapter 10: Shakespeare’s Sexual Language and Metaphor: a Cognitive-Stylistic Approach José L. Oncins-Martínez

215

Chapter 11: Cognitive Interplay: How Blending Theory and Cognitive Science Reread Shakespeare Amy Cook

246

Index

269


Figures 1.1 2.1 5.1

7.1

Proportion of EEBO and other provenances in sample batch. A comparison of normalized word type frequency profiles for Shakespeare and the King James Bible. Syntax of subject pronoun after an initial non-subject constituent: percentage of VSpro order overall, and percentage of VSpro order when the pronoun was in an even (ictic) syllable. Defining wordplay at a crossroads of four continuums.

16 44

108 150

Tables 2.1 2.2 3.1 3.2 3.3 3.4 3.5 3.6 5.1

Three vocabulary tests: Shakespeare and eight others. Types-tokens, new-to-the-group words, total inferred vocabulary, Shakespeare and other writers (appendix). Word-types and tokens in Shakespeare’s plays. The top ten ranked-ordered collocates of ‘good’ within a five-word span. The top ranked-ordered three-word lexical bundles in Shakespeare and other texts. Rank-ordered keywords for Romeo and Juliet. The semantic categories used (derived from McArthur 1981). Love comedies and tragedies: characteristic semantic categories (rank-ordered). Syllable position of subject pronouns accompanying monosyllabic verbs.

42 55–7 65 67 73 75 77 77 106

viii

5.2

5.3

5.4

9.1 9.2 9.3

9.4


Frequencies of VS and SV syntax with main clause subject pronouns in inversion contexts featuring a monosyllabic finite verb. Frequencies of inverted subject pronouns in even and odd syllables with monosyllabic verbs in Shakespeare’s dramatic verse. Frequencies of main clause VSpro and SproV syntax in inversion contexts with monosyllabic verbs, with subject pronouns in even syllables. Corpora descriptive statistics. Mean, standard deviation, and F scores for LIWC values (Hamlet’s versus Ophelia’s suicidal discourse). Mean, standard deviation, and F scores for LIWC values (Ophelia’s suicidal discourse versus non-suicidal discourse). Mean, standard deviation, and F scores for LIWC values (Hamlet’s suicidal discourse versus non-suicidal discourse).

106

107

107 200 201

202

204


Thomas Anderson is Assistant Professor of English at Mississippi State University, USA. His book, Performing Early Modern Trauma from Shakespeare to Milton (Ashgate 2006), looks at the way historical crises such as the Reformation, regicide and royal death are registered in early modern literature. He has published articles on Shakespeare, Marvell and Milton and has recently co-edited with Ryan Netzley a collection of essays on John Foxe’s Actes and Monuments entitled Acts of Reading: Interpretation, Reading Practices and the Idea of the Book in John Foxe’s Actes and Monuments (University of Delaware Press 2010). Amy Cook is Assistant Professor at Indiana University, Bloomington, USA. Her book, Shakespearean Neuroplay: Reinvigorating the Study of Dramatic Texts and Performance through Cognitive Science (Palgrave Macmillan 2010) uses Hamlet’s mirror held up to nature as a test case and provides a methodology for applying conceptual blending theory to Shakespeare’s textual theatrics. She has published essays in Harold Bloom’s Modern Critical Interpretations series, Theatre Journal, TDR, and SubStance, among others. Scott Crossley is Assistant Professor of Applied Linguistics and ESL at Georgia State University, USA, and he maintains a close affiliation with the Institute for Intelligent Systems at the University of Memphis, USA. His interests include computational linguistics, corpus linguistics, discourse processing and discourse analysis. He has published articles in genre analysis, multi-dimensional analysis, discourse processing, speech act classification, psycholinguistics, second language acquisition and text linguistics. Jonathan Culpeper is Professor of English Language and Linguistics at Lancaster University, UK. He is particularly interested in written (re)presentations of spoken interaction. He has published works in the area of stylistics, especially the stylistics of drama. His most recent major publication, co-authored with Merja Kytö, is Early Modern English Dialogues: Spoken Interaction as Writing (Cambridge University Press 2010).

x


Dirk Delabastita is Professor of English literature and literary theory at the University of Namur, Belgium. He wrote his PhD on Shakespeare’s wordplay in Hamlet and the problems of translating it (There’s a Double Tongue, 1993). He edited Wordplay and Translation (1996) and Traductio. Essays on Punning and Translation (1997) and co-authored a Dutch-language dictionary of literary terms (7th edition, 2007), translated into French. He also co-edited European Shakespeares (1993), Fictionalizing Translation and Multilingualism (2005) and Shakespeare and European Politics (2008). Ward E. Y. Elliott is Professor of American Political Institutions at Claremont McKenna College, California, USA. He and his colleague, Robert J. Valenza were co-advisors of the Claremont Shakespeare Clinic, 1987–1994. They have published articles on Shakespeare authorship, notably in Shakespeare Quarterly, Computers and the Humanities, Literary and Linguistic Computing, among others. Ward Elliott is co-dedicatee of Brian Vickers’s Shakespeare Co-Author (OUP 2002). Giles Goodland is Senior Editorial Researcher at the Oxford English Dictionary, UK. He specializes in the use of databases of early modern English (such as EEBO and ECCO) to provide information in the revision of OED entries, helping to supervise a small team of Research Assistants in this task. He has a D. Phil. in English Literature from Oxford University. He has had several books of poetry published, most recently What the Things Sang (Shearsman 2009). Peter Groves did his PhD on Shakespeare’s prosody at Cambridge University, UK. He is a Lecturer in Shakespeare, Renaissance literature and poetry at Monash University, Melbourne, where he co-edits the e-journal Versification: an Electronic Journal of Literary Prosody. He has published a book and a number of articles on metre and versification in English, the most recent being ‘Shakespeare’s Pentameter and the End of Editing’, in Shakespeare, the Journal of the British Shakespeare Association. Richard Ingham is a Senior Lecturer in English at Birmingham City University, UK, where he teaches language acquisition, English grammar and the linguistic history of English. His research interests are in language acquisition and change, with special reference to English. He has published in a large number of international refereed journals. His most frequently cited publications are on Middle English negation and pronoun use in English child language. His current research focus is on language in medieval England.


xi

Michael Ingham is Associate Professor in the Department of English, Lingnan University, Hong Kong, where he has been teaching English Studies since 1999. His areas of professional expertise are speech and drama, literature and literary linguistics and performance studies, particularly Shakespeare in performance. He is a founder member of Theatre Action, a Hong Kong based drama group that specializes in action research on literary drama texts. Peter Kanelos is Assistant Professor of dramatic literature, theatre history and dramaturgy at Loyola University’s Department of Fine and Performing Arts, Chicago, USA. He is the editor of Much Ado About Nothing (The New Kittredge Shakespeare), associate editor of Twelfth Night (The New Variorum Shakespeare) and co-editor of the forthcoming, Thunder at a Playhouse: Essaying Shakespeare and the Early Modern Stage. He has lectured internationally on Shakespearean drama, and is currently working on two book projects on Shakespeare. José L. Oncins-Martínez is Lecturer in English language and literature at the University of Extremadura, Spain. He is the author of a volume on the translations of Shakespeare’s Timon of Athens into Spanish and co-editor of three collections of essays published in Spain. His academic interests include topics such as translation, cognitive linguistics, pragmatics and stylistics, on which he has published several articles. At the moment he is engaged in a research project on Shakespeare’s phraseological language. Mireille Ravassat is a Senior Lecturer in the Department of English Studies at Valenciennes University, France, where she teaches English literature – mainly Shakespeare – stylistics and translation from English into French. Her main fields of interest and investigation are Shakespeare’s language and stylistics. She has contributed many articles and book chapters on this topic and regularly organizes international seminars on Shakespeare, notably at ESSE conferences. She is currently working on a monograph on Shakespeare’s style in the Sonnets. Robert J. Valenza is the Dengler-Dykema Professor of Mathematics and the Humanities at Claremont McKenna College, California, USA, and author of books on linear algebra, abstract algebra and number theory. With Ward E. Y. Elliott, he was faculty advisor to the Claremont Shakespeare Clinic from 1987 through 1994. They have published widely on the issue of Shakespeare authorship.

Introduction Mireille Ravassat and Jonathan Culpeper

Even if recent times testify to a renewed surge of interest in the issue of Shakespeare’s language and style, the following statement by M. C. Bradbrook is still arresting and valid in the early twenty-first century: ‘There is no question relating to Shakespeare as a writer which does not involve his style . . . . Yet on this central problem comparatively little has been written. It is too vast and intimidating. . . ’(Shakespeare Survey 1954: I)1. Some four decades later, in an issue of the same journal also devoted to Shakespeare and Language, Stephen Booth chimes in somewhat provokingly: ‘Shakespeare is our most underrated poet’. He adds: ‘The reason it is necessary to point out Shakespeare’s poetic superiority to competing poets is, I think, that we have so long, so industriously ignored the qualities in literature that drew us to it in the first place’ (1997: I).2 In other words, there is definitely still room for far-reaching explorations of Shakespeare’s multifarious artistry and mind-boggling inventiveness in matters of language. Moreover we decided to edit the present volume because we felt such an enterprise was a necessary, a timely and appropriate one, with respect to bridging the gap between Shakespeare and language studies, while taking into account new approaches like computational, corpus based and cognitive studies. In addition, we intend this book to be a synthesis of linguistic and literary criticism, too often alien approaches despite the in-between development of the concept of linguistically-oriented ‘new stylistics’.3 Here the following statement by Geoffrey Leech and Michael Short in Style in Fiction (2007 [1981]) is most relevant: One major concern of stylistics is to check or validate intuitions by detailed analysis, but stylistics is also a dialogue between literary reader and linguistic observer, in which insight, not mere objectivity is the goal . . . . [l]iterary expression is an enhancement, or a creative liberation of the resources of language which we use from day to day. Correspondingly, stylistics builds on linguistics, and in return, stylistics challenges our linguistic

2

Stylistics and Shakespeare’s Language frameworks, reveals their deficiencies, and urges us to refine them. In this sense, stylistics is an adventure or discovery for both the critic and the linguist.4

Stylistics has actually been ‘something like a revolution in the relation between linguistic and literary studies’ to quote Randolph Quirk in his foreword to Style in Fiction (x). We are happy to edit this volume at a time when stylistic studies have shown significant development. The success of the Poetics and Linguistics Association (PALA) conferences testifies to this trend, for example, the 2006 one in Finland devoted to ‘The State of Stylistics’5. In France the growth of the Société de Stylistique Anglaise is yet another indicator. It is also a time when going beyond boundaries is no more a dream, as, for example, the 2008 Queen’s University Belfast conference on ‘Text and Context: Bringing Early Modern Literature and Linguistics Together’ has shown. The increasing number of participants at sessions on Shakespeare and style in which we are involved, notably at ESSE conferences, is also very encouraging. The idea of tackling Shakespeare’s language and style has really involved a stimulating cooperation. Through it, as a literary scholar and a linguist ourselves, we are glad to offer the following transdisciplinary studies which demonstrate that there is new life in traditional approaches as well as entirely new ones, charting the map of most innovative research in the joint areas of language and literature. The chapters in the collection capture a rich diversity of points of view and cover fields such as lexicography, versification, rhetorical analyses, cognitive and computational corpus-based stylistic studies. The perspective is deliberately a broad one, the purpose being to confront ideas and visions at the intersection of various techniques of textual investigation. We wish to warmly thank all the authors of this volume for their contributions and also all the team of specialists who kindly accepted to review parts of this book and did it with close attention and expertise. *

*

*

The opening chapter of this volume by Giles Goodland questions the originality of Shakespeare’s lexicon. Perhaps due to the prevalence of first citations by Shakespeare in the Oxford English Dictionary, first known as New English Dictionary, it is a commonly held item of belief that Shakespeare was a great coiner of words, but this has often been asserted in an impressionistic and peremptory manner. Now, in the last few years it has become

Introduction

3

possible to address this issue with greater confidence since a large body of Early Modern printed material has become available for searching on Early English Books Online and similar databases. In this chapter, Giles Goodland focuses on some sections of the dictionary and traces to what extent Shakespeare’s antedatings in the OED are true first uses, and to what extent they are a product of the prominence and availability of Shakespeare’s texts at the time at which the dictionary was originally written. His demonstration paves the way for a useful reassessment of Shakespeare’s lexical creativity. Chapter 2 by Ward Elliott and Robert Valenza also tackles this specific issue. Indeed bardolatry lives on, notably in the still-common notion that Shakespeare’s vocabulary dwarfed all others. The authors debunk this myth by analysing Renaissance texts and using three tests of vocabulary size and richness: type/token ratios, well-known measures of variety, new-to-thegroup words, a test measuring distinctiveness and inferred vocabulary, a new test based on the top half of inferred frequency curves. All three tests say that Shakespeare’s vocabulary was larger and richer than John Fletcher’s, smaller and poorer than Milton’s, and about equal to those of five other leading Elizabethan playwrights, and to those of most college-educated people today. The authors demonstrate the problem has not been overestimation of Shakespeare, but underestimation of everyone else. In that, their research most usefully complements Chapter 1. Jonathan Culpeper’s study concludes the lexicography section of this volume. The language of Shakespeare’s plays has received substantial treatment in various ‘dictionaries’, ‘glossaries’, ‘lexica’ and ‘concordances’. However, the classic works are written in the philological tradition that characterized the Oxford English Dictionary. This chapter explores how modern principles and techniques developed in Corpus Linguistics can be deployed in the creation of a radically new kind of dictionary. In particular, this involves a focus on usage and frequency. A further innovation is that the proposed dictionary will be comparative, making both internal comparisons (e.g. female characters compared with male) and external comparisons (e.g. Shakespeare’s usage compared with that of contemporary plays and other genres). The bulk of this chapter is made up of case studies, involving discussion of the words ‘horrid’, ‘good’, ‘ah’ and ‘and’, multiword units, and linguistic profiles for characters and plays. Through these, the aim is to demonstrate the characteristics of the dictionary and raise pertinent issues, including, for example, how many and what kind of words to include in the dictionary, whether the dictionary should include only words (and how they should be defined), how word-senses should be distinguished, how

4


stylistic and social meanings should be captured, and what approach to grammar should be taken. This study is not narrowly focused on dictionary design, but taps into key stylistic issues in Shakespeare’s plays, including characterization, genre and theme. The three following chapters deal with another important topic, namely Shakespeare’s verse. Peter Kanelos, a professor and editor of Shakespeare, with long-standing experience of working with actors and stage directors, probes into one of the hallmarks of Shakespeare’s mature verse, namely the issue of his increasingly enjambed style. Indeed, the more Shakespeare’s career progressed, the more his line patterning became ‘irregular’, thereby reflecting his more complex vision of the universe. Relying on his dramaturgical background, Peter Kanelos focuses on the issue whether Shakespeare advocated a flowing diction unimpeded by any pause in the phrasing, or conversely meant to mimic the metrical arrangement by uttering the enjambed lines in succession. By getting to the heart of this question, he sheds light on the development of Shakespeare’s poetics. In Chapter 5, Richard and Michael Ingham point out the fact that subjectverb inversion was still common as a stylistic device in late sixteenth-century English and they investigate here what interaction there may have been between ictic stress placement (i.e. the stress is placed on an even-numbered syllable) and the use of syntactic inversion in Shakespeare’s verse drama. The frequencies of subject-verb inversion with respect to the placement of ictic stress in blank verse pentameters are compared in 15 non-comedy plays from Titus Andronicus to Cymbeline. Shakespeare’s practice changed very markedly with respect to ictic stress and subject pronoun inversion between the earlier and later plays. In the later plays, in cases when iambic metre places ictic stress on a pronoun, the syntax strongly tends to avoid inversion. Their findings notably highlight that the decline in ictic stress placement reflects an independent stylistic choice on the author’s part and is not simply a by-product of a decline in inversion. Their study of syntax and metre is relevant to debates about the value of iambic metre to modern-day actors speaking Shakespearean verse. Awareness of stylistic change in versification over the course of Shakespeare’s career should enhance understanding and informed choice among the wider discourse community of those who perform Shakespeare’s language. In Chapter 6, Peter Groves scrutinizes Shakespeare’s innovative use of lacunae (empty slots in the metrical codified pattern of iambic pentameter). Lacunae represent either ‘silent beats’ or ‘silent off-beats’, produced in our negotiation as performers between written line and known phonological and metrical constraints. The silent off-beat comes itself in two

Introduction

5

distinct flavours: the ‘jolt’, which occurs between intonational phrases, and which has the effect of emphasizing discontinuity, and in mimetic terms of suggesting surprise, alarm, anger, urgency and so on; and the much rarer ‘drag’, which occurs within a phrase and has the effect of locally slowing down the tempo and forcing pitch-accent emphasis on the syllable that precedes it. The silent beat or ‘rest’, on the other hand, functions as a way of cueing gesture and action in the theatre, and in addition has a number of interesting deictic uses, marking and drawing attention to features of the interaction between characters. The chapter explores the ontology and epistemology of lacunae – how as readers we recognize them, and as performers produce them – and investigates in detail the sorts of aural and experiential effects they produce in performance. In the following chapter, Dirk Delabastita presents a comprehensive, flexible and radically historical model of Shakespeare’s wordplay. A systematic set of critical distinctions is proposed that needs to be considered in the discussion of puns. In this way, various types of wordplay can be distinguished from various types of non-wordplay. Quite crucially, it is argued that these are gradual and very much context-sensitive distinctions rather than radical ones. Moreover, by permitting the description of diachronic shifts along the definition’s various continuums, such distinctions can be viewed from a historical angle, which takes into account factors such as phonetic and semantic evolution but also the historicity of interpretation (after all, it is our interpretative and discursive strategies as spectators, readers and critics that make or mar the puns that could potentially be attributed to Shakespeare’s texts). The model is original inasmuch as it tries to transcend the limitations inherent in the narrowly ‘linguistic’ or ‘literary’ approaches that have dominated the field until now: it reconciles the need for maximum descriptive rigour with an acknowledgement of the hermeneutic indeterminacies and the role of interpretative strategies and interpretative communities inevitably involved in the identification and understanding of wordplay. Examples are taken from a range of Shakespeare plays and the chapter enters into a dialogue with wordplay critics as wide apart as the Victorian F. A. Bather and our contemporary Patricia Parker. In Chapter 8, Mireille Ravassat focuses on the use of oxymoron in Shakespeare’s Sonnets. In a contextualized perspective, she offers a comprehensive and systematic study of this ‘present absent’ figure of speech and thought ruling the various intertwined plots of this poetic collection. She argues that a unified perception of difference within similarity and similarity within difference is the doubly chiastic and oxymoric pattern generally defining Shakespeare’s bifocal vision in his dramatic and poetic production.

6


She further demonstrates that in the Sonnets, which are at the crossroads of the baroque vein of ontological vacillation and of a modernized vision of the concept of psychomachia, oxymoron violates its ethos of reconciliation in favour of unresolved division. Far from an ideal of oxymoron as ‘heavenly mingle’ (Antony and Cleopatra 1.5.62), the figure of the Crosse-couple, in the Sonnets, turns out to spawn ‘compounds strange’ (76.4) and incongruous lexical and semantic creations. Ultimately, systematically undoing what it does, oxymoron maps out a discordant pattern of dissolution in fusion that signals the birth of the modern self. This study reassesses oxymoron as a very rich and complex stylistic device in Shakespeare which, by resorting to clashes of opposites and the explosion of semantic categories, leads the reader very deep into the human psyche. In Chapter 9, Thomas Anderson and Scott Crossley examine the rhetoric of suicide in Hamlet and demonstrate how computational stylistic methods enrich literary approaches to the play. Using the computational tool LIWC (Pennebaker et al. 2001), they conduct a lexico-semantic and corpus analysis of the play to examine the ‘semantic prosodies’ present in the suicidal and non-suicidal discourses of Hamlet and Ophelia. A stylistic and literary interpretation of the play demonstrates that the lexical variation between the characters supports the interpretation that the two ‘suicides’ are discrete events with different cultural valences. Statistical analyses demonstrate that while Ophelia’s apparent ‘suicide’ registers the rhetoric of religious sin that offers a narrative of appropriate punishment by Renaissance standards, Hamlet’s suicidal rhetoric highlights the evolving cultural meaning of suicide, removing it from its religious narrative and placing it within an emerging discourse of affect and societal disengagement. The representation of Ophelia’s ‘suicide’ includes the threat of legal and theological condemnation while the rhetoric of Hamlet’s death embodies a new cultural understanding of suicide that avoids the stain of criminality and sin. In Ophelia’s words, the two suicidal characters ‘rue with a difference’ (4.5.178). In Chapter 10, José L. Oncins-Martínez proposes a cognitive stylistic approach to Shakespeare’s sexual language. He offers a panoramic survey of the field and, more specifically, he sets out to explore the metaphors that Shakespeare uses for the expression of sexual matters in his plays and poems and their stylistic import. As eminent critics have demonstrated, much of Shakespeare’s imagery is informed by this particular topic, but since the 1980s, this notion of imagery has been pointed out as lacking terminological rigour. It has been notably reassessed within Conceptual Metaphor Theory (CMT), already fruitfully used, since Lakoff and Johnson’s (1980) influential work, for the analysis of Shakespeare’s style on the one hand and

Introduction

7

of sexual language in non-literary contexts on the other hand, but not for exploring Shakespeare’s sexual language itself, as proposed here by José L. Oncins-Martínez, with a specific focus on the ‘sex is war’ source domain. In the final chapter, Amy Cook argues that cognitive linguistics – Conceptual Blending Theory (CBT) in particular – illuminates the dynamic interplay between language, cognition, and the body in ways that impact on practical and theoretical issues in performance. The key endeavour here is to explain how a meaning comes about, not simply what the meaning is. She neatly demonstrates this with reference to Laura Bohannan’s (1995) article ‘Shakespeare in the bush’ accounting for how diverse interpretations of Shakespeare’s Hamlet arise. She goes on to unpack the first sentence of Richard III, using CBT as a tool in her analysis. Finally, she turns to the embodied interplay between actor, character and spectator. The brain can re-write the actor’s sense of self, so that the distinction between their own sense of self and that of the character becomes blurred. This can even apply to the actor’s sense of their own body, and Amy Cook backs this up with reference to cognitive research on phantom limbs (i.e. limbs which are missing but the brain believes still to be there in some respect).

Notes 1

2

3

4 5

‘Fifty years of the criticism of Shakespeare’s style: a retrospect’ in: Allardyce Nicoll (ed.), Shakespeare Survey 7 – Style and Language. Cambridge University Press, 1954, I–11. ‘Shakespeare’s Language and the Language of Shakespeare’s Time’, in: Stanley Wells (ed.), Shakespeare Survey 50 – Shakespeare and Language. Cambridge University Press, 1997, I–17. A term coined by Roger Fowler and used in a collection of essays he edited: Style and Structure in Literature: Essays in the New Stylistics, Blackwell, 1975. Second edition. Longman, introduction, 4–5. See the proceedings of the conference: The State of Stylistics, 2008, edited by Greg Watson. Amsterdam and New York: Rodopi.

Chapter 1

‘Strange deliveries’: Contextualizing Shakespeare’s First Citations in the OED Giles Goodland

1. Introduction This chapter examines the use that the Oxford English Dictionary has made of Shakespeare’s works, referring in particular to new electronic databases that offer a wider and more easily searchable sample of published English than was available to James Murray and his team. The large number of quotations in which Shakespeare is the first recorded user, has in the past led to some criticism of the dictionary from the discipline of historical linguistics; while, in the field of literary studies, it has led to the occasional but persistent impression that Shakespeare either was an extremely prolific coiner of words, or that he possessed an unusually large vocabulary1. Historical linguists on the one hand, and literary specialists, on the other, can be characterized as demanding different kinds of accuracy from the OED: an accurate reflection of English as it developed, versus a full and explicatory treatment of a number of writers deemed to be great or signifi cant. These demands can be difficult to reconcile because a full treatment of a single author, including atypical or rare uses of language, distorts the overall picture of language in that writer’s period. Shakespeare is the most obvious, but not the only instance. In the past, it has been hard to see how this irreconcilability can be squared. Databases such as ‘Early English Books Online’ should help to redress this imbalance, both by antedating and contextualizing many of Shakespeare’s first citations, and by isolating a number of instances in which Shakespeare was demonstrably using language in an atypical way. In certain cases, words used by Shakespeare have neither been antedated nor post-dated. I will argue that these probable hapax legomena in the language present a special case of vocabulary, explicable within a

Shakespeare’s First Citations in the OED

9

specific literary context, but largely inappropriate for historical-linguistic analysis, because they are not recorded diachronically. Their only meaning resides in their single known use. This distinction should benefit both those who use the OED as a resource in historical linguistics, and the community of OED users who seek clarification of specific literary senses and uses. In this chapter, some of the terminology will need explanation, since much of it is specific to lexicography, and even to the OED. A lemma is any lexical item, such as a word or compound, defined or listed in a dictionary. An antedating is a recorded instance of the use of a word earlier than the previous first use recorded in the NED. A nonce-word is a (usually rare) word used ‘for the nonce’, that is, according to the demand of a specific occasion, such as Coleridge’s ‘Mammonolatry’ for the worship of Mammon. A slip is a small sheet of paper on which an illustrative quotation is written, containing also the necessary bibliographical details and the lemma which the quotation illustrates. A quotation paragraph is the section of an OED entry in which quotations for a lemma are given, in chronological order. A coinage is a word that is formed deliberately by a known person, usually with some attribution to prove it.2 A first citation, following the use of Jürgen Schäfer (1980), is the earliest citation in an NED quotation paragraph. In this chapter also I shall use the phrase hapax legomenon (shortened to hapax) in a slightly unconventional way, to mean a lemma of which only one contextual instance is recorded within a historical period. A further comment is necessary on the way in which the OED is referred to. When describing the original aims and the fi rst publication of the dictionary, up to its fi rst full publication in 1928, I use ‘NED’ for New Oxford Dictionary. In 1933 the full set was republished with a Supplement, under the title Oxford English Dictionary (OED). In the 1970s and 1980s, four more volumes of Supplements were produced, containing new or previously unrecorded lemmas or senses, and these were integrated into the earlier OED in a new edition in 1989. This edition is what is now referred to as the OED. It was not a full revision of the NED, but the old text with the supplementary material added, so it does not affect the data in this chapter, since attempts were not made to antedate NED entries. In the year 2000, publication started of what is now called OED Online. Sections of the OED are thoroughly researched and revised, and these are published online at quarterly intervals. Work started at the letter M, and now proceeds both alphabetically, and, more recently, in out-of-sequence batches.

10


2. Shakespeare in the New English Dictionary It is perhaps unfortunate timing that Murray and his team started work on the NED before historical linguistics was a developed science, while by the time of publication of the last fascicle of the NED, many of its premises would have seemed questionable from a linguistically-based point of view. Murray’s Preface to Volume I (1888), in describing the original plan of the dictionary, states how it was resolved to ‘extract . . . typical quotations for the use of words, from all the great English writers of all ages’. To the linguist, it is the word ‘great’ that might give pause. If it is only the great writers who are being quoted, it may well seem doubtful that the dictionary really can, as claimed, furnish an adequate account of the meaning, origin, and history of English words now in general use, or known to have been in use at any time during the last seven hundred years [illustrated] by a series of quotations ranging from the first known occurrence of the word to the latest. (Preface, iv–v) Considering the limited resources available to nineteenth and early twentieth-century lexicographers, however, the choice of the great writers must have seemed reasonable, for at least two reasons. First, the texts of ‘great’ writers were more readily available. In the case of Shakespeare, Alexander Schmidt’s 2-volume Shakespeare Lexicon (1874–1875) provided a template for the treatment of Shakespeare’s vocabulary, with each headword, however obscure, provided with a definition. Likewise, Murray and his team would have been able to consult concordances of the 1611 Bible, and the works of Milton. Resources such as these made it hard to ignore even the most atypical words from canonical texts. Secondly, ‘great’ writers were (then as now) the ones that are more likely to continue to be read: there was a cultural imperative to cover the canonical writers. Failure to record a difficult or obscure word used by Shakespeare or Milton would be less excusable than the omission of a word used by a less well-known writer. In fact, obscurities by Shakespeare positively called for explication, whereas obscurities in minor or non-literary works tended to remain invisible because there was no necessity to include them. The OED archive contains a long sequence of slips that are marked as ‘Not In’: that is, after some procedure of evaluation, it has been decided that the lemma that the slips illustrate are not important or common enough to merit inclusion in the OED. To a certain extent this is an


11

inevitable tendency, given the need to filter out noise in order to make sense of the apparent chaos of new and unrecorded words. Even up to the 1980s, Robert Burchfield (1992), editor of the OED Supplements, advocated the inclusion of all vocabulary from certain twentieth-century canonical authors. However, compared to the other canonical texts, Milton and the Bible, Shakespeare still illustrates a proportionally higher number of first citations. This is because Milton and the Bible tended to use a more restricted set of linguistic contexts than Shakespeare. As a playwright, Shakespeare employed representations of speech as well as literary language. Scenes in his plays (as with other playwrights of the period) moved rapidly from the discourse of kings to that of commoners, with a high representation of classes, types and trades between. Indeed, linguistic register often provided a shorthand that signalled location, class and character to the audience. In addition, characters onstage manipulated forms and conventions of language according to specific situations, which for intrinsic reasons of stagecraft are more various in a play than in, for instance, a narrative poem. The nature of the dramatic action often demanded rapid changes of context. Choice of vocabulary in plays could be a marker of social station, dialect or personality. Different characters were often distinguished by idiolect. All of these factors contribute to a greater linguistic variety than in other types of text. There are also instances (which I explore more fully in section 4), in which Shakespeare does appear to have been deliberately coining, or, as his character Berowne puts it, creating ‘fire-new words’ (Loves Labour’s Lost 1.1.176). This tends to be a more subdued activity in the other canonical texts which the NED used while building up its quotation paragraphs. Other, non-canonical literary writers such as Thomas Nashe or Gabriel Harvey, do seem to have coined words with an energy as great, or greater than, that of Shakespeare. However, deliberate coinage of words in literary texts is very hard to measure and quantify. Dictionaries have no way of assuming intentionality on the part of a writer, so tend to ignore the issue. For instance, it is not uncommon for a writer to state that they are coining a word, when in fact earlier evidence of this word can be found. In a literary text such unmediated statements of intent would be rare (a character in a play, such as Don Armado or Berowne, may declare that a word is new, but for the audience it would be unimportant if this were true or not). In this chapter I will follow OED practice, and not make assumptions about intentionality. There are of course cases in English in which words have been deliberately invented, and this conscious invention is on record as a first citation,

12


but in general it is impossible to infer with certainty if OED Online’s first citations are cases of a word being invented for the first time. Even if we could be certain that we are seeing evidence of the first written occurrence of a word, this would not in itself be evidence of coinage. Often, perhaps predominantly, a word exists in speech, or in ephemeral writings such as letters and journals, before it is published. OED Online follows NED in not claiming that first citations are the actual first occurrence of a word. As ever, there are also complicating factors involved. Unlike in the cases of Milton and the Authorized Version, Shakespeare did not (at least in most cases) supervise the publication of his own works; hence, there was more room for transmission errors and compositorial misreadings or suppositions than in other bodies of canonical texts. Even without making any assumptions about Shakespeare’s greatness, there are clearly factors – some extrinsic to him – tending to place him as a high-vocabulary writer. The NED’s high proportion of Shakespeare first citations was already noted by the historical linguist Otto Jespersen in 1909: In turning over the pages of the New English Dictionary . . . one is struck by the frequency with which Shakespeare’s name is found affixed to the earliest quotation for words or meanings. In many cases this is no doubt due to the fact that Shakespeare’s vocabulary has been registered with greater care. (234) The case was exposed more particularly after the publication of Jürgen Schäfer’s (1980) Documentation in the OED. Before the age of databases or complex computer technology, using statistical analysis based on concordances and the research of a number of graduate students, Schäfer demonstrated the extent to which Shakespeare was over-represented in the OED compared to similar writers, both in Shakespeare’s period (Nashe) and in earlier periods (Malory and Wyatt). Schäfer did not choose to question the central role of literary texts in the OED. His chosen counter-examples were all literary writers (but not playwrights); in the case of Nashe, a unique, pre-Joycean one even more prone to neologizing and linguistic playfulness than Shakespeare. A non-literary text would have significantly changed the statistical picture that Schäfer developed; but his work remains as significant independent evidence that Shakespeare was relied on disproportionately, compared to other writers of the period, in the NED. Essentially, Schäfer showed that Murray and his team allowed themselves to stray from the procedure outlined in the NED Preface to extract ‘typical quotations for the use of words’. For Shakespeare, and a small number of other writers,


13

atypical quotations were used to illustrate atypical lemmas. Because of this, a statistical account of language in the period of Shakespeare, basing itself on the un-reedited OED’s data, runs the risk of being skewed. The obvious answer to these criticisms would be a utilitarian one, based on the profiles of those who use the OED as a source of definitions and factual information: as in Murray’s time, and perhaps more so, accurate information on most, if not all, of Shakespeare’s words is required by literary scholars. Shakespeare, in this argument, was, and remains, a special case, in which a coverage which includes atypical uses and hapax would seem essential. In some cases there is a gap of 120 years or more between the entries edited for the NED, and the text as re-edited for OED Online. Literary scholarship, influenced by varying modes of historicism, has changed. It does not just want explication of what Shakespeare meant by particular words. The fact that almost all of Shakespeare’s language is presented somewhere in the OED becomes a point of interest in itself. As editing on OED Online progresses, a more modulated answer becomes possible; it is not that there was too much Shakespeare, it is that there was too little of other texts and writers. Previously unrecorded lemmas from this period are being drafted as part of the OED revision process. In addition, added information supplied by OED Online in the form of extra quotations, fuller definitions, more detailed etymologies and clearer labelling help us to contextualize the Shakespeare first citations.

3. Shakespeare in OED Online In the last few years, electronic corpora and websites such as Literature Online (LION) have become readily and immediately available to lexicographers and researchers. With fully searchable databases, one of the most difficult aspects of lexicography in the era of Murray was resolved: examples of any lemma could be retrieved without human error, especially without the tendency to miss a significant word. Working on paper, only a small amount from any text would have been converted into slips, often only the more visible and obvious uses, at the expense of apparently less significant early uses. Full searchability removed this tendency to miss significant evidence: each text suddenly had a greater ease of use, and the same evidence-value as any lexicon or concordance. Literature Online, as its name suggests, is a source of literary texts, so part of the same general field of discourse that the OED covers with Shakespeare,

14


Milton and its many other literary sources. However, it includes a large number of writers not generally considered to be canonical. A further and more significant change in the nature of the searchable texts has come more recently with Early English Books Online (EEBO). A considerable amount of Early Modern material is viewable: this incomparable collection now contains about 100,000 of over 125,000 titles listed in Pollard & Redgrave’s Short-Title Catalogue (1475–1640) and Wing’s Short-Title Catalogue (1641–1700) and their revised editions, as well as the Thomason Tracts (1640–1661) collection and the Early English Books Tract Supplement. and a sizable proportion of this is searchable: To accompany the citations and page images, a separate initiative, the Text Creation Partnership (TCP), is in the process of creating SGML coding for the full text of 25,000 EEBO works, so users can search the full ASCII text of the documents and view both the text and the corresponding original page images. (EEBO website information, accessed May 2010) EEBO is still unfinished, but at last it has become thinkable that all printed Early Modern English-language texts can be searched. There are two important factors in this: the first is the sheer size of the EEBO database, dwarfing any other resource from the period. The second, no less significant, is the nature of the EEBO texts. They are not restricted to the literary register, as was the earlier Literature Online. Because of this, large numbers of texts that were never considered for the original NED reading programme because of their recondite, turgid or inaccessible nature are now instantly searchable. It is worth remembering that the NED ‘Reading Programme’ was a voluntary endeavour, appealing to the general public, and thus limiting itself to books that were in some degree interesting or appealing to its unpaid readers. In contrast, recent antedatings of Shakespeare on OED Online that were found by means of EEBO include sources such as a 1550 translation of Thucydides’ History of the Peloponnesian War (PRACTISANT n.), A Treatise concerning Statutes (PRECURSE n.), a 1612 commentary on an Epistle of St. Paul (PREPAREDLY adv.), a 1541 defence of the marriage of priests (PRODIGIOUSLY adv.), a 1549 translation of Erasmus’ Praise of Folly (PROMETHEUS n.) and a 1566 book with


15

the title Answeare to M. J. Fekenham (RATIFIER n.). In particular, works of theological controversy and debate have become a fruitful source of antedatings, perhaps because they were an unattractive proposition for a method of research that involved the close perusal of the full text. In short, electronic searchability removes the task of having to read a text in order to extract lexical information from it. The graph and data below show that, in the sample of the edited portion of OED Online, in a large proportion of cases the Shakespearian first citations can be antedated and contextualized. However, there are also a significant number of lemmas that have not been antedated. This large remaining substrate of first citations might appear to justify the criticism of some historical linguists that Shakespeare is over-represented in the OED. I will contextualize some of these senses in greater detail below, but it might be worth initially clarifying the nature of this material, with reference to the field of academia in which historical linguistics and literary studies meet: that of poetics. The Russian Formalist and linguist Viktor Shklovsky argued that one of the distinguishing features of literature is atypicality. Literature, he argued, is in fact at its most typical when it is being unconventional in its use of language.3 This bold argument was taken up and modified by the Prague School of linguists. Jakobson and Mukarovsky, among others, distinguished literary discourse from ordinary language because it contained or consisted of types of foregrounding that functioned by producing effects of estrangement on its audience, thus giving literature its aesthetic quality. Linguists of the Prague School linked this effect to that of markedness in general linguistics: having a positive content, against a (more expected) unmarked form. They argued that language in literature will tend in general to have a higher content of marked than unmarked forms than non-literary language: The novelty, e.g. of poetic language in contrast to spoken language, the patterning imposed by metre, the tension of plot, all ‘devices’ . . . for the aim of art, which . . . is conceived as a shock to our ordinary indifference, as a heightening of awareness, as ‘making strange’. (Wellek: 8) Shakespeare’s poetry and drama are literary, and hence distinguished from functional or purely communicative or non-literary discourses by effects of foregrounding; most relevantly for us, in atypical use of language. In a non-literary text, language is a medium with which to express some purpose or argument with clarity. In many literary texts, words, the medium of expression, are foregrounded.

16


4. Graph and data For this chapter I have chosen to focus on the NED Shakespeare first citations in the alphabetical sequence from P to Ra-. These sections of the NED were originally published between 1903 and 1904. The corresponding OED Online sections were edited and published online in the period just before, and during the first uses of the EEBO databases, from March 2005 to June 2008 (see Figure 1.1). Other electronic databases, notably ‘Literature Online’, had been in use for some time. In addition, OED revision involved the searching of OED itself, the Middle English Dictionary, material from the OED’s in-house library, the OED (15) 12%

SL (3) 2%

EEBO (15) 12%

NOT ANTEDATED (64) 53%

HC (1) 1% HRP (3) 2% LIB (1) 1% LION (6) 5%

MED (8) 7% MICH (6) 5%

Figure 1.1 Proportion of EEBO and other provenances in sample batch. Parenthetical values represent associated raw frequencies. SL: slips; HC: historical corpus, an old database of early modern material; HRP: Historical Reading Program; LIB: from the OED library; LION: from Literature Online; MED: Middle English Dictionary; MICH: slips from University of Michigan’s abandoned Early Modern dictionary project.


17

OED’s electronic ‘historical corpus’ of quotations, and the use of slips from its own collection of quotations, as well as from the Michigan collection that it inherited from an abandoned project to assemble an Early Modern English Dictionary. All of these sources contributed to the antedating of Shakespeare first citations in this period. An ensemble of sources and databases continues to be most useful; to cover not just printed works, but also diaries, notebooks and manuscripts (when printed later), and unprinted archival material. Thus provenances such as HRP (for Historical Reading Program) will continue to be significant for OED revision in the foreseeable future. More recent tranches of OED Online as they are published online are highly likely to show a higher proportion of Shakespeare antedatings, because (in contrast to the sample discussed in the following sections) it does not include a period before the use of EEBO in research and editing.

5. Interpreting the data 5.1 A full list of the Shakespeare NED first citations in the range P-RA: (117) Those antedated are in italics. Those citations which have been re-analysed by OED Online are in square brackets. pageant v.; pageantry n.; paiocke (pajock) n.; palate v.; pale-faced a.; palliament n.; palmy a.; pander v.; parkward adv.; parling a.; partner v.; passado n.; paternal a.; pauser n.; peaking a.; pebbled a.; pedant n.; pedantical a.1; peeping n.2; pellet v.; pelting n.; pendulous a.; peregrinate a.; perplex v.; persistency n.; persistive a.; personating n.; perusal n.; petition v.; phantasime n.; Phoebe n.; phraseless a.; picked-hatch a.; pig-nut n.; pilcher n.; pioned a.; pious a.; pip n.2; [placcate (now placate n.)]; plantage n.; pleached a.; please-man n.; plighter n.; plodder n.2; plodding n.; plumpy a.; poniard n.; pooh int.; poppering n.; portage n.2; portcullis v.; posied a.; poster n.1; pouncet-box n.; prabble n.; practisant n.; preceptial a.; precipit n.; precurrer n.; precurse a.; predecease v.; predict n.; preformed a.; premeditated a.; [prenzie a.]; preparedly adv.; presented a.; press n.2; prevailment n.; preyful a.; pribble n.; priceless a.; primogenitive n.; primy a.; printless a.; prison-gate n.; probal a.; prodigiously adv.; profitless a.; prologue v.; Promethean a.; Prometheus n.; promising a.; prompture n.; proof a.; propertied a.; property v.; prophetic a.; proposer n.; protester n.; protesting n.; protractive a.; published a.; pudency n.; pugging a.; puke v.; [pulpiter n.]; pulsidge n.; pupil age n.; puppy-dog n.; purr n.3; push n.3; push-pin n.; qualifying a.; quarrelsome a.; quartering a.; quatch n.2 (now adj.); queen v.; questant n.;

18


questrist n.; quillet n.2; radiance n.; rancorous a.; ransomless a.; rant v.; ranting a.; rat-catcher n.; rated a.; ratifier n. NED’s ‘placcate’ has been re-analysed by OED Online as sense 3 of PLACATE n., which is attested earlier. ‘Prenzie’ is now explained by OED Online as ‘a compositorial misreading of precise’. ‘Pulpiter’ was the NED’s nowrejected emendation of ‘Iupiter’; it remains as a small-type note, since PULPITER is attested as a lemma from 1681. Thus the crude statistics are, out of 112 lemmas, 2 are excluded. Fifty-six have been antedated, and 54 remain unantedated. 5.1.1 Antedated Shakespeare first citations, P-RA: (41) First dates refer to the NED’s dating, usually based on records of first performance. Second dates refer to OED Online’s more conservative dating, based on publication dates. The third date refers to the year of the antedating quotation. In each case, OED Online can be consulted for more details.

pageant v. 1606/1609–1606 pale-faced a. 1592/1593–1570 palmy a. 1602/1604–1440 paternal a., 1605/1608–a1450 peaking a. 1598/a1616–1595 pedant n. 1588/1598–a1586 perplex v. 1595/a1616–1477 perusal n. 1600/1604–?1589 Phoebe n. 1590/1600–a1393 pious a. 1602/1604–c1450 placcate (now placate n.) 1588/ 1598–1567 please-man n. 1588/1598–1570 plodder n.2 1588/1598–1584 poniard n. 1588/1600–1533 pooh int. 1602/1604–1593 poster n.1 1605/a1616–1538 practisant n. 1591/a1616–1550 precurse a. 1602/1604–a1591 premeditated a. 1590/1600–1583 press n.2 1596/1598–?1592

prison-gate n. 1590/1600–1560 probal a. 1604/a1616–1439 prodigiously adv. 1595/a1616–1541 profitless a. 1599/1600–1574 Prometheus n. 1588/1594–1549 promising a. 1601/a1616–1594 prophetic a. 1595–a1616–c1484 proposer n. 1602/1604–1566 protester n. 1601/a1616–1591 protesting n. 1599–1582 protractive a. 1606/1609–1596 published a. 1605/1608–a1400 qualifying a.1606/1609–1582 quarrelsome a. 1596/a1616–1582 quillet n.2 1588/1598–1576 radiance n. 1601/1608–a1593 rancorous a. 1590/1597–?1517 ransomless a. 1588/1594–a1420 rat-catcher n. 1592/1597–1565 rated a. 1595/1598–1487 ratifier n. 1602/1604–1566.


19

It should be noted here that each antedating represents a slight equalizing in the over-representation of Shakespeare noted by Schäfer and others. Each antedated entry also represents an added contextualization of a lemma that was previously a Shakespeare first citation, and hence this is significant information for the Shakespeare scholar. For instance, in the case of PROTESTER n., NED had a citation from Julius Caesar as the only evidence for this to mean ‘one who makes a protestation or solemn affirmation’, which OED Online is now able to redefine as ‘A person who makes an avowal or solemn declaration, esp. a person who makes a protestation of love’. OED Online now has six quotations for this sense, ranging from 1591 to 1742. On certain occasions, these antedatings expose a possible borrowing or influence. For instance, at PLODDER n.2, OED’s antedating is interesting because of the similarity of the quotations. This from Loves Labour’s Lost (1.1.86): Small haue continuall plodders euer wonne, Saue base aucthoritie from others Bookes. Ironically (given the meaning of this sentence), John Lyly (Sapho and Phao i. iii. sig. Bv [1584]) is now shown to have used this word in 1584, in a sentence with a similar meaning, and the same last word: Wee silly soules are onely plodders at Ergo, whose wittes are claspt Vppe with our bookes. Ergo (Latin for ‘therefore’) is here presumably referring to scholastic use of Logic.

5.1.2 Lemmas antedated due to re-dating of NED quotes: (16) palliament n. 1588/1594–1593 pedantical a.1 1588/1598–1592 peeping n.2 1593/1594 (equidated with Nashe)–1593 personating n. 1607/a1616–1615 petition v. 1607/a1616–1607 picked-hatch a. 1598/1602–1598 pip n.2 1596/a1616–1604 prabble n. 1598/a1616–1603

preparedly adv. 1606/a1616–1612 prevailment n. 1590/1600–1599 pribble n. 1598/a1616–1603 Promethean a. 1588/1598–1594 proof a. 1592/1597–1583 puke v. 1600/a1616–1601 rant v. 1598/1604–1602 ranting a. 1598/a1616–1609.


20

This is a more problematic set of data.4 In these cases, it was possible that the word was used earlier by Shakespeare, but the only documentary evidence supporting this refers to the date of performance of the play (except for PEEPING n. 2 , in which the text of Shakespeare’s Lucrece was re-dated from 1593 to 1594 to follow the dating of the English Short Title Catalogue). In some cases, as with PUKE v., the entry has been restructured on the basis of this redating. The proximity of the two dates suggests in general that as a playwright, Shakespeare may have been using words that were common in speech, but had not yet made it into print. The antedated Shakespeare first citations are thus 57 out of the 117 lemmas, or 49 per cent.

6. Un-antedated Shakespeare first citations 6.1 Followed by non-influenced later citations within 200 years: (38) pageantry n. palate v. pander v. parling a. passado n. pauser n. pebbled a. pelting n. pendulous a. persistency n. persistive a. pig-nut n. pilcher n.

pioned a. plodding n. plumpy a. poppering n. portcullis v. posied a. preceptial a. predecease v. preformed a. presented a. preyful a. priceless a. primy a.

printless a. prologue v. prompture n. propertied a. property v. pudency n. pupil age n. puppy-dog n. purr n.3 push-pin n. quartering a. queen v.

In these cases, Shakespeare remains the recorded first user of a word; but later evidence recorded in OED Online shows that the word may have had some currency at a similar time, or not long after. In these cases it is impossible to say confidently if these words were coined or first used in Shakespeare’s texts; but the evidence would suggest that many of them were probably current, at least in speech, at that time.


21

Often, the information that the Shakespeare first citation for a word cannot be antedated is itself interesting. When Caliban in The Tempest says ‘I with my long nayles will digge thee pig-nuts’ (2.2.167), OED Online has no evidence that this word had not until then been used in print. This says something about the register that Caliban is using: a regional or idiomatic speech far removed from the literary language of the courtiers in The Tempest. OED Online’s following quotation, from 1693, describes ‘The Roots . . commonly call’d Kepper-Nuts, Pignuts and Gernuts . . in the North’. Colloquiality can perhaps explain the use of ‘puppi-dogges’ instead of puppies: Here’s a large mouth . . That spits forth death, and mountaines, rockes, and seas, Talkes as familiarly of roaring Lyons, As maids of thirteene do of puppi-dogges. (King John 2.1.461–4) Here the suggestion is that this is the kind of word a ‘maid of thirteen’ might use. Similarly, ‘push-pin’ in this list is a childhood game. Then, as now, colloquial and childhood words have a greater tendency to exist in speech before they are used in print. In some cases, reference to the OED Online etymology will provide useful information. For instance, POPPERING n., a type of pear, is first recorded in Romeo and Juliet. The term comes from the name of a town in Belgium that was famous for its pears. It seems to be just chance that Shakespeare is mentioning this in English for the first recorded time, as a modern writer might mention a new type of soft drink, perhaps. PASSADO, a term in fencing, from the same play, is thought to have come from Italian. Again, Shakespeare was unlikely to have been the first person to have borrowed this term: more likely, the word was in speech at the time. Other items in this list derive from Latin, and a distinction might usefully be made between Latin-derived terms (such as PENDULOUS a. and PUDENCY n.), and those formed from English stems and affixes. A more detailed chapter than this might distinguish between Shakespeare’s foreign, Latin and English-derived first citations. In other cases, these words show a distinctive literary provenance. For instance PLUMPY a., a synonym for plump, occurs in Antony and Cleopatra (2.7.111) in song, where ‘plumpy’ fits the partly trochaic metre in a way that plump (or other synonyms) would not: ‘Come thou Monarch of the

22


Vine, / Plumpie Bacchus, with pinke eyne’. Here ‘eyne’ is a grammatical archaism, suggesting a context for the unusual ‘Plumpie’. There are other cases in which exigencies of metre, stress or rhyme may have determined the choice of an unusual word: ‘primy’ instead of ‘prime’, perhaps to avoid two strong stresses in ‘A Violet in the youth of primy nature, Forward, not permanent, sweete, not lasting.’ (Hamlet 1.3.7) (but producing a hendecasyllabic line); although (as with plump and plumpy), primy and prime are of course not necessarily synonymous, and the NED entry for the -y1 suffi x has an interesting note: ‘In the 15th cent., if not earlier, certain monosyllabic adjs. were extended by means of this suffi x, app. with the design of giving them a more adjectival appearance . . . In this application the suffix has not infrequently come to express much the same notion as -ish.’ These words are also recorded as being used by other writers, so it is possible that they had an extra-literary provenance, or were part of a fund of literary words that could be called upon at will (in a time before English dictionaries, it is important to remember that a sense of what was verbally possible would have been more fluid). The situation with many of these words is probably that, even though Shakespeare may well be the first user of a word, he was not neologizing. 6.2 Cases in which a Shakespeare first citation is either followed by a gap in citations of more than 200 years, or in which all subsequent citations refer to the Shakespeare quote in some way: (10) paiocke (pajock) n. parkward adv. partner v. pellet v. peregrinate a.

phraseless a. plantage n. pleached a. plighter n. pouncet-box n.

These are in effect Shakespeare hapax, but such is the unique canonical status of Shakespeare that his unusual usage has inspired later imitators. 6.3 Shakespearian hapax (lemmas not diachronically attested): (12) phantasime n. portage n.2 precipit n. precurrer n.

predict n. primogenitive n. pugging a. pulsidge n.

push n.3 quatch n.2 (now adj.) questant n. questrist n.


23

It is categories 6.2 and 6.3 that are of most interest: although smaller than the previous categories, they are statistically significant, and repay study both from a literary and linguistic point of view. In a sense (and this is from someone whose job it is to do this) antedating an OED headword by using EEBO can be a banal activity; there is a probability that many, if not most headwords can be antedated. But as Schäfer notes, any antedating simply opens up the possibility of a further antedating from a yet earlier source. The OED quotation paragraphs, although it is not their intention, tend to reinforce an illusory sense of the primacy of the written, and especially the printed word. The OED’s aim is to give the first recorded instance of a usage; for many hundreds of years, this could only have been through the medium of writing. The NED, we should remember, did not claim that its first citations were first uses. An antedating is more primarily interesting for the information that it conveys, which would include contextual data about the meaning or register of the word; dating changes that enable a change in the structure of an OED entry; or information about etymology (for instance, if the earlier quote turns out to be from a translation). It is in this sense, of conveying information, that not being able to antedate a word can also be significant. These words in sections 6.2 and 6.3 can be shown to be explicable in terms of either a conscious neologistic poetics occurring at certain times, and for certain contextual reasons, within Shakespeare’s work; or else may be cases of either errors at some point in the transmission of the text, or else editorial over-interpretation. Of the 22 words in 6.2 and 6.3 above, some are 6.3.1 Possible transmission errors POUNCET BOX n. from 1 Henry IV (1.3.37): Twixt his finger and his thumbe he helde A pouncet boxe [1623 Pouncet-box], which euer and anon He gaue his nose, and tookt away againe. The etymology section of OED Online cannot say with certainty where this first element, ‘pouncet’, originates and states that it may be a misprint for pounced box. PRECIPIT n. from Henry VIII (5.1.140): Go too, You take a Precepit for no leape of danger, And woe your owne destruction.

24


is, as noted in OED Online’s etymology, probably a transmission error for PRECIPICE n. (OED Online also mentions the French precipite, meaning ‘precipice’). PUGGING a. occurs in The Winter’s Tale (4.3.7) in a song by Autolycus: The white sheete bleaching on the hedge, ... Doth set my pugging tooth an edge, For a quart of Ale is a dish for a King. OED Online defines this as ‘Of uncertain meaning; perhaps: “that pulls or tugs, thieving” ’; the etymology section notes that early comments sometimes interpreted it as a transmission error for prigging. With words like this, there is not enough context to define it with certainty.

6.3.2 Possible editorial over-interpretation In cases where the apparent hapax is a later homonym of an established word, editorial over-interpretation may be suspected. In the two cases in our sample, we have PORTAGE n.2, treated in section 6.3.5, and PUSH n.3 from Much Ado about Nothing (5.1.38): There was neuer yet Philosopher, That could endure the tooth-ake patiently, How euer they haue writ the stile of gods, And made a push at chance and sufferance. OED gives this a separate homonym and relates it to the interjection ‘push’ (similar to PISH int.). This sounds persuasive (as editorial interpretations should; otherwise they would not have survived); but it is asking us to believe that Shakespeare wanted us not to think of the common noun (from the even commoner verb) that is PUSH n.1 In this case, this begs the question, why did Shakespeare not use the more common form, ‘pish’? Shewmaker (1996) simply defines ‘make a push at’ as ‘to defy or scorn’. The remaining terms in the list can be accounted for largely in terms of a poetic, rather than a referential meaning. These are the words that sit in the dictionary with least happiness, because their semantic meaning is less important than the role they play in the literary text.


25

6.3.3 Humorous, malapropistic or comically pedantic nonce-formations These have two functions: as a source of humour, and as idiolect, a marker of character. These uses tend to be confined to a proportionally small number of characters and plays. In our sample, they are largely from his earlier comedies: two from Loves Labour’s Lost, one from As You Like It, one from The Merry Wives of Windsor and one from 2 Henry IV. Shakespeare here may be drawing on the humorous rhetorical trope of ‘bomphiologia’, which Sherry describes as verborum bombus – ‘when small & triflying thynges are set out wyth great grasyng words’ – (Treat. Schemes & Tropes [1555] sig. D.vii.), or perhaps ‘cacozelia’ or mala affectatio: Mala affectatio, euyll affectacion or leude folowyng, when the wytte lacketh iudgement, and fondlye folowyng a good maner of speaking, runne into a faute, as . . . we fall into a vaine bablynge, or laboryng to be brief, wax bare & drye. Also if we shuld saye: a phrase of building, or an audience of shepe, as a certen homely felow dyd. PARKWARD adv. from The Merry Wives of Windsor (3.1.5): Marry Sir, the pittie-ward, the Parke-ward: euery way: olde Windsor way, and euery way but the Towne-way. Here the -ward suffix is clearly being employed by the character Simple in nonce-fashion. PEREGRINATE a. from Loves Labour’s Lost (5.1.14): Ped: He is too picked, to spruce, too affected, to od as it were, too peregrinat as I may call it. Curat: A most singuler and choyce Epithat. This in context is clearly neologistic: so much so that after Holofernes (Pedant in the Quarto’s directions) says this, Nathaniel (Curate) in the stage-directions is instructed to draw out his table-book and write it down. OED Online derives it from classical Latin peregrinatus. PHANTASIM occurs only in Loves Labour’s Lost, but twice, 4.1.98 and 5.1.18. It is clearly related to the foreignness of the braggart Don Armado: This Armado is a Spaniard that keepes here in court, A Phantasime a Monarcho, and one that makes sport To the Prince and his Booke-mates. ...

26

Stylistics and Shakespeare’s Language I abhorre such phanatticall phantasims, such insociable and poynt deuise companions.

This second citation is again by Holofernes, immediately after his speech in which he used the word ‘peregrinate’, above. The word is deliberately strange because the thing it is describing is strange. OED Online’s etymology is from the Italian fantasima. PULSIDGE in 2 Henry IV (2.4.21), is a humorous blunder for ‘pulse’, and hence is idiolectal. This is Mistress Quickly in a comedic passage: You are in an excellent good temperalitie. Your pulsidge beates as extraordinarily as heart would desire. QUATCH a. from All’s Well that Ends Well (2.2.17) is apparently a humorous variant (though earlier) of ‘quat’ or ‘squat’, in the phrase ‘quatchbuttock’: A Barber’s chaire, that fits all buttockes, the pin buttocke, the quatchbuttocke [etc.]. 6.3.4 Possible metrical determinants A number from this list may be determined by exigencies of metre, although in all of them there must be other considerations also at play, as I shall attempt to uncover. QUESTANT n. from All’s Well that Ends Well (2.1.16) is from the dignified language of the King. OED Online cross-references this to the defi nition ‘A person or hound who quests (in various senses).’ The ‘est’ of questant chimes with the ‘est’ of bravest, while also echoing the ‘e’ sound in ‘wed’: You come Not to wooe honour, but to wed it, when The bravest questant shrinkes. 6.3.5 Possible literary estrangement effects The last category, occurring mainly but not exclusively in the later plays, seems to be one in which the effect is one of strangeness, of rhetorical flourish occurring at moments of high dramatic tension, highlighting for the audience a state of unease in character or situation:


27

PAJOCK n. (NED PAIOCKE) from Hamlet (3.2.272): For thou doost know oh Damon deere This Realme dismantled was Of Ioue himselfe, And now raignes heere A very very paiock. This is glossed as a possible transmission error, perhaps meaning ‘peacock’, although it occurs in a speech where Hamlet is in a heightened, perhaps crazed, state, so may show the disjointedness of his thought. David Crystal, among others, suggests that the meaning is ‘savage’. It follows a previous rhyming quatrain, and Shewmaker (1996) suggests that ‘[T]he expected end-rhyme would have been ass’ (388). PELLET v. from A Lover’s Complaint (in Sonnets sig. Kv) which the OED defi nes in a small-type note as ‘to send or supply in the form of pellets’: Laundring the silken figures in the brine, That seasoned woe had pelleted in teares. Since the ‘silken figures’ on a napkin are the object of the verb, this is a possible transmission error, or perhaps a suggestive and metrically demanded elongation of the verb ‘pelt’. It is hard to see how a silken figure could be sent or supplied in the form of pellets. The sense could also be: to hit or pelt with pellets. In Antony and Cleopatra (3.13.163) there is a similar phrase: ‘By the discandying of this pelletted storm, Lie Graveless’. PHRASELESS a. from A Lover’s Complaint (Sonnets sig. L) has a gloss in OED Online ‘apparently: “that there is no phrase to describe” ’. Apart from the semantic sense and the metre, there may be a conscious use of assonance and alliteration in repeated ‘a’ and ‘s’ sounds: Oh then aduance (of yours) that phraseles hand, Whose white weighes downe the airy scale of praise. PLANTAGE n. in Troilus and Cressida (3.2.173) occurs in a passage where Troilus is describing the language that future swains will want in order to describe him; the neologized word (perhaps a metrical lengthening of ‘plants’) reinforces the mock-heroic clichés that become ludicrous through repetition: As true as steele, as plantage to the moone. As sunne to day: as turtle to her mate.

28


PLEACHED a. occurs twice in Much Ado about Nothing (1.2. 316 and 3.1.7), and also occurs in a different sense in Antony and Cleopatra (4.15.73). OED defines as ‘Interlaced, intertwined, tangled’, but it may in fact mean something more like, as David Crystal suggests, ‘bound together’. It derives from the much earlier verb ‘pleach’; in all occasions in ‘poetic’-sounding verse passages: The prince and Count Claudio walking in a thicke pleached alley in mine orchard (Much Ado about Nothing 1.2.8) Bid her steale into the pleached bowere, Where hony-suckles ripened by the sunne, Forbid the sunne to enter. (3.1.7–9) and from Antony and Cleopatra: Would’st thou . . see Thy Master thus with pleacht Armes, bending downe His corrigible necke? (4.14.73) PLIGHTER n. occurs in Antony and Cleopatra (3.13.127), again in a passage of heightened verse. OED Online defines this as ‘a person who makes a pledge or promise’, and gives the etymology as from PLIGHT v.1; it is tempting, but not inevitable, to see a pun with PLIGHT n.1, or indeed with an earlier sense of PLIGHT v.1, in the sense of putting someone in danger: My play-fellow, your hand; this Kingly Seale, And plighter of high hearts. PORTAGE n.2 occurs in Henry V (3.1.10) in the famous ‘once more unto the breech’ speech: Lend the Eye a terrible aspect: Let it pry through the portage of the Head, Like the Brasse Cannon. OED Online defines this as ‘Provision of ports or portholes; fig. in quot’. This certainly fits the sense, but we have to remember that if we accept this, we have to also accept that Shakespeare is coining a new and figurative sense, when he could mean the existing sense of PORTAGE n.1, ‘An amount of space or weight on board a ship allowed to a mariner for his own


29

cargo in lieu of wages’, in the sense that Shakespeare also used in Pericles (3.1.35–6): Thy losse is more then can Thy portage quit, with all thou canst find heere. It may also be a substitution for ‘porthole’, or indeed a transcription error for the more obvious ‘portals’. More important here than the specific dictionary definition, is the heightened rhetorical context that the unusual word implies. PRECURRER n. from Phoenix (in R. Chester Loves Martyr 170); is perhaps easier in diction and sounds less Latinate than ‘precursor’ (although OED Online gives it a Latin etymology): ‘Thou shriking harbinger, Foule precurrer of the fiend’. PREDICT n. from Sonnet 14 (sig. B3v); it occurs in some lines with a clipped, abbreviated tone, with no words longer than two syllables: Nor can I fortune to breefe mynuits tell; Pointing to each his thunder, raine and winde, Or say with Princes if it shal go wel By oft predict that I in heauen finde. PRIMOGENITIVE n. from Troilus and Cressida (1.3.106) is cross-referenced in OED Online to PRIMOGENITURE, sense 3, ‘The right of the firstborn child of a family, esp. a son, to succeed or inherit property or title to the exclusion of other claimants’. There is an earlier word in English, PRIMOGENIT, a firstborn child. It occurs in a long rhetorical speech by Ulysses, in which he asks what happens when ‘Degree is shak’d’: the long and legalistic Latinate words perhaps suggest that there is something over-determined in his pleading: How could Communities, Degrees in Schooles, and Brother-hoods in Cities, . . The primogenitiue [1609 primogenitie], and due of Byrth, Prerogatiue of Age, . . (But by Degree) stand in Authentique place? QUESTRIST n. is from King Lear (3.7.15). This is perhaps an idiolectal example: the Steward who speaks these words has a tendency to insincere long-windedness; it occurs just after the ‘Plucke out his eyes’ scene: Thirtie of his Knights Hot questrits [1623 Questrists] after him, met him at gate.

30


7. Conclusion If we exclude the hapax from the percentage, the percentage of antedatings change. 57 antedated citations out of 95 ‘genuine’ words gives us a rate of 60 per cent of Shakespeare fi rst citations antedated. But the figures also tell us that 19 per cent of Shakespeare’s lemmas in the sample batch were probably coinages, which seems like quite a high rate. The category of strange or genuinely atypical words brings us back to the field of poetics: these are all words peculiarly hard to defi ne from a linguistic point of view, but effective and unsettling aesthetically. The literary theorist Michael Riffaterre (1983) has something very pertinent to say here: Literary neologisms are profoundly different from neologisms in everyday language. The latter are coined to express a new referent or signified. They are, first and foremost, the bearers of a new meaning, and are not necessarily perceived to be unusual forms. A literary neologism, by contrast, is always perceived as an anomaly, and is employed precisely because of that anomaly, sometimes even without regard for its meaning. (iv. 62) Riffaterre had in mind more modernist modes of literature. Neologism has been an engine of poetry in many types of modernism, and of course had progenitors in nonsense poetry. But Riffaterre is perhaps not saying anything that would have surprised Shakespeare or his contemporaries. The grammarian Richard Mulcaster, writing in 1582, wanted to give some historical justification for the increasing use of new words in his period. He wrote ‘new occasions brought furth new words, as either more cunning made waie to more terms, or as strange deuises did seke strange deliueries’ (1st Pt. Elementarie xxii. 154). Mulcaster is making the same distinction as Riffaterre, between ordinary neologisms, called for as occasion requires (much like the OED’s nonce-words), and the alienating or unsettling devices that called for expression with ‘strange deliveries’: the same that occur, at certain points, in literature. If we consider the Shakespeare hapax to be essentially sui generis, lexically speaking, making sense only within the language game of a poem or play, and set apart from communicative discourse, we can exclude them from the statistics of general Shakespeare fi rst citations. In this case, the number of OED Online Shakespeare fi rst citations falls to a much more reasonable 41 per cent of NED Shakespeare fi rst citations,


31

and one that promises to decrease still further as EEBO becomes more widely searchable. This is a useful distinction in that it – along with the many antedatings of Shakespeare tabled above – helps us to square the circle between the demands of historical linguistics upon the OED, and those of philological or literary studies. By excluding the Shakespeare hapax from their statistics (for instance by ignoring OED Online senses or headwords labelled as rare), the ground is evened. Shakespeare is made more equal with the other writers of his period, such as Nashe, whose atypical uses were often silently ignored, or not noticed, by the NED editors. At the same time, the isolation of Shakespeare’s rare uses is of significance for literary scholars, since it shows those passages where he was aiming for effects of strangeness and extremity; where the action of choosing a word was replaced by the act of creating one. The fact that a healthy number of first citations, despite the efforts of the OED Online team, still remain unantedated, may still perhaps continue the impression that Shakespeare had a large vocabulary. We can now look at this in a different light. If you look closely at the list of unantedated Shakespeare quotations, and for each of them, ask the question, under what exigency did Shakespeare choose to use this word, one realizes that the often-posed question, how many words did Shakespeare coin?, is largely false. Like other poets and playwrights of the period, he was simply using language to create effects. In a time period before practical, or even remotely comprehensive dictionaries, it would not have seemed like a good question at the time, to ask if a word was in the language. Perhaps, in the Early Modern period, we should look at writers as not so much possessing a vocabulary, but having an ability to generate one. No one would advance Finnegans Wake as evidence that Joyce possessed a large vocabulary. It shows that he was able to generate a vocabulary. The high number of hapax in our sample shows a remarkable linguistic creativity. Any argument about Shakespeare’s exceptionality, however, would need to compare these figures with the number of hapax in other writers of the same genre and period. Clearly, no other writer shares exactly the same background and characteristics as Shakespeare. But there are a few poetplaywrights for the period for whom comparison might be more enlightening than obscuring. It may be that Shakespeare (or any other author whose verbal creativity we wish to analyse) was simply using a word that was already known in common speech, and it is simply chance that, after searching many databases, this word remains the first that can be cited. It may be a nonce-

32


word, formed for the occasion, perhaps because of the demands of metre or rhyme. Or it may be a strange word. These are the ones that can now – with the help of EEBO, or as mediated by OED Online – be shown to be hapax. Isolating this last category will, I believe, be a useful tool in the continued analysis of Shakespeare’s vocabulary.

Notes 1

2 3 4

The most recent I am aware of is in Lerer: ‘[Shakespeare] coined nearly six thousand new words’ (129); ‘[Shakespeare] coined words and phrases at a rate unmatched by any previous or subsequent author’ (135). In a more popular format, see also McQuain and Malless. See Berg (1993) for more detail on these, and a much wider glossary. See Shklovsky (1998), especially Chapter 1, ‘Art as Device’. This issue is discussed in more detail in Durkin (2002: 68).

References All Shakespeare citations in this chapter are given as cited by OED Online. NED refers to the New English Dictionary fascicles published in 1904–1905. Berg, Donna Lee 1993. A Guide to the Oxford English Dictionary. Oxford and New York: Clarendon Press. Burchfield, Robert 1992. Points of View: Aspects of Present-Day English. Oxford and New York: Oxford University Press. Durkin, Philip 2002. ‘Changing Documentation in the Third edition of the Oxford English Dictionary: Sixteenth-century vocabulary as a test case’ in: Teresa Fanego, Belén Méndez-Naya and Elena Seoane (eds), Sounds, Words, Texts and Change. Amsterdam: John Benjamins, 65–81. Jespersen, Otto 1909. A Modern English Grammar on Historical Principles. Heidelberg: Carl Winter. Lerer, Seth 2007. Inventing English: A Portable History of the Language. New York and Chichester, West Sussex: Columbia University Press. McQuain, Jeffrey and Malless, Stanley 1998. Coined by Shakespeare: Words and Meanings First Penned by the Bard. Springfield, MA: Merriam-Webster. Murray, James 1888. New English Dictionary. Oxford: Oxford University Press. Riffaterre, Michael 1983. Text Production. New York: Columbia University Press. Schäfer, Jürgen 1980. Documentation in the O.E.D.: Shakespeare and Nashe as Test Cases. Oxford: Clarendon Press. Schmidt, Alexander 1874–1875. Shakespeare Lexicon and Quotation Dictionary (Vols 1 and 2). Mineola, N.Y. Dover Publications.


33

Shewmaker, Eugene 1996. Shakespeare’s Language: A Glossary of Unfamiliar Words in His Plays and Poems. New York: Facts on File. Shklovsky, Viktor 1998. Theory of Prose, trans. Benjamin Sher. Urbana-Champaign: University of Illinois. Wellek, René 1969. The Literary Theory and Aesthetics of the Prague School. Ann Arbor: University of Michigan.

Chapter 2

Shakespeare’s Vocabulary: Did it Dwarf All Others? Ward E. Y. Elliott and Robert J. Valenza

1. Size Did Shakespeare’s vocabulary dwarf all others?1 Many respected scholars have said that it did and does. F. Max Müller, the great Victorian philologist and Sanskrit scholar, said it thus: Shakespeare, who displayed a greater variety of expression than probably any writer in any language, produced all his plays with about 15,000 words. Milton’s works are built up with 8,000; and the Old Testament says all that it has to say with 5,642 words. By contrast, he added, English country labourers of the day supposedly ‘had not 300 words in their vocabulary;’ ‘a well-educated person in England, who has been at a public school and at the university, who reads his Bible, his Shakespeare, and the ‘Times’, . . . seldom uses more than about 3,000 or 4,000 words in actual conversation; . . . and eloquent speakers may rise to a command of 10,000’ (Müller 1862: 266–7, 1891: 377–9). Müller’s magnum opus, The Science of Language (from which these numbers are taken), is still in print. He was widely followed in subsequent centuries; and he is still widely cited. Notable twentieth-century supporters of the Müller thesis were Ernest Weekley, who published five editions of The Romance of Words and ten other widely read books on words and names; Louis Marder (1962), long-time editor of the Shakespeare Newsletter and Robert McCrum, William Cran and Robert MacNeil, authors of the companion text to the 1986 Public Broadcasting System series, The Story of English, now in its third, revised edition (2002). In 1928 Weekley wrote: ‘Of Shakespeare it may be said without fear of exaggeration that his contribution to our phraseology is ten times

Shakespeare’s Vocabulary: Did it Dwarf All Others?

35

greater than that of any writer to any language in the history of the world’ (Weekley 1952 [1928]: 55). Alfred Hart, the leading mid-20th-century authority on Shakespeare’s use of ‘new words’, though less sweeping than Müller and the others, thought that Shakespeare’s vocabulary and coinages far outstripped those of his early contemporaries, Marlowe, Greene, Peele, Kyd, and Chapman. In 1942 he wrote: [S]hakespeare was a lifelong and insatiable word-collector. . . . Shakespeare’s unmistakable sign-manual in a play is the presence of plenty of words peculiar to it alone. . . . [A]n author who is credited by the compilers of the Oxford English Dictionary with being the first user of about 3,200 words . . . has verbal riches compelling the employment of superlatives in describing them. . . . Shakespeare’s vocabulary was fuller, more varied and expressive of finer shades of meaning than that of any of his early contemporaries . . . . [Marlowe’s (and Chapman’s)] inferiority in vocabulary is manifest. (Hart 1942: 22, 28, 450–1) In 1986 McCrum et al. said this to a Public Broadcasting System audience of millions: ‘Shakespeare had one of the largest vocabularies of any English writer, some 30,000 words. (Estimates of an educated person’s vocabulary today vary, but it is probably about half this, 15,000)’ (2002 [1986]: 102). The Müller thesis marches on in this century in McCrum’s third edition, and others heartily concur: ‘The average educated man today . . . has a working vocabulary of less than half that of Shakespeare’(Bragg 2003: 135); ‘Take a look at Shakespeare’s enormous vocabulary . . . 27,780 are different words. The average person . . . has a recognition vocabulary of about 5,000 words. Some of the greatest writers may have twice this capability’ (Ravi 2003). It should not be forgotten, when one is thinking of superlatives for Shakespeare’s vocabulary, that English itself is said to have the richest vocabulary in the world today, with at least a million words, depending on what gets counted (Crystal 1995: 119),2 and growing. According to McCrum et al., English is five to ten times richer in words than German or French (2002: 10). Shakespeare lived and wrote during the language’s most glorious and formative years, years when its vocabulary was growing explosively at rates unmatched before or since (Nevalainen 1999: 338–9), years when, driven by patriotism, Protestantism and the mass markets both permitted and demanded by the printing press, it grew from a collection of base, rude, rustic, spoken local dialects to replace Latin as the default language of national

36


literary discourse (Hussey 1992 [1982]: ch. 2, McDonald 2001: ch. 1). It wasn’t just that Shakespeare had the most superlative English vocabulary of his day – if, in fact, he did. His day was the most superlative period of vocabulary growth for English, and English vocabulary today leaves all others in the dust. How much more superlative could you get? Everyone in Shakespeare’s day was adding furiously to the language, and judging from listings in the Oxford English Dictionary and the Chronological English Dictionary, Shakespeare’s new-word coinages seem to have led all the rest, even after many corrections for extravagant overcounts in the early twentieth century. We argue in Part 2 that Shakespeare’s coinages are still probably overcounted by a factor of at least two. For now, it is enough to note that coinages were an important component of the persistent belief that Shakespeare’s vocabulary dwarfed all others in size, variety and creativity. Shakespeare’s supposedly outsized vocabulary and coinage rates are frequently invoked in authorship controversies. Hart’s studies of Shakespeare’s new-word usage figured prominently in MacDonald Jackson’s and Kenneth Muir’s landmark ascriptions of A Lover’s Complaint to Shakespeare, helping explain why A Lover’s Complaint’s abundance of words new to Shakespeare strengthened, rather than contradicted the Shakespeare ascription ( Jackson 1965: 8–19, Muir 1973 [1964]). The powerfully documented Jackson and Muir studies overturned the consensus of the day that A Lover’s Complaint was not Shakespeare’s; thanks to their work, the consensus today (which we and others now think is probably wrong, see Elliott and Valenza 1997 and 2004, Vickers 2007) is that Shakespeare did write A Lover’s Complaint. Moreover, if mainstream scholars savoured the idea of Shakespeare’s verbal pre-eminence with their morning coffee, anti-Stratfordian insurgents devoured it for breakfast, lunch and dinner, providing, as it seemingly did, compelling evidence that Shakespeare was far too learned to have been the lowly Stratford grain-dealer. ‘Never’, said Oxfordian doyen Charlton Ogburn Jr. after a long, exclamatory discourse on Shakespeare’s learning, citing Müller, Marder and others, ‘has such verbal prodigality as Shakespeare’s been approached’ (Ogburn 1984: 291–2). To be sure, there was at least one notable oldtime Müller sceptic, not so much of the richness of Shakespeare’s vocabulary as of the supposed poverty of others’ vocabulary. The sceptic was Otto Jespersen (1860–1943). His classic Growth and Structure of the English Language was first published in 1905 and last revised by him in 1938. In three tightly-packed pages on the subject, Jespersen conceded to Shakespeare his customarily estimated vocabulary of 20,000+ words, but he thought that Müller’s notion that a farm-labourer uses only 300 words was ‘obviously wrong’. So, it seems,


37

was his notion that educated persons used not more than 4,000–10,000 words. Two-year-olds with doting, documenting parents were known to use 489–1,121 words; Swedish peasants seemed to have more than 26,000 words; and enterprising college professors E. S. Holden and E. H. Babbitt, using randomly selected pages from a dictionary, estimated that most of their students recognized a little below 60,000 words (Jespersen 1982 [1938/1905]: 200–1). Jespersen added: These statements are easily reconciled with the ascription of 20,000 words to Shakespeare. For it must be remembered that in the case of each of us there is a great difference between the words known . . . and the words actually used in conversation. . . . If Milton as a poet uses only 8,000 as against Shakespeare’s 20,000 words, this is a natural consequence of the narrower range of his subjects, and it is easy to prove that his vocabulary really contained many more than the 8,000 words found in a Concordance of his poetical works. We have only to take any page from his prose writings, and we shall meet with a great many words not in the Concordance (202, listing 21 such words on a single page of Milton’s Areopagitica). The greatness of Shakespeare’s mind is therefore not shown by the fact that he was acquainted with 20,000 words, but by the fact that he wrote about so great a variety of subjects . . . that he needed this number of words in his writings.(ibid.) Jespersen’s cautionary book went through ten editions and must have been read and admired by thousands of people – and you can still find his cautions reflected and updated in the writings of at least two modern English-language encyclopaedists, Tom McArthur (1992: 1092), and David Crystal (1995: 123, 2004: 317). But Jespersen (unlike Müller) was out of print for decades; not enough people have paid attention to McArthur and Crystal; and it would seem from the continued prevalence of Müller’s view that the old-line bardolatry still holds sway. But is it right? Our short answer, based on our analysis of the Claremont Archive of Renaissance Texts and on type-token figures for other writers graciously supplied to us by Professor David Hoover, of New York University, is that Müller and company were right that Shakespeare had a big vocabulary, but wrong in supposing that it was bigger or better than other writers’ vocabularies, either in his own day or since. Much of Shakespeare’s apparent pre-eminence over others is due to the greater accessibility of his writing. He wrote more than most others and was better recorded, catalogued

38


and anthologized. The people who wrote the Oxford English Dictionary could get to him in ways that they could not get to other writers. Müller, for example, could choose between three Shakespeare concordances, notably Mary Cowden Clarke’s Complete Concordance to Shakespeare (1881 [1845]), which had been 16 years in the making. His estimates for works with lexicons – Shakespeare, the Bible and parts of Milton – were in the right ballpark; but, where he lacked a lexicon, as with farm labourers, he was way off; and, where, as with Milton, he had a lexicon based on a much shorter, less comprehensive corpus than Shakespeare’s, he had no way to correct for it. What is true of Müller was largely true of his uncomputerized successors, even the best of them, Hart, who was well aware of the limitations of handmade lexicons but could only partially compensate for them. It takes a lexicon, an inventory of word types, to count someone’s total manifest vocabulary; handmade lexicons took years to produce and were unavailable for most other writers; and, even when available, they offered no easy way to correct for differing corpus size, which is crucial for vocabulary comparisons since, other things equal, larger corpora yield larger vocabularies of distinct words. The most authoritative modern Shakespeare lexicons are Marvin Spevack’s (1968, 1973), a long one and a short one, both meticulously compiled on a computer from the meticulously edited Riverside Shakespeare (Evans et al. 1997 [1974]) – though it would be surprising if even the Riverside preserved every countable distinction between one archaic spelling and another – burthen/burden; murther/murder; fadom/fathom; vild/vile; Dolphin/ Dauphin; crocadile/crocodile; wreck/wrack, and so on. Every Shakespeare edition, in a sense, is a translation from Early Modern to contemporary English, and variations among different editors’ commonizing zeal produce defendable variations among the countable different words to be found in each one’s canon (Taylor 1989: 254, 316). Spevack machine-counted 29,066 different words – ‘types’ – from a Riverside Shakespeare canon of 884,647 total words – ‘tokens’ (Spevack 1973: v). His is considered the best available count of Shakespeare’s manifest vocabulary, taken from words Shakespeare actually recorded, but even his count has limitations typical of machine counts, our own included. It is utterly reliant on the editor’s orthographical judgement; it includes some words supplied by editors, and it includes words from parts of the canon – maybe as much as a tenth of it – which were not written by Shakespeare solo but by some combination of Shakespeare and a co-author (see Elliott and Valenza 1996, Vickers 2002b, Wells 1987). Most important, it counts every different word inflection found, not just each word’s root (also called its lemma, or headword). ‘Horse’ and ‘horses’


39

are two words by this convention, not two variants of one dictionary headword.3 Both of these conventions take a large view of Shakespeare’s vocabulary, especially using an inflected lexicon, which is much larger – maybe two-thirds larger – than the lemmatized vocabularies used by hand-counters like Hart.4 If the two-thirds correction factor is good, Shakespeare’s lemmatized, manifest vocabulary is closer to 17–18,000 than to 29,000 words (Hart 1943a = 17,677; Scheler 1982: 89 = 17,750; WordHoard = 18,0905) – close enough, given the many arguable ways of counting words, to make Müller’s estimate of 15,000 look remarkably accurate. On the other hand, Spevack’s count, like ours, understates Shakespeare’s vocabulary in three ways, two small, one large. The first small one is multiple meanings of homographs, different words with the same spelling. Our Webster’s New Collegiate Dictionary (which was new in 1950) lists eight meanings for spring as an intransitive verb, eight more for spring as a transitive verb, nine for spring as a noun, and three for spring as an adjective, 28 different possible meanings in all, and up to 28 possible different words to the context-sensitive hand counter – but only one word to most computers, including Spevack’s and ours. Correcting for homographs, if you could do it, would modestly increase the total, probably by at least the 700+ words listed in Spevack (1968, vol. IV, App. D). The second small understatement is linked word phrases like ‘come on’ or ‘grown up’, lexical units which are functional equivalents of a single word and mean something distinguishable from ‘come’ and ‘grown’, yet normally don’t get caught and counted by computers. The largest understatement of any kind of vocabulary count, machine or hand, is the one touched on by Jespersen, that they don’t measure how many words an author knows – his latent vocabulary – but only those he has used and recorded – his manifest vocabulary. Some writers use the term ‘recognition vocabulary’ or ‘passive vocabulary’ as near-synonyms for ‘latent’, and ‘use vocabulary’ or ‘active vocabulary’ as near-synonyms for ‘manifest’. Latent /passive /recognition vocabularies are always larger than manifest/ active/use vocabularies. Works which cover many different subjects, as Jespersen noted, should show larger vocabularies than works with fewer subjects. And large corpora (other things equal) should normally show larger vocabularies than small ones because each new token (token = countable word, including every repetition) added to a given corpus offers a chance of adding a new type (type = distinct word type, doesn’t count repetitions) as well, by transferring the type from the author’s latent vocabulary to his manifest vocabulary. Hart well understood this last point and, with the help of concordances, calculated Shakespeare’s rates of introducing once-used,

40


or ‘new’ words for Shakespeare’s poems and a few plays. Such new-toShakespeare words amounted to 10 per cent of the total word types in King Lear and appeared about once every ten lines of text in Lear and Hamlet, once every twenty lines in earlier plays (Hart 1943b: 128–31, summarized in Jackson 1965: 11–2). Hart did his best to control for date and text length, but he was badly hampered by the limited concordancing available to him, which was skimpy for everybody but Shakespeare, and which could not easily control for corpus or sample size. Like Müller and everyone else between the nineteenth century and the computer era, Hart had the nineteenth-century deconstructing genius for breaking down Shakespeare’s and others’ writing into their constituent components – words, grammar, metrics and so on (see Taylor 1989: 193) – but he couldn’t apply it as freely and systematically as you can with computers today. All these stalwart old stylometricians were roadbound and oxcart-borne, amazingly resourceful and proficient in using the slow, judgemental tools they had, and hugely better than their predecessors who lacked even roads (lexicons) and oxcarts, but gravely limited, nonetheless, by the technology of their time. Computers and sizeable machine-readable text archives changed this situation profoundly and, in 1976, made possible an astonishingly sophisticated, landmark estimate of Shakespeare’s total recognition vocabulary. However, the landmark appeared in Biometrika, a distinguished statistical journal unfamiliar to most number-shy Shakespeare and language scholars. It was a landmark all but invisible to regular Shakespeare scholars but of surpassing visibility and importance to us. Even without their own text archive, and using only the then-new Spevack concordance, plus a clever methodology first propounded by Sir Ronald Fisher in the 1940s, and massive crunching with an IBM 360, master statisticians Brad Efron and Ronald Thisted, were able to make a wholesale extrapolation of Shakespeare’s newword productivity. They concluded that Shakespeare, had he doubled his historic output of 884,647 tokens, would have added 11,460 new types, plus or minus 150, to his known 31,564. Even more astonishingly, they showed that Shakespeare’s latent vocabulary had to be at least 35,000, besides the ‘known’ 31,564, for a total vocabulary of at least 66,000 inflected words (Efron and Thisted 1976: 435). If David Crystal is right in supposing that the English language contained about 150,000 lemmas in Shakespeare’s day (Crystal 2004: 317), and if our rough estimate of Shakespeare’s inflected-tolemma ratio of 1.67:1 is close to correct, it means that Shakespeare knew about 40,000 lemmas, more than a quarter of words then in the English language.6


41

Like Fisher’s original study of how to forecast the discovery of new butterfly species from known past discovery rates (1943), the Efron-Thisted study is still seen as a triumph of statistical ingenuity, a landmark demonstration of how accurately you can guess the unknown from the known, with the right techniques. No one has tried more elaborately than we have to retest Efron’s and Thisted’s main techniques, assumptions and conclusions from the ground up. We found some limits for microlevel applications (below) but concluded that their findings for Shakespeare in the large are still sound and well-deserving of their landmark status among statisticians (Valenza 1990). This chapter is, in part, an effort to make it a landmark for language and literature scholars as well. A decade after their unheralded breakthrough, responding to Gary Taylor’s ascription of Shall I Die? to Shakespeare, Thisted and Efron attempted to apply the same methods they had applied wholesale in 1976 at retail to 500-word Shakespeare and non-Shakespeare samples comparable in length to Shall I Die? They concluded that Shall I Die? had enough Shakespeare-new words to be Shakespeare’s, but other sample poems by Marlowe, Jonson and Donne did not (Thisted and Efron 1987).7 Though it was the second of the two Thisted-Efron articles (1987) that started the Shakespeare Clinic’s work, it may well be the first (1976) which finishes it. Working from the second article, we found that ‘new words’ were one of the Clinic’s best tests − but not in the way that the Müller thesis would have led us to expect. Shakespeare did not invent and use new words so furiously as to dwarf his peers; quite the contrary. After we calculated Shakespeare’s normal, expected new-word ranges, using our Intellex software to isolate 1,500-word blocks from the rest of the Shakespeare corpus and to make some necessary corrections for type-token ratios, we found (unsurprisingly, since we drew our Shakespeare profiles to keep false negatives at or below 5 per cent) that none of our 3,000-word blocks from Shakespeare’s poems, and only 5 per cent of those from his plays, exceeded Shakespeare’s expected new-word range. The surprise was that 64 per cent of such blocks by other poets, and 56 per cent of such blocks by other playwrights exceeded Shakespeare’s normal expected range of words new to him. In other words, most of other people’s poems and plays have more Shakespeare-new words than 95 per cent of Shakespeare’s. Also surprisingly, not a single other-authored block fell below Shakespeare’s expected new-word range. How could this be, when everyone knew from Müller and company that Shakespeare’s vocabulary and new-word coinage rates dwarfed everyone else’s? Our first thought was that the other writers must have different


42

inventories of favourite words, so different as to overbalance Shakespeare’s supposedly overwhelming advantage in vocabulary richness. Our second thought was to wonder whether Shakespeare’s supposed advantage was as overwhelming as everyone believed. Both of these thoughts are probably true. The second one is what prompted us to write this chapter. We didn’t and don’t doubt that he knew 29,000 words; they were all there in the 884,000word canon. Or that he coined hundreds, if not thousands of words. But did it actually set him above his peers and posterity? Only if the traditional low estimates for other writers were accurate. We believe they were not. To get a better estimate of how Shakespeare’s vocabulary compared with pertinent others, we first controlled for size, taking large blocks of about 40,000 words from Shakespeare, eight of his contemporaries and Milton. We gave them three tests, one traditional, two novel and doable only with our own program, Intellex, to test raw variety, distinctiveness and total inferred vocabulary. Table 2.1 and Table 2.2 (appendix), from which it is drawn, show the results: no matter which test you use, Shakespeare comes out toward the middle of the pack. If anyone towered over the others, it was Milton and perhaps Spenser, not Shakespeare. Test One, Types to Tokens, is a well-known but imperfect one. It measures raw variety. Milton used an average of 6,500 different words (or ‘types’) in each of the two 40,000-word (= 40,000-token) halves of Paradise Lost. Shakespeare used an average of 5,470 words per block in four 40,000-word blocks of plays and one of poetry. Types-to-tokens was too variable, overlapping and sensitive to sample size for us to have used it as a reliable author identifier for the Shakespeare Clinic, but, with sample size very large and well controlled, it says clearly enough that, in terms of raw variety of words used, Milton was at the top, Fletcher at the bottom, and Shakespeare and the others in between.

Table 2.1

Three vocabulary tests: Shakespeare and eight others

Author: number of blocks

Types

Milton: 2 Shakespeare: 5 6 others: 9 Fletcher: 2

6,500 5,470 5,223 4,444

New words 1,691 905 839 430

Inferred vocab. 131,953 62,569 66,132 49,983

Three vocabulary tests, Shakespeare, seven contemporaries and Milton, 18 40,000-word blocks averaged for each author or author-group in each category. Milton leads the group; Fletcher trails; Shakespeare and six others – Jonson, Dekker, Marlowe, Middleton, Chapman and Greene – fall in between. All these tests measure inflected, not root vocabularies. ‘Types’ and ‘New Words’ are per block, averaged.


43

Test Two is New-to-the-Group Words, employed for the first time in this chapter. If types-tokens measures variety, New-to-the-Group Words measure distinctiveness. Using Intellex, we compiled a baseline lexicon from a corpus of 600,000 words, 160,000 by Shakespeare, the rest by seven other authors. The 600,000 tokens yielded 28,747 different types, 12,265 of them ‘new words’ found in only one of the fifteen 40,000-word blocks. Shakespeare, with 27 per cent of the group’s tokens, contributed 31 per cent of the group’s new words, a bit more than his share, but far less per block than Milton, who was, again, at the top, with Fletcher at the bottom, and Shakespeare and others in between.8 Test Three, Total Virtual Inferred Vocabulary (TVIV), is our newest and most ambitious. From earlier work (Valenza 1990), we had two pieces of solid-looking ground, to which we now add a new way of inferring from them a writer’s total vocabulary, from a much smaller baseline and with much less elaborate computation than it took Thisted and Efron to do similar calculations for all of Shakespeare in 1976. The first piece of solid ground was Thisted and Efron’s own ‘bootstrap’ calculations, which, using two completely different methodologies, came up with the same low-range estimate for Shakespeare’s latent vocabulary of 35,000 words. This, with his manifest vocabulary of 31,000 words, added up to a well-supportable estimated total vocabulary of at least 66,000 words – or perhaps something closer to 64,000, if added to Shakespeare’s corrected manifest vocabulary of 29,000 words per Spevack’s 1974 count. Either would do for our purposes, but, for consistency with Thisted-Efron’s earlier article, we use the original higher one, of 66,000 words. The second piece of solid ground was Valenza’s own massive recalculations of Thisted-Efron’s data in 1990 for all of Shakespeare and all of the King James Bible. Both of these large corpora, though differing greatly in provenance, genre, subject matter and vocabulary richness, turned out to have normalized word frequency curves which tracked each other almost exactly across the entire frequency range (Figure 2.1). If two corpora as different from each other as Shakespeare and the Bible track each other that closely, you should be able to infer a plausible comparable frequency curve for other poets and playwrights. From that, you should be able to infer their total vocabularies simply from the number of word types they have to go through to get through half the tokens in their corpus or sample. Another way of putting this is that (1) if an author has a big vocabulary of different word types, an analyst should go through more types before exhausting all the author’s tokens; and (2) if different authors’ word-frequency curves do track one another, the analyst should


44 4.00

Shakespeare

3.50

King James Bible

3.00 2.50 2.00 1.50 1.00 0.50 0.00 −0.50 −1.00 0

10

20

30

40

50

60

70

80

90

100

Figure 2.1 A comparison of normalized word type frequency profiles for Shakespeare and the King James Bible. Shakespeare introduces new types at roughly twice the rate at which they occur in the Bible across the entire profile. Source: Valenza, 1990: 32.

also go through more types before exhausting half of each author’s tokens, and this adjustment makes the counting task much easier. If you accept the Thisted-Efron figure of 66,000 for Shakespeare’s total inferred vocabulary as reasonably solid, as most do, and suppose that most word frequency curves track one another, as is certainly so of Shakespeare and the King James Bible and could well be so for others, you can then make a plausible guess about a given author’s total inferred vocabulary simply from counting the types it takes to get through half the tokens and applying a correction factor to place the block’s frequency curve relative to Shakespeare’s total vocabulary of 66,000 words. See Table 2.1, last column; in Table 2.2 (appendix) the 18 blocks by nine authors are ranked by TVIV. The correction factors are such that taking 77 types to go through 20,000 tokens would put a given block’s TVIV just about at the level of Shakespeare’s, 66,000. Higher or lower type counts at the halfway mark would put a block’s TVIV proportionally higher or lower, as can be seen by comparing Milton1, with 127 types at the 20,000-token halfway mark, implying a TVIV of 140,000 types, and Fletcher1, with only 63 types at the halfway mark, implying a TVIV of 49,000 types. The extra ‘V’, for ‘virtual’, is our way of underscoring the speculative nature of our assumptions that word frequency curves track each other.


45

Do these tests make any sense? Our first move was to scan Table 2.1 and Table 2.2 (appendix) for consistencies and discrepancies. The consistencies are many: both Milton samples cluster at the top by all three tests. Both Fletcher samples cluster at the bottom by all three tests. Everyone else falls in between by all three tests. Shakespeare himself, with five blocks measured, shows the greatest range of inter-block discrepancy, with three blocks of tragedies and romances very tightly clustered around a TVIV midpoint of 60,000 or so, but with two outliers, poems at 84,000 and early comedies at 49,000. Neither of the two Shakespeare outliers is shockingly distant from Shakespeare’s mean, the Thisted-Efron-estimated TVIV of 66,000: the poems are +36 per cent, the comedies –26 per cent. Neither the long outlier nor the short one clashes with common-sense expectations, that poems would be richer than plays, and tragedies richer than comedies – an expectation which is also confirmed by the other two tests. The mean of all five Shakespeare blocks tested is 62,569 words, just 5 per cent lower than the 66,000 calculated for the entire canon by Efron and Thisted (1976), and only 2 per cent lower if Shakespeare’s actual vocabulary is 64,000 per Spevack’s later, lower estimate. Most two-block samples by the remaining authors, three out of the four remaining pairs, are separated by no more than ± 5,000 words (± 7 per cent) from the two-sample midpoints. In short, though the three tests measure richness in very different ways – one (types-tokens) with no reference at all to frequency curves, one (new words) based solely on the bottom 1 per cent of the frequency curve (least frequent), and one (TVIV) based solely on the top, most frequent, half – they are remarkably consistent with one another and with prior work by Thisted and Efron – and in remarkable agreement that Müller and company were wrong about other writers’ vocabularies relative to Shakespeare’s. Once you remove the gross biases of corpus size from the calculations, it becomes clear that, if anyone’s vocabulary dwarfed others in size, it was Milton’s, and maybe Spenser’s, not Shakespeare’s. If so, we would conclude that Shakespeare’s pre-eminence was much less a matter of how many words he knew, than of how he used them. The spreadsheet also has a scattering of further comparative statistics on other writers, including a few willing members of the Claremont McKenna College faculty, to the same general effect: the first five colleagues we could get to give us recognition percentages of three sample pages from the dictionary all turned out, after extrapolation, to have vocabularies five or ten times larger than the 5–10,000 words confidently and wrongly assigned to educated people by Müller and his modern followers. It also turns out,

46


not surprisingly, that we were not the first to think of this simple, obvious, dictionary-sampling methodology, which was equally available to Müller and the others, but apparently never tried by them. Jespersen and Crystal both mention such tests, and with similar outcomes, that is, crediting runof-the-mill college-educated moderns with Shakespeare-sized or larger inflected vocabularies of 50–80,000 words (Crystal 1995: 123, Jespersen 1982 [1938/1905]: 201). TVIV’s project even larger inflected vocabularies for other people than Shakespeare, including two moderns well known to us who are not otherwise to be compared with Shakespeare: 132,000 words for Spenser, 96,000 for Melville, 81,000 for the Iliad, almost 100,000 for Elliott and Valenza’s manuscript on Shakespeare, and 177,000 for Elliott’s alone on population policy. Jespersen and Crystal were in no position to test people’s TVIVs the way we have, but our results are entirely consonant with two of Jespersen’s views, that modern vocabularies equal or exceed Shakespeare’s, and that vocabulary richness should increase with the variety of subjects addressed (Jespersen 1982 [1938/1905]: 202); hence the much greater TVIV for Elliott’s wide-ranging population manuscript than for our more tightlyfocused Shakespeare book. People who write about Shakespeare often use much fancier language than Shakespeare did himself. One need only open a late twentieth-century Shakespeare journal to find words far more abstruse than Shakespeare’s – fetishisation, commodification, poststructuralism, inferred virtual vocabulary and enclitic microphrases, for example. Some of these inkhorn polysyllables are our own, and there is a place for them, but they are all far, far from Shakespeare. Shakespeare managed to do some of his best work with short, concrete words like ‘seething brains’, or ‘cool reason’, artfully combined: . . . I never may believe These antic fables, nor these fairy toys. Lovers and madmen have such seething brains, Such shaping fantasies, that apprehend More than cool reason ever comprehends. (A Midsummer Night’s Dream 5.1.2–6) The result would be the same with any set of Shakespeare’s most famous lines – ‘Friends, Romans, countrymen’, ‘All the world’s a stage’, ‘To-morrow, and to-morrow, and to-morrow’, and many others. In terms of sheer numbers of words known, and by several different measures, many of us can match or exceed Shakespeare – unsurprisingly,


47

perhaps, given the vast expansion of the language and the whole human enterprise since Shakespeare’s time. So could a number of his own contemporaries who drank from the same Castalian fountain of words that Shakespeare did. Quantitatively, Müller and company were completely wrong. Qualitatively, in terms not of how many words Shakespeare knew, but of how well he used them, we would suppose that there is still plenty of room for bardolatry. Shakespeare learned early how to strike deep, not with an outsize inventory of long, inkhorn words, but with a par-for-the-course inventory, mostly of plain words, surpassingly well chosen and put together. It’s not too soon, or too late, for the rest of us to learn it too.

2. Shakespeare’s word coinages What about the other Shakespeare commonplace, that he invented more new words than anyone else? That, too, could well be a myth, but we do not yet have the evidence that would settle the question. Some of today’s authorities still rely on old, extravagant overcounts of Shakespeare’s newword coinages. Seth Lerer tells us, for example, that Shakespeare ‘coined nearly six thousand new words’ ‘at a rate unmatched by any previous or subsequent author’ (2007: 129, 135). Most other modern estimates are much lower, clustering around 1,700,9 still high enough for Shakespeare to surpass all others. But, as Giles Goodland’s ‘Strange deliveries’, in this volume, Chapter 1, shows, current estimates could still be twice too high, and there is little evidence that the deflationary process begun in ‘Strange deliveries’ has run its full course. Bryan Garner, using the Chronological English Dictionary, counted 10,302 loanword neologisms added to the English language between 1580 and 1619, a 39-year span (1982: 151). Shakespeare’s writing life, which ran from about 1590 to 1613, encompassed almost 60 per cent of this period, good for, say, 6,000 of the whole period’s new, mostly Latinate loanwords. Garner thought that Shakespeare contributed just over 600 of them, about a tenth. Latinate loanwords are only a fraction of all newly coined words, but Garner’s one-tenth estimate for loanwords both matches and cites estimates by Joseph T. Shipley (1977: 28), and is based on tables in Jürgen Schäfer (1973: 206–20) that Shakespeare was the first user of ‘well over 1,700 words’ ‘one new word in every ten’. ‘Shakespeare’, Shipley concluded, ‘was the greatest word-maker of them all’ (cited, Garner 1982: 153). David Crystal, having scanned the electronic OED for coinages, and subtracted nonsense words like gratillity – but not ‘parallel’ words used by others at about the

48


same time – found 2,035 lexeme coinages for Shakespeare, about a tenth of Shakespeare’s 20,000 or so total headwords, and many more than Nashe (around 800 coinages), Spenser (c. 500), Sidney (c. 400) or Marston (c. 200); (Crystal 2004: 327–8). Following Crystal and Goodland, we would guess that the 2,000+ coined words estimate is probably still too high for Shakespeare’s actual contributions of coined words relative to others’, for some of the same reasons we would guess that his overall vocabulary is still overestimated relative to others’: Shakespeare’s words were well recorded and catalogued, and readily accessible to the compilers of the original Oxford English Dictionary; other people’s words were not (Crystal 2004, Jespersen 1982 [1938/1905]: 211, Schäfer 1980). His ‘first-uses’ would routinely be dated from the first recorded mention of the play whose later publication turned out to contain it. Lesser writers’ new words were generally dated from first publication, not first mention. The OED2 now dates everyone’s first-uses from first publication, not first-mention, and, as Goodland has shown, fixing that accounting bias has cost Shakespeare many first uses. Shakespeare gets credit for the first recorded mention of words like ‘sblood and Newgate, but it’s hardly likely that he originated these terms. He also gets credit for putting together composite and compound words like unreal, worthless, well-read or worm-hole, whose constituent elements were not his invention (Crystal 2004: 325). Others likewise get credit for such composites, but one wonders whether Shakespeare doesn’t profit disproportionately from such accounting. He gets more credit for malapropisms, nonce-words and proper names than, say, Thomas Nashe. Shakespeare had many such words, and it seems odd for us to count them as coinages if Shakespeare did not intend them as such. Fewer of his new words were overlooked in the OED; and more ‘parallel citations’ of new words from the same year are credited to him than to others (Goodland 2011, Schäfer 1980: ch. 2). According to Crystal, up to 644 headwords used by others within 25 years of Shakespeare’s first use could be described as ‘parallel citations’. If we divide these half and half between Shakespeare and others, it gives him 1,713 coinages, about the going rate today for Shakespeare coinage estimates (Crystal 2004: 326). Further whittling of the total will be done as more of other people’s writings gets digitized, and as co-authored parts of the canon get more clearly distinguished from Shakespeare’s parts. Both of these are happening apace – see Goodland’s ‘Strange deliveries’, Chapter 1 in the present volume, and Vickers 2002a and 2002b. If someone else wrote A Lover’s Complaint or the first two acts of Pericles, Shakespeare should not get the


49

credit for the coinages these contain. Moreover, over 900 of Shakespeare’s ‘coinages’ were words like adoptious, which never caught on with others and hence, as coinages, are more like private IOU’s or confederate money than like actual current coin of the realm (Crystal 2004: 326; see the discussion of nonsense and nonce-words above). When George W. Bush comes up with ‘Bushisms’ like subsidation, analyzation, hopefuller, more few and explorationists, we suppose that he is struggling to follow accepted rules of word formation (see Crystal 2004: 314–5) but has gotten in over his head. Everyone sniffs at such gaffes, and no one praises them as additions to the language (http://slate.msn.com/id/76886/). If Bush gave us words like insultment, omittance, opulency, revengive, thoughten or casted, these would likewise be gathered and laughed at as ‘Bushisms’. But it was not Bush who gave us the second set, it was Shakespeare – and his gaffes are hailed as brilliant landmarks of ‘linguistic daring’, fresh evidence of his peerless mastery of the language, 24-carat coinages for Shakespeare that would be dismissed as pot-metal if they came from someone else. It is probably true that the lines between obvious gaffes and permissible creative license are clearer now than they were in Shakespeare’s time,10 but a coinage still in circulation today, like acutely, carries more weight than one like insultment that never caught on and would be a gaffe or a malapropism if it came from anyone else. Again, we suspect that Shakespeare’s coinage list would be considerably shorter if he were held to the same standard as others. Counting coinages has most of the problems of counting other words – making due allowance for corpus size, latency, inflections, multiple meanings and so on – plus special problems of its own, which are harder to do with computers and have been less explored: sorting out what counts as a word, whether it is one person’s real coinage, and not another’s, and deciding whether it has caught on. The bigger, better and more searchable the corpus of available comparison texts, the more and better such sorting can be done. David Crystal has pushed this kind of analysis farther than most, but it is still in its earliest stages (Crystal 2004: 317–29). We do suspect from a Jürgen Schäfer estimate (1980: ch. 4), that Shakespeare’s rate of ‘innovation units’ per thousand words of corpus is lower than Nashe’s, but we don’t know yet how it compares with, say, Jonson, whose total corpus and inferred vocabulary seem similar in size to Shakespeare’s, nor with Milton, whose inferred vocabulary looks much larger than Shakespeare’s, but who wrote too late to get in on the EarlyModern orgy of word-creation. After Crystal’s initial whittling down (but not Schäfer’s or Goodland’s), Shakespeare still retained an impressive

50


residue of core coinages which exceeded what would be left of Nashe and the others similarly whittled (Crystal 2004: 327) – but not nearly as other-dwarfingly as one would have thought from reading Müller, Marder or Shipley. Schäfer’s (1980) computations of authorial innovation rates per thousand published words would put Shakespeare behind Nashe, just as our three different estimates of vocabulary richness per 40,000-word block put him behind Milton and Spenser. Over the years, as analysis has grown more carefully focused, and the impacts of bardolatry better controlled, Shakespeare’s coinage edge over his peers has not grown, but steadily shrunk. Giles Goodland shows us that much more shrinkage can be expected. He started with what looks like a baseline of 122 words beginning with P, Q and Ra-, about 7 per cent of Shakespeare’s estimated 1,700 coinages, and re-examined them against just 40 months of then newly-recorded Renaissance e-texts. From the baseline he subtracted 41 words antedated by the new texts (about one a month); 16 words redated by rules for the first time applied equally to Shakespeare and others; and 17 words that looked like intended nonce-words, not coinages. These reductions cut the baseline by 60 per cent, with no allowance for further antedatings or for the three surviving ‘Shakespeare coinages’ from A Lover’s Complaint, which Shakespeare may not have written – ‘posied’, ‘pellet’, and ‘phraseless’. If all of Shakespeare’s estimated 1,700-word coinages were also deflated by a comparable 60 per cent, it would leave only 680 coinages, with an overall ongoing shrinkage of about 14 words a month. That would be less than Crystal’s 2004 estimate for Nashe (2004: 327–8) and consistent with Schäfer’s conclusion that Nashe had higher rates of innovation than Shakespeare (1980: ch. 2). We would expect Nashe’s coinages, like Shakespeare’s, to suffer some attrition from recent and future additions to Early-Modern e-text databases but, following Schäfer, would be surprised if his coinage counts turned out to be as grossly inflated as Shakespeare’s. In fairness to Shakespeare, again following Crystal, we should note that our simple word-counting does not give Shakespeare (or others) due credit for what could be hundreds of cases where he (or they) added a new sense or meaning to an already-known word (Crystal 2004: 318–21). It is possible that Shakespeare would do better than most by this hard-to-count, qualitative standard and might yet be found to surpass the others even after further massive deflation – but this is a different way of stating the theme of this chapter: that Shakespeare’s genius is not so well measured by the number of words he used or coined, as by the way he used them.


51

Notes 1

2

3

4

5 6

7

8

Our thanks to David Crystal, David Hoover, MacDonald Jackson, and three anonymous readers for helping us try to answer this question. And our congratulations to Hugh Craig, of the University of Newcastle, Australia, for his impressive study, ‘Shakespeare’s Vocabulary: Myth and Reality’, forthcoming in the Shakespeare Quarterly, which we have had a chance to read in manuscript. He arrived independently at conclusions very similar to ours: ‘Shakespeare is in fact no different from his contemporaries in the number of different words he uses.’ Some encyclopaedists say that English could have as many as a billion words, counting proper names, abbreviations, and scientific terms (McArthur 1992: 1091). Horse can also be a morpheme, the smallest linguistic unit of a ‘generated’ or compounded headword (like horsemanship) that has semantic meaning. Horsemanship has three morphemes, horse, -man, and -ship. In principle, a dictionary stripped of compounded headwords found in the language and reduced to nothing but morphemes would be even shorter than conventional dictionaries. But we know of no one who has tried to de-compound all of Shakespeare’s words into morphemes, and we have not tried it ourselves. Hart’s handcounting also permitted him to eliminate proper nouns from his word counts, a task which is harder for machine counts. http://wordhoard.northwestern.edu/userman/index.html. Efron and Thisted’s estimate had a problem which we consider a small one: it seems to be based on a preliminary 1968 Spevack estimate of 31,654 Shakespeare word types, and not on their final 1974 estimate of 29,066 types. No one can remember what caused the difference between the two estimates. Possibly some of the poems from The Passionate Pilgrim were assigned to Shakespeare in the first estimate but not the second. Given the variety of ways a word or a corpus could legitimately be defined, and the fact that both estimates are in the same ballpark with generations of prior estimates, we consider the 7 per cent difference between the two estimates of manifest vocabulary of minor importance compared to the major breakthrough that the two statisticians made in inferring a large latent vocabulary from a smaller manifest one of either size. After many years of crunching, we (and most others, such as Foster 1987, Pendleton 1989, and Vickers 2002a) doubt that Shall I Die? is Shakespeare’s; but Thisted and Efron, in fact, had only claimed that it was a ‘could-be’ for Shakespeare, not a ‘must-be’. We also doubt that Thisted-Efron tests are valid for samples as short as 500 words. But we consider these tests well-validated for samples of 1,000 words or more and copiously validated for the entire Shakespeare corpus of 884,000 words. Thisted-Efron’s 1987 article was the primary inspiration for our Claremont Shakespeare Clinic, a series of student teams using computers to test the works of 37 ‘Shakespeare Claimants’, along with 30-odd poems and plays of the Shakespeare Apocrypha and Dubitanda, for Shakespeare authorship (Elliott and Valenza 1996). One could argue that Milton’s Paradise Lost had a different subject-matter vocabulary from those of the playwrights tested, and would therefore be expected to add more to the group’s new-word inventory than a sample with the same subjectmatter as the others – or that Shakespeare, who contributed at least a quarter of the group’s tokens, might have contributed fewer words per block that were


52

9

10

new to himself. Both could be so, but probably not enough so to bring Shakespeare to Milton’s level. Shakespeare’s own poems score higher on all three of our tests than his tested tragedies, which, in turn, score higher than his tested comedies. But even his poems score below Milton’s, and also below two Dekker plays. Further testing with other combinations of texts could explore these possibilities, but we believe that our comparison of large, standardized blocks from a variety of different authors is far more indicative of their relative vocabularies than the traditional alternative of comparing whole corpora of vastly different sizes. Schäfer (1973: 206–20; Shipley (1977: 28); Garner (1982: 151); Crystal (2004: 326); Hitchings (2008: 124); McQuain (1998 viii [1,500]). It is also probably true that the Shakespeare list of ‘Bushisms’ if read in context, could be discounted in several ways. ‘Revengive’ KL 2.1.48, is found in two Lear Quartos, but the Folio has it as ‘reuenging’. ‘Omittance’ AY 5.5.133; ‘insultment’ Cym., 5.5.141; ‘opulency’ Tim., 5.1.387; and perhaps even ‘casted’ H5 4.1.23, are spoken by characters of whom Shakespeare was making fun for trying to speak beyond their abilities: respectively, Phoebe, Cloten, Poet, and Henry V. It is doubtful that Shakespeare himself intended them as serious words, yet lexicographers count his ‘Bushisms’ but not Bush’s, at full value. It seems to us a double standard.

References All Shakespeare references are to Evans, G. Blakemore and J. J. M. Tobin (eds), (1997 [1974]). The Riverside Shakespeare. Boston: Houghton Mifflin Company. Bragg, Melvyn 2003. The Adventure of English. New York: Arcade Publishing. Clarke, Mary Cowden 1881 [1845]. The Complete Concordance to Shakespeare: Being a Verbal Index to All the Passages in the Dramatic Works of the Poet, Revised Edition. London: Bickers & Son. Crystal, David 1995. The Cambridge Encyclopaedia of the English Language. Cambridge University Press. Crystal, David 2004. The Stories of English. Woodstock and New York: The Overlook Press. Efron, Brad and Ronald Thisted 1976. ‘Estimating the number of unseen species: how many words did Shakespeare know?’, Biometrika 63: 435–47. Elliott, Ward E. Y. and Robert J. Valenza 1996. ‘And then there were none: winnowing the Shakespeare claimants’, Computers and the Humanities 30: 191–245. Elliott, Ward E. Y. and Robert J. Valenza 1997. ‘Glass slippers and seven-league boots: C-prompted doubts about ascribing A Funeral Elegy and A Lover’s Complaint to Shakespeare’, Shakespeare Quarterly 48: 177–207. Elliott, Ward E. Y. and Robert J. Valenza 2004. ‘Did Shakespeare write A Lover’s Complaint? The Jackson ascription revisited’ in: Brian Boyd, (ed.), Words That Count: Essays on Early Modern Authorship in Honor of MacDonald P. Jackson. University of Delaware Press, 117–40.


53

Finkenstaedt, Thomas, Ernst Leisi, and Dieter Wolff 1970. A Chronological English Dictionary Listing 80,000 Words in Order of their Earliest Known Occurrence. Heidelberg: C. Winter, Universitätsverlag. Fisher, Ronald A. et al. 1943. ‘The relation between the number of species and the number of individuals in a random sample of an animal population’, Journal of Animal Ecology 12: 42–58. Foster, Donald W. 1987. ‘ “Shall I die?” post mortem: defining Shakespeare’, Shakespeare Quarterly 38 (Spring): 58–77. Foster, Donald W. 1989. Elegy by W. S.: A Study in Attribution. Newark, Delaware: University of Delaware Press; London and Toronto: Associated University Presses. Garner, Bryan A. 1982. ‘Shakespeare’s latinate neologisms’, Shakespeare Studies 15: 149–70. Goodland, Giles 2011. ‘ “Strange deliveries”: Contextualizing Shakespeare’s First Citations in the OED’, in this volume, Chapter 1. Hart, Alfred 1942. Stolne and Surreptitious Copies, a Comparative Study of Shakespeare’s Bad Quartos. Melbourne and London: Melbourne University Press. Hart, Alfred 1943a. ‘Growth of Shakespeare’s vocabulary’, Review of English Studies 19: 242–54. Hart, Alfred 1943b. ‘Vocabularies of Shakespeare’s plays’, Review of English Studies 19: 128–40. Hitchings, Henry 2008. The Secret Life of Words: How English Became English. New York: Farrar, Straus, and Genoux. Hussey, Stanley S. 1992 [1982]. The Literary Language of Shakespeare. Harlow, Essex: Longman Group Limited. Jackson, MacDonald P. 1965. Shakespeare’s A Lover’s Complaint: Its Date and Authenticity. Auckland: University of Auckland. Jespersen, Otto 1982 [1938/1905]. Growth and Structure of the English Language. Chicago: The University of Chicago Press. Lerer, Seth 2007. Inventing English: A Portable History of the Language. New York: Columbia University Press. Marder, Louis 1962. His Exits and Entrances: The Story of Shakespeare’s Reputation. Philadelphia and New York: J.B. Lippincott Company. McArthur, Tom (ed.) 1992 [1902]. The Oxford Companion to the English Language. Oxford and New York: Oxford University Press. McCrum, Robert, Robert MacNeil and William Cran 2002 [1986]. The Story of English. New York: Penguin Books. McDonald, Russ 2001. Shakespeare and the Arts of Language. Oxford: Oxford University Press. McQuain, Jeffrey and Stanley Malless 1998. Coined by Shakespeare: Words and Meanings First Penned by the Bard. Springfield, MA: Merriam-Webster. Muir, Kenneth 1973 [1964]. ‘ “A Lover’s Complaint”: a reconsideration’ in: Shakespeare the Professional and Related Studies. London: Heinemann, 204–19. Müller, F. Max 1862. Lectures on the Science of Language. New York: Charles Scribner. Müller, F. Max 1891. The Science of Language, Founded on Lectures Delivered at the Royal Institution in 1861 and 1863. New York: Charles Scribner’s Sons.

54


Nevalainen, Terttu 1999. ‘Early Modern English Lexis and Semantics’, in: Roger Lass (ed.), The Cambridge History of the English Language, 1476 to 1776. Cambridge: Cambridge University Press, volume 3, 771. Ogburn, Charlton, Jr. 1984. The Mysterious William Shakespeare: The Myth and the Reality. New York: Dodd, Mead, & Company. Pendleton, Thomas A. 1989. ‘The Non-Shakespearean Language of “Shall I Die?” ’, Review of English Studies 40 (August): 323–51. Ravi, V. S. 2003. ‘No doubts about his genius’, The Hindu, 31 August 2003. Schäfer, Jürgen 1973. Shakespeares Stil: Germanisches und Romanisches Vocabular. Frankfurt am Main: Athenaeum Verlag. Schäfer, Jürgen 1980. Documentation in the O.E.D.: Shakespeare and Nashe as Test Cases. Oxford: Clarendon Press. Scheler, Manfred 1982. Shakespeares Englisch: Eine sprachwissenschaftliche Einführung. Berlin: Schmidt. Shipley, Joseph T. 1977. In Praise of English: The Growth and Use of Language. New York: Times Books. Spevack, Marvin1968. A Complete and Systematic Concordance to the Works of Shakespeare. Hildesheim: Georg Olms. Spevack, Marvin 1973. The Harvard Concordance to Shakespeare. Cambridge, MA: Belknap Press of Harvard University Press. Taylor, Gary 1989. Reinventing Shakespeare: A Cultural History from the Restoration to the Present. New York: Weidenfeld & Nicholson. Thisted, Ronald and Brad Efron 1987. ‘Did Shakespeare write a newly-discovered poem?’, Biometrika 74(3): 445–55. Valenza, Robert J. 1990. ‘Are the Thisted-Efron authorship tests valid?’, Computers and the Humanities 25: 27–46. Vickers, Brian 2002a. ‘Counterfeiting’ Shakespeare: Evidence, Authorship, and John Ford’s Funerall Elegye. Cambridge, UK, and New York: Cambridge University Press. Vickers, Brian 2002b. Shakespeare, Co-author: A Historical Study of Five Collaborative Plays. Oxford and New York: Oxford University Press. Vickers, Brian 2007. Shakespeare, A Lover’s Complaint, and John Davies of Hereford. Cambridge and New York: Cambridge University Press. Weekley, Ernest 1952 [1928]. The English Language. London: Andre Deutsch Limited. Wells, Stanley W., Gary Taylor, John Jowett, and William Montgomery 1987. William Shakespeare: a Textual Companion. Oxford and New York: Clarendon Press, Oxford University Press.


55

Appendix Table 2.2 Types-tokens, new-to-the-group words, total inferred vocabulary, Shakespeare and other writers

By all three measures, Shakespeare’s poems had richer vocabularies than his tragedies and romances, which, in turn, were richer than his early comedies. On average, Shakespeare’s writings were about as rich as those of most of his contemporaries tested, and richer than Fletcher’s, but not as rich as Milton’s in Paradise Lost or Spenser’s in The Faerie Queene, Book 1, or the last part of Moby Dick, or two modern non-fiction texts by Elliott and Valenza. Name

Includes 40,000 words each from

Tokens

Types

NW

1/2 tokens @

Total Virtual Inferred Vocab.

MILTON1 MILTON2

Paradise Lost, 1-6 Paradise Lost, 7-12

40,000 38,886

6,671 6,328

1,790 1,591

127 117

140,415 123,490

Whore of Babylon, Honest Whore SHPOEMS* Venus, Lucrece, Sonnets MARLOWE1* Tamburlaine, I & II, DF16 JONSON1* Sejanus, Volpone

40,000

5,742

1,093

92

85,567

40,000

6,054

1,147

91

84,180

40,000

5,444

936

83

73,449

40,000

5,452

576

81

70,867

GREENE*

Alphonus, Friar Bacon, James IV MIDDLE1* Phoenix, Michaelmas, Chaste Maid MIDDLE2* Witch, Hengist, Women Beware Women JONSON2* Alchemist, Bartholomew Fair SHTEMPWT* Tempest, Winter's Tale SHCORCYM* Coriolanus, Cymbeline SHOTHLR* Othello, Lear

40,000

5,042

667

79

68,324

40,000

4,761

883

77

65,822

40,000

5,006

833

76

64,586

40,000

5,857

1,076

74

62,144

40,000

5,841

964

73

60,938

40,000

5,341

879

72

59,742

40,000

5,221

865

71

58,556

MARLOWE2* Jew of Malta, Edward II, Massacre CHAPMAN* Gentleman Usher, Bussy D'Ambois FLETCHER2* Barnavelt, Island Princess SH3COMS Shrew, TGV, Comedy of Errors FLETCHER1* Woman's Prize, Valentinian

40,000

4,892

720

66

52,777

40,000

4,811

767

65

51,652

40,000

4,460

477

64

50,536

40,000

4,892

672

63

49,430

40,000

4,427

382

63

49,430

DEKKER*

Continued

56 Name

Stylistics and Shakespeare’s Language Includes 40,000 words each from

Tokens

Milton Average Shakespeare Average 6 Other Playwrights Average Fletcher Average * = in baseline composite lexicon for eight authors

Types

NW

1/2 tokens @

Total Virtual Inferred Vocab.

6,500

1,691

122

131,953

5,470

905

74

62,569

5,223

839

77

66,132

4,444

430

64

49,983

Sh NW

3,855

=31%

All NW

12,265

18 blocks by 9 authors, 15 blocks by 8 authors in baseline lexicon; 600,000 words, 4 blocks by Shakespeare, 160k words = 27% Total composite lexicon: 28,747 words, 12,265 of them NW (43.1%) All calculations done with Intellex Spenser Molière Elliott Elliott & Valenza Melville Iliad 3-40 Odyssey 2-40

Faerie Queene, Book 1, old spelling L'Avare, Don Juan, in French Politics of Population, 40k words Shakespeare Book 40k words Moby Dick Last 40k words Iliad Block 3 First 40k words Odyssey Block 2 First 40k words

40,000

6,235

122

131,827

39,253

5,003

69

56,215

40,000

7,067

147

177,275

40,000

5,810

102

99,984

40,000

7,056

99

95,553

40,000

3,985

89

81,437

40,000

4,023

58

44,053

David Hoover Type-Token figures, all done with TACT Paradise Lost 1

40,000

7,005

Paradise Lost 2

40,000

6,860

40,000

6,098

Moby Dick 1

40,000

6,554

Moby Dick 2

40,000

7,339

Moby Dick 3

40,000

6,855

Moby Dick 4

40,000

7,171

Moby Dick 5

40,000

6,937

Lewis, Main St. 1 Lewis, Main St. 2

40,000

7,285

40,000

7,119

Sh. Poems

Sonnets, Venus, Lucrece

Continued

Shakespeare’s Vocabulary: Did it Dwarf All Others? Name

Includes 40,000 words each from

Tokens

Types

Lewis, Main St. 3

40,000

6,996

Lewis, Main St. 4 Lewis, 367K

40,000

6,381

367,609

24,105

612,647

31,979

884,647

29,066

Iliad & Odyssey, Samuel Butler Translation

269,779

13,041

Entire King James Version Para Lost; Para Regained, Samson; young Paradise Lost

789,955

12,803

117,693

14,899

80,779

11,244

Babbitt, Main St., Our Mr. Wrenn Melville, 612 K Moby Dick, Pierre, Redburn, Benito C., Omoo

NW

1/2 tokens @

57 Total Virtual Inferred Vocab.

Other large corpora Shakespeare 884K entire canon per Spevack Homer

King James Bible Milton, all

Milton, Para Lost

Sh. Average, 6x40,000–word blocks

62,569

Thisted-Efron estimate of Shakespeare's Total Vocabulary, minimum

66,000

Chapter 3

A New Kind of Dictionary for Shakespeare’s Plays: an Immodest Proposal1 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

A New Kind of Dictionary for Shakespeare’s Plays

59

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. Moreover, my objective in elaborating these case studies is not simply the design of a dictionary, but to show how light can be thrown on some general issues in stylistics, such as characterization and theme, as well as style in Shakespeare.

2. Labels and contents of current general Shakespearean ‘dictionaries’ I refer to general Shakespearean ‘dictionaries’ in order to exclude ‘dictionaries’ focusing on specific registers, such as legal, military or informal language (see the Athlone Shakespeare Dictionary Series). However, even with this exclusion, identifying what might count as a general Shakespearean dictionary is far from easy. We find various labels for books with contents characteristic – at least to some degree – of dictionaries, notably, ‘dictionary’, ‘glossary’, ‘lexicon’ and ‘word-book’. To these one might wish to add ‘concordances’, in recognition of the fact that such works contain a complete word list and (statistical) information about those words – aspects that might characterize a dictionary. Moreover, what these works contain varies greatly. It is possible to identify three groups. One is strongly linguistic in content, typically containing information about the existence of a word-form, as well as its meaning (conveyed with a brief definition) and illustrative quotation(s) and part-of-speech (e.g. Crystal and Crystal 2002, Foster 1908, Onions 1986 [1911], Schmidt 1971 [1902]). Another group is strongly non-linguistic in content, typically containing play summaries (largely plot focused),

60


character descriptions, cultural information and biographical information (e.g. Boyce 1996, Wells 1998). Note that, although non-linguistic, both of the examples cited are entitled ‘Dictionary of Shakespeare’. The final group is strongly focused on (frequency of) occurrence information, typically containing an index of all words (plus textual location) and the frequency of word-forms (absolute and relative), (e.g. Spevack 1968–1980, Howard-Hill 1969–1972). There is a little slippage between these groups – for example, Schmidt (1971 [1902]) contains a complete index of words and Crystal and Crystal (2002) was constructed with frequency information in mind – but in the main they are separate. My proposal involves bringing together the three areas in a more comprehensive and systematic fashion. This will clearly involve a broad scope. Consequently, the label ‘Dictionary of Shakespeare’ may not be the best. A better alternative might be ‘Encyclopaedia of Shakespeare’s Language’.4

3. General Shakespearean ‘dictionaries’ and present-day English language dictionaries compared The majority of present-day dictionaries of English contain pronunciation information (typically, a broad phonetic transcription with an indication of syllable stress). Doing the same for Shakespeare would require a significant research programme, and there would be thorny issues, such as whose accent to represent. Consequently, this is not currently part of the dictionary proposal. Many present-day dictionaries contain spelling variants, and the OED, of course, excels in this respect. Shakespearean dictionaries do not note more than the occasional spelling variant. Perhaps spelling variants are assumed not to be part of the ‘real’ Shakespeare, given that they are produced by compositors and printers. Nevertheless, spellings are the prism through which we receive Shakespeare, and Shakespearean texts represent a source of information about spelling in the Early Modern period. Moreover, quantifying spelling variation would be relatively easy to do with the computational methodology supporting the proposed dictionary (see section 12). Other differences in content include the fact that corpus-based dictionaries of present-day English, notably, the Collins Cobuild Dictionary, include definitions that are more contextualized and information about multiword units, as I will illustrate in sections 4 and 9. Perhaps even more significant than differences in the kinds of information that might be included are differences in policies for including


61

or excluding words and for prioritizing meanings. Shakespearean ‘dictionaries’, notably, Foster (1908) and Onions (1986 [1911]), as well as the even more recent corpus-informed dictionaries such as Crystal and Crystal (2002), tend to include only those words considered difficult or ‘hard’ for readers. In contrast, corpus-based dictionaries typically include all the words in the corpus (though that may not, in fact, be the best thing to do for a Shakespearean dictionary; see section 5). Furthermore, present-day dictionaries, particularly corpus-based dictionaries, take a different approach to the way meanings are prioritized within particular entries. Dictionaries in the philological tradition exemplified by the OED (e.g. Foster 1908 and Onions 1986 [1911]) take etymology as a guiding principle. This is most obviously reflected in the way that (1) word definitions gravitate towards etymological meanings, and (2) the organization of the senses of polysemous words is based on etymological priority (i.e. the earliest sense is listed first). In contrast, corpus-based dictionaries capture meanings based on usage in context, and organize those meanings according to frequency (usually the most frequent is placed first).

4. Towards a contextualized definition: the case of ‘horrid’ The OED gives three senses for the word ‘horrid’: (1) ‘bristling, shaggy, rough’, (2) ‘causing horror or aversion; revolting to sight, hearing or contemplation; terrible, dreadful, frightful; abominable, detestable’ and (3) ‘colloq. in weakened sense. Offensive, disagreeable, detested; very bad or objectionable. Noted in NED as especially frequent as a feminine term of strong aversion’ (here, and in all quotations from dictionaries in this chapter, accompanying quotations are generally excluded for brevity). The first sense corresponds with that of the Latin term horridus from which the English word is derived, and, judging from the illustrative quotations, was still current in Shakespeare’s period. The second sense, and one that is contemporary with Shakespeare, is a metonymic development of the first, and the final sense is apparently a ‘weakened’ development of the second. The fact that the first quotation given to illustrate the second sense is from Shakespeare should alert us to a major problem in using the OED to interpret Shakespeare – the problem of circularity, given that Shakespeare plays such a large role in determining the entries in the OED for the period in question. The third sense developed after Shakespeare. Note that the OED does at least supply a modicum of stylistic information, noting that the third sense is colloquial, and very occasionally some social

62


information, here noting that the third sense is ‘especially frequent as a feminine term’.5 Turning to three Shakespearean dictionaries, we find the following definitions: Foster (1908): (1) Awful, hideous, horrible. (2) Terrific.6 (3) Horrified, affrighted. Onions (1986 [1911]): No entry. Crystal and Crystal (2002): Horrifying, frightful, terrifying. Foster’s (1908) first definition seems to shade into the third sense given in the OED. This is odd because the first citation date for that sense given in the OED is 1666. The single illustrative quotation given by Foster is from Macbeth: ‘If good, why do I yield to that suggestion / Whose horrid image doth unfix my hair’. This quotation includes a classic reaction to fear – the unfixing of hair. This does not necessarily support the sense given by Foster, which need not involve fear, just as in the third sense given in the OED. In fact, the usage here falls within the scope of the OED’s second sense, as indeed do Foster’s second and third definitions. Note that the strongly overlapping array of synonyms given in the definitions do little to pin down the sense of ‘horrid’ in Shakespeare. What is being described as horrid? Who is using this word? In what circumstances are they using it? Is Shakespeare using it in a way that his contemporaries would not? And so on. We can look at a computer concordance (a list of the occurrences of the word along with their local co-text) and the distribution of a word, in order to answer such questions. Here is the entire concordance of ‘horrid’ (the head noun to which it refers is underlined):7 Appear in forms more horrid, – yet my duty, As doth a Rock Up Sword; and know thou a more horrid hent8: When he is drunk asleep And cleave the general ear with horrid speech, Make mad the guilty heard and seen, Recounts most horrid sights seen by the watch shall break his wind With fear and horrid flight. I will meditate the while upon some horrid message for a Challenge. Not in the legions of horrid hell, can come a devil more damned Proper deformity seems not in the fiend So horrid as in woman. And what a beard of the general’s cut and a horrid suit of the camp Presented then unto the gazing moon So many horrid ghosts.


63

Crammed with distressful bread; Never sees horrid night, the child of hell all the sparks of nature, To quit this horrid act. Out treacherous Such sheets of fire, such bursts of horrid thunder, Such groans of couriers of the air, Shall blow the horrid deed in every eye, Of thy dear husband, than that horrid act Of the divorce he’ld make I yield to that suggestion, Whose horrid image doth unfix my hair It can be seen that ‘horrid’ is used to describe acts, sights and sounds, but not just any such things – most have a strong supernatural or unnatural connection. This seems to have been overlooked in all dictionary definitions, despite the fact that it is quite obvious in the concordance. We can deepen our understanding of the word by considering its distribution both within Shakespeare and without. Putting the results together, a dictionary entry might be as follows (All = all Shakespeare’s plays, T = tragedies, C = comedies, H = histories, M = male speakers, F = female speakers, Pla = other EModE plays, Fic = EModE prose fiction, Tr = EModE trial proceedings, Ha = EModE handbooks in dialogue form, Sc = EModE scholarly works; the figures in brackets are normalized per 100,000 words): HORRID. Something that is horrid causes fear; typically, it refers to supernatural or unnatural acts, sights and sounds. Distribution: All = 16 (1.8); T = 10 (3.9), C = 2 (0.6), H = 4 (1.5); M = 14 (1.9), F = 2 (1.4). Comparisons: Pla = 187 (0.17), Fic = 0, Tr = 0, Ha = 0, Sc = 1 (0.14). E.g. ‘Whose horrid Image doth vnfixe my Heire’, ‘I wil meditate the while vpon some horrid message for a Challenge.’9 The above is no more than an indication as to the direction a dictionary entry might take. Note that the first sentence offers a contextualized definition of the type used in the Collins Cobuild, rather than a handful of synonyms. However, going beyond the Collins Cobuild, the figures following offer a broader discoursal contextualization. They give some indication as to the social and stylistic meanings the word might have acquired on account of being to some degree ‘contextually bound’ (Leech 1981: 14–5; see also Enkvist 1964: 29–35). Focusing on the more meaningful normalized figures, one might note, for example, that the word ‘horrid’, which appears much more densely in tragedies than either histories or comedies, is used slightly more frequently by male characters compared with female; and that Shakespeare uses it considerably more than his contemporary playwrights did, and also that it is most characteristic of Early Modern plays and, surprisingly, scholarly literature.

64


However, this particular example is severely hampered by frequency limitations: the strongest finding revealed by the figures simply being that ‘horrid’ is rare. I will focus on frequency limitations in the next section. Here, I will briefly indicate four ways in which the above entry could be improved: z z

z z

The definition was derived from collocational information and some of this information could have been included in the entry (see section 6). Sociolinguistic information could be enriched by the inclusion of other sociological variables (e.g. status, age) and also comparative data (e.g. addressing questions such as: is X word associated with male or high status speakers in Shakespeare specifically or is this a more general feature of Early Modern English?). A statistical measure could be employed in order to indicate whether differences in distribution are significant. The presentation of information could be improved (e.g. the use of graphs, or a verbal description instead of figures).

5. Frequency limitations A corpus-based dictionary typically includes all words in the corpus. However, this presents two problems: (1) how to treat rare or infrequent words, and, from the more practical point of view of publication, (2) how to fit all the words into one volume.10 As is clear from the sample entry of ‘horrid’ above, low frequency words lead one to the mere conclusion that they are low frequency, as the more robust and informative distribution patterns fail to materialize.11 The corpus-based methodology is not best suited to investigating low frequency words (cf. Biber et al. 1998: 30; Meyer 2002: 15); instead, we need to look towards alternative methodologies, such as the philological approach that already underpins most current Shakespearean dictionaries. A partial solution to these problems is simply to adopt a frequency cut-off point such that words below a certain frequency are not considered for inclusion in the dictionary. But what would be the implications of such a cut-off point for the coverage of Shakespeare’s vocabulary? Onions (1986 [1911]) supposedly covers some 3,000 words, according to Crystal and Crystal (2002: Introduction), who also claim to include 21,263 entries under 13,626 headwords in their own volume. Table 3.1 displays the consequences of various cut-off points. For example, the first row considers words (word-types) which are repeated at least 100 times (they have at least 100 tokens). As can be seen from the following columns, 998 words

A New Kind of Dictionary for Shakespeare’s Plays Table 3.1

65

Word-types and tokens in Shakespeare’s plays How many word-types (different words) are accounted for

Word-types with more than 100 tokens Word-types with more than 50 tokens Word-types with more than 16 tokens Word-types with less than 10 tokens All instances of all word-types

How many word-tokens (instances of words) are accounted for

998

706,974

1,564

761,472

4,652

835,925

7,753

37,260

24,842

899,092

(word-types) are repeated at least 100 times (have more than 100 tokens), and, if these 998 word-types and all their repeats are added up, they account for 706,974 of the words that make up Shakespeare’s plays. As the bottom row shows, there is a total of 899,092 word-tokens in Shakespeare and 24,842 different word-types (in other words, a smaller number of different words are repeated a number of times to make up the total vocabulary of Shakespeare’s plays). ‘Horrid’ occurs 16 times. If we only consider for the proposed dictionary word-types that occur more than 16 times, then, potentially, we would only need to have 4,652 different entries in our dictionary, and yet we would still cover most of the language of Shakespeare’s plays: 835,925 word-tokens. However, I pointed out above that 16 occurrences are too few for our purposes. If a cut-off point of, say, 50 word-tokens for any word entry were imposed, resulting in a potential − and certainly more manageable − 1,564 word-form entries, then that still would account for the vast bulk of the words in Shakespeare (761,472 out of 899,092 word-forms). Note that 7,753 word-types occur less than 10 times, accounting for a mere 37,260 word-tokens. Yet it is precisely here that the current Shakespearean dictionaries tend to focus, as these rare items tend to be considered ‘hard’. In my view, there is no justification for excluding more frequently occurring vocabulary items. From a linguistic perspective, we know that all words change meaning: even the most frequent of items have incurred shifts of meaning which present-day readers must take on board. From a literary perspective, we should beware of letting more unusual vocabulary distract our attention from the more usual. As John F. Burrows (1987: 1) eloquently puts it: ‘It is a truth not generally acknowledged that, in most discussions of works of English fiction, we proceed as if a third, two-fifths, a half of our material were not really there.’

66


6. Polysemy and collocates: the case of ‘good’ Current Shakespearean dictionaries give definitions for the word ‘good’ (as an adjective) such as these (illustrative quotations are excluded): Foster (1908): (1) Not bad, worthy of praise; (2) Fit, adapted; (3) Trustworthy, genuine; (4) Kind, benevolent; (5) Proper, right; (6) Substantial, safe, solvent, able to fulfil engagements; (7) Real, serious; (8) Favourable, propitious; (9) Abundant, rich; (10) Skilful, clever; (11) Adequate. (Notes phrases and compounds). Onions (1986 [1911]): (1) Conventional epithet to titles of high rank; (2) Comely; (3) Financially sound, (hence) wealthy, substantial. (Notes quasiadverbial usage, e.g. ‘good easy man’, and phrases and compounds). Crystal and Crystal (2002): (1) [intensifying use] real, genuine (‘love no man in good earnest’); (2) kind, benevolent, generous; (3) kind, friendly, sympathetic; (4) amenable, tractable, manageable; (5) honest, virtuous, honourable; (6) seasonable, appropriate, proper; (7) just, right, commendable; (8) intended, right, proper; (9) high-ranking, highborn, distinguished; (10) rich, wealthy, substantial. (Notes phrases and compounds). Lists of synonyms – in some cases overlapping – do not always provide the reader with assistance in discriminating the various senses. For example, in Foster’s (1908) definitions, how does ‘genuine’ in sense 3 differ from ‘real’ in sense 7? Similarly, ‘fit’ in sense 2 can uncomfortably overlap with ‘proper, right’ of sense 5. Onions’s (1986 [1911]) definitions are fairly discrete, while in contrast Crystal and Crystal (2002) seem to have gone for a deliberate policy of overlap (note that ‘kind’, ‘proper’ and ‘right’ appear in more than one definition), perhaps indicating that indeed senses do overlap. We might also note that each dictionary orders the senses in a different way, and that some, rather worryingly, contain senses that others do not (note, for example, Onions’s first sense). A simple technique in corpus linguistics for investigating the meaning of a word is to examine a concordance and note the words with which the word in question co-occurs, something which we have already demonstrated with the word ‘horrid’. It is the collocates of a word – ‘the company it keeps’ (cf. J. R. Firth 1957) – that may help distinguish different senses (see, for example, Partington 1998: 33–46). Frequent collocating words to the right of ‘good’ include: ‘(my) good friend(s)/sir/Lord/master/man/Lady/Madam/etc.’, ‘good old man/friend/etc.’, ‘good morrow/night/even’, ‘(in) good faith’,


67

‘good will/wish(es)’, ‘good god(s)’, ‘good luck / hap’, ‘good news/report/ words’, ‘good now’, and ‘(in) good time’. Even without further elaboration, seeing such collocations helps make accessible distinct senses, and so they should be included within a dictionary entry. Also, the frequency of such collocations can feed into the ordering of senses within the entry. However, with a dizzying 2,711 instances constituting a concordance of ‘good’, the human can only identify some collocational patterns, and cannot accurately assess the strength of those patterns and thus come to a principled decision about which to include in the dictionary. One possible solution is to calculate the statistical likelihood with which particular words and ‘good’ co-occur to form a collocation. Using z-scores, a statistical measure, the top ten rankedordered collocates five words to the left and right of ‘good’ are: morrow, Lord, my, do, sir, good, your, have, be and you, as can be seen in Table 3.2.12 These collocational patterns point to sentences like the following (constructed) example: ‘Good morrow, my good lord, you have . . . ’ This evidence clearly underpins Onions’s (1986 [1911]) first sense, a sense that is not clearly represented in the other dictionaries, and underscores the role of ‘good’ as a politeness marker. Such an investigation could be extended in three ways: (1) collocational patterns (and ones not limited to single word collocates) can be identified with other statistical procedures (including the methodology in section 9), (2) collocational patterns in Shakespeare can be compared with collocational patterns in other Early Modern texts (e.g. is Shakespeare peculiar in his usage of ‘good’ as a politeness marker?) and (3) grammatical relations can be explored via collocations (e.g. as is transparent for the concordance of ‘horrid’ the items immediately to the right are nearly always nouns – something that confirms the status of ‘horrid’ as a typical adjective in English). Table 3.2 The top ten ranked-ordered collocates of ‘good’ within a five-word span Collocate (±5/–5) Morrow Lord My Do Sir Good Your Have Be You

Frequency 6 11 21 8 6 6 11 9 9 15

Z-score 18.0 4.8 2.8 2.5 2.1 2.1 2.0 1.5 1.1 0.7


68

7. The inclusion of pragmatic/discoursal words: the case of ‘ah’ Interjections, onomatopoeic sounds, hesitation phenomena, discourse markers and so on have received scant attention in Shakespearean dictionaries (of course, this is not true of specialist dictionaries, notably Blake 2004). For example, there is no entry for ‘ah’ in Foster (1908), Onions (1986 [1911]) or Crystal and Crystal (2002). The issue is whether such items are considered words, and that depends on your definition of a word. Corpus linguistics favours an orthographic definition, such as ‘a string of uninterrupted non-punctuation characters with white space or punctuation at each end’ (Leech et al. 2001: 13–4). In which case, ‘ah’ is clearly a word. Does ‘ah’ have meaning? That depends on your definition of meaning. If meaning is associated with ideational meaning, to use Halliday’s (e.g. 1978) terminology, and not textual or interpersonal meanings, then words like ‘ah’ do not have meaning. One of the reasons such words are not generally included in Shakespearean dictionaries is that words that reflect some aspect of the world are privileged above words that help organize other words or words that help organize people. In my view, this approach is entirely inappropriate for a dictionary of Shakespeare’s plays because those plays are made up of dialogue. What lies at the heart of dialogue are those pragmatic and discoursal words that structure and mediate the interaction between characters. Let us consider the pragmatic and discoursal meanings of ‘ah’, and also its social and stylistic meanings. If a concordance of ‘ah’ is scrutinized, one can discern the three key pragmatic meanings following (an illustrative example is provided of each): Speaker attitude/state communicated = sorrow, emotional distress

desdemona othello desdemona

To whom, my lord? with whom? how am I false? Ah Desdemon! away! away! away! Alas the heavy day! Why do you weep? Am I the motive of these tears, my lord? (Othello 4.2.40–3)

Speaker attitude/state communicated = pity

gloucester

Canst thou blame him? . . . His daughters seek his death; ah, that good Kent! He said it would be thus, poor banished man!


69

Thou say’st the king grows mad; I’ll tell thee, friend. I am almost mad my self: . . . (King Lear 3.4.161–5) Speaker attitude/state communicated = surprise, realization

[Enter Adriana and Luciana.] adriana Ah, Luciana, did he tempt thee so? (The Comedy of Errors 4.2.1) And one can discern the two key discoursal meanings following: Discourse marker: preface to the correction/rejection of the previous speaker’s proposition(s), emotions or actions

menas

These three world-sharers, these competitors, Are in thy vessel: let me cut the cable; And, when we are put off, fall to their throats: All there is thine. pompeius Ah, this thou shouldst have done, And not have spoke on’t! In me ‘tis villainy; In thee’t had been good service. . . . (Antony and Cleopatra 2.7.70–5) Discourse marker: reinforces elicitation

leonato All thy tediousness on me, ah? dogberry Yea, an ‘twere a thousand pound more than ‘tis; for I hear as good exclamation on your worship as of any man in the city; and though I be but a poor man, I am glad to hear it. (Much Ado about Nothing 3.5.21–5) Turning to stylistic and social meanings, consider the distribution of ‘ah’: Distribution:

All = 179 (19.9); T = 54 (21.3), C = 32 (8.9), H = 93 (35.4); M = 121 (16.1), F = 59 (41.9). Comparisons: Pla = 1,573 (14.4), Fic = 9 (10.9), Tr = 1 (2.9), Ha = 11 (11.2), Sc = 0. Within Shakespeare, ‘ah’ is characteristic of the histories, to some extent the tragedies, but to a much lesser extent the comedies. This distribution may reflect the frequent functions of ‘ah’ in signalling emotional distress and pity. Interestingly, the distribution across genders is far from even: it is more than twice as dense in female dialogue. Compared with other playwrights


70

of the period, Shakespeare can be said to be fairly fond of this item. Also, we can see that it is more characteristic of plays than other contemporary genres. Moreover, there is evidence that ‘ah’ is a strong colloquial marker. It does not appear at all in scholarly works, the genre that is far removed from colloquial genres; it hardly appears in trial proceedings, a genre that – influenced by the formal setting, legal routines and need to create an official document – tends to be remote from colloquial language; while on the other hand, it appears in fictional prose (the choice of prose for this data set being specifically geared towards more colloquial prose) and handbooks in dialogue form. Interestingly, and remarkably, the density of ‘ah’ in a sample of five present-day plays is 94.27 (contrasting with Shakespeare’s 19.9), something which presumably reflects the drift of genres, including plays, towards more colloquial language (see, for example, Biber and Finegan 1992).

8. The inclusion of grammatical words: the case of ‘and’ The most frequent words in any body of texts are closed-class. Yet Shakespearean dictionaries do not, or do not adequately, treat such grammatical items, despite – or may be because of – their high frequency of occurrence. For example, the entries for the second most frequent word in Shakespeare, the word ‘and’, in general Shakespearean dictionaries are as follows: Foster (1908): Cross-references Abbot’s Shakespearean grammar. Onions (1986 [1911]): (1) Coordinating conjunction (nouns, adjectives and phrases); (2) Subordinating conjunction: if, even if, though, as if, whether. Crystal and Crystal (2002): [also spelling variant ‘an’] (1) if, even if; (2) as if; (3) if, whether. As can be seen, it is not treated at all in Foster (1908), while Crystal and Crystal (2002) only mention conditional ‘and’ (used as a subordinate conjunction introducing a conditional clause with the sense ‘if’). Conditional ‘and’ is likely to be the focus of attention in Shakespearean dictionaries, because of editorial policies to select items with which the modern reader is assumed to be unfamiliar and thus likely to experience difficulty. Examples of conditional ‘and’ include the following: bawd

What would you have me be, an I be not a woman? (Pericles 4.2.81)

A New Kind of Dictionary for Shakespeare’s Plays romeo

71

Noting this penury, to myself I said ‘An if a man did need a poison now, Whose sale is present death in Mantua, Here lives a caitiff wretch would sell it him’. (Romeo and Juliet 5.1.49–52)

Only Onions (1986 [1911]) acknowledges the fact that words such as ‘and’ play an important grammatical role. It is the coordinating function of ‘and’ that accounts for the overwhelming majority of instances in Shakespeare. ‘And’ makes a significant contribution to textual meaning in Shakespeare in the way it conjoins nouns, adjectives, (nominal or adjectival) phrases and clauses, and it is also used as a pragmatic connective. Compare the following two extracts in which instances coordinating clauses are underlined and instances coordinating words/phrases are emboldened: vincentio

She should this Angelo have married; was affianced to her by oath, and the nuptial appointed: between which time of the contract and limit of the solemnity, her brother Frederick was wracked at sea, having in that perished vessel the dowry of his sister. But mark how heavily this befell to the poor gentlewoman: there she lost a noble and renowned brother, in his love toward her ever most kind and natural; with him, the portion and sinew of her fortune, her marriage-dowry; with both, her combinatehusband, this well-seeming Angelo. (Measure for Measure 3.1.213–23)

second gentleman first gentleman second gentleman first gentleman second gentleman first gentleman

Who’s that that bears the sceptre? Marquess Dorset: And that the Earl of Surrey, with the rod. A bold brave gentleman. That should be The Duke of Suffolk? ‘Tis the same: high steward. And that my Lord of Norfolk? Yes. (King Henry VIII 4.1.37–44)

The density of lexical/phrasal coordination in the first extract contrasts with the density of clausal coordination in the second. This grammatical

72


difference reflects differences in style and communicative purpose. Lexical/ phrasal coordination in Vincentio’s speech helps create a high rhetorical style, underscoring the seriousness of what he is saying. More specifically, the conjoins of coordinated pairs tend to be closely related in meaning. Thus, ‘noble’ and ‘renowned’ overlap in meaning (reflecting the rhetorical figure of ‘pleonasm’), and ‘kind’ and ‘natural’ could be viewed as being in a hierarchical relationship such that one is subordinate to the other (reflecting the rhetorical figure of ‘hendiadys’, that is amounting to ‘naturally kind’). In contrast, the clausal coordination of the second extract creates a low rhetorical style, underscoring the casual conversation, a style which is, of course, reinforced by the ellipsis. In fact, in this particular case, ‘and’ is not merely coordinating clauses but also acting as a pragmatic connective. Specifically, the second instance is used to create a series of questions, or, as Schiffrin puts it, to ‘link questions in a question agenda’ (1994 [1987]: 146). As a consequence of their rather different functions, lexical/ phrasal coordination tends to correlate with rather different genres compared with clausal coordination. I cannot prove this claim with regard to Shakespeare, as the computational analysis of Shakespeare’s grammar is not yet sufficiently accurate or sophisticated; indeed, one of the aims of my dictionary project is to solve this (however, see Culpeper and Kytö 2002, which provides supporting evidence for this correlation with regard to four Early Modern genres). In sum, my argument is that such grammatical items should be included in a dictionary of Shakespeare, and that dictionary should focus widely on the contribution of those items to meaning.

9. Multiword units John Sinclair (e.g. 1991), among other linguists, has argued that words may belong to semi-fixed phrases that constitute single lexical choices (e.g. ‘of course’, where the individual words cannot be assumed to produce the sense of the phrase). Current Shakespearean dictionaries pay scant attention to these. An empirical way of retrieving lexical items that tend to bunch together is to run an n-gram analysis. Essentially, the computer works through the text, recording the co-occurrence of every word with its neighbours, and then calculates which groups of words most frequently cooccur. Multiword units, thus defined, may be considered a kind of extended collocational unit and are frequently referred to as lexical bundles or clusters. The results for Shakespeare, retrieved by WordSmith Tools (Scott 1999), are included in Table 3.3 along with the results for three other data sets for


73

Table 3.3 The top ranked-ordered three-word lexical bundles in Shakespeare and other texts Shakespeare

EModE plays

EModE trials

Present-day plays

I pray you I will not I know not I am a I am not my good lord there is no I would not it is a and I will

it is a what do you and I will it is not I have a I will not in the world I tell you I know not I warrant you

do you know I did not did you see I do not he told me at that time out of the I told him he did not there was a

I don’t know what do you I don’t want do you think do you want I don’t think to do with do you know going to be don’t want to

comparison (the underlining, italics and emboldening show that a particular lexical bundle is used in another data set; no lexical bundle is used in more than two data sets). It has been noted in the literature that lexical bundles are good discriminators of different styles (e.g. Stubbs and Barth 2003). The bulk of the items in Table 3.3 are unique to the specific data sets. Lexical bundles in Early Modern trials reflect the fact that discourse is made up of question-answer routines (e.g. ‘do you know’, ‘did you see’ vs. ‘I did not’, ‘I do not’) and crimenarrative report (e.g. ‘he told me’, ‘at that time’, ‘out of the’, ‘I told him’). Lexical bundles in present-day plays seem to gravitate towards questions and assertions to do with knowing, wanting and thinking – perhaps reflecting the essence of present-day drama in which plot and character development is conveyed through highly interactive character-to-character dialogue (in other words, what is said between characters is partly designed to inform the audience of character and plot). A characteristic of both Shakespeare and other Early Modern plays is that many of the bundles begin with the first person pronoun ‘I’, perhaps reflecting the essence of Early Modern drama with its more direct presentation of characters and plot to the audience (the epitome of this being the use of soliloquies and asides). Shakespeare’s lexical bundles are distinguished by the fact that his top five most frequent bundles begin with the first person pronoun. Also, it is interesting to note that the most frequent three-word unit in Shakespeare’s plays, ‘I pray you’, is something that is not characteristic of other Early Modern plays, other genres or, of course, of present-day plays. The kind of distributional stylistic information I have been discussing here could, of course, be recorded along with the entries for the most frequent lexical bundles in Shakespeare in the dictionary. Perhaps even more

74


importantly, such n-gram analysis can feed into the grammatical description contained in the dictionary. I will attend to this issue in the following section.

10. A note on grammatical description Linguists like Sinclair (e.g. 1991, 2004) emphasize that grammar is in the lexicon and not in some a priori set of abstract categories (e.g. parts of speech) imposed on the language. A way into describing the lexico-grammar of Shakespeare would be to describe the grammatical frames or patterns, revealed through collocational analyses, as discussed in section 6, and multiword analyses, as discussed in section 9 (see Hunston and Francis 2000, for this approach). I have already hinted that collocational analyses could be deployed in the exploration of grammatical relations, noting the case of ‘horrid’ (and ‘good’ is similar).13 In fact, my discussion of ‘and’ was very much geared towards the grammatical relations of co-occurring units. Regarding multiword units, ‘I pray you’, for example, is a grammatical pattern partly consisting of a first person pronoun (i.e. either ‘we’ or ‘I’), a verb in the present tense and a second person pronoun (i.e. either ‘you’ or ‘thee’). While the items that can occur as pronouns are relatively restricted, a much wider range of verbs can occur in the middle of this particular pattern. However, not any verb is used: the set is also restricted. One subset of those verbs is comprised of speech act verbs such as ‘advise’, ‘arrest’, ‘assure’, ‘beseech’, ‘charge’, ‘tell’, ‘thank’ and ‘warrant’. Such verbs occur when the grammatical pattern is used in isolation or parenthetically to a matrix clause. Making the step from an n-gram analysis to the description of grammatical patterns or frames is not necessarily straightforward. N-gram analysis results in units which are not necessarily complete idioms or grammatical structures. Nevertheless, such analysis offers a way into it identifying grammatical frames, and the results can be complemented by collocational analyses. I would not argue for quite as radical an approach to grammar (i.e. ditch all abstract grammatical categories) as Sinclair, for four reasons. First, my analysis of ‘and’ already demonstrated that grammatical categories can be useful. Knowing the grammatical status of the conjoins (i.e. lexical/phrasal versus clausal) helps us account for textual meanings. Second, grammatical categories can provide a useful way of tracking variation and change in the language; specifically, in the case of the dictionary, it can help provide a way of understanding how language varied in Shakespeare’s time (e.g. from


75

register to register, from person-to-person) and how language has changed since Shakespeare. For example, the proposed dictionary could quantify parts of speech, particularly in cases where an item can function as more than one part of the speech, and thereby reveal differences in distribution (e.g. the distribution of verbal versus nominal usages of the lexeme ‘love’ used to be weighted in favour of nominal but is now approximately even). Third, supplying such information about words would enable researchers to compare and contrast with extant research. Fourth, supplying such information can simply be one additional means by which a dictionary can help users understand words.

11. Character and play profiles Some Shakespearean dictionaries contain non-linguistic descriptions of characters and plot summaries. I propose providing a description of the idiolect of each major character. This can be done by conducting a statistical comparison between the vocabulary of one character and that of the other characters in the same play, in order to reveal words that are statistically characteristic of particular characters. Those words are ‘keywords’. As an illustration, consider some of the results relating to characters in Romeo and Juliet (see Culpeper 2002, for a more detailed discussion). Table 3.4 contains the keywords of Romeo and Juliet (rankordered in terms of the statistical ‘keyness’) produced by the program WordSmith Tools. This reveals, for example, the predictable result that Romeo’s two most unusually frequent words (or ‘keywords’) are ‘beauty’ and ‘love’, as well as the less predictable – and thus possibly more interesting – result that Juliet’s two most unusually frequent words are ‘if’ and ‘be’. Although the results for Juliet are less predictable, they can readily be explained by a qualitative analysis of the text (i.e. they are motivated). Furthermore, and Table 3.4 Rank-ordered keywords for Romeo and Juliet (raw frequencies in brackets) romeo

juliet

Beauty (10), Love (46), Blessed (5), Eyes (14), More (26), Mine (14), Dear (13), Rich (7), Me (73), Yonder (5), Farewell (11), Sick (6), Lips (9), Stars (5), Fair (15), Hand (11), Thine (7), Banished (9), Goose (5), That (84)

If (31), Be (59), Or (25), I (138), Sweet (16), My (92), News (9), Thou (71), Night (27), Would (20), Yet (18), That (82), Nurse (20), Name (11), Words (5), Tybalt’s (6), Send (7), Husband (7), Swear (5), Where (16), Again (10)

76


following the line of argument articulated above, although many of Juliet’s keywords are grammatical in nature, they are no less meaningful. Upon closer inspection of Juliet’s keywords, one can see that keywords such as ‘if’, ‘be’ (often subjunctive), ‘or’, ‘would’ and ‘yet’ reflect Juliet’s anxieties and worries about Romeo’s intentions and welfare, as the following examples illustrate: . . . if he be married, / My grave is like to be my wedding-bed (1.5.132–3) If they do see thee, they will murder thee. (2.2.70) . . . But if thou mean’st not well, (2.2.150) Is thy news good, or bad? answer to that; Say either, and I’ll stay the circumstance: Let me be satisfied, is’t good or bad? (2.5.35–7) ‘Tis almost morning; I would have thee gone: And yet no further than a wanton’s bird, (2.2.176–7) The key point about such analysis is that, although a reading of the play would obviously have resulted in an understanding of Juliet’s anxieties and worries, such a reading would not necessarily have led to the identification of the linguistic source of that very understanding. Indeed, no ‘manual’ critical analysis to date, literary or linguistic, has fully accounted for the source. Regarding plays, plot summaries tend to include information about the plays’ ‘themes’. Such information relies on the intuitions of the editor. I propose something more empirical: providing a description of the semantic categories (or lexical fields) characterizing each play. This can be done by getting the computer automatically to assign each word in the plays to a semantic category (this assignment can, of course, be recorded in the dictionary entry for each word). The dominance of categories within plays can be statistically compared. For example, in an earlier study I conducted with Dawn Archer and Paul Rayson (Archer et al. 2009), we compared three ‘love tragedies’ (Othello, Antony and Cleopatra and Romeo and Juliet) with three ‘love comedies’ (A Midsummer Night’s Dream, The Two Gentlemen of Verona and As You Like It). Each word was assigned to the categories in Table 3.5 using the USAS suite of programs (for further details, see section 12). Then a statistical comparison was conducted in order to establish which semantic categories were characteristic of each data set (each semantic category has several subcategories). Our findings are displayed in Table 3.6.

A New Kind of Dictionary for Shakespeare’s Plays Table 3.5

77

The semantic categories used (derived from McArthur 1981)

A general and abstract terms F food and farming K entertainment, sports and games O substances, materials, objects and equipment T time

B the body and the individual G government and public L life and living things P education

Q world and environment

C arts and crafts

E emotion

H architecture, housing and the home M movement, location, travel and transport Q language and communication

I money and commerce in industry N numbers and measurement S social actions, states and processes

X psychological actions, states and processes

Y science and technology

Z names and grammars

Table 3.6 Love comedies and tragedies: characteristic semantic categories (rank-ordered) Most overused categories in comedies relative to tragedies

Most overused categories in tragedies relative to comedies

S3.2 = intimate/sexual relationship L2 = living creatures L3 = plants S1.2.6- = (not) sensible X3.1 = sensory: taste E2+ = liking T3- = old, new, young: age

G3 = warfare, defence, and the army L1- = (lack of) life/living things Z2 = geographical names E3- = (not) calm/violent/angry M4 = movement (by sea/through water) S9 = religion and the supernatural S7.1- = (lack of) power/organizing

It is love comedies that are characterized by the most obviously love-related category, ‘intimate/sexual relationship’. The love tragedies, by contrast, are characterized by categories far removed from love: ‘warfare, etc.’, ‘lack of life, etc.’, and so on. Closer inspection of the results in the context of the plays reveals many points of interest. For reasons of space, I will just comment on a few. The appearance of ‘plants’ as highly characteristic of comedies may seem puzzling. In fact, there is a connection with love, as the following extract illustrates (Silvius explains why he loves Phoebe despite


78

the fact that she is a prostitute) (words assigned to the ‘plants’ semantic category are emboldened): silvius

So holy and so perfect is my love, And I in such a poverty of grace, That I shall think it a most plenteous crop To glean the broken ears after the man That the main harvest reaps: loose now and then A scattered smile, and that I’ll live upon. (As You Like It 3.5.98–103)

More precisely, the connection is a metaphorical one. As Oncins-Martínez (2006) has pointed out, the underlying cognitive metaphor here is sex is agriculture and its sub-mappings include a woman’s body is agricultural land. Similarly, metaphor accounts for the presence of the semantic category ‘sensory: taste’, as illustrated in the following example: julia

Nay, would I were so angered with the same! O hateful hands, to tear such loving words! Injurious wasps, to feed on such sweet honey And kill the bees that yield it with your stings! (The Two Gentlemen of Verona 1.2.104–7)

The underlying cognitive metaphor here is love is food (see Barcelona 1995: 672–3; see also Oncins-Martínez 2006).

12. Conclusions The main features of my proposed dictionary can be summarized as follows: z z z z z z

All ‘words’ will be treated equally (e.g. not just ‘hard’ words or ‘content’ words). Meanings will not be restricted to semantic or ideational meaning. Meanings will be based on usage in context (e.g. not etymology). Context will include linguistic co-text (e.g. collocations) and nonlinguistic context (e.g. social properties of the speaker). Linguistic description will be relative, that is, it will compare Shakespeare’s usage with that of contemporary texts. The dictionary will include linguistic profiles of characters and plays.


79

Perhaps the most important question to raise at this stage is: to what extent is this agenda feasible? In fact, the reason why I am proposing this kind of dictionary now is that until recently it would have been impossible. With developments in both corpora and computational techniques, we are now at a point when it can be realized. Here, I will briefly list some methodological problems and indicate the extent to which they have been solved: z

z

z

z

z

z

There used to be a lack of comparative textual data in electronic form. However, this has been partially solved by, for example, the Helsinki Corpus of English Texts, the Corpus of English Dialogues 1560–1760, and so on. Early Modern spelling variation has been perhaps the major stumbling block for historical corpus linguistics, and hitherto the major stumbling block for the proposed dictionary, for the reason that one cannot search for a particular word-spelling and assume that all the relevant words will be retrieved. However, this problem has largely been solved by the Variant Detector (VARD), primarily devised by Dawn Archer, University of Central Lancashire and Paul Rayson, Lancaster University (see, for example, Archer and Rayson 2004, Archer et al. 2003). Studying abstract grammatical patterns in a corpus requires grammatical annotation. The Lancaster-developed CLAWS part-of-Speech annotation system works fairly well for present-day English (for descriptions of how CLAWS works, see Leech et al. 1994 or Garside 1987). It has been recently adapted at Lancaster for Early Modern English. However, it is not sufficiently accurate for the dictionary and manual correction is required (once this is done, of course, a powerful resource will be created). Semantic annotation has received attention from generations of researchers at Lancaster University, including Geoffrey Leech, Jenny Thomas, Roger Garside, Andrew Wilson, Paul Rayson and Dawn Archer. The USAS semantic annotation system has been adapted for Early Modern English, and demonstrated to have value (see, for example, Archer et al. 2003). However, it is not sufficiently accurate for the dictionary and would require a further round of development. There is also the thorny problem of what ‘world view’ the system should adopt. Social annotation, information about, for example, gender, status, age, has not yet been comprehensively and systematically applied to Shakespeare, but the methodology has been developed and applied to Early Modern English texts (see Archer and Culpeper 2003), and so it would be fairly straightforward to extend this to Shakespeare. A final problematic area to note, and one that is philological and not computational, is that the paper-based dictionary will need to be pri-


80

marily based on one particular edition of Shakespeare, and this will involve an evaluation of available editions to arrive at a final choice. Having said that, the dictionary could incorporate important information arising from a comparison of different editions. Moreover, the project would deliver not just a paper-based dictionary, but also a website, which would allow researchers to pursue their own searches in whatever editions that could be made available. In fact, such searches are already possible via the Shakespeare Database Project (see: www. shkspr.uni-muenster.de/index.php). However, this database only includes Shakespearean texts (it does not have the wider comparative capability of the dictionary project described here) and is not publicly available.

Notes 1

2 3

4 5

6

7

8 9

This chapter is a revised version of: Culpeper, J. (2007) ‘A New Kind of Dictionary for Shakespeare’s Plays: an Immodest Proposal’, SEDERI (Yearbook of the Spanish and Portuguese Society for English Renaissance Studies), 17: 47–73. Onions was in fact one of the editorial team of the OED. This kind of approach is akin to the corpus-based grammar produced by Biber et al. (1999). I am grateful to Antony Warner for this suggestion. Although the evidence is thin, explorations in the present-day British National Corpus suggest that women do tend to use the term ‘horrid’ more than men. In the period Foster was writing, this could have the earlier sense of ‘causing terror’. Of course, a concordance of a word will vary in terms of how many instances it contains according to the edition of Shakespeare used (and occasionally according to how good the search software is). The particular Shakespeare edition used in this chapter is outlined in note 9. ‘Hent’ means ‘clasp’. The Shakespeare frequencies given in this chapter are based on The Nameless Shakespeare (2003), a joint project of the Perseus Project at Tufts University, The Northwestern University Library, and Northwestern University Academic Technologies. It is derived from The Globe Shakespeare, the one-volume version of the Cambridge Shakespeare, edited by William G. Clark, John Glover, and William A. Wright (1891–1893). There is no claim here that this constitutes the ideal edition of Shakespeare. It is searchable via ‘WordHoard’ (the concordance in section 4 was derived by this). The comparative ‘Pla’ corpus is the ‘Korpus of Early Modern Playtexts in English’ (KEMPE), initially compiled by Lene B. Petersen and Marcus X. Dahl, University of Bristol, 2001–2003. It is searchable via ‘Corpuseye’. Note: a particular problem with the Corpuseye search engine is that it only searches the whole corpus and that corpus includes Shakespeare. Nevertheless, given the great


10

11

12

13

81

size of the corpus – 10.7 million words – the results will still mean something. The samples for Early Modern prose fiction, trial proceedings and handbooks are sourced from the Corpus of English Dialogues 1560–1760, and the scholarly works comprise half history writing and half science writing, sourced from the Helsinki Corpus of English Texts. Schmidt’s (1971 [1902]) complete treatment of Shakespeare’s lexicon stretches over two volumes of small print and thin paper, yet only contains the briefest of definitions. There are also difficulties in applying statistical significance tests to differences in distribution that involve low frequencies. It is a matter of debate as to which statistical measure to use. Mutual information scores are frequently used, some use t-scores and some argue for the Fisher exact test. These results were produced using the software Xiara. As I have already indicated in this chapter, a highly accurate part-of-speech tagged corpus of Shakespeare does not exist. Also, there are issues to do with the compatibility of tags and software, as well as devising software to assess adequately grammatical relations. One possibility to be explored is SketchEngine (see Kilgarriff et al. 2004), used for lexicography by Oxford University Press, for example.

References Shakespeare frequencies and examples given in this paper are based on The Nameless Shakespeare (2003); see note 9. Archer, Dawn and Jonathan Culpeper 2003. ‘Sociopragmatic annotation: new directions and possibilities in historical corpus linguistics’, in: Andrew Wilson, Paul Rayson and Anthony M. McEnery (eds), Corpus Linguistics by the Lune: A Festschrift for Geoffrey Leech. Frankfurt/Main: Peter Lang, 37–58. Archer, Dawn, Tony McEnery, Paul Rayson and Andrew Hardie 2003. ‘Developing an automated semantic analysis system for Early Modern English’, in: Dawn Archer, Paul Rayson, Andrew Wilson and Tony McEnery (eds), Proceedings of the Corpus Linguistics 2003 conference. UCREL technical paper number 16. UCREL, Lancaster University, 22–31. www.comp.lancs.ac.uk/computing/users/paul/public.html. Archer, Dawn and Paul Rayson 2004. ‘Using an historical semantic tagger as a diagnostic tool for variation in spelling’. Presented at Thirteenth International Conference on English Historical Linguistics (ICEHL 13) University of Vienna, Austria 23–29 August, 2004. (Also available at: www.comp.lancs.ac.uk/computing/users/paul/ public.html). Archer, Dawn and Jonathan Culpeper 2009. ‘Love – “a familiar or a devil”? An exploration of key domains in Shakespeare’s comedies and tragedies’, in: Dawn Archer (ed.), What’s in a Word-List? Investigating Word Frequency and Keyword Extraction. London: Ashgate, 136–57. Barcelona Sánchez, Antonio 1995. ‘Metaphorical models of romantic love in Romeo and Juliet’, Journal of Pragmatics 24: 667–88.

82


Biber, Douglas and Edward Finegan 1992. ‘The linguistic evolution of five written and speech-based English genres from the 17th to the 20th centuries’, in: Matti Rissanen, Ossi Ihalainen, Terttu Nevalainen and Irma Taavitsainen (eds), History of Englishes: New Methods and Interpretations in Historical Linguistics. Berlin: Mouton de Gruyter, 688–704. Biber, Douglas, Susan Conrad and Randi Reppen 1998. Corpus Linguistics: Investigating Language Structure and Use. Cambridge: Cambridge University Press. Biber, Douglas, Stig Johansson, Geoffrey N. Leech, Susan Conrad and Edward Finegan 1999. Longman Grammar of Spoken and Written English. Harlow: Pearson Education Limited. Blake, Norman Francis 2004. Shakespeare’s Non-standard English: A Dictionary of his Informal Language. Athlone Shakespeare Dictionary series. London and New York: Continuum. Boyce, Charles 1996. Dictionary of Shakespeare. Ware, Hertfordshire: Wordsworth Editions Ltd. Burrows, John F. 1987. Computation into Criticism: A Study of Jane Austen’s Novels and Experiment in Method. Oxford: Clarendon Press. Crystal, David and Ben Crystal 2002. Shakespeare’s Words: A Glossary and Language Companion. London: Penguin. Culpeper, Jonathan 2002. ‘Computers, language and characterisation: an analysis of six characters in Romeo and Juliet’, in: Ulla Melander-Marttala, Carin Ostman and Merja Kytö (eds), Conversation in Life and in Literature: Papers from the ASLA Symposium, Association Suédoise de Linguistique Appliquée (ASLA), 15. Uppsala: Universitetstryckeriet, 11–30. Culpeper, Jonathan and Merja Kytö 2002. ‘Lexical bundles in Early Modern English: a window into the speech-related language of the past’, in: Teresa Fanego, Belén Méndez-Naya and Elena Seoane (eds), Sounds, Words, Texts, Change. Selected Papers from the Eleventh International Conference on English Historical Linguistics (11 ICEHL), Current Issues in Linguistic Theory. Amsterdam: John Benjamins, 45–63. Enkvist, Nils Erik 1964. ‘On defining style’, in: Nils Erik Enkvist, John Spencer and Michael Gregory (eds), Linguistics and Style. Oxford: Oxford University Press, 1–56. Firth, John R. 1957. ‘A synopsis of linguistic theory 1930–1955’, in: Studies in Linguistic Analysis, Philological Society, Oxford. Reprinted in F. Palmer (ed.) 1968. Selected Papers of J. R. Firth. Harlow: Longman, 168–205. Foster, John 1908. A Shakespeare Word-Book: Being a Glossary of Archaic Forms and Varied Usages of Words Employed by Shakespeare. London: Routledge. Garside, Roger 1987. ‘The CLAWS Word-tagging System’, in: Roger Garside, Geoffrey N. Leech and Geoffrey Sampson (eds), The Computational Analysis of English: A Corpus-based Approach. London: Longman, 30–41. Halliday, Michael A. K. 1978. Language as Social Semiotic: The Social Interpretation of Language and Meaning. London: Edward Arnold. Howard-Hill, Trevor H. 1969–1972. Oxford Shakespeare Concordances. Oxford: Oxford University Press. Hunston, Susan and Gill Francis 2000. Pattern Grammar: A Corpus-Driven Approach to the Lexical Grammar of English. Amsterdam: John Benjamins.


83

Kilgarriff, Adam, Pavel Rychly, Pavel Smrz and David Tugwell 2004. The Sketch Engine. Proc EURALEX, Lorient: France: www.sketchengine.co.uk. Leech, Geoffrey N. 1981. Semantics. Middlesex: Penguin Books. Leech, Geoffrey N., Roger Garside and Michael Bryant 1994. ‘CLAWS4: The tagging of the British National Corpus’. Proceedings of the 15th International Conference on Computational Linguistics (COLING 94) Kyoto, Japan, 622–8. www.comp.lancs. ac.uk/computing/research/ucrel/papers/coling.html. Leech, Geoffrey N., Paul Rayson and Andrew Wilson 2001. Word Frequencies in Written and Spoken English Based on the British National Corpus. Harlow: Pearson Education. McArthur, Tom 1981. Longman Lexicon of Contemporary English. Longman: London. Meyer, Charles F. 2002. English Corpus Linguistics: An Introduction. Cambridge: Cambridge University Press. Oncins-Martínez, José Luis 2006. ‘Notes on the metaphorical basis of sexual language in Early Modern English’, in: Juan G. Vázquez González, Montserrat Martínez Vázquez and Pilar Ron Vaz (eds), The Historical Linguistics-Cognitive Linguistics Interface. Huelva: Universidad de Huelva, 205–24. Onions, Charles T. 1986 [1911]. A Shakespeare Glossary (3rd edition enlarged and revised by Robert D. Eagleson). Oxford: Clarendon Press. Partington, Alan 1998. Patterns and Meanings. Amsterdam: John Benjamins. Schiffrin, Deborah 1994 [1987]. Discourse Markers. Cambridge: Cambridge University Press. Schmidt, Alexander 1971 [1902]. Shakespeare Lexicon and Quotation Dictionary (Vols 1 and 2). Mineola, N.Y.: Dover Publications. Scott, Mike R. 1999. WordSmith Tools. Oxford University Press. www.liv. ac.uk/~ms2928. Sinclair, John 1987. Collins Cobuild English Language Dictionary. London: Harper Collins. Sinclair, John 1991. Corpus, Concordance, Collocation. Oxford: Oxford University Press. Sinclair, John 2004. Trust the Text: Language, Corpus and Discourse (edited with R. Carter). London and New York: Routledge. Spevack, Marvin 1968–1980. A Complete and Systematic Concordance to the Works of Shakespeare. Hildesheim: Georg Olms. Stubbs, Michael and Isabel Barth 2003. ‘Using recurrent phrases as text-type discriminators: a quantitative method and some findings’, Functions of Language 10(1): 65–108. Wells, Stanley 1998. A Dictionary of Shakespeare. Oxford: Oxford Paperbacks.

Chapter 4

‘If I break time’: Shakespearean Line Endings on the Page and the Stage Peter Kanelos

1. The ends of Shakespeare’s lines For scholars surveying the development of Shakespeare’s poetic style, the question of his progressively more enjambed and knotty verse is a vital one. As Frank Kermode (2000) and Russ McDonald (2001) have argued, Shakespeare’s increasingly complex and irregular poetics developed concurrently with his increasingly complex and irregular representations of the world. Yet this evolution must have also had dramaturgical purpose, serving the need for Shakespeare’s characters to speak in a particular manner. By attending to Shakespeare’s dramatic texts as the working documents of players, scholars of performance studies have helped us to gain a more nuanced understanding of Shakespeare as a dramatic poet (see Palfrey 2005, Palfrey and Stern 2005, Stern 2004, Tucker 2002, Weimann 2000). In this spirit, I will apply to Shakespeare’s dramatic verse questions of critical import to our understanding of Shakespearean performance. Did Shakespeare expect players speaking his verse to follow cues laid out by punctuation? Did he intend players to be guided by lineation, indicating the independent status of the line by speaking that line through, perhaps in a single breath, and marking its end prosodically (via intonation, stress, a pause, a breath, etc.)? Or did he anticipate that actors, when encountering an enjambed line, would not mark it in the phrasing? In other words, when lines are enjambed by Shakespeare, did Elizabethan and Jacobean players signify through some gesture of the voice that break as a break, or did they follow organically the shape of the sentence? We are aware how finely Shakespeare tuned the rhythms of his characters’ speech through textually-signalled distinctions – short lines, feminine endings, overlapping cues and so forth. Were line endings also understood as a sort of punctuation?

Line Endings on the Page and the Stage

85

To address these questions, it seems prudent to draw on experience from both literary study and theatrical practice. As a professor of English, my work is primarily of a textual order. In my Introduction to Shakespeare class, howsoever often we might read lines aloud, our attention is invariably drawn to the page. I teach as well in the Old Globe Master of Fine Arts in Acting program and although my focus when working with actors is also textual, given their profession, a slate of concerns emerges of an entirely different order. Reading The Comedy of Errors, for example, when students in a literature course discover a clever bit of enjambment, such as the ‘Sprung’ that springs up at the beginning of line 10 in Act 1 (see below), they tend first to notice this phenomenon spatially, their eyes registering diagonally the shift from right to left and the scrolling movement down the page (what I call the ‘hypotenuse factor’), as well as the capital ‘S’, all ensuring the visual impact of springing ‘sprung’ to their attention. Students of acting, however, thinking ahead to performance, immediately begin to consider how, when speaking these lines, one might replicate this effect. Of course, ‘Sprung from’ is a trochee, and ‘Sprung’ will be lifted in pitch and/or lengthened; yet is that enough? Or should line 9 end with a pause, breath or some other mark of emphasis (even though we are mid-phrase), after which ‘Sprung’ presents itself with greater force and, perhaps, a mild sense of surprise? Merchant of Siracusa, plead no more. I am not partiall to infringe our Lawes; The enmity and discord which of late Sprung from the rancorous outrage of your Duke, To Merchants our well-dealing Countrimen, Who wanting gilders to redeeme their liues, Haue seal’d his rigorous statutes with their blouds, Excludes all pitty from our threatning lookes: (TLN 7–14) The lines above are representative of the early plays, the great majority end-stopped, so that enjambment is highly conspicuous. They compel us to ask whether Shakespeare expected the end of lines to be marked prosodically that were not marked grammatically (lines where the syntax continues, without intervening punctuation), achieving one effect; or whether he expected an actor to continue through the enjambment, lengthening the sentence over the line, achieving another effect.

86


How to answer such a question responsibly? Peter Hall demands that the line always be honoured: In order to phrase naturally, there is an absolute need to feel where the end of the verse line occurs. To run lines together risks incoherence and often necessitates a pause in the middle of the next line for breath or sense. This wrecks the structure of the verse. (Hall 2003: 29) Neither John Barton nor Cicely Berry is as insistent about capping off the ends of verse line, but both ask actors to attend to the end of the line in most cases (see Barton 1984: 40–1; Berry 1997: 52–81). Many directors I have worked with (and several of my academic colleagues) maintain that actors should simply follow the given punctuation when it makes sense to do so and abandon it when it does not. Rather than trying to be original and/ or definitive, I will attempt to lay out coherently the question of line breaks in Shakespeare and to suggest possible avenues of response in the hope of generating further discussion. I do want to take seriously, however, the possibility of recapturing some sense of practice on the Elizabethan and Jacobean stage, and thus I will be turning primarily to the First Folio as a control text, with reference to several quartos, to achieve some sort of grounding.

2. The question of enjambment The issue of enjambment, of course, is in the nature of the beast when dealing with dramatic verse. Formal poetry, as Northrop Frye notes, has always utilized the tension between form and syntax: In every poem we can hear at least two distinct rhythms. One is the recurring rhythm . . . a complex of accent, metre, and sound-pattern. The other is the semantic rhythm of sense, or what is usually felt to be the prose rhythm. (Frye 1957: 263) The friction between these separate rhythms is felt acutely at the end of enjambed verse lines: ‘Viewed as competitors for the rhythmic control of the poem’, writes Mary Kinzie, ‘the line is the conservative force and the sentence the anarchist as it pulls attention away from meaning and line, which seek to contain it’ (Kinzie 1999: 76). The interplay between line and syntax is readily available on the page. Yet, verse on the stage, ephemeral, subordinate to stylistic interpretation, maintains its structure


87

less evidently. The place to start to examine whether or not an Elizabethan or Jacobean actor would have tried to preserve, when speaking lines of verse, the sense of each line as a fixed unit seems to me to be the evolving nature of Shakespeare’s use of enjambment. In the early plays, thought unit and line almost always correspond, and each of these units fits rather neatly into the larger grammatical structure dictated by the sentence. Here are two examples of the openings of plays from the First Folio: Cease to perswade, my louing Protheus; Home-keeping youth, haue euer homely wits, Wer’t not affection chaines thy tender dayes To the sweet glaunces of thy honour’d Loue, I rather would entreat thy company, To see the wonders of the world abroad, Then (liuing dully sluggardiz’d at home) Weare out thy youth with shapelesse idlenesse. But since thou lou’st; loue still, and thriue therein, Euen as I would, when I to loue begin. (The Two Gentlemen of Verona, TLN 1–13) Hvng be y heauens with black, yield day to night; Comets importing change of Times and States, Brandish your crystall Tresses in the Skie, And with them scourge the bad reuolting Stars, That haue consented vnto Henries death: King Henry the Fist, too famous to liue long, England ne’re lost a King of so much worth. (The First Part of King Henry the Sixth, TLN 9–15) End-stopping seems to ensue as a matter of expressing the sense of the phrase, buttressed by the need to preserve the shape of rhetorical figures. Caesuras in these early works are rare, indicating that each line was generally conceived of as a whole, even if part of a greater syntactical structure. The composition of verse in this manner seems to pattern itself on the way in which the lines would be delivered; of course, conventions for speaking dramatic verse were also established by the text that made that verse available to the players. It may have been a closed, self-reinforcing system of composition and delivery that Shakespeare worked within in his early career; and it may have been those neat boundaries that he worked against as he gained ever-greater mastery of his own technique.

88


Tudor poetry, which utilized an iambic pentameter line, was generally composed of two phrases, the first of four syllables, the latter of six: as George Gascoigne advises, ‘There are also certayne pauses or restes in a verse whiche may be called Ceasures . . . in a verse of tenne [syllables] it will best be placed at the ende of the first foure sillables’ (in Sylvester 1984: 325). In such an arrangement, as Thomas Wyatt’s Sonnet 33 illustrates, the sentence is always subordinate to the line: ‘What rage is this? what furour of what kynd? / What powre, what plage, doth wery thus my mynd?’ (in Sylvester 1984: 155). Elizabethan poets played a bit looser with this structure, often alternating the syllabic groupings, but they maintained the division of the line into two phrases and the priority of the line over the sentence, as in Philip Sidney’s Sonnet 36: ‘Stella, whence doth this new assault arise, / A conquerd yelden, ransackt heart to winne?’ (in Sylvester 1984: 437). The voices of Shakespeare’s characters in the early plays, when speaking in verse, have the timbre of such poetry. Compare the linear structure of the passages above to the opening of Venus and Adonis: Even as the sunne with purple-colourd face, Had tane his last leaue of the weeping morne, Rose-cheekt Adonis hied him to the chace, Hunting he lou’d, but loue he laught to scorne: Sick-thoughted Venus makes amaine vnto him, And like a bold fac’d suter ginnes to woo him. (Venus and Adonis, Quarto 1593) Shakespeare’s early dramatic characters, like the speakers in his narrative poems, express themselves in formal, structured language, a language in which the explicit expression of the form itself is required and in which the line is the primary vehicle for organizing complex thought. There is self-evident tension at the opening of Richard III between the first and second lines of verse; that is to say that, for Richard, ‘Now is the winter of our discontent’, is a distinct articulation of his interior world, set in contradistinction to the effacement of this winter (and his identity) by the son/sun of York. Syntactically speaking, there is no break between the first two lines; therefore modern editors nearly always leave the opening line’s end without punctuation: Now is the winter of our discontent Made glorious summer by this son of York; (Wells and Taylor Oxford, 2nd edition)


89

Yet this is actually a rather bold instance of editorial prerogative. From the first extant quarto (1597) through five other quarto editions (1598, 1602, 1605, 1612, 1622) to the First Folio (1623), and in two later quartos (1629, 1634) a comma follows ‘discontent’. Now is the winter of our discontent, Made glorious summer by this sonne of Yorke: (Q1 1597) Now is the Winter of our Discontent, Made glorious Summer by this Son of Yorke: (Folio 1623) Whereas modern editors prioritize the integrity of the sentence, Early Modern ‘editors’ were quite willing to retain a syntactically unnecessary comma in order to retain, one would imagine, the implied break at the end of the line. Were the original compositors/printers/publishers working from a scribal copy (foul/fair papers, prompt book) that included the comma? If Shakespeare wanted a break, would he indicate it through punctuation, sense of the sentence be damned? Were they indicating a pattern of delivery that had become customary on the stage? Or was some conflation of these two factors at play? The opening lines of Richard III, then, as now, likely lodged themselves in the collective cultural memory, so that the comma may have denoted simply what audiences heard again and again during the play’s performance: Richard commenting on his despondency – pause for effect – followed by the ironic turn towards the general state of delight pervading the kingdom. Yet a question arises from this train of argument: if it was conventional for actors to mark every line as a whole before continuing on, and if printers/publishers were willing to use commas to indicate such markings, why would any lines at all be left without some sort of punctuation mark? Or, conversely, if it was the custom to break at every line, why would the first lines of Richard III need a seemingly anomalous comma to signify something that would have been marked anyway? If we accept Peter Hall’s contention that, ‘The end of each line is in fact a punctuation often more crucial than the regular punctuation itself’ (Hall 2003: 28), we may begin to find an answer. Assuming that to preserve the sense of the whole line one must call attention to the close of the line in some fashion, then there must have been a qualitative difference in the sorts of ways that lines were brought to a close, perhaps indicated by punctuation

90


marks. According to Renaissance manuals of rhetoric, a comma, semi-colon, colon, period, question mark, exclamation point, each indicated a particular quality of marking – of length, pitch, tone, etc. In Elementarie (1582), Richard Mulcaster writes that the comma ‘warneth us to rest there, and to help our breath a little’, while the period demands a longer pause ‘to help our breath at full’, and a parenthesis signifies a ‘lower & quicker voice’ (Hunter and Lichtenfels 2005: 20–1). Thus it was not simply a break at the end of Richard’s first line that was indicated, but a certain sort of break. I would further suggest that although every verse line ended with a break, those that ended with a punctuation mark indicated a sort of intentionality in the break. A punctuation mark assumed a self-conscious break on the part of the speaker as he or she accommodated grammatical necessity. Speech and sense coincided. In contrast, a line without punctuation was still marked as a complete line even as it broke, in order to help the auditor gain or retain a sense of meaning; yet that meaning may or may not be entirely under the control of the character speaking. If we remind ourselves that the line is the conservative force and the sentence the chaotic, in that open space at the end of unpunctuated verse lines, we can see how Shakespeare began to explore something like what we in modern terms call subtext.

3. Shakespeare’s dramaturgical development As Shakespeare’s career progressed, he used enjambment with greater frequency and fluidity, breaking lines in an ever-more radical fashion. Given that Shakespeare was deviating progressively more from contemporary practice and given that Shakespeare is always writing for the stage, we must question the dramaturgical purpose here. At least two possibilities seem to present themselves: (A) that there was on the stage persistent, habitual end-stopping, inured by practice and custom, and that Shakespeare used this convention to fragment units of thought and meaning, or (B) that Shakespeare was encouraging, even demanding, his actors to allow the sentence to trump the line, replicating more natural speech patterns. Looking at the first scene in Twelfth Night, we can return to the question of intentionality in pausing: O she that hath a heart of that fine frame To pay this debt of loue but to a brother, How will she loue, when the rich golden shaft


91

Hath kill’d the flocke of all affections else That liue in her. When Liuer, Braine, and Heart, These soueraigne thrones, are all supply’d and fill’d Her sweete perfections with one selfe king: Away before me, to sweet beds of Flowres, Loue-thoughts lye rich, when canopy’d with bowres. (Twelfth Night, TLN 39–47) Surely part of the emotional and psychological agitation of this moment is communicated by Orsino, lost in self-indulgent erotic reverie, lingering on ‘shaft’ and ‘fill’d’. To achieve this frisson, Shakespeare would have had to have some confidence in his actor’s proclivity to mark the end of the verse line. Orsino’s narcissistic fantasy is most clear to us if we hear him mark suggestively ‘shaft’ and ‘fill’d’; yet if he does so consciously, even intentionally, his sentiments are reduced to the vulgar. As John Barton argues, ‘You breathe at the end of verse-lines . . . [thus Shakespeare] catches our trick of often pausing in ungrammatical places and running on at full stops’ (Barton 1984: 42). The purpose of this practice is always to reveal something critical about the speaker’s state of mind and/or heart. Jan Gist, vocal coach for the Old Globe Theatre, widens the field of possibilities for marking line endings beyond just taking a breath. In addition to or instead of inhaling air into the body after the last word of each line, or taking a ‘catch breath’, or quick inhale, before the thought continues, one might ‘lengthen or strengthen the vowels and consonants of the last word, finalize the last sound of the last word, inflect the pitch on the last word to relate it forward, or pause, that is, take a moment of silence after the last word’. Alternatively, an actor ‘can commit to and fill out’ the idea or image of the final word, or emphasize the last word as ‘operative’ by separating it from the flow of the whole sentence (Gist 2008). The range of possibilities available to modern actors, even if not organized according to the terms of formal rhetoric, was available to Early Modern players as well, and Shakespeare’s increasing experimentation with verse-endings indicates that he was experimenting foremost with the sound of language on the stage. By working sometimes with, sometimes against syntactical structures and the conventions of playing, Shakespeare suggests the thought, emotion or impulse that lies beneath, behind or between what is said. Sometimes a character is aware of and in control of what is revealed, very often he or she is not. Thus naturalness of speech, paradoxically, emerges from honouring the artificial constraints imposed by the line.

92


4. Line endings in late Shakespeare By the late phase of Shakespeare’s career enjambment had become radical, often splitting subject from verb, even articles from their referents. In the example below, we see the first lines literalizing the action of ‘hanging’, as the eye drops from one line to the next, similar to the earlier example of ‘Sprung’ from The Comedy of Errors. Yet a quick glance over the passage indicates that the line has loosened in extreme ways: ‘I’ is split from ‘Had’, ‘that’ from ‘Which’, ‘thou’ from ‘His’. To maintain the integrity of the line by breathing or pausing at the line’s end seems to risk losing entirely the sense of the sentence. Why he that weares her like her Medull, hanging About his neck (Bohemia) who, if I Had Seruants true about me, that bare eyes To see alike mine Honor, as their Profits, (Their owne particular Thrifts) they would doe that Which should vndoe more doing: I, and thou His Cup-bearer, whom I from meaner forme Haue Bench’d, and rear’d to Worship, who may’st see Plainely, as Heauen sees Earth and Earth sees Heauen, How I am gall’d, might’st be-spice a Cup, To giue mine Enemy a lasting Winke: Which Draught to me, were cordiall. (The Winter’s Tale, TLN 404–15) This passage is taken admittedly from someone in a rather acute state of emotional distress – Leontes in The Winter’s Tale – yet from this extreme example it seems we can discern what it is that Shakespeare has come to learn about verse and what he is trying to achieve. In breaking the line, he breaks thought, not into distinct units but into fragments. Trying to piece those fragments together is what speech itself is about – attempting to articulate both for others and for oneself the truth of one’s experience. The need to maintain some sort of whole sense of the self is often at odds with our fragmentary, mercurial experience of the world. On the one hand, we have often imperfect traces of information, on the other, we are compelled to shape this provisional sense of ourselves and our environment into a whole, as if trying to piece together a mosaic in the dark. This is certainly how Leontes must feel looking at Hermione’s


93

‘bastard’ son: Ynch-thick, knee-deepe; ore head and eares a fork’d one. Goe play (Boy) play: thy Mother playes, and I Play too; but so disgrac’d a part, whose issue Will hisse me to my Graue: Contempt and Clamor Will be my Knell. Goe play (Boy) play, . . . (The Winter’s Tale, TLN 268–72) In three consecutive lines, Leontes splits subjects from their actions: I/play, issue/will hisse, Clamor/will be. George T. Wright suggests that Shakespeare at this stage in his writing has moved to prioritize the sentence over the verse line: ‘Such endings lead us without pause over the line-ending’ (Wright [1988], in McDonald 2004: 873). Yet, it is precisely a split – the unexpected, yet absolute division he now feels between father and ‘son’ – that Leontes is experiencing. Patrick Stewart, breathing at the end of each Leontes line above, described the effect: ‘It gave me the sensation of somebody who was beginning to spiral out of control’ (Barton 1984: 43). And in the same way that, breathing at, pausing for, or attending to the end of each line, he takes a moment to find the verb to connect to its preceding subject, so, too, does Leontes, given the unanticipated turn life has presented to him, find what he feels to be an appropriate action. Each turn of the line allows for a discovery. In the brief moment it takes to make that discovery, we witness Leontes’ attempt at self-cohesiveness and judge for ourselves his success. A glance at Posthumus in Cymbeline will illustrate the same principle in action. Believing that he is responsible for the death of Imogen, his words are whirling as he seeks his purpose: Yea bloody cloth, Ile keep thee: for I am wisht Thou should’st be colour’d thus. You married ones, If each of you should take this course, how many Must murther Wiues much better then themselues For wrying but a little? Oh Pisanio, Euery good Seruant do’s not all Commands: No Bond, but to do iust ones. Gods, if you Should haue ’tane vengeance on my faults, I neuer Had liu’d to put on this: so had you saued The noble Imogen, to repent, and strooke Me (wretch) more worth your Vengeance . . . .

94

Stylistics and Shakespeare’s Language . . . so Ile fight Against the part I come with: so Ile dye For thee (O Imogen) euen for whom my life Is euery breath, a death: . . . (Cymbeline, TLN 2858–68/2881–4)

We find through a series of line breaks Posthumus discovering what course of action he must take. From wishing he ‘neuer / Had liu’d’ and that vengeance had ‘strooke / Me’, to resolving to ‘fight / Against’ the Romans and ‘dye / For thee’, we witness the quicksilver nature of Posthumus’ emotions and the progressively cogent plan he is shaping. Yet, if these lines are spoken without some sort of break at each line’s end, then Posthumus’ speech is smoothed out so completely as to have the quality of resolve before that quality is discovered. In other words, at each turn of the line, an actor, whether he or she chooses to indicate this through a breath, a pause, emphasis or some other vocal gesture, has the opportunity to represent the active thought processes of a character, formed in the crucible of circumstance. The more radical the disjuncture, that is, the more that a line splinters a unit of thought where we would not conventionally have that thought broken, the more evident the fracture and the more aware we are that, from a myriad of possibilities, the words that continue the thought as the next line continues are chosen in the moment, reflecting a living, breathing intelligence. It is curious that as Shakespearean lineation becomes increasingly broken and irregular, his plays are increasingly concerned with the overarching power of art and artifice and its ability to give structure to a seemingly disordered world. Consider the nine scenes of The Tempest – first balanced against last, second against penultimate, so that the play folds out from its centre – the betrothal of Miranda and Ferdinand, engineered by Prospero – like the halves of a paper snowflake. This play begins with a raging storm, yet ultimately bears witness to our ability to control what is outwardly unruly and untamed. Prospero is consistently confident that he is in control of his ‘Art’. Yet in the moment when Prospero chooses mercy over vengeance, we find in his speech some rather interesting slippage: Hast thou (which art but aire) a touch, a feeling Of their afflictions, and shall not my selfe, One of their kinde, that rellish all as sharpely, Passion as they, be kindlier mou’d then thou art? Thogh with their high wrongs I am strook to th’ quick,


95

Yet, with my nobler reason, gainst my furie Doe I take part: the rarer Action is In vertue, then in vengeance: . . . (The Tempest, TLN 1971–8) Prospero’s decision to forgive the trespasses of those who have wronged him is much more engaging if we witness it in process. Ariel’s ‘feeling’, which is what prompts Prospero’s forgiveness, is highlighted by its placement at the end of the line and, more critically, by the unexpected (syntactically speaking) break that follows. Prospero’s ‘furie’ must be relinquished; but its hold lingers with the line. Finally, if his conclusion is read without interruption – ‘the rarer Action is in vertue, then in vengeance’ – Prospero’s sentiment is rather sententious. If, however, ‘the rarer Action is . . .’, and we observe a moment filled with something before Prospero proceeds – resignation, enlightenment, relief – then his commitment to ‘vertue’ over ‘vengeance’ is active, in the moment, and most importantly, purposeful. We see order, imposed by humane sentiment, overcoming the potential for chaotic vengeance (Bacon’s ‘wild justice’); and as this is a dramatic performance, we find it much more dramatically compelling to watch this happen than to feel that it had been concluded before the action even commenced. The radical turns of the line in Shakespeare’s late plays mirror the radical turns in the narrative of these stories. Each line is a horizon on a voyage. The unexpected is discovered around corners. Yet we cannot see where the turns are taken if the horizon has not been marked. Poetry ultimately engages in the play between expectation and surprise, between pattern and variation. As I am constrained to advise the actors that I work with, I find it my obligation to make as clear as possible the patterns of Shakespeare’s verse, not to impose those patterns rigidly, but to do so firmly so that when an actor swerves, he or she swerves with purpose and intent, retaining control even where chaos seems to be imminent. In fact, The Tempest is the endpoint, begun in Pericles, working through The Winter’s Tale and Cymbeline, of art’s conquest over chaos. In Pericles, the vicissitudes of chance are overcome through the direct intervention of the goddess Diana; in Cymbeline, Jupiter intercedes. Yet in The Winter’s Tale, the oracle of the gods fails to avert disaster and must be supplemented by artifice in the end. By The Tempest, Shakespeare has dispensed with any sort of divine interference and has located in the hands of a human character the power to use art to supersede chaos. Insofar as Prospero is an avatar of Shakespeare, one who selfconsciously uses his art to order the world, there seems to be a concession

96


to the playwright’s own role in shaping out of even the most wildly irregular verse something close to a cohesive vision.

References Quotations from Shakespeare’s plays are taken from Charlton Hinman (ed.), The Norton Facsimile: The First Folio of Shakespeare (New York: W.W. Norton & Co., 1968), unless otherwise noted, with line numberings corresponding to the Through-Line Number system (TLN). Other Shakespeare references are to: Wells, Stanley and Gary Taylor, with John Jowett and William Montgomery (eds) 2005 [1986]. William Shakespeare: The Complete Works. Oxford: Clarendon Press. Facsimile of the 1593 Quarto edition of Venus and Adonis accessed from Internet Shakespeare Editions at http://internetshakespeare.uvic.ca/ Library/plays/Ven.html. Facsimiles of quarto editions of Shakespeare’s plays accessed from the British Library online at www.bl.uk/treasures/shakespeare/homepage. html Barton, John 1984. Playing Shakespeare. New York: Anchor Books. Berry, Cicely 1997. The Actor and the Text. New York: Applause Theatre Books. Frye, Northrop 1957. The Anatomy of Criticism. Princeton: Princeton University Press. Gist, Jan 2008. ‘Inhaling Shakespeare’. Presented at the Voice and Speech Trainers Association Conference, New Orleans, Louisiana, July 2008. Hall, Peter 2003. Shakespeare’s Advice to the Players. New York: Theater Communications Group. Hunter, Lynette and Peter Lichtenfels (eds) 2005. Shakespeare, Language and the Stage. London: Thomson Learning. Kermode, Frank 2000. Shakespeare’s Language. New York: Farrar, Straus and Giroux. Kinzie, Mary 1999. A Poet’s Guide to Poetry. Chicago: The University of Chicago Press. McDonald, Russ 2001. Shakespeare and the Arts of Language. Oxford: Oxford University Press. Palfrey, Simon 2005. Doing Shakespeare. London: Arden. Palfrey, Simon and Tiffany Stern 2005. Shakespeare in Parts. Oxford: Oxford University Press. Stern, Tiffany 2004. Making Shakespeare. London: Routledge. Sylvester, Richard (ed.) 1984. English Sixteenth-Century Verse: An Anthology. New York: W.W. Norton & Co.


97

Tucker, Patrick 2002. Secrets of Acting Shakespeare: The Original Approach. New York: Routledge. Weimann, Robert 2000. Author’s Pen and Actor’s Voice. Cambridge: Cambridge University Press. Wright, George T. 1988. ‘The play of phrase and line’, from Shakespeare’s Metrical Art. Berkeley: University of California Press, 207–28, in: Russ McDonald (ed.) 2004. Shakespeare: An Anthology of Criticism and Theory 1945–2000. Oxford: Blackwell Publishing: 861–79.

Further reading Adamson, Sylvia, Lynette Hunter, Lynne Magnusson, Ann Thompson and Katie Wales (eds) 2001. Reading Shakespeare’s Dramatic Language. London: Arden. Hope, Jonathan 2005. Shakespeare’s Grammar. London: Arden. Hyland, Peter 2003. An Introduction to Shakespeare’s Poems. New York: Palgrave Macmillan. Prince, F. T. (ed.) 1960. The Poems. London and New York: The Arden Shakespeare, Methuen.

Chapter 5

Subject-Verb Inversion and Iambic Rhythm in Shakespeare’s Dramatic Verse Richard Ingham and Michael Ingham1

1. Introduction Studies of the interaction of metre and syntax in Shakespearean verse, though uncommon, are of great potential interest for the prosodic aspects of his work in performance. Culpeper (2001: 202) has drawn attention to the often underrated significance of syntactic features in Shakespearean texts. His discussion is restricted to instances where syntactic features relate to cognitive organization of speech, but can be usefully complemented by considering what choices syntax makes available to the writer, and what their prosodic consequences are for the performer. One significant type of syntactic choice is the inversion of main constituents. Examples of inverted word order occur in a range of Shakespearean contexts, both verse and prose, with a range of textual impacts. The device is commonly assumed to have provided English verse writers, especially in the pre-contemporary period, with a ready means of complying with metrical exigencies. In this chapter, using an empirical database of 998 target instances sampled across Shakespeare’s career, the validity of such an assumption is examined with respect to one particular syntactic option: the construction in which a subject could optionally be inverted round the verb in the context of an initial non-subject constituent. We address the issue by means of a dual approach: first a linguistic analysis across Shakespeare’s dramatic verse career, and secondly a comparison of interpretative styles by contemporary Shakespearean performers. The subject-verb inversion option can be simply illustrated from one of Shakespeare’s main sources, Holinshed’s Chronicles. The examples in (1) invert the subject, while those in (2) do not: (1)a. Then longed the duke sore to heare what he would haue said, . . . Holinshed, Chronicles (405)

Subject-Verb Inversion and Iambic Rhythm

99

(1)b. . . . [t]his did he to driue time, and to put his enimies out of all suspicion . . . Holinshed, Chronicles (266) (2)a. Then all the companie sware to him fealtie, and did to him homage . . . Holinshed, Chronicles (420) (2)b. The headlesse trunk he commanded to bée hoong vp upon an high paire of gallowes. Holinshed, Chronicles (266) The inversion of subject and verb in (1) may be seen as a relic of the inversion of subject and verb in Old and Middle English2 after clause-initial adverbials (cf. 1a) and direct objects (cf. 1b), a rule which continues to characterize present-day Germanic languages. It was a fairly common minority pattern in fourteenth- to sixteenth-century English (Baekken 1998, 2000), and seems to have typified a more literary register. Later in the Modern English period, most types of subject-verb inversion fell out of use except in archaizing verse styles3. An English author writing in the late sixteenth century, such as Holinshed, is likely to have perceived it as a viable stylistic option, to be used with discretion, rather than an out-and-out archaism. The issue considered here was how the availability of inversion interacted in Shakespeare’s verse drama with the presence of iambic rhythm, in particular how metre and syntax comported when the subject was a personal pronoun (henceforth Spro). Unless emphatic or contrastive, a subject pronoun never bears stress in natural English speech rhythm. Therefore, a regular iambic pentameter, with its alternation of weak and strong syllables, would often not map directly onto the speech rhythms of English, cf. the following examples from Shakespeare’s prose: (3)a. Yet you began rudely. (Twelfth Night 1.5.203) (3)b. . . . for truly I think you are damned. (The Merchant of Venice 3.5.5) In the sequences you began and truly I the subject pronoun either precedes or follows a phonologically weak syllable, so were this to be recast as blank verse two weak beats would occur in succession, a problem for regular iambic rhythm. Shakespeare’s verse appears to show numerous cases where the inversion of verb and subject pronoun respects iambic metre, for example, (4) The other part reserved I by consent, (Richard II 1.1.128) The uninverted order The other part I reserved would not be metrical, since the sequence I reserved fails to fit an iambic foot. It is therefore plausible to

100


see subject pronoun inversion as a useful option available to an author of iambic verse, since in principle the respective positions of verb and Spro could be varied, depending on the syntactic context, to fit the metre. In fact, however, we show in this study that this is very unlikely to have been Shakespeare’s practice, and that in earlier works he tended to place an inverted subject pronoun in an ictic position,4 seemingly contrary to a strategy driven by the needs of regular iambic rhythm. Shakespeare’s verse is known to be characterized by a striking flexibility in the interaction between syntax and metre (see for example, Wright 1983), and it is of course not argued here that the iambic scheme formed an absolute requirement on the wording of the line. Nevertheless most studies of Shakespeare’s metre consider that iambic rhythm is present as a factor providing a counterpoint to the speech rhythms of English. This will present an interpreter of a Shakespearean line with clear alternatives as to whether iambic rhythm is used or not, for example: (5) And are you yet to your own souls so blind/That you will war with God by murd’ring me? (Richard III 1.4.247–8) In line 248 the three function words that you will succeed each other with no intervening content word. A regular iambic rhythm would place a beat on the subject pronoun you, but speech rhythm normally would not, as it is a function word. We shall see later in this article how different actors react to this alternative, in interpretatively sensitive ways. It is interesting to consider how Shakespeare may have changed his practice with respect to the fit between syntax and metre over his authorial career, and then to ask what approaches to rhythmic variation are taken by performers of his verse drama. Here we seek to provide a preliminary assessment of the major aspects of Shakespeare’s use of syntax and metre in cases where these two components – inverted syntax for stylistic effect, and respect for iambic rhythm – may have created a mismatch. We shall examine how subject pronoun inversion evolved over Shakespeare’s writing career, and then how rhythmic delivery in contemporary performance has engaged with it. Although our study comprises a linguistic analysis and an interpretative analysis which clearly differ in methodological approach, it is intended to function as a single line of enquiry in which the initial linguistic analysis provides the empirical basis for discussion of earlier versus later practice by the poet, and of alternative interpretations by performers. The study is organized as follows: in section 2 we briefly review previous findings relevant to the issue discussed in this enquiry. In section 3, we discuss the


101

sources and methodology we have used to track the evolution of subject pronoun placement in Shakespeare’s verse drama, and its relationship with iambic rhythm. Section 4 presents the results of these analyses. In section 5 we turn to the issue of how recent recorded performances of the plays studied have handled the tension between syntax and metre, drawing attention to the subtle interpretative effects derived by particular performers. Section 6 concludes the chapter with some observations on the earlier as compared with the later authorial style of Shakespeare as to the link between speech rhythm and metrical rhythm.

2. Research background Shakespeare’s use of inversion constructions has been studied by Houston (1988), who found that verb-subject (VS) order tended to decline in plays thought to have been written from the end of the 1590s onwards. However, whether this decline affected inversion with subject pronouns (henceforth VSpro) to the same extent as inversion with full nominal subjects was not stated. Since the same study also found increasing frequency of SubjectObject-Verb patterns from roughly the same time-point, Shakespeare cannot be said to have moved away from all types of inverted syntax in his later plays. Hence VSpro could have remained constant as a stylistic option throughout Shakespeare’s artistic career. Hope (1994) studied syntactic patterns in Shakespeare’s work and found that do-support fell within a percentage frequency band, which was distinct from that of other contemporary authors such as Marlowe and Middleton. His work showed the value of considering the use of a variable feature of syntax in particular writers of this period. The use of a syntactic variable is seen to be not merely a random feature but clearly under the control, conscious or not, of the author. Hope did not, however, investigate inversion patterns, the focus of the present study. To our knowledge there has not been any empirical investigation of how the interplay of metre with syntactic variation was handled by Shakespeare in the domain of subject-verb inversion. The present study is a modest attempt to remedy this deficiency. Shakespeare’s metre has received a great deal of critical attention which can only briefly be summarized in the space available here. Abbott (1870), as well as clarifying the notions of emphatic stress and regular, metrical stress, addressed the problem of tension between iambic rhythm and speech rhythm, which continues to cause difficulty for performers today, and indeed for critics too, judging by the vigorously polemical online debate

102


in 2002 on the website Shaksper.net between David Wallace and David Evett (Wallace 2002). Wallace argued that a polysyllabic word was not rhythmically free in Shakespeare: its stress pattern was respected by the rhythmic structure of the line in which it appeared, insofar as its primary stressed syllable had to be in an ictic position. A monosyllabic word, however, was not affected by this requirement and so could occur either in or out of an ictic position. The empirical prediction was thus that in the sample of plays used here polysyllabic verbs would always map their primary stressed syllable on to an ictus of iambic rhythm, while monosyllabic verbs would show no association with an ictic beat. Evett, on the other hand, was sceptical as to whether metre is determined by polysyllables alone, and of the claims of metrical theorists in general. For the purposes of this study, Wallace’s position offers the advantage of being a hypothesis to be empirically investigated, by distinguishing between monosyllabic and polysyllabic contexts in which subject pronouns occur. This point will be taken up again in section 4 below. Their exchanges in some ways echoed the disputes in defence of alternative representations of the iambic pentameter in the late 1960s and early 1970s, informed by the then recent work in the field of Chomskyan structural linguistics. Here too the leading theoretical model presented the notion of a correspondence between the abstract metrical pattern and the surface features of any given line. For Halle and Keyser (1971) the concept of ‘stress maximum or maxima’ allows the Abbott distinction between iambic rhythm and emphatic accent to be handled in terms of the phrasal or surface stress in relation to an underlying metrical pattern. They postulate that such syllables bearing a ‘stress maximum’ can only be preceded and followed by unstressed syllables. Wright (1983) traces the profound changes and radical experiments with form undertaken by Shakespeare, from the more conventional metrical structures of the early plays to the rupture with regular metre evident in late plays like Cymbeline and The Tempest. He differentiates between what he calls the underlying ‘deep structure’ of the iambic rhythm and the ‘surface form’ of the increasingly irregular distribution of syllables and semantic components in Shakespeare’s mature verse. By the end of Shakespeare’s writing career, according to Wright, the formal restrictions of earlier tradition-bound iambic poets such as Gascoigne had been discarded. Instead of the variables of beat (ictus), phrasal stress and major words coinciding, as had been customary previously, in Shakespeare’s late plays there is a deliberate play of counterpoint (polyphony as Wright refers to it), by which the conventional ‘equilibrium’ is upset and the line ‘more and more cast into


103

structural doubt’ (Wright 1983: 155). Scepticism about the power of smooth rhetoric and the ‘deceitful uses and users of language’ (Wright 1983: 156) is now much more evident in the frequently irregular or disruptive distribution of phrasal stresses, according to Wright’s insights. At this stage of the development of his versification Shakespeare is prepared to take risks and depart from tried and tested structures: ‘The hallmark of the new verse is the autonomy of the phrase, no longer obedient to the line or the linesegment but with its own authority well established’ (Wright 1983: 156). For Wright the polyphony allied to the metre is the key to Shakespeare’s dramatic verse: ‘semantic content and emotional design ride on the meter and the syntax’ (Wright 1983: 157). The notion of a counterpoint between underlying iambic rhythm and surface speech rhythm was pursued empirically by Tarlinskaja (1983), who conducted a detailed examination of how English phrasal stress, falling on content words but not function words, interacted with the ictic beat. She found a clear evolution over the course of Shakespeare’s writing career in the likelihood with which phrasal stress would be placed on the sixth and eighth syllables. This finding is relevant to the concerns of the present study in that the placement of phrasal stress on an ictic beat in Shakespeare was not random, but appears to have varied in a coherent way across time. In this study we use the convenient term ‘ictic’ to refer to placement on an even number syllable of a pentameter line, without prejudice to the issue of where syllables bearing stress maxima in speech rhythm occur.

3. Source texts and method of analysis For purposes of investigating language change, studying a particular idiolect over time is informative, as compared with analysing samples from a large number of individuals, in which particular trends may get blurred. As is well known, however, the textual transmission of the works ascribed to Shakespeare is far from straightforward, which complicates the reliability of its attribution to a single idiolect. Accordingly, the analysis we present in this study cannot be claimed to represent more than that of a corpus of works plausibly attributed to the same individual. In the case of a literary author, stylistic variation across genres presents further complications. We therefore opted to address the issues raised in the preceding sections by analysing a subset of the verse dramas of Shakespeare, namely those based on historiographical sources, in which it was hoped to observe a greater number of relevant data-points5. For this purpose, we excluded comedies

104


and also tragedies lacking what might be called a public narrative, that is, a succession of events in which a community beyond the protagonists’ own immediate circle participates, for example, the wider public domain in which King Lear is played out as compared with the more private sphere of action in Romeo and Juliet. We thus identified 16 dramas written principally or exclusively in verse, which were divided into the following chronological periods in terms of the dates conventionally attributed to them: Period I (1592–1597): Richard III, Titus Andronicus, Richard II, King John, Henry IV 1 Period II (1598–1603): Henry IV 2, Henry V, Julius Caesar, Troilus and Cressida, Othello Period III (1606–1610): The History of King Lear, Macbeth, Antony and Cleopatra, Coriolanus, The Winter’s Tale, Cymbeline Note that these temporal categories do not necessarily correspond to those of previous studies of Shakespeare, but were chosen as a means of roughly dividing the available data into three periods determined by their known or presumed dates of composition. The reference edition used was the Oxford Shakespeare (Wells and Taylor 1986). The syntactic construction targeted in these texts was the inversion of the finite verb and subject pronoun in the context of a preceding non-subject constituent. The latter was defined as an adverbial, a direct or indirect object, or a subject/object complement. Examples of these constructions from Shakespeare’s prose are provided below6: (6)a. Now will not I deliver his letter, . . .

(Twelfth Night 3.4.181)

(6)b. The maid will I frame and make fit for his attempt. (Measure for Measure 3.1.257–8) (6)c. Bawd is he doubtless, . . .

(Measure for Measure 3.2.65)

Because line-final metre poses particular issues in Shakespeare’s verse, cases where the relevant verb or subject pronoun stood in the fifth foot were not included. We also excluded cases where a subject pronoun had an auxiliary contracted to it, as well as incomplete or hanging half lines and other clear cases of irregularity as regards the iambic pentameter. The variables analysed, then, were the syntactic order of subject pronoun and verb in declarative main clauses with an initial non-subject constituent,


105

and the positioning of a subject pronoun in ictic position. All types of verbs were counted for the purpose of the analysis, regardless of whether they were main or auxiliary verbs. Monosyllabic verbs and verbs of two or more syllables were analysed separately, in view of Wallace’s argument that stress placement in polysyllabic verbs in Shakespeare respects ictic beats. A more refined analysis could subdivide our categories of analysis, but for the purposes of an initial examination of the field it was judged best to maximize the data frequencies to see how common the phenomena were and what was revealed by a quantitative investigation conducted on this basis.

4. Results of frequency analyses The issue addressed in this part of the study is how Shakespeare negotiated the relationship between word order and iambic rhythm, specifically in contexts where syntax allowed him flexibility in positioning a subject pronoun. To recall the proposal of Wallace (2002) that iambic rhythm was established by the presence of a polysyllabic word in the pentameter line is adopted in this study as a starting hypothesis. In the texts analysed, of 47 polysyllabic verb tokens co-occurring with a subject pronoun, in every case the naturally stressed syllable of the word fell on an evennumbered syllable. Wallace’s claim is thus entirely upheld. Unless the poet had been following a metrical scheme of some kind, this outcome would have been an inexplicable coincidence. Since polysyllabic verbs occurred in varying positions across the ten syllables of the line, that scheme must therefore have been applicable to the ten-syllable line as a whole. An iambic rhythm clearly constituted a default option for the tensyllable line, in accordance with the traditional stance adopted by adherents of the iambic pentameter analysis of Shakespeare’s dramatic verse. The position taken here thus endorses both the recently proposed claim of a rhythmic constraint associated with polysyllabic verbs, and the traditional assumption that iambic rhythm is at least underlyingly present in Shakespearean verse. Monosyllabic verbs, however, are a different matter. On the approach followed here, single-syllable words, including verbs, were free to appear in either odd or even syllable positions in the line. There is no expectation that, in pairs of monosyllabic verbs and subject pronouns, one or the other will be favoured in, for example an odd syllable position. Shakespeare may nevertheless have shown a preference for placing a particular form, pronoun or verb, in a particular position. In terms of natural speech rhythm,

106


non-contrastive personal pronouns are always unstressed, whereas verbs, including monosyllabic ones, may bear stress. With polysyllabic verbs, an underlying iambic rhythm was evidently at work. A preference for placing a single-syllable verb in an even (ictic) syllable position and a subject pronoun in an odd (non-ictic) position would tell us that Shakespeare was broadly tending to align the rhythm of the line on iambic rhythm with monosyllabic verbs as well. If this turned out to be the case, the question can then be addressed whether Shakespeare manipulated the possibility of subject-verb inversion as a device for keeping subject pronouns out of ictic positions. An initial analysis was therefore carried out, investigating whether in the 16 target plays a monosyllabic verb fell on an odd or an even syllable. The results are shown in Table 5.1. This established that Shakespeare was four times more likely to position a pronoun in a weak position than in a strong position, clearly indicating that he wrote with an awareness of an underlying weak-strong rhythmic alternation influencing the disposition of monosyllabic verbs and pronouns. This feature of his verse did not change over the three periods studied. Turning now to the question of how frequently Shakespeare deployed the VSpro option, Table 5.2 first presents the overall frequencies of inverted versus non-inverted syntax. In inverted contexts with a monosyllabic verb, therefore, SV and VS orders were almost equally weighted options for Shakespeare in both

Table 5.1 Syllable position of subject pronouns accompanying monosyllabic verbs Odd

Period I Period II Period III Average

Even

N

%

357 158 274 789

84.8 74.2 86.4 83.0

N 64 55 43 162

T % 15.2 25.8 13.6 17.0

421 213 317 951

Table 5.2 Frequencies of VS and SV syntax with main clause subject pronouns in inversion contexts featuring a monosyllabic finite verb

Period I (1592–1597) Period II (1598–1603) Period III (1606–1610) Average

VS

SV

Total

214 (50.8%) 109 (51.8%) 95 (30.0%) 418

207 104 222 533

421 213 317 951


107

periods I and II. It was only from 1606 onwards that a clear preference emerges for the non-inverted order. So far, then, it can be seen that VSpro was a freely available option for Shakespeare at least up to and including Othello, but that he strongly favoured putting the subject pronoun in an odd (non-ictic) position. Was VSpro used as a way of complying with the latter preference? Table 5.3 shows whether subject pronouns following monosyllabic verbs fell on an odd (weak) or an even (ictic) syllable. Although the majority of the time Shakespeare tended to place an inverted subject pronoun on an odd (weak) syllable, it can be seen that in a substantial minority of cases, especially in periods I and II, inversion caused the pronoun to stand in an even (ictic) syllable, contrary to our initial hypothesis that VSpro syntax was used to obtain conformity with an underlying iambic rhythm. Indeed, when all subject pronouns in even syllables are considered, both in VS and in SV order, a rather striking development emerges. In over three quarters of cases in period I, Shakespeare appears to have used VS order to position the subject pronoun in an even (ictic) syllable (see Table 5.4).

Table 5.3 Frequencies of inverted subject pronouns in even and odd syllables with monosyllabic verbs in Shakespeare’s dramatic verse

Period I (1592–1597) Period II (1598–1603) Period III (1606–1610)

Inverted subject pronoun in even syllable

Inverted subject pronoun in odd syllable

Total of inverted subject pronouns

N

%

N

%

N

49

22.9

165

77.1

214

29

26.6

80

73.4

109

10

10.5

85

89.5

95

Table 5.4 Frequencies of main clause VSpro and SproV syntax in inversion contexts with monosyllabic verbs, with subject pronouns in even syllables

Period I (1592–1597) Period II (1598–1603) Period III (1606–1610) Average

VSpro

SproV

Total

49 (76.6%) 29 (52.7%) 10 (23.3%) 88 (54.3%)

15 26 33 74

64 55 43 162


108 90 80 70 60 50

VS pro

40

VS pro ictic

30 20 10 0 Per I

Per II

Per III

Figure 5.1 Syntax of subject pronoun after an initial non-subject constituent: percentage of VSpro order overall, and percentage of VSpro order when the pronoun was in an even (ictic) syllable.

That is, VSpro order was adopted proportionately far more often if the pronoun was ictic than the overall average rate of VSpro order. The comparison is displayed graphically in Figure 5.1. Whereas in periods II and III, pronouns in an even syllable (ictic) are about as frequent as we would expect in terms of the overall rate of VSpro, there was clearly a strong preference in the period I plays for VS order. Why this should have been the case is an intriguing question which awaits further exploration; it may be related to expressive dimensions such as characterization in the particular plays, adopting the perspective of Culpeper (2001). Our analyses show, in any case, that inversion did not constitute a device for managing compliance with iambic rhythm. Shakespeare’s use of the two traits was quite clearly unrelated: on the one hand inversion shows a monotonic decline in subject pronoun inversion from about half to about a quarter of contexts (see Table 5.2), while on the other (see Table 5.1) we found a fairly low incidence of ictic inversion showing non-monotonic variation around the 20 per cent mark over the period studied. The foregoing data analysis has thus established three initial points. First, inversion of subject pronouns declines over the course of Shakespeare’s work. Secondly, VSpro order is not associated with odd syllable placement of a subject pronoun: Shakespeare placed subject pronouns on an even syllable about as often in his later plays as in his earlier ones, but now mostly in SproV order. Finally, Shakespeare’s earlier verse dramas strongly associate


109

ictic position of the subject pronoun with VSpro order. These findings, though clearly of interest as regards authorial practice in the interaction of syntax and metre, may further be enhanced by consideration of how the verse is performed. It is, after all, in the interpretation of the lines that the subtlety of the author’s craft, as well as that of the actor, is brought most keenly into contention. In the next section, therefore, we examine how established actors deal with the tension between iambic rhythm and syntax with particular reference to the cases of inverted subject pronouns in the early plays.

5. Contemporary interpretations in performance Performing Shakespeare’s verse poses considerable problems in terms of understanding his sometimes complex metrical patterns, and especially in respect of stress placement. The relationship between syntax and metre and between natural speech rhythm and metrical rhythm is conceptually challenging. Native-speaker intuition alone is inadequate to explain the complexities. In the teaching and directing of performance to nonnative speakers, invoking simplistic iambic rules for the delivery of lines is essential, yet, at the same time, frustratingly inadequate. Likewise, for native speakers we cannot assume competence in discerning appropriate patterns of phrasing in the enunciation of Shakespearean verse or, for that matter, of other verse dramatists. Our contemporary auditory sensibilities are ill-attuned to the prosodic structure of metrical verse in performance, accustomed as we are to free verse with fairly arbitrary patterns or to the frequently monotonous beats of much popular music. The subtleties of syncopation in English verse and lyrics can still be perceived in modern rap and other verse forms. However, sensitivity to metrical regulation and variation in contemporary Shakespeare performance so challenges, or is assumed to challenge, the contemporary ear that many film directors of Shakespeare adaptations virtually dispense with the verse altogether in favour of a proselike, ‘naturalistic’ speech delivery. For this reason BBC recordings of the Shakespeare canon between 1978 and 1983 have been used as a yardstick for assessing the sensitivity of contemporary actors to metrical features in general and iambic rhythm in particular. The rationale for selecting a sample of the BBC recordings is that the actors represent universally acknowledged specialists in Shakespearean performance, of the calibre of John Gielgud and Derek Jacobi, whose handling of verse drama may be considered authoritative. Recorded performances

110


were essential for the purposes of the investigation, since impressionistic recollections of live performances cannot be empirically studied. The consistently high standard of enunciation in the BBC series facilitated close textual/aural analysis with the unaided ear and the assistance of pause and playback commands.7 Furthermore, the BBC versions enjoyed wide distribution and many, though by no means all, were critically acclaimed. By comparison with the hit-and-miss approach of feature film directors and actors to Shakespearean versification, full-length stage-like performances, as opposed to abbreviated and truncated filmic versions of the works, were felt to be more appropriate. In the BBC versions filmic devices are used discreetly in such a way that they underscore rather than dominate the verbal delivery. In some respects the BBC series can be seen as taking up the baton of the masterful film performances of Laurence Olivier as Hamlet, Henry V, Richard III and King Lear, which represented for many the apogee of Shakespearean diction and the perfect fusion of sound pattern and semantic signification in the phrasing of Shakespearean verse. Since these authoritative recorded performances, however, the decline in the use of RP English in the electronic media and the tendency of film directors and illustrious contemporary actors, such as Jeremy Irons,8 to blur the distinction between Shakespearean verse and prose, is palpable. Perhaps there is a preconception on the part of certain directors and performers concerning the reception by contemporary audiences of Shakespearean pentameter. Perhaps also, fashionable recent adaptations of Shakespeare and Jane Austen, inter alios, into contemporary linguistic idiom inhibits any mass-audience Shakespeare performance that would be perceived as adhering to standard, ‘old-fashioned’ verse delivery style. All of the above critical debates concerning the complexities of the iambic pentameter as formal unit inevitably have a strong bearing on Shakespeare in performance. Where the iambic rhythm and the phrasal stress pattern of normal speech coincide, the trained actor will have no difficulty achieving a sense of rhythm in performance. What is more complicated, however, is the increasingly frequent use of disparities between metrical rhythm and phrasal or contrastive stress, quite apart from rhetorical emphasis in Shakespeare’s dramatic verse. Research into the delivery techniques of the actors in the BBC sample suggests an idiosyncratic approach to metre and rhetorical emphasis, with certain actors adhering largely to an underlying iambic pattern in the modern tradition of Olivier, while others prefer a more prose-like enunciation of lines with very limited observance of line-pauses and verse cadences or, indeed, metrical rhythm.


111

Our findings above reveal that in the earlier use of iambic rhythm, Shakespeare’s syntactic predilection for subject-verb inversion resulted in many instances of a non-contrastive subject pronoun occurring in even syllable position. In his later works, he produced verse that tended to avoid subject-verb inversion with minimal additions to the tally of noncontrastive subject pronouns. The present study set out to determine to what extent sensitivity to these factors was in evidence in the performance of leading Shakespearean actors of recent decades, whose interpretations were sampled from the series of BBC productions available in digital video format. Five plays were chosen from the BBC series, namely Richard III, Richard II, Julius Caesar, Troilus and Cressida and Cymbeline, so as to represent a broad span of the plays analysed in section 4 above. The actors included Ron Cook (Richard III), Zoe Wanamaker (Lady Anne), Jon Finch (Bolingbroke), John Gielgud (John of Gaunt), Richard Pasco (Brutus), Sam Dastor (Casca), Anton Lesser (Troilus), Suzanne Burden (Cressida), Helen Mirren (Imogen), Robert Lindsay (Iachimo) and Michael Pennington (Posthumus). Those who broadly adopted the phonological stress rules of natural speech in order, presumably, to render their dialogue more modern in feeling included Michael Byrne (Buckingham in Richard III), Rowena Cooper (Queen Elizabeth in Richard III), Charles Gray (Duke of York in Richard II, Julius Caesar in Julius Caesar and Pandarus in Troilus and Cressida), Vernon Dobtcheff (Agamemnon in Troilus and Cressida), Keith Michell (Antony in Julius Caesar) and Richard Johnson (Cymbeline). An analysis of their stress placement in cases where Spro was positioned on an even syllable showed that in the majority of cases in the BBC samples the pronoun receives light but distinct stress. There is a high correlation between actors who observe metrical rhythm more closely throughout their performance as a whole and the discernible enunciation of stress on Spro placed as even syllables in these specific cases. Conversely, those actors who are inclined to ‘play more and more astride the verse instead of in it’9, to cite Granville-Barker’s admonition to the young Gielgud, and treat the verse as if it were prose, tend to ignore the question of subject pronoun stress entirely unless it is strictly contrastive. Examples of both types are given below (our own emphasis shown in bold type): Pronoun in even position enunciated with stress (VS): (7) A cockatrice hast thou hatched to the world, Richard III 4.1.54 (Annette Crosbie: Duchess of York) (8) Again shall you be mother to a king, Richard III 4.4.317 (Ron Cook: Richard III)

112


(9) Come I appellant to this princely presence. Richard II 1.1.34 (Jon Finch: Bolingbroke) (10) Hast thou tapped out and drunkenly caroused. Richard II 2.1.128 (John Gielgud: John of Gaunt) (11) I wish we may. But yet have I a mind Julius Caesar 3.1.145–6 That fears him much; . . . (David Collings: Caius Cassius) Pronoun in even position enunciated without stress (VS and SV): (12) More bitterly could I expostulate, Richard III 3.7.182 (Michael Byrne: Buckingham) (13) To Bolingbroke are we sworn subjects now, Richard II 5.2.39 (Charles Gray: Duke of York) (14) From your own mouth, my lord, did I this deed. Richard II 5.6.37 (Desmond Adams: Sir Piers of Exton) (15) For certain she is dead, and by strange manner. Julius Caesar 4.2.241 (Brian Coburn: Messala) A variable approach to delivery of the pentameter is evident in all five of the productions assessed. To some extent speech style and delivery probably depended on directorial preference. Jonathan Miller’s Troilus and Cressida deploys a prose-like delivery for much of the play, with little discernible attempt at suggesting an underlying iambic rhythm, apart from in the romantic exchanges of the lovers’ scenes. Nevertheless in all five productions it was clear that some actors were palpably more rhythmically faithful in their spoken verse technique. However, all of the actors studied (both the iambic rhythm-oriented and the non-iambic rhythm-oriented) showed awareness of verse scansion in telling instances. For example, the past tense and adjectival -ed endings which are non-syllabic in modern English are frequently sounded as a separate syllable in Shakespearean verse, the extra syllable thereby contributing to the syllable count of a particular line. Unlike a number of non-Shakespearean specialists in film adaptations, the BBC series actors flawlessly observe the metrically required articulation of -ed, or non-articulation of -‘d. Probably the most interesting case in the comparison between iambicoriented and non-iambic oriented actors was Derek Jacobi, who adopted a deeply subtle approach to metre in his interpretation of the role of Richard II. In the first two acts of the play the capricious nature of the King’s character is conveyed by his capricious, almost throw-away treatment of the verse with the notable exception of a more ceremonial use of regular rhythm in his decrees


113

of banishment on Bolingbroke and Mowbray. As Richard’s dignity and power are gradually stripped away in Acts 3 and 4, he exhibits a serious, contemplative and sorrowful attitude with more measured tones, which is reflected in his greater observation of iambic regularity. In Richard’s Act 5 pre-death soliloquy Jacobi reverts at one point to a more syncopated approach to rhythm to express his character’s agitated state of mind, but what precedes and follows is in many places more conventionally iambic. Below are four examples from different stages of the play with stress/emphatic accent bolded: (16) We will ourself in person to this war, And for our coffers with too great a court And liberal largess are grown somewhat light, We are enforced to farm our royal realm, The revenue whereof shall furnish us For our affairs in hand . . . . Richard II 1.4.41–6 (Jacobi) Prosodic features in Jacobi’s delivery observe the stress and tempo of natural speech rhythm with additional optional rhetorical emphasis. Iambic rhythm is not evident, although the rising and falling intonation, very distinct diction and alliterative emphasis all confer a poetic quality on the delivery. The actor’s choice in respect of pace and pause accentuates this non-iambic attribute10: (17) My crown I am, but still my griefs are mine. You may my glories and my state depose, But not my griefs; still am I king of those. Richard II 4.1.181–3 (Jacobi) The trochaic interpretation of the third foot causes am to be unstressed, but then the subject pronoun I receives stress. In this more ceremonial setting Jacobi deploys iambic rhythm for the first two lines, but, his character still torn between dignified resignation and bitter sorrow at his usurpation, he departs from iambic rhythm again in the second half of the third line. After which, he intones what amounts to a formal abdication speech in regular pentameter: (18) Now mark me how I will undo myself. I give this heavy weight from off my head, And this unwieldy sceptre from my hand, The pride of kingly sway from out my heart. Richard II 4.1.193–6 (Jacobi) (19) Music do I hear. 11 O, Oh keep time! How sour sweet music is

114


When time is broke and no proportion kept. So is it in the music of men’s lives. And here have I the daintiness of ear To check time broke in a disordered string; But for the concord of my state and time, Had not an ear to hear my true time broke. I wasted time, and now doth time waste me, Richard II 5.5.41–9 (Jacobi) Here, and in particular in line 45 ‘here have I ’ there is recourse to iambic rhythm, though not in all lines, as if to underline metrically the semantic content, in particular the reference to broken time and the counterpoint with the strict musical tempo of the lute which is discordant to Richard’s thought processes. Clearly, Jacobi’s choices of when to employ iambic rhythm and when to accentuate specific syllables for rhetorical purposes are far from arbitrary. His interpretation of the lines sets the character apart as he intends for legitimate dramatic reasons, resulting in a definitive and critically acclaimed performance. To add a further example, this time from the 1981 BBC version of Richard III, the interchange between Richard III and Buckingham in Richard III Act 3, scene 7 contrasts an iambically regular delivery by Ron Cook (Richard) with a prose-like delivery from Michael Byrne (Buckingham). The exchange between Cook and Rowena Cooper (Queen Elizabeth) in Act 4, scene 4 sets up a powerful contrast between the smooth-tongued hypocrite Richard (a deceitful user of language to refer to Wright 1983: 156) and the less intelligent and ultimately outmanoeuvred Duke and Queen respectively. Cook rarely deviates from the predominantly iambic rhythm, but again delivers a performance of calculated brilliance. He prefers to ‘ride on the verse’ to great effect, achieving a delivery of which Granville-Barker would, presumably, have approved. It can be seen from these instances that in Cook’s personal interpretation of Shakespearean prosody, subject pronouns can accordingly be stressed in even (ictic) position in the line, whether or not they function in a contrastive sense.

6. Conclusions The findings of our linguistic analysis have already been set out in section 4. They indicate that the option of inverting a subject pronoun around the verb in Shakespeare’s dramatic verse had nothing to do with respecting


115

iambic pentameter. Rather it was a freestanding stylistic variant used by the poet in order to achieve a high style effect, possibly even a monumental quality. The later plays show a decline in the frequency with which this variant is deployed. They display no tendency to favour or avoid placement of a pronoun on an odd or even syllable, depending on whether inversion or non-inversion of the subject pronoun was adopted. Unexpectedly, an early period has been identified of tension between versification and syntax, that of the earlier plays, up to King John. In these, performers will often have to disregard iambic rhythm if they wish to avoid stressing a subject pronoun contrary to sense, that is, where no emphatic or contrasted meaning is intended. The performer seeking to avoid this unwanted effect will be drawn towards a speech rhythm delivery in which the effect of dramatic verse will be diminished, whereas the performance of later plays will very rarely encounter mismatches between speech rhythm and iambic rhythm occasioned by VSpro ictic12. This finding goes against a strict interpretation of the position taken by Wright (1988: notably 220–3), to the effect that it was only in Shakespeare’s late period that he achieved independence of metre and speech rhythm. As regards the syntax of inversion, disruption of iambic rhythm is not a distinctive hallmark of the late plays, but rather of the earlier ones. It would be reasonable to assume, based on the work of Wright and others, that the structural, metrical and characterizational complexities of Shakespeare’s later plays would always pose more problems of stress placement for performance practitioners. Paradoxically, the findings of our study suggest that the high incidence in the earlier plays of inverted subject pronouns in even syllable positions may induce some actors to adopt greater natural speech rhythm strategies, contrasting markedly with those actors whose preference is for more traditional iambic rhythms. More generally, it is to be noted that, although this research was undertaken with the intention of addressing controversies on the status of iambic rhythm in Shakespeare’s dramatic verse, our findings have uncovered further support for the position that he used it as an underlying default scheme. There is clear evidence of his preference for placing pronouns in odd syllables, as well as corroboration of earlier arguments for the presence of iambic rhythm on the basis of his alignment of iambic rhythm and word stress in the case of polysyllabic verbs. Together, these findings constitute strong support for the view that Shakespeare skilfully managed the relationship between syntactic position, rhythm and metre, and in distinct ways, at successive stages of his play-writing career.

116


Notes 1

2

3

4 5

6

7

8

9

10

11

12

We gratefully acknowledge the generous support of the British Academy’s SinoBritish fund in the preparation of this research. In fact, inversion of a subject pronoun, as in (1)b, was not generally the rule in Old English and Early Middle English. It may plausibly be attributed to the influence of Anglo-Norman (Haeberli 2010). Baekken (1998), using the Helsinki corpus, shows that the terminal decline of inversion post-dates 1630. That is, on an even-numbered syllable, cf. Tarlinskaja (1983). We followed this approach on the basis of the first author’s earlier work on inversion in Late Middle English chronicles (Ingham and Grohman 2008), in which the presence of a narrative timeline favours the clause-initial placement of nonsubject constituents, notably temporal expressions. We repeat the point that inversion was not yet an entirely archaic trait in later sixteenth-century English, and not confined to verse. A reviewer suggests that using an oscilloscope for more scientific measurement would give a set of uncategorized gradient-based readings. To categorize the readings into different levels of stress is a task that the human ear and brain can do, but the oscilloscope cannot, especially if the aim is to categorize the readings according to metrical or rhythmic criteria. Objective acoustic measurement methods, in addition to the present subjective ones, would be useful if the present study targeted purely the performance pragmatics of Shakespearean dramatic verse, but this is not our aim. Irons’ interpretation of the role of Antonio in Michael Hoffman’s 2004 film of The Merchant of Venice exemplifies the self-conscious avoidance of iambic stress patterns in contemporary performance and the tendency to simplify the verse by turning it into what is effectively prose. H. Granville-Barker, letter to John Gielgud, cited in The New Penguin Shakespeare, Richard II, p. 9. The shift in rhythm and stress patterning on the second hemistich of the third line of this extract can be said to constitute a ‘choriambic cluster or segment’, as an anonymous reviewer has pointed out; that is a mixture of trochaic and iambic rhythms with an intervening spondee – at least in Jacobi’s interpretation referred to here. Most versions of the text have ‘Ha, ha!’ but Jacobi’s performance employs the variant given here. In fact the 13 examples of VSpro ictic in the Period III plays studied all occur in the first foot of the line, where Shakespeare often uses a trochee anyway, so they will not necessarily be perceived as disrupting the metre.

References William Shakespeare: The Complete Works, Stanley Wells and Gary Taylor, with John Jowett and William Montgomery (eds) 2005 [1986]. Oxford: Clarendon Press.


117

Holinshed (Richard III): www.r3.org/bookcase/holinshed/h-408.htm (Macbeth): www.clicknotes.com/macbeth/Holinshed/ The Shakespeare Collection – BBC Shakespeare series (1978–1985) on DVD. www.en.wikipedia.org/wiki/BBC Television Shakespeare (for online reference) Wallace, David ‘Accents’, SHK 13.1285 posted 9th May 2002: Shaksper: The Global Electronic Shakespeare Conference. Accessed 12 December 2007 www.shaksper.net/archives/2002/1252.html

Abbott, Edwin A. 1870. A Shakespearian Grammar – An Attempt to Illustrate some of the Differences between Elizabethan and Modern English. New York: Dover Editions. Re-issued 2003. Originally published by Macmillan: London. Baekken, Bjørg 1998. Word Order Patterns in Early Modern English. Oslo: Novus Press. Baekken, Bjørg 2000. ‘Inversion in Early Modern English’, English Studies 81(5): 393–421. Culpeper, Jonathan 2001. Language and Characterisation: People in Plays and Other Texts. Harlow: Pearson. Halle, Morris and Samuel Jay Keyser 1971. ‘Illustration and defence of a theory of the iambic pentameter’, College English 33(2): 154–76. Hope, Jonathan 1994. The Authorship of Shakespeare’s Plays: A Sociolinguistic Study. Cambridge: Cambridge University Press. Houston, John Porter 1988. Shakespearean Sentences: A Study in Style and Syntax. Baton Rouge: Louisiana State University Press. Tarlinskaja, Marina 1983. ‘Evolution of Shakespeare’s metrical style’, Poetics: Journal of Empirical Research on Literature, the Media and the Arts 12.6: 567–87. Wells, Stanley 1969. ‘ Introduction’, Stanley Wells (ed.) Richard II. London: Penguin. Wright, George T. 1983. ‘The play of phrase and line in Shakespeare’s iambic pentameter’, Shakespeare Quarterly 34(2), Folger Shakespeare Library, 148–58. Wright, George T. 1988. Shakespeare’s Metrical Art. Berkeley, Los Angeles and London: University of California Press.

Further reading Groves, Peter L. 1998. Strange Music: The Metre of the English Heroic Line. ELS Monograph Series 74. Victoria, British Columbia: University of Victoria. Haeberli, Eric 2010. ‘Investigating Anglo-Norman influence on Late middle English syntax’, in: Richard Ingham (ed.), The Anglo-Norman Language and its Contexts. Woodbridge: Boydell & Brewer, 143–63.

118


Ingham, Richard and Kleanthes K. Grohman 2008. ‘On the post-finite misagreement phenomenon in Late Middle English’, in: Marina Dossena, Richard Dury, and Maurizio Gotti (eds), English Historical Linguistics 2006 (ICEHL) 14. Volume I: Syntax and Morphology (Bergamo, 21–25 August 2006). Amsterdam: John Benjamins Publishing, 125–40.

Chapter 6

Shakespeare’s ‘Short’ Pentameters and the Rhythms of Dramatic Verse Peter Groves

1. The ‘short’ pentameter Just before the much anticipated joust in Richard II, the Marshal abruptly calls a halt to the proceedings: ‘Stay, the King hath throwne his Warder downe’ (1.3.118). Because this line has only nine syllables it has worried many of Shakespeare’s editors, who have generally assumed that metricality in the pentameter requires a standard – or at any rate a minimum – number of ten syllables, and who for three centuries have been conscientiously tidying up his versification, anxious to defend him against the traditional charges of artlessness and lawlessness.1 Eighteenth-century editors simply added words ad libitum to fill in the apparent gaps: they ‘repaired’ this line, for example, by arbitrarily inserting a ‘But’ before ‘Stay’. Later editors have usually been more cautious, and have tried to remove apparent gaps by relineation where possible. Where it is not, a note of desperation sometimes creeps in: the Arden 2 editor of Richard II, for example, suggests pronouncing the Q1 spelling throwen to ‘regularize the metre’ of the line (Ure 1961: 28n) thus creating a highly irregular sequence of three successive reversed feet (a reversed foot – traditionally a ‘trochee’ – is one where the beat or ictus precedes rather than follows the off-beat or nonictus). More recent editors have tended to pay less attention to the metre, but have (oddly) retained many previous relineations designed solely to remove apparent gaps (see Groves 2007). What I wish to suggest in this chapter is that many examples of syllabic deficit in the early texts of Shakespeare represent not compositorial botching, metrical licence or authorial inadvertence, but rather choices within the signifying system of the metre, a way of ‘pointing’ performance for actors (and readers) through the use of what I term ‘short’ pentameters, or lines that have the requisite five beats but fewer than ten syllables because they

120


include silent off-beats or silent beats.2 Take the Marshal’s headless line: the surprise of the silent initial off-beat – performatively a little like treading in the dark on a step that isn’t there – arises from our perception that the first position should (by default) be an off-beat. We experience the same surprise in the case of an initial reversal or trochee: with the headless line, however, the surprise is prolonged as the metre fails to re-stabilize itself with the familiar POM-titty-POM or ‘O for a Muse’ opening. This silent off-beat draws attention to Richard’s coup de théâtre and mimetically it conveys the abruptness of the Marshal’s injunction. I am not, of course, the first to suggest that Shakespeare’s metrical gaps or lacunae might represent a rhythmical device rather than a lapse or a licence – see, for example, Sicherman (1982, 1984) and Wright (1988) – but there has been little attempt to theorize the phenomena in terms either of English phonology or of a coherent theory of metricality. For Sicherman, for example, they simply represent ‘metrical pauses’ (1982: 175) which may occur in succession, so that ‘He’s tane’ becomes a pentameter when followed by eight metrical pauses (1984: 185), as does ‘No, my Lord’ when preceded by seven (1984: 193). In the interests of theoretical rigour I shall use a system of phonologically based prosodic and metrical analysis which I call Base and Template scansion, which can distinguish metrical and unmetrical lines; it is set forth fully in Groves (1998), but I will need to sketch in some salient features here. It begins from the fact that metre is an organization of speech, and that English speech is itself structured (and, in part, timed) around periodic recurrences of articulatory intensity called ‘beats’: they are not directly features of sound but rather of the muscular organization of articulation, sympathetically perceived by the listener, and distinct both from lexical stress, the inherent prominence that obtains in at least one syllable of every lexical or ‘dictionary’ word, and accent, the pitch-slide that marks contrast in speech (‘ON the fridge, not IN the fridge!’). Those who accompany their speech with gestures, for example, will tend to time those gestures on the beats. Beats occur most frequently on lexically stressed syllables, but not on all: we say /three blind /mice, not /three /blind /mice (unless we want to emphasize both the number and the blindness). They can, moreover, fall upon unstressed syllables to break up a long run of them, and (since they are not a form of sound, but of muscular activity, a means of organizing utterances in time), they may (unlike stress and accent) fall (and be perceived) upon silence, like a rest in music. The silent beat, represented as ‘’, distinguishes (for example) between the restrictive and non-restrictive relative clauses in a pair like the following: ‘I /spoke to the /waiter who /brought my /soup’

Shakespeare’s ‘Short’ Pentameters

121

(bringing precedes speaking); ‘I /spoke to the /waiter, / who /brought my /soup’ (bringing follows speaking).3 As Derek Attridge (1982) has suggested, silent beats are a typical structural feature of chantable English demotic verse – nursery rhymes, protest chants and so on – where they represent a beat-long pause, a ‘rest’, in the chant. It is easy to clap on the eight beats of an appropriate performance of the following, for example, including on the final rest: /What do we /want? /Ten per cent! /When do we /want it? /Now! / A once well-known football chant goes: /Georgie /Best! /Georgie /Best! /Georgie! / /Georgie /Best! Metrical structure in English is produced through the placement of beats in the spoken line, and this is enabled and limited by three prosodic phenomena: by the disposition of lexical and syntactic stress, by the location of syntactic junctures, and by the speaker’s (contextually motivated) placing of pragmatic accent (used for pointing contrast or highlighting information) within the utterance. To simplify somewhat the argument of an entire book, an iambic pentameter is an utterance or part-utterance that can be accommodated (by rule-governed elisions where necessary) to five feet (pairs of syllable-positions), each of which contains at least one syllable capable of carrying a beat, normally the second syllable of the foot. A syllable is prevented from carrying a beat when it is ‘dominated’ by a fullystressed neighbour within the same intonational phrase; a syllable not so dominated may be called ‘independent’; in ‘my apartment’, for example, fully stressed (and thus independent) -part- dominates both a- and -ment but leaves unstressed my independent. An independent syllable is thus one capable of carrying a beat, and it must be either (a) a fully stressed syllable, (b) an accented syllable, or (c) an unstressed syllable that is not dominated by a neighbouring stressed syllable (i.e. it is protected by an intervening syntactic break or liberated by accent). Where the independent occurs on the first syllable we say the foot is reversed, and enclose it in angle brackets: ‘<Wanton> as youth|ful goats|, <wild as> young bulls|’ (1 Henry IV 4.1.103); each reversal must be followed by a normal foot. Syncopated syllables are printed in superscript to suggest that they should be pronounced quickly and lightly: ‘<Myriads> of ri|vulets hu|rrying through| the lawn|’.

122


If (to illustrate) I enclose dominated syllables in brackets and bold all independent ones, the following line can be shown to have five normal feet (or iambs), because ‘by’, though unstressed, is not dominated: ‘(I ùn|der) stánd| (your méa|ning) by| (your éye|)’. As George Gascoigne, the earliest English metrist, points out, this verse ‘may passe the musters’ metrically, whereas the following re-arrangement unmetrical, or ‘neyther true nor pleasant’ (1575: 295); this is because the second pair of syllables has no independent and cannot, therefore, function as a foot: ‘(Your méaning) (I ùnder)(stánd by) (your éye)’. The phonological conditions for silent off-beats and silent beats may be simply stated: a silent off-beat can only occur either initially or in a position that is flanked by fully independent syllables, and a silent beat can only occur at the site of an obligatory intonational break (normally marked by terminal punctuation). The adjacency of beats that in metre may generate the perception of a silent off-beat is something that English generally tries to avoid (three successive stresses in the same phrase will typically produce only two beats (/three blind /mice). Adjacent beats will be produced in one of two ways: by the occurrence of an intonational phrase-break between two stressed syllables (/Run, /Spot!) or by the contiguity of pragmatically accented syllables within a phrase (Not /white /coffee but /green /tea), and each of these produces a distinct kind of silent off-beat, which I term respectively the ‘jolt’ (^) and the ‘drag’ (~).

2. The jolt A silent off-beat that is mapped onto a potential or obligatory intonational break will accentuate that break (where – in general – regular metre by its insistent continuity tends to bridge and smooth over such gaps), and this is why I call it a ‘jolt’. It can have the expressive effect of suggesting urgency and spontaneity, most commonly at the beginning of the line, where it often underscores abrupt attention-seeking imperatives and vocatives: ^ Goe|, take hence| that Tray|tor from| our sight|, ^ Come| my Lord|, Ile leade| you to| your Tent|. ^ Proue| it Hen|ry, and| thou shalt| be King. ^ Gen|tlemen|, impor|tune me| no far|{ther}, ^ Iay|lor, take| him to| thy cus|todie. ^ Ti|tus, I| am come| to talke| with thee|,

(2H6 2.3.100) (1H4 5.4.9) (3H6 1.1.131) (SHR 1.1.48) (ERR 1.1.155) (TIT 5.2.16)


123

Such headless lines constitute the most common kind of short pentameter, and do not represent an innovation on Shakespeare’s part. He probably learned the trick of buttonholing vocative headlessness from Marlowe (105): ^ Bar|barous| and bloo|dy Tam | burlaine |, ^ Trea|cherous| and false| Theri | damas |, ^ Bloo|dy and| insa|tiate Tam | burlain |.

(1TAM 2.7.1) (1TAM 2.7.3) (1TAM 2.7.11)

Shakespeare’s innovation consists in extending the possibility of lacuna to every position in the line (though there are precedents for this in Wyatt – see Groves (2005) – which Shakespeare may have been familiar with). A jolt, for example, may draw attention to imperatives and vocatives within the line: my Batt|lements|. ^ Come| you Spi|{rits}, (MAC 1.5.40) Good Mar|garet|[,] ^ runne| thee to| the par|{lour}, (ADO 3.1.1) Good Ma|dam pa|tïence|. What say| you? | Strikes him (ANT 2.5.63)


129

Less violently, a beat may be signalled by a touch: the two following are marked by a touch on the shoulder, from the king’s sword or the bailiff’s hand: king john:

downe Phi|lip, | but rise| more great|, Arise Sir Richard, and Plantagenet. (JN 1.1.161–2) antonio: <But be| of com|fort. second officer: | Come sir| away|. (TN 3.4.338) A piece of stage-business may be cued by a rest: Bassanio’s opening of the fateful casket or the flourishing of a prop: portia:

I feele too much thy blessing, make it lesse, For feare| I sur|feit. bassanio: | What finde| I here|? (MV 3.2.113–4) sicinius: Haue you a Catalogue Of all the Voices that we haue procur’d, → <set downe| by’th Pole|? aedile: I haue|:’tis rea|dy |. (COR. 3.3.10) jeweller: And rich|: a Wa|ter looke| ye |. (TIM 1.1.18) The rest may cue a gestural pointing to or indication of some object or person: england: And there|upon| your Daugh|ter |. (H5 5.2.347) lear: . . . Death on my state: wherefore <Should he| sit he|ere |? This act| perswades| {me}, (LR 2.3.112–3) hamlet: My fa|ther, | me thinkes| I see| my fa|{ther}. (HAM 1.2.184) Horatio’s alarm (‘Oh where, my Lord?’) makes more sense if the rest in Hamlet’s line cues a gesturing, pointing or turning towards the imaginary father. A second function of the silent beat is to cue an expressive bodily gesture that represents a character’s response or emotion. Take Polonius’ ‘you’re-a-man-of-the-world, work-it-out-for-yourself’ shrug as he ekes out


130

his list of suggested ‘slips’ on Laertes’ part to Reynaldo, who has tentatively suggested ‘gaming’: polonius: ^ I|, or drink|ing, fenc|ing, swear|ing, | ling, dra|bbing. You| may goe| so farre|. (HAM. 2.1.25) It can equally cue a mock-shrug: King Henry’s sarcastically simulated bafflement (this silent beat was first pointed out by George Wright 1988: 181) or Flavius’ mimicry of the embarrassed prevarications of Timon’s ungrateful parasites: worcester: I haue not sought the day of this dislike. king: |, Something hath beene amisse; a Noble Nature May catch a wrench; would all were well; tis pitty, (TIM 2.2.204–9) Other possible gestures include the shudder or grimace: ophelia:

As if he had been loosed out of hell, To speake| of ho|rrors: | he comes| before| {me}. (HAM 2.1.80–1) claudius: Oh my offence is ranke, it smels to heauen, It hath the primall eldest curse vpon’t, A Bro|thers mur|ther. | I not|, (HAM 3.3.36–8) The sigh: portia: I must go in. Aye me! How weake a thing The heart| of wo|man is|? O Bru|tus, | The Heauens speede thee in thine enterprize! (JC 2.4.40) aegeon:

In Syracusa was I borne, and wedde Vnto a woman, happy but for me, And by| me; | | | Dismisse| {me}? (OTH 4.3.14) Or a gesture of exasperation: hotspur:

Sicke now? droope now? this sicknes doth infect The very Life-blood of our Enterprise, He writes me here, that inward sicknesse, | And that his friends by deputation (1H4 4.1.27–8/30–1) othello: And he that is approu’d in this offence, ... Shall loose| me. | <What[,] in> a Towne| of warre|, ... To Manage priuate, and domesticke Quarrell? (OTH 2.3.211/213/215) othello: The Handkerchiefe. desdemona: A man that all his time Hath founded his good Fortunes on your loue; Shar’d dan|gers with| you. othello: | The Hand|kerchiefe|. (OTH 3.4.93–6) A silent beat may cue a theatrical registration of dismay or alarm at being suddenly at a loss for an answer. One delightful example arises when


132

Cleopatra, ‘wrinkled deep in time’ (1.5.29), confronts her messenger (still smarting from their earlier interview) with a dangerous request: cleopatra: The Fellow ha’s good iudgment. charmian: Excellent. cleopatra: her yeares|, I pry|thee. messenger: Ma|dam, |→ she was a widdow. cleopatra: Widdow? Charmian, hearke. (ANT 3.3.25–7) He is caught in a cleft stick: if he makes her too young, he risks another beating, but if he makes her too old for the ‘ne’er lust-weary’d Antony’ she will think he’s lying. His momentary alarm, registered in the facial gesture that signals the silent beat, is effaced by an inspired answer, both satisfying and non-committal. Some further examples of embarrassment: olivia:

But would you vndertake another suite, I had rather heare you, to solicit that, Then Mu|sicke from| the spheares|. viola: Deere La|dy |. (TN 3.1.108–10) clarence: Who sent you hither? Wherefore do you come? second murderer: To, to|, to – | clarence: To mur|ther me|? both murderers: I, I|. clarence: You scarcely haue the hearts to tell me so, (R3 1.4.171–5) George Wright (1988: 181) has drawn attention to a line (printed as two fragments in most editions) in which Iago just hints a fault, and hesitates dislike of Cassio: othello: iago:

, my Lord|? (OTH 3.3.103)

Bob Hoskins, in the BBC Othello, takes advantage of the intimacy of television to half-raise a sceptical eyebrow on the third beat. In addition to these mimetic and expressive applications of the silent beat, we also find a couple of interesting deictic uses, marking and drawing attention to features of the interaction between characters. One common


133

one, for example, is to mark the end of a question, perhaps with a conventional head-tilt or some similar item of body language: king richard: What sayes| he |? northumberland: Nay no|thing; all| is said|: (R2 2.1.148–9) king richard: <Shall I| obtaine| it |? bolingbroke: faire Cou|{sin}. (R2 4.1.304) gloucester: <Would he| deny| his Le|tter, said| he |? (LR 2.1.78) Another kind of rest is the interruption, where the silent beat involves the speaker breaking off in mid-sentence, as when the Jeweller breaks into the Osric-like verbosity of the Merchant’s praise of Timon with an intrusive gesture such as showing the jewel, or Gertrude interrupts her son’s intolerable tirade about Claudius by (perhaps) throwing up her hands: merchant: A most incomparable man, breath’d, as it were, To an vntyreable and continuate goodnesse: He pa|sses. [―] jeweller:

| I haue| a Ie|well heere|.7

(TIM 1.1.10–2)

hamlet: . . . A Cutpurse of the Empire and the Rule. That from a shelfe, the precious Diadem stole, And put| it in| his Po|cket. queen:

| No more|!

(HAM 3.4.99–101)

And then there’s the absolutely final interruption: hotspur: . . . O, I could Prophesie, But that the Earth, and the cold hand of death, Lyes on my Tongue: → No Pe|rcy thou| art dust| / And food| for – |8

(1H4 5.4.83–6)

Like the jolt, the rest can deictically underscore an imperative, presumably with a supporting gesture. In the following, York perhaps points to or lays a hand on the desired dagger; in King Richard’s case, the first injunction ‘leave me’, marked by a jolt, doesn’t seem to work, so the repetition is accompanied by an intensifying silent beat (perhaps an impatient wave of the hand or a thump on the table): york:

I pray| you, Vn|ckle, | thìs Da|{gger}. (R3 3.1.110) king richard: ^ Set| it downe|. Is Inke| and Pa|per rea|{dy}?


134

ratcliff: It is| my Lord|. king richard: Guard watch|. ^ Leaue| {me}. Ratcliffe, → about the mid of night come to my Tent And helpe| to arme| me. | I say|. (R3 5.3.75–9) ghost: Lord Has|tings: | dispaire|, and dye|. (R3 5.3.156) Another deictic use of the rest is in turning towards a new interlocutor (perhaps cuing an actual turning of head or body): bolingbroke: Thankes gen|tle Vn|ckle: | | For your| sake (Ie|{well}) I am glad at soule, I haue no other Child, (OTH 1.3.193–6) Sometimes exits are punctuated by a silent beat: martius: Your va|lour puts| well forth|: Pray fo|llow. |[Exeunt] (COR 1.1.251) imogen: I am bound| to you|. belarius: And shalt| be e|ver. | [Exit Imogen] (CYM 4.2.46) brutus: Let’s . . . carry with us Eares and Eyes for th’ time, But hearts| for the| euent|. sicinius: you. | [Exeunt] (COR 2.1.268–70) As are entrances: lady:

How my good name? or to report of you <What I> shall thinke| is good|[?] The Prin|cesse |. [Enter Imogen] (CYM 2.3.85) coriolanus: Shall I be tempted to infringe my vow
Shakespeare’s ‘Short’ Pentameters cassio:

135

I humbly thanke you for’t. I neuer knew A Flo|rentine| more kinde|, and ho|nest. | [Enter Emilia] (OTH 3.1.39–40)

An interesting variant is when the entrance-rest coincides with interruption of the speaker. Macbeth doesn’t want his wife to find him contemplating tergiversation, any more than Iago wants his general to find him gossiping about him, and so the words side and Desdemona are not just cut off by the entrances but replaced by gestural responses of surprise or alarm: macbeth:

Vaulting Ambition, which ore-leapes it selfe, And falles| on th’ o|ther |. [Enter Lady.] How now|? What Newes|? (MAC 1.7.27–9) cassio: I do not vnderstand. iago: He’s married. cassio: To who? iago: <Marry> to – | [Re-enter Othello] Come Cap|taine, will| you go|? (OTH 1.2.52–3) To illustrate the short pentameter at work in an extended passage consider the following, in which Emilia hears for the first time of her husband’s slander of Desdemona. It contains four occurrences of the phrase ‘my husband’, all of which are placed by their prosodic and pragmatic contexts in different relations to the metre. The first (l.138) is simply a puzzled query (‘Why mention him? What has he got to do with all this?’); the second (l.142) repeats that query after a silent beat that signifies a kind of stunned double take (the penny is beginning to drop). The third utterance (l.145) has a drag, cueing an incredulous, horrified accent on my; by the fourth (l.148) accentuation has returned to normal as she desperately seeks a contradiction from Othello (whose exasperation provides the jolts in this line): othello: . . . thy Husband knew it all. emilia: My Hus|{band}? othello: Thy Hus|{band}. emilia:
Notes 1 2

3

4

The process begins as early as the First Folio itself (see O’Connor 1977). It is true that metrical and prosodic variations cannot mean directly, but they can certainly ‘seem an Eccho to the Sense’ (Pope, Essay on Criticism 365). In what follows I shall use to indicate a silent beat, ^ to indicate a jolt (a silent off-beat that falls in a syntactic break), ~ for the drag (a silent off-beat that comes between words not separated by a syntactic break), | to indicate the end of a foot, to enclose a reversed foot, [] to enclose a swap (see below), { to indicate a ‘feminine ending’, superscript letters (e) to indicate syncopation or elision, and → to indicate where a metrical line-ending is for some reason (such as compositorial convenience) not registered in F. This foot has two independents and so may be realized either as a normal or as a reversed foot, which is the meaning of the < | notation.

Shakespeare’s ‘Short’ Pentameters 5

6

7

8

137

Compare ‘That croakes| the fa|tall en|t[e]rance| of Dun|{can}’, Macbeth 1.5.39. Four out of the nineteen verse-occurrences of entrance in Shakespeare are trisyllabic. Macbeth 2.1.5–7 as it appears in F1 needs relineation; since Rowe, however, editors have routinely (and mistakenly) relineated the entire sequence 2.1.2–7. Most editions print a full stop after passes, as in F1, but while the verb could be used intransitively to mean ‘excels’ this leaves the silent beat unexplained (and in any case ‘He passes.’ seems far too laconic for this speaker when he could say something like ‘He passes the extolment of all virtues, / And . . . ’). Editors follow F1 in placing ‘No, Percy, thou art dust’ after ‘Lies on my tongue’, and stranding ‘and food for’ – as a fragment. My arrangement arguably makes better sense, with an expressive pause after ‘Lies on my tongue’ and the grunt or death-rattle expressed as the final rest of Hotspur’s final line.

References Quotations are from Hinman, Charlton (ed.) 1968. The Norton Facsimile: The First Folio of Shakespeare. New York: W. W. Norton & Co., Inc., except where otherwise indicated; where a line is shared between speakers, all segments of that line after the first are indented, as in modern edited texts, in order to clarify the metrical structure of the whole line. Line-numberings refer to Evans, G. Blakemore and J. J. M. Tobin (eds) 1974. The Riverside Shakespeare. Boston: Houghton Mifflin Company. Attridge, Derek 1982. The Rhythms of English Poetry. London: Longman. Bowers, Fredson 1980. ‘Establishing Shakespeare’s text: notes on short lines and the problem of verse division’, Studies in Bibliography 33: 74–130. Furness, Howard H. (ed.) 1873. Macbeth, vol. 2 of A New Variorum Edition of Shakespeare. 3rd edition. London & Philadelphia: J. B Lippincott Co. Gascoigne, George 1575. ‘Certayne notes of instruction concerning the making of verse or ryme in English’ in: The Posies of George Gascoigne Esquire. London: For Richard Smith. Groves, Peter L. 1998. Strange Music: The Metre of the English Heroic Line. ELS Monograph Series 74. Victoria, B.C.: University of Victoria. Groves, Peter L. 2005. ‘Finding his feet: Wyatt and the founding of English metre’, Versification: An Electronic Journal of Literary Prosody 4 (www.arsversificandi.net/current/groves.html) Groves, Peter L. 2007. ‘Shakespeare’s pentameter and the end of editing’, Shakespeare (Journal of the British Shakespeare Association) 3: 126–42. Marlowe, Christopher 1973. The Complete Works of Christopher Marlowe, 2 vols, Fredson Bowers (ed.), vol. 1. Cambridge: Cambridge University Press. O’Connor, John 1977. ‘A qualitative analysis of compositors C and D in the Shakespeare First Folio’, Studies in Bibliography 30: 57–74. Sicherman, Carol 1982. ‘Meter and meaning in Shakespeare’, Language and Style 15: 169–92.

138


Sicherman, Carol 1984. ‘Short lines and interpretation: the case of Julius Caesar’, Shakespeare Quarterly 35: 180–95. Ure, Peter (ed.) 1961. King Richard II, 5th edition. The Arden Shakespeare, London: Methuen & Co. Ltd. Wells, Stanley and Gary Taylor, with John Jowett and William Montgomery (eds) 1986. William Shakespeare: The Complete Works. Oxford: Clarendon Press. Wright, George 1988, Shakespeare’s Metrical Art. Berkeley: University of California Press.

Further reading Groves, Peter L. 2001. ‘ “What music lies in the cold print”: Larkin’s experimental metric’. Style 35: 703–23.

Chapter 7

Wholes and Holes in the Study of Shakespeare’s Wordplay Dirk Delabastita1

1. Introduction Nineteenth-century philologists felt confident enough to put an exact number on Shakespeare’s puns (in 1887 Dr F. A. Bather counted exactly 1,062 of them) and to contain their often anarchic abundance within strictly systematic taxonomies (a game at which Leopold Wurth’s 1895 study still remains to be beaten). More recently the study of puns has moved a very long way indeed from restrictive positions like these. Some of the new critical paths have led as far as the Derrida-inspired type of position which holds wordplay to be a principle of unbridled plurality and free linguistic play. Regardless of whether one agrees with them or not, the historical reality and indeed the intellectual interest of such radically opposed critical positions need to be acknowledged. It is easy enough to dismiss them, from your own vantage point, as the fallacies of a naively self-confident past or those of a perversely over-theorized present. But we make our work irrelevant to the historical reality of how many of us are now interpreting and how we used to interpret Shakespeare’s wordplay if our theoretical models of the pun cannot somehow accommodate positions as wide apart as those of Dr Bather and modern poststructuralism. In such a spirit of comprehensiveness, the approach to Shakespeare’s wordplay advocated in this chapter aims to be wide-ranging, flexible and multiperspectival. We shall propose a whole set of critical distinctions that need to be considered in the discussion of puns, and, crucially, we shall argue that these are gradual rather than radical distinctions. No less crucially, our approach is historical and thus fairly relativistic. Among other points, we shall emphasize the historicity of interpretation, since, after all, it is our interpretative strategies that make or mar the puns that could potentially be attributed to Shakespeare’s texts.

140


2. Defining wordplay I am offering my definition as a comprehensive ‘whole’ that covers various types of wordplay and links them meaningfully and systematically to various types of non-wordplay. But, far from being totalitarian or deterministic, the model will also provide for the existence of the many ‘holes’ – gaps, questions, indeterminacies, doubts – that keep appearing when we try to understand and conceptualize Shakespeare’s punning and its history. In their legitimate search for descriptively rigorous and theoretically coherent frameworks or ‘wholes’, linguistic approaches to the pun have tended to turn a blind eye to such problematic hermeneutic ‘holes’. Conversely, in its acute and equally legitimate sensitivity to the latter, postmodern literary criticism has tended to forget about the importance of the former. If all goes well, the integrated model presented below should amount to a ‘whole’ showing the ‘holes’ in their right perspective. Our basic descriptive parameters derive from the following definition of wordplay. I suggest that wordplay is best seen as an umbrella term covering the various discursive phenomena in which certain features inherent in the structure of the language(s)2 used are mobilised to produce a communicatively significant, (near-)simultaneous confrontation of at least two linguistic units with more or less dissimilar meanings and more or less similar forms. As in many definitions of this type, the wording is dense, technical and inelegant, but a few examples will help to clarify the definition’s key points and implications. The following short fragment from the punning match in scene 2.4 of Romeo and Juliet provides us with a suitable starting point: (1) benvolio Stop there, stop there! mercutio Thou desirest me to stop in my tale against the hair. (Romeo and Juliet 2.4.93–5)3 There are actually three distinct puns here working together to produce the joke: (1a) stop: s1 ‘discontinue’ stop in: s2 ‘put in, insert (in order to close or plug an opening)’ (1b) tale: s1 ‘account, story’ tail: s2 ‘penis’ (1c) against the hair: s1 (figuratively) ‘against my inclination or sentiment’, ‘against the grain’ against the hair: s2 (literally) ‘against the (female pubic) hair’.

Wholes and Holes in the Study of Shakespeare’s Wordplay

141

We shall now in the following paragraphs revisit the different conditions stipulated in the definition above and apply them one by one to this example, beginning perhaps with the issue of communicative significance, which in the case of this integrated sequence of puns is perfectly obvious. Their function is quite evidently to sustain Benvolio’s and Mercutio’s bawdy banter. The puns must have been intended by the author. They are also meant to be seen as deliberate on the part of the characters within the fictional world of the play, which is not always or necessarily the case. Sometimes a pun is intended by the author ‘over the head’ of the character who delivers it blissfully unaware of the double meaning, which may or may not be picked up by the other characters, either immediately or later. This produces an effect known as dramatic irony. In their own, less subtle way malapropian puns such as in example (7) below fall under this category too. There may thus be different embedded levels of intentionality and understanding which can produce a range of ironic effects. Let us continue the exemplification of the definition. The respective units involved in sequence (1) are formally identical in the cases of puns (1a) and (1c). They are not identical but highly similar in the case of (1b), where the components ‘tale’ and ‘tail’ show identical sounds but different spellings. In this respect pun (1b) presents an instance of what is called homophony, which is traditionally opposed to homography (same spelling but different sound), homonymy (same sound, same spelling) and paronymy (both sound and spelling show certain differences against a background of similarities). Needless to say, the exact significance of these distinctions requires further scrutiny when we consider semiliterate cultures, let alone oral ones, or when we attend to linguistic situations where both pronunciation and spelling are subject to change and variation, as happened to be the case in Shakespeare’s England. The formal similarities in each of the three puns serve to highlight the very different meanings contrasted in each of them. Witness the semantic distance between the respective meanings indicated as ‘s1’ and ‘s2’ in the glosses under (1a), (1b) and (1c) above. But what really matters of course is the cumulative effect they have in producing two very different readings of the same sound sequence [stɔp ɪn maɪ tɛɪl]. The more or less striking combination of different meanings contained within the same or nearly the same forms is widely seen as the main distinctive feature of wordplay. The confrontation of forms and meanings can be simultaneous, as in (1b) and (1c), where the two meanings are superimposed, one on top of the other so to speak, within a single formal sequence. Using a spatial metaphor, such wordplay may be called vertical wordplay. Invoking the Saussurean

142


distinction between paradigmatic and syntagmatic relations in language, we might also call it paradigmatic wordplay.4 Pun (1a), on the other hand, presents the formal constituents of the pun in a near-simultaneous manner. There are different occurrences of the word ‘stop’ involving a shift of meaning; they come at the heels of each other rather than appearing together. While the pun components appear consecutively, they have to do so within a textual span that is sufficiently short to still enable their cognitive linkage to take place. This may be called horizontal or syntagmatic wordplay.5 The linking of forms and meanings brought about by the pun is made possible by the mobilization of certain linguistic principles, alone or in combination. Thus, pun (1a) exploits the polysemy of the verb ‘to stop’, with ‘stop in’ possibly being on its way to becoming a phrasal verb. Pun (1c), on the other hand, presents a literal, compositional reading (s2) of an idiom (‘against the hair’), which is set off against the meaning that the idiom has as the indivisible unit which it normally is (s1). A comparable example of an idiom is ‘to fall in love’, which we understand as a single phrase with a meaning of its own which is different from the sum total of the meanings of its components, but when Julia asks Lucetta in The Two Gentlemen of Verona if she would ‘counsel [her] to fall in love’, she receives a punning reply which invokes the literal sense of the idiom: ‘Ay, madam, so you stumble not unheedfully’ (1.2.2–3). Linguists would argue that polysemy and idiom-formation are closely related, both being lexical-semantic phenomena. Pun (1b) depends on phonological structure and that is an altogether different level of language. More exactly, pun (1b) owes its existence to the fact that English, like any other language, operates with a limited number of phonemes, which can moreover occur in certain combinations only. Very similar or even identical phonological pairs such as ‘tale’ and ‘tail’ are therefore bound to occur, despite their being totally independent words in terms of meaning and etymology. At yet another level of language structure, puns may furthermore exploit the loopholes of grammar. Some possible uses of morphology and syntax are illustrated below by examples (2) and (3) respectively. In the case of morphology-based wordplay, words deriving from compounding or derivation are decomposed or analysed in an ‘incorrect’ manner, as in: (2) antipholus of syracuse Where stood Belgia, the Netherlands? dromio of syracuse O, sir, I did not look so low. (The Comedy of Errors 3.2.137–8) s1: name of a country s2: the nether regions (of her body).


143

Syntax-based puns involve phrases that can lend themselves to two possible parsings, as in the following double-edged prophecy: (3) The duke yet lives that Henry shall depose, s1: Henry will depose a duke s2: a duke will depose Henry.

(2 Henry VI 1.4.30)

These linguistic mechanisms which are creatively exploited by punsters are known by all language users inasmuch as we need them to produce verbal utterances and to understand those produced by others. But they normally operate under the radar of our consciousness. It is only in metalingual statements – in language about language, for example, grammar books or comments such as these – that we become fully aware of phenomena such as grammatical ambiguity, polysemy or phonemic structure. However subtly, wordplay too could be said to raise our awareness of the mechanisms of language and of the verbal materiality of the specific phrase in question. In other words, wordplay endows language with a certain level of metalingual information. It causes language to bend back upon itself, reminding us that meaning comes from language and that languages have their own stubborn ways of putting things. Last but not least, for all their dependence on the structural properties of the language system as recorded in grammars and dictionaries, puns are discursive phenomena. This means that puns exist as language-in-action. They occur in concrete utterances, and their meanings and effects depend crucially on their interplay with many specific factors – the surrounding text, genre conventions, pragmatic rules and contexts, intertextual networks, ideologies, psychological dispositions and so on – which condition meaning-formation at both the point of production and the point of reception. The fact that the two points of production and reception are so far removed from each other with older writers such as Shakespeare makes it doubly important to adopt a strictly historical and context-sensitive approach.

3. Areas of transition As we have just suggested, there are no context-free puns floating free in abstracto. Puns exist in discourses, or they don’t exist at all. This part at least of the definition given above leaves no room for doubt or disagreement. But for all the other criteria that make up the definition, it is possible and indeed unavoidable to describe puns in terms of degrees to which they meet its several conditions. In this way, wordplay turns out to be gradually

144


different from various sets of other, more or less related, discursive phenomena, rather than constituting a totally autonomous and homogeneous domain. In this sense we could see wordplay and its various types in terms of prototypical categories. This fundamental fact is worth emphasizing, as it has all too often been obscured by the tendency of rhetoricians and linguists alike to pigeonhole wordplay and its various possible textual manifestations in systems of discrete categories and neat distinctions.6 Take the following question: how far can we stretch the ‘near’ in the condition ‘(near-)simultaneous confrontation of at least two linguistic units’ before the breaking point is reached and the components of a horizontal pun have drifted too far apart for their ‘dissimilar meanings and similar forms’ to still refer back to each other? Cases such as pun (1a) above, or such as Pistol’s notorious (4) To England will I steal, and there I’ll steal;

(Henry V 5.1.86)

present us with textbook examples of horizontal wordplay; the proximity of their respective pun components plays an important part in making their confrontation strikingly direct and effective.7 Now compare these puns with the following example taken from M. M. Mahood’s (1957) classic study of Shakespeare’s wordplay. Mahood refers to the end of Macbeth’s great monologue in Act 1 scene 7: (5a) . . . I have no spur To prick the sides of my intent, but only Vaulting ambition, which o’erleaps itself And falls on th’other – (Macbeth 1.7.25–8) Mahood argues that ‘[t]he final image from horsemanship is so vivid that it makes possible a kind of long-distance pun in Macbeth’s words after the murder’ (144): (5b) The wine of life is drawn, and the mere lees Is left this vault to brag of. (2.3.93–4) There may be a vertical pun in the last line on ‘vault’ as meaning ‘winecellar’ or ‘sky’, but Mahood’s point here specifically concerns the possibility of horizontal wordplay linking the ‘vault’ in this line to the much earlier use of ‘vaulting’ as ‘leaping’ in line 1.7.27. This reading is very attractive. But then, there is no denying that the solid 292 lines which separate the


145

‘vaulting ambition’ (5a) from ‘this vault to brag of’ (5b) create a very long textual distance indeed – the kind of distance which can only be spanned by a reading strategy that is prepared to stretch the notion of horizontal wordplay from an inch narrow to an ell broad. Mere textual distance is not all that counts, though. See the following instance: (6) The more shame for ye. Holy men I thought ye, Upon my soul, two reverend cardinal virtues – But cardinal sins and hollow hearts I fear ye. Mend’em for shame, my lords. . . (Henry VIII 3.1.102–5) Queen Katherine applies the word ‘cardinal’ both to the ecclesiastic status of the men she is addressing and to their virtuousness or rather lack thereof. Her speech also contains a horizontal pun on the words ‘holy’ and ‘hollow’ and this is what we are now concerned with. The components of the pun are separated by 15 intervening words and roughly the equivalent of two lines. This is much less than the 292 lines in Mahood’s example above, but surely more than in the average horizontal pun. Moreover, the actual sound similarity between the pun components ‘holy’ and ‘hollow’ is far from being complete. The pun’s strong semantic effect, then, is largely due to the complex and tight structural organization of the entire passage, which places both words in strongly equivalent positions. Consider the following features: the adjectival character of ‘holy’ and ‘hollow’ (both belong to the same word-class); the semantic and syntactic parallelism of ‘holy men’ and ‘hollow hearts’ (as well as of ‘cardinal virtues’ and ‘cardinal sins’); the syntactic and metrical parallelism of ‘holy men I thought ye’ and ‘hollow hearts I fear ye’, and the similar relative position of each word group in its respective line; and, finally, the chiastic construction of the entire passage (shame-holy-cardinal versus cardinal-hollow-shame). Critical attention given to such structural patterning is easily tarred with the brush of ‘formalism’ these days, but such attention seems vital to understand the operation of many a pun. If examples (4) (steal/steal), (6) (hollow/holy) and, more controversially, (5) (vaulting/vault) show instances of horizontal wordplay, we cannot be blind to their mutual differences. Distances between the pun components in this set of examples range widely between three or four intervening words to 292 lines; we sense that when the distance gets too long, the horizontal pun is liable to dissolve into non-punning discourse; but we have also seen how grammatical patterning may play a crucial part in foregrounding

146


a punning connection. In other words, distinctions between puns and nonpuns are gradual and careful contextualization is necessary. I contend that what is true in these examples for the distances between punning components in horizontal wordplay applies across the board to the other distinctive features in the definition above, and thus to wordplay generally. It is worth spelling out the major implications of this argument for what is the ultimate condition in the definition, namely the communicative significance of puns. As we have seen in example (5), the distance between pun components in a horizontal wordplay can be measured quite accurately. But it is an altogether more delicate interpretative manoeuvre to assess whether and to what extent the repetition of ‘vault’ in Macbeth would have struck Shakespeare himself and/or the people around him as being in any way punningly significant. It might be safest to assume that the jury is still out on this one, as well as in the case of many hundreds of other potential puns in Shakespeare. The jury would at best comprise sensitive and intelligent critics of the calibre of William Empson, M. M. Mahood, Stephen Booth or Patricia Parker. But, to continue the legal metaphor, I would argue that very often there is simply not enough hard evidence to settle the case for or against wordplay, leaving us somewhere in the zone of incertitude with different shadings of grey between the obvious ‘puns’ and the obvious ‘non puns’. Explicit metalingual signals such as ‘Can sick men play so nicely with their names?’ (Richard II 2.1.84) will as a rule leave little doubt as to the communicative pertinence of puns. But then, such overt cuing is comparatively rare. Horizontal puns such as illustrated by examples (4) and (6) also tend to belong to the category of fairly ‘obvious’ puns inasmuch as they advertise themselves and their rhetorical intent by the formal repetitions involved in them.8 Other ‘unusual’ features in a text besides repetition may similarly acquire signal value and help us recognize a meaningful wordplay when we see one. Such telltale features would include idiosyncratic spellings or pronunciations, strange collocations, apparent problems with textual cohesion and so on. For example, the reply of Dromio of Syracuse in fragment (2) above introduces what appears to be a non sequitur until, after a split second and much to our gratification, we realize that ‘Netherlands’ lends itself to a double reading that perfectly matches the double context of geography and sexual anatomy. Similarly, along with the wider dramatic setting, it is the logical-semantic and collocational incongruity of Bottom’s famous malapropism (7) Thisbe, the flowers of odious savours sweet – (A Midsummer Night’s Dream 3.1.77)


147

that alerts us to the underlying punning substitution of ‘odious’ for ‘odorous’. However, not all textual anomalies have sufficient signal value to guarantee the level of communicative significance we tend to require of wordplay. Sometimes an anomaly in the text turns out to be just an infelicity of expression or perhaps some error in the production or transmission of the text, with no ulterior significance attached to it. And sometimes seeming anomalies or incongruities may be interpreted either way, that is as either having signalling value or not. It is worth recalling here that Shakespearean text editors have hit upon several instances of textual variants being formally similar and contextually appropriate, quite in the manner of vertical wordplay. In such cases one may then wonder if the textual hesitations in the transmission of the play could be regarded as traces of deliberate wordplay. Or is the textual crux no more than that – an accident in the copying and printing history of the text? A notorious example of this type of problem is Hamlet’s (8) O that this too too ?sullied ?solid ?sallied flesh would melt, (Hamlet 1.2.129) in which ‘sullied’, ‘solid’ and ‘sallied’ are not only textual variants vying for editorial validation; following Mahood’s suggestion (1957), we could also consider the possibility that all the versions of the phrase may participate in the complex semantics of Hamlet’s famous monologue, no longer as mere textual variants of course, but as intrinsic components of a subtle but artistically effective ‘subdued form of wordplay’ (Mahood: 15) or ‘portmanteau word’ (16). In the final analysis such critical hesitations between Shakespearean ‘puns’ and ‘textual cruxes’ are not very numerous. Even though they may constitute a subset of it, they are certainly very far indeed from making up the largest category of potential wordplay where the question of communicative significance is problematically at stake. That troublesome category consists mainly of the multitude of potential vertical puns that come without explicit metalingual cues (compare with the quote from Richard II above), that are monological (compare with example (1) from Romeo and Juliet, where the pun serves as a pivot in a dialogical exchange) and that are not prompted by any particular semantic, cohesive, collocational, stylistic or other anomaly in the text (compare with Bottom’s blatant verbal blundering in example (7) above), so that, all things considered, the double meanings come as a semantic surplus to a textual sequence that already appears quite well-formed and meaningful without it. Let us look at one example

148


that could stand for many. This is Othello speaking, expressing his trust in Iago: (9)

Why did I marry? This honest creature doubtless Sees and knows more – much more – than he unfolds. (Othello 3.3.245–7)

Patricia Parker (1996: 236–7) relates this passage to the wider context of the language that was used to refer to the sexual parts of women: The privity or lap itself was understood as something folded and hence needing to be unfolded in order to be available for show—a sense of unfolding exploited . . . in Othello’s suspicion that his informer “Sees and knows more, much more, than he unfolds”. Without actually using the term pun or wordplay, Parker suggests that the semantics of ‘unfold’ in example (9) may be unpacked as follows: s1: (figuratively) reveal, disclose (a secret) s2: (literally) open, display (a woman’s private parts). The critic embeds this reading within an intriguing network of intertextual references – taken from the same play and from other plays (especially Hamlet) but also from a range of anatomical and other contemporary discourses outside the Shakespeare canon – all of which seem to agree in conceptualizing the female genitals as something hidden, closed and secret, to be opened, dilated and displayed. True, the convergence of all these discursive contexts as they are made to crowd in on the above fragment from Othello lends considerable credibility to the pun that Parker reads in ‘unfold’. But when one (re)situates the phrase in the play’s immediate logical, grammatical and dramatic context rather than in the persuasive momentum of Parker’s erudite and sophisticated criticism, one may find the case for s2 somewhat less convincing. Isn’t Othello’s phrase perfectly meaningful and coherent – linguistically, dramatically, situationally and psychologically – if one simply confines the meaning of ‘unfold’ to s1? Doesn’t perhaps projecting s2 onto the phrase create more problems than it solves? How to account for the mid-sentence shift from the naive and sentimental indefiniteness of Othello’s belief in Iago’s innate goodness and tactful discretion (‘this honest creature’) to the lurid anatomical precision of a reference to vaginal dilation? And which woman is it anyway that Iago would have sexually ‘unfolded’ to Othello to the dissatisfaction


149

of the latter, who would want to see ‘much more’ of it? When would such a scene have taken place? Some readers at least will be left wondering if this passage in Shakespeare’s play is really elucidated by the references to the discursive networks that prompt s2 in Parker’s reading. Could it be that the elucidation rather happens in the opposite direction, with the virtual but unrealized ambiguity of the word ‘unfold’ in the fragment being used to give substance and credence to the critic’s argument about discourses and conceptualizations of female sexuality in Early Modern England? My intention in asking these questions is not per se to challenge Parker’s reading of this passage. Rather, I have wanted to illustrate how potential wordplay of this type – vertical, monological, lacking explicit metalingual cues and occurring in contexts that show no special textual anomalies whatever that would have required resolution by some double meaning to be unearthed – is bound to be more vulnerable to critical doubt and controversy. That critical decisions about the significance of wordplay are taken along a spectrum of possibilities and increasing probabilities rather than in terms of simple ‘either/or’ positions is already indicated by the subtle discriminations we find in the phraseology of our best wordplay critics. To quote a short sample of variously hedged pronouncements on the communicative strength of puns, Mahood speaks of ‘subtle and subdued forms of poetic wordplay’ (12–3), ‘unconscious wordplay’ (13), punning which is ‘unintentional’ (17) or ‘not wholly unintentional’ (21), ‘submerged puns’ (23), ‘unspoken puns, whether conscious or unconscious’ (25), puns ‘too dependent on the reader’s fickle responsiveness to be counted and catalogued’ (51), ‘concealed pun[s]’ (81), and so forth and so on. Empson in his Seven Types of Ambiguity identifies ‘latent puns’ and Stephen Booth (1977) finds a range of ‘ideational puns’ in the Sonnets, as well as many ‘overtones’ and ‘echoes’ (adding that, when he uses words like these, ‘I mean no more than I say and have no intention of denying or offering a substitute for the immediately available sense that most of the sonnets make’, 369). Patricia Parker, too, in her Shakespeare from the Margins (1996) has recourse to a wide range of subtly modulated phrases to acknowledge that not all double meanings in Shakespeare will strike us as unambiguously pertinent because they may not involve sufficiently ‘direct resonances’ or ‘acknowledged subtext[s]’ (10). Indeed, very often the talk in Parker’s book is of mere ‘resonances’, ‘affinities’ and ‘echoes’ (many of which are heard ‘sotto voce’), or of ‘unwitting echoes’ (59), ‘marginal allusions’ (75), ‘unnoticed interstitial or marginal links’ (187), and so on. Taking care not to overplay her critical cards, Parker identifies many of the Shakespearean puns she has spotted in the rather uncommittal language of semantic effects becoming ‘part of a suggestibility’ (3) and not ‘necessarily

150


enabling a particular deciphering’ (15). Critical circumspection may thus entail a circumlocutionary approach allowing puns to play hide-and-seek with the reader, as on page 252, where a phrase is said to ‘hid[e] a language that has lurking within it’ other meanings (my emphasis). The various other conditions included in our definition need to be taken on board, because the question of the significance of potential wordplay is predicated on them, but it is this last issue that is ultimately both the most decisive and also the most elusive one. We shall therefore have to return to it once more at the end of this chapter, where a further complication will emerge. But meanwhile, the argument about the need for multiple continuums in defining and identifying wordplay should be clear. Much of this argument may be visualized by the following diagram: polyptoton and other figures of word repetition

semantic dissimilarity

rhyme

irony speech-act ambiguity

alliteration assonance consonance jingle and other figures of sound repetition

formal similarity

WORD PLAY

allusion

based on language structure

communicatively significant

allegory metaphor metonymy referential ambiguity referential vagueness

accidental ambiguities, unintentional sound echoes, Freudian slips... single-reading sentences

Figure 7.1 Defining wordplay at a crossroads of four continuums

I am emphatically not arguing that rhymes, word repetitions, irony, accidental ambiguities, etc. have to be regarded as instances of wordplay, only that there tend to be grey zones between the prototypically clear cases of wordplay in the centre and the various clear instances of non-wordplay on the border. This is so because the distinctive features summed up in the definition of wordplay (in the diagram printed in italics) operate in a gradual and context-sensitive manner. Thus, on the side of ‘formal similarity’, the diagram accommodates the notion that wordplay may in some instances become hard to distinguish from rhyme, assonance and the like. On the


151

side of ‘semantic dissimilarity’, the diagram reflects our experience that Shakespeare sometimes plays with semantic nuances which may become so subtle as to be barely distinguishable. And on the side of ‘communicative significance’ it acknowledges the difficulties inherent in assessing the relevance of perceived potential wordplay. All of this should be familiar ground by now, but on the side of ‘language structure’ the diagram’s insistence on the idea of a gradual scale of possibilities challenges what was for a very long time, and may still be for some linguists, a basic tenet, backed by the authority of Saussure’s programmatic separation of langue and parole, namely that a firm line can and should be drawn between language structure on the one hand and language use on the other. To be sure, there are many convincing examples on either side of the distinction which suggest that such a clear separation makes perfect sense. We have already indicated how the threefold pun in sample (1) above depends quite decisively on linguistic structure. At the other end of the spectrum, one finds countless instances of ‘doubleness’ of meaning whose special effect owes nothing to linguistic structure as such. One such example is the Third Apparition’s riddling prophecy in Macbeth: (10) Macbeth shall never vanquish’d be, until Great Birnam wood to high Dunsinane hill Shall come against him. (4.1.92–4) s1: Great Birnam Wood s2: soldiers having camouflaged themselves with boughs cut in Great Birnam Wood. For all its ambiguity, this phrase indisputably exceeds our definition of wordplay, as the Apparition’s equivocation hinges on a misleadingly original metonymic extension of the word ‘wood’ and the name derived from it (‘Great Birnam Wood’), and certainly not on the intrinsic linguistic properties of that word or name itself. Or consider the famous metaphor in the fourth line of Sonnet 73: (11) That time of year thou mayst in me behold, When yellow leaves, or none, or few do hang Upon those boughs which shake against the cold, Bare ruined choirs where late the sweet birds sang; s1: (vehicle) choir as that part of a church where the service is sung s2: (tenor) branches of a tree.

152


Needless to say, there is neither polysemy, nor any other lexical, grammatical or phonological mechanism at work to generate the ‘ambiguity’ between s1 and s2. The latter does not depend on language structure and therefore the ‘ambiguity’ cannot qualify as a pun. We are dealing with a quite unique metaphorical blend which results from creative language use.9 The ambiguities in examples (10) and (11) result from metonymy and metaphor respectively and it would be misleading to describe them as wordplays. The picture that emerges from clear examples such as these connects wordplay quite specifically with language structure (of which it cleverly exploits certain intrinsic features, such as polysemy, etc.), while the effects of phenomena such as metaphor or metonymy are situated in the realm of language use ‘only’ (since, beyond the general productivity of metonymic or metaphorical association, no specific linguistic structures are triggered into operation to achieve the interpretative hesitation). It appears that this binary logic works well in a good many cases, and certainly well enough to satisfy the taxonomically minded. But historical semantics has of course demonstrated one or two fundamental facts about polysemy that spoil the binary neatness of the picture. Polysemy as an intrinsic feature of lexemic structure usually results from meaning shifts such as metaphorical or metonymic extension, and these shifts tend to start as more or less occasional or creative instances of language use before progressively becoming more common and being integrated into the structures of the lexicon. This process of lexicalization is given something like an ‘official’ confirmation when the word in question is declared to be polysemous by lexicographers and starts appearing in dictionaries as a complex entry with its diverse senses spelt out distinctly. This is not necessarily the final stage, as in some cases the process of semantic differentiation will go further and lead to the point where linguistic consciousness will reinterpret (historical) polysemes as (semantically unrelated) homonyms. Thus, we treat the words ‘pupil’ (ward, disciple, child receiving tuition) and ‘pupil’ (circular opening at the centre of the iris in the eye) as unrelated homonyms, whereas they can be traced back historically, via French and Latin, to a common polysemous root. In fact, Shakespeare may well have intended a slight pun on ‘pupil’ in Romeo and Juliet: (12) romeo Thou chid’st me oft for loving Rosaline. friar laurence For doting, not for loving, pupil mine. (2.3.77–8) s1: disciple s2: opening in eye; (literally but also figuratively) apple of the (my) eye.


153

The main context that would support the secondary sense s2 would be Friar Laurence’s preceding speech, which enlarges quite elaborately on the all too common mistake of young men to love ‘in their eyes’ rather than ‘truly in their hearts’ (line 64). Another example would be Shakespeare’s more frequent and also more obvious puns on ‘metal’ and ‘mettle’, two words that have even come to be differentiated by diverging spellings but which developed from a common lexical root via metaphorical extension. In cases like these we should beware of underplaying the earlier sense of semantic connectedness in Early Modern English between such latter-day homonyms or homophones, reading the semantic realities of Shakespeare’s generation through the lexical grid of early twenty-first-century English. The crucial point about these lexical processes is precisely that they are processes occurring along the gradual distinction between language use and language structure rather than happening overnight at the flick of some switch. Without wishing to collapse the distinction, we have to conclude therefore that there is some essential continuity between phenomena such as creative metaphors or metonymies, on the one hand, and wordplay, on the other. This continuity was noted by critics such as William Empson (1984 [1930]), partly because they were hardly interested in the orthodoxies and classificatory niceties of linguistic description anyway, but also because they had a keen intuitive sense of the deeper cognitive effects of language. Empson’s beautifully cadenced comment on Shakespeare’s ‘bare ruined choirs’ metaphor – see (11) above – is worth quoting here: the comparison holds for many reasons; because ruined monastery choirs are places in which to sing, because they involve sitting in a row, because they are made of wood, are carved into knots and so forth, because they used to be surrounded by a sheltering building crystallised out of the likeness of a forest, and coloured with stained glass and painting like flowers and leaves, because they are now abandoned by all but the great walls coloured like the skies of winter, because the cold and Narcissistic charm suggested by choir-boys suits well with Shakespeare’s feeling for the object of the Sonnets, and for various sociological and historical reasons (the protestant destruction of monasteries; fear of puritanism), which it would be hard now to trace out in their proportions; these reasons, and many more relating the simile to its place in the Sonnet, must all combine to give the line its beauty, and there is a sort of ambiguity in not knowing which of them to hold most clearly in mind. (Empson: 2–3) For Empson meaning-formation depends on a combination of many factors over and beyond dictionary meanings, as auditive and visual sensations

154


become part of the semantic equation as well as a range of contexts and ‘encyclopaedic’ (e.g. biographical, historical, political) frames. As a result the process of understanding can hardly be other than complex, dynamic and ultimately characterized by a degree of doubt and hesitation; therefore, and also as a result of sheer historical distance (‘hard now to trace out in their proportions’), it may at times defeat the critic’s best efforts. Cognitive linguistics has done much to enhance our understanding of semantic evolution and polysemy, emphasizing how the multiple meanings of words – and thus the use of polysemy in wordplay – presuppose complex forms of cultural or ‘encyclopaedic’ knowledge and not just the features of a supposedly autonomous ‘lexical’ structure. It could even be argued that ‘monosemy and polysemy cannot be definitively separated – like all categories, these also have fuzzy boundaries’ (John Taylor, paraphrased in Crane 2001: 28). The full implication of this may be that our constant recourse to that most valuable of lexicographical resources – the OED – may not quite suffice for an historical understanding of Shakespeare’s words and wordplay. We also need to look at Shakespearean words in the many other discursive contexts in which they occur in order to have a better grasp of what is at stake beyond their dictionary definition when Shakespeare puts them to punning use: who else besides the Bard was using them; how often; in which senses; within which contexts and institutional settings; with which range of social or cultural implications, etc.? Such a cognitively inspired attention to social and cultural meanings conveyed by words and puns characterizes Mary Thomas Crane’s Shakespeare’s Brain. Reading with Cognitive Theory (2001). Without the explicitly cognitive framework, it is also a key feature of Patricia Parker’s work on wordplay, which firmly broadens language into discourse and discursive contexts, and therefore routinely oversteps the descriptive perimeter of the OED – curtly described as ‘useful in certain instances, in spite of its obvious biases and critical omissions’ (Parker 1996: 17) – to consider intertextual semantic networks that emerge in language use. Needless to say this type of research is facilitated and further encouraged by the increased availability of large varieties of Early Modern English texts in linguistic corpora permitting context-sensitive searches.

4. Historical variations The model presented here with its multiple descriptive scales helps us see both differences and continuities between various forms of ‘wordplay’ and various forms of ‘non-wordplay’. Crucially, the model may also heighten our awareness of historical change and help us in modelling it, since it is


155

along the descriptive axes we have identified that various types of shifts may occur which have the effect of obscuring wordplay or of changing its impact for later generations of readers and spectators. Thus, phonetic changes – some of them associated with the final stages of the Great Vowel Shift – may play a large part in blocking our access to Shakespearean wordplay. Classic examples include the names Ajax (Love’s Labour’s Lost) and Jaques (As You Like It); the phonetic resemblance of both with ‘jakes’ (privy, latrine) enabled bawdy jokes to be made which are no longer directly available to our generation. Here is another textbook example: (13) [jaques] And so from hour to hour, we ripe, and ripe, And then from hour to hour, we rot, and rot, And thereby hangs a tale . . . (As You Like It 2.7.26–8) Not until Helge Kökeritz’s Shakespeare’s Pronunciation (1953: 58–9) did it become clear why these lame pseudo-philosophical contemplations would have induced Jaques to ‘laugh, sans intermission, / An hour by his dial’ (lines 32–3). It turns out that the passage involves a bawdy subtext involving puns on ‘hour’/‘whore’, ‘ripe’ (meaning either ‘ripen’ or ‘search’), ‘rot’/‘rut’ and ‘tale’/‘tail’. A crucial key to unlock this sustained double entendre lies in the phonetic likeness of ‘hour’ and ‘whore’, which has subsequently been weakened by sound evolution beyond unprompted recognition. The student of the Shakespearean pun will therefore benefit from having frequent recourse to works such as Helge Kökeritz’s aforementioned book or Fausto Cercignani’s Shakespeare’s Works and Elizabethan Pronunciation (1981). Phonetic dictionaries or philological footnotes will admittedly be of little use to theatregoers. This could be a reason to perform Shakespeare today in Elizabethan English pronunciation, or as close a modern approximation of it as we can manage. Coached by linguist David Crystal, who prepared phonetic scripts for the actors, the London Globe theatre did experiments precisely along those lines with Romeo and Juliet and Troilus and Cressida in their 2003 and 2004 seasons, and one of the interesting side effects of this exercise in historical couleur locale was indeed to bring some of lost rhymes and puns to life again. The effect of semantic change upon Shakespeare’s wordplay is even more decisive than that of phonetic evolution. In an article on the puns in The Comedy of Errors and Twelfth Night which are ‘now falling on deaf ears’ as a result of linguistic (mainly semantic) evolution, Gillian West (1990: 6) suggests that the scale of the loss is sufficient to ‘cause the reader some consternation’. This encouraged West to extend her research and publish her Dictionary of

156


Shakespeare’s Semantic Wordplay in 1998. Ambiguities and subtleties in Early Modern English which are no longer directly understandable today fill up a large proportion of the indispensable annotations in our critical editions, while semantic change over the past four centuries has also created a market for reference works such as David and Ben Crystal’s Shakespeare’s Words (2002) and Eugene F. Shewmaker’s Shakespeare’s Language (2008 [1996]). None of these works, however, suffices and neither does the OED, to cover the lexical underbelly of Early Modern English – that enormous field of bawdy slang and allusion, which always features prominently in punning. The semantics of Shakespeare’s taboo language is explored in time-honoured studies such as Eric Partridge’s Shakespeare’s Bawdy (1968 [1947]) and E. A. M. Colman’s Dramatic Use of Bawdy in Shakespeare (1974). More recent guides, which generally show even fewer inhibitions towards sexual taboo, include Frankie Rubinstein’s Dictionary of Shakespeare’s Sexual Puns and their Significance (1984), Gordon Williams’ Dictionary of Sexual Language and Imagery in Shakespearean and Stuart Literature (1994) and Pauline Kiernan’s juicily titled Filthy Shakespeare: Shakespeare’s Most Outrageous Sexual Puns (2006). These phonetic and semantic issues should always be viewed within a proper historical sociolinguistic perspective. The social values ascribed in our own epoch to certain regional varieties of English, to foreign accents, to the use of certain foreign languages, to the observance of a standard of ‘correct’ English, or to ‘rude’ or ‘obscene’ language must not be projected back onto the sociolinguistic context in which Shakespeare worked. The English language was in a state of rapid flux, much like London itself, which was attracting many foreigners as well as provincials coming from Britain’s various regions, to produce a fairly unregulated Babylonian mix of accents, registers, speech varieties and even foreign tongues. All of this matters for our purpose. For example, in order to build certain puns Shakespeare used regional words and pronunciations that nowadays have dialectal or social connotations attached to them that would not have applied in his age. Similarly, he frequently puns on ‘hair’, ‘heir’ and ‘air’, but it would be a sociolinguistic anachronism to attribute any social significance to such cases of aitch-dropping, as the absence of the word-initial ‘h’ would not have been the marker of social inferiority it was to become in later periods. Historical sociolinguistics and dialectology are therefore relevant disciplines for the student of Shakespearean wordplay, who may already benefit from an older and now largely forgotten work such as Morgan’s Study in the Warwickshire Dialect (1900). Conversely, historical change can also produce puns for a later generation which may not have been intended or meaningful originally. Thus, the semantics of modern English will highlight the anatomical, scatological


157

or sexual meanings in words such as ‘bottom’, ‘excrement’, ‘incontinent’, ‘gay’ or ‘make love’, but such meanings were either absent in Early Modern English or much less prominent than today, so that we have to mistrust these words as historical ‘false friends’ liable to lead us astray in the historical interpretation of wordplays. Finally, the need for an historical approach applies also, at a very fundamental level, to the way in which language generally connects ‘forms’ and ‘meanings’, since, here too, the danger of linguistic anachronism is always just around the corner. We routinely understand Shakespeare through a linguistic prism that involves a fairly clear sense of individual words and their ‘correct’ forms and meanings. This modern awareness of sharply delineated words is the combined result of several developments which started to take full effect after Shakespeare’s age: the growing impact of a print culture with increasingly uniform spellings and printing conventions; rising levels of literacy; the progress of linguistic standardization, leading to clearer distinctions between (and class-bound stratifications of) language varieties and involving a clearer sense of what was ‘good’ and ‘bad usage’; the drive towards greater semantic order, both symbolized and boosted by the work of lexicographers such as Dr Johnson, possibly as a way to reflect and strengthen the striving for social order; the consolidation of all the above by the education system; and so on. But we should try to keep in mind that in the greater fluidity of Shakespeare’s verbal universe, associative linkages could still happen more freely. There was perhaps even a lesser sense that something specific like ‘wordplay’ was taking place, inasmuch as there was a lesser awareness to begin with of the discreteness of the language elements that entered into the punning game. In the terms of our definition above, the ‘confrontation’ of the different words and phrases may thus on the whole have produced a less pointed effect that it might for us today.

5. Interpretative communities As Rosaline reminds us in Love’s Labour’s Lost (5.2.852–4), ‘[a] jest’s prosperity lies in the ear / Of him that hears it, never in the tongue / Of him that makes it’. These words could make us think again about the question of the significance of wordplay and wonder if the notion of a scale of possibilities which we have advocated is actually good enough to get a grip on the wide range of responses to Shakespeare’s punning. Rosaline brings to mind an idea that centuries later became a cornerstone of reader-response theory, namely that meanings – be they single ones, jestingly double ones or hesitations between

158


single and double ones – are always realized at the point of reception, and the full implications of this for the understanding of Shakespeare’s puns are too often overlooked. Yet, their importance already leaps to the eye when one looks at that most democratic and least sophisticated form of literary criticism these days, namely Amazon.com’s online customer reviews. Commenting on Pauline Kiernan’s Filthy Shakespeare, one very appreciative reader applauds the author’s discussion of sexual puns as a ‘serious dramatic device for important issues such as morality, politics, and war’, calling this ‘an important book’ which will be appreciated by playgoers, readers, actors and directors alike (posted by CB on 2 November 2007). A rather less enthusiastic reader begs to differ: the book appears to be well-researched and accurate if you’ve never read any of these plays. Some of the puns are obvious enough, but Kiernan really goes over the top on the rest (mostly taking passages out of context) . . . The book is self-consciously vulgar, attempting to surprise rather than to inform. Sure, Shakespeare had an incredibly filthy mind, but it seems that Kiernan has one far filthier (posted by Enter Pirates on 21 November 2008).10 Both readers are referring to the same text corpus, and it is surely not, or not primarily, the level of their Shakespearean expertise that distinguishes them. Rather, they seem to have read their Shakespeare with very different semantic seismometers, one being super-sensitive and responding eagerly to the slightest hint of a possible sexual vibration below the textual surface, while the other will register nothing short of semantic shocks. The idea of the pertinence of double meanings forming a spectrum between obvious puns and obvious non-puns does not stop making sense, but, clearly, the added complication is that we often fail to agree on how to calibrate the scale. Most critics naturally assume that they occupy the neutral ground between the two extremes of ‘over-reading’ and ‘underreading’. They believe to be alert and receptive to the punning that is really there in Shakespeare, while wisely refraining from historically irresponsible pun-hunting. This self-calibration at the sensible centre shows even in the theoretically refined wordplay criticism of someone like Patricia Parker. Parker lucidly acknowledges an awareness of her own critical position, methodological assumptions and theoretical allegiances, but she also seems quite adamant that her approach to Shakespeare’s puns is a truly historical one and free from anachronistic bias. In her actual readings of Shakespeare she lets the puns mean and sound for themselves: however subtle and sotto voce their resonance may be, the wordplays are ‘already there’, waiting to be picked up by the critic’s finely tuned ear.


159

It is decidedly not my argument that we should give up trying to read Shakespeare’s puns historically and from a maximally neutral or balanced position. The point is rather that for many potential puns no amount of historical attention lavished on them will enable us to settle the question of their significance, because neither the texts themselves nor their contexts provide sufficiently clear information on the matter. The author’s conscious intentions and his/her more hidden or unconscious motives are lost in the mists of time. The same applies to levels of awareness and responsiveness in the contemporaneous readers as well as in the actors on the stage and the spectators in the playhouses. Indeed, given the virtual absence of theatre criticism or other detailed records, the theatrical realization of the potential puns is poorly documented. How were the lines delivered in terms of intonation and body language, and how, if at all, did they become part of the dynamics of the scene?11 Therefore, for the countless puns in the grey zone along the scale of communicative significance, critics are well advised to adopt a due measure of modesty, caution and scepticism, even if these hardly count as strong selling points in the academic marketplace. The point where we run up against the intrinsic limitations of the available evidence is also the point where the decisive importance of our interpretative choices becomes fully visible and where we can therefore no longer avoid addressing the historicity of interpretation. In other words, studying puns historically also has to mean recognizing the historicity of our interpretative strategies. Puns do not simply either exist or fail to exist; they have a history, and that history depends on reading strategies. Certain generations or groups of readers (‘interpretive communities’, as Stanley Fish might say) are more responsive to semantic slippage or doubleness than others, and will more readily rediscover, discover or perhaps invent Shakespearean puns by endowing potential double readings and verbal associations with a semantic substance, a communicative value, a productivity and/or a form of intentionality they did not possess before, perhaps not even in the mind of the original author or of his most immediate audiences. Other interpretative communities may again be much less alive to semantic plurality, if not downright hostile to it. Against the backdrop of such broad patterns of readerly habits and discursive preferences, a single influential critical edition or commentary may then accomplish the feat of initiating or changing the course of a particular pun’s history, for example by securing a place for hitherto undetected double readings in the canon of ‘orthodox’ interpretations, or conversely (and more rarely) by dismissing an accepted ambiguity from that canon as the result of contextually unfounded pun-hunting. To quote one example among hundreds, Helge

160


Kökeritz has to be and is widely credited for the recognition of the pun in example (13) above. The task of self-critically taking on board our own interpretative strategies adds an extra dimension of interest and complexity to discussions of Shakespeare’s wordplay. I would like to conclude by briefly mentioning two arguments that merit further consideration in this context. First, responses to Shakespeare’s puns are bound to show great variety not only if we consider diachronic evolutions in the understanding and appreciation of wordplay, but even if we make the point from a synchronic viewpoint, looking, for example, at Shakespeare’s own age. Even intuitively it seems evident that the potential puns must have worked out very differently for Shakespeare’s original readers (on the one hand) and for those who saw the plays in performance (on the other), with theatrical shows potentially offering a greater density of situational context but also imposing a much faster textual pace than any reading act. For the readers a further difference may perhaps have to be made between those who saw the shows and bought the quarto editions, and the later readers of the 1623 Folio, a book which presents the plays at a further remove from the theatrical environment which had initially engendered them, giving them a somewhat greater autonomy as ‘literary’ texts. If we consider the spectators in the playhouses, they, too, must have shown a wide range of responses, even if we agree to make abstraction of the uniqueness of every single performance and the singularity of each individual’s theatrical experience. Efforts have been made to analyse the composition of Shakespeare’s audiences in terms of their social background and tastes, and then to connect certain types of wordplay with the interests of the subgroups in the audience; witness the commonly heard argument that Shakespeare would have indulged in bawdy punning and knockabout humour to please the palate of the ‘groundlings’, while his more refined poetic puns would have been destined for the educated members occupying the more expensive places. However, such sociological reconstructions of Shakespeare’s audience turn out to be quite problematic, being based on rather unreliable sources and often reflecting the critic’s own political bias (Boecker 2008). What we are left with is the sobering conclusion that any statement on Shakespeare’s ‘original audience’ and on how it would have perceived the wordplay implies a great deal of speculative abstraction. Second, and finally, we need to consider the discursive materiality of our responses to the wordplay in Shakespeare. A reader’s or spectator’s response to a pun may happen with different levels of awareness and of physical and verbal explicitness. It may, or may not, be externalized by a nod of recognition,


161

a smile, a chuckle or a fit of belly-laughter. The acknowledgement of a pun may, or may not, be articulated verbally. When it occurs, such metalingual awareness may be expressed privately in one’s own thoughts or more overtly in some public form. In an absolutely overwhelming majority of cases, our responses remain unreported, but when an understanding of a pun is captured and expressed in some published critical report, the latter is not necessarily an accurate representation of the initial mental reality that it pretends to put into words. There may indeed be a significant gap between the critic’s initial understanding of Shakespeare’s puns and his/her more reflective critical writing on the puns, which is done in accordance with the discursive norms that happen to prevail in the critic’s intellectual and social milieu. Thus, it is too easily assumed that Molly Mahood was not a perceptive reader of Shakespeare’s bawdy wordplay, on which she has admittedly little specific to say. But let us not confuse her critical writing with the intellectual reality of her understanding of the puns. As Mahood (1996–1997: 136) was to indicate in a reappraisal of her Shakespeare’s Wordplay four decades later, ‘I was up against another limitation of time and place: young women tutors, around 1950, did not explicate Shakespeare’s bawdy, which could be left to Eric Partridge’. Wordplay critics today work under totally different strictures. ‘Filthy puns’ stopped being a source of embarrassment years ago and have actually become a strong pole of intellectual and even commercial attraction. This is one of the reasons why it is more likely nowadays for extra puns to be produced by the discursive demands of critical practice than for puns to be suppressed by them (as happened with Mahood). There are more such reasons. Consider the more flexible way in which many wordplay critics have come to define ‘relevant’ contexts for double meanings, extending the sense of context far beyond short individual dramatic and verbal sequences, possibly to include all of Shakespeare’s works and even their biographical, historical and discursive backgrounds; in other words, the lesser extent to which recent wordplay critics feel that the onus is on them to demonstrate the immediate contextual plausibility of a pun. Consider furthermore: their lesser concern with psycholinguistic realities restricting the amount of vertical ambiguity that the mind can handle, or with the limits that our memory span imposes on long-distance horizontal punning; the zest with which critics mingle their self-produced puns with the Shakespearean wordplay under discussion; the critics’ self-granted freedom to overrule the linearity of ‘normal’ readings and performances and to zigzag across fragments, texts and contexts, thereby assuming simultaneities of allusion and reference. Let us accept that these are more than mere interpretative strategies enabling us to find wordplay in Shakespeare, as they

162


can also become, in variable degrees, discursive strategies that generate the puns within the essential context of the critic ’s writing.

Notes 1

2

3 4

5

6

7

8

9

Many thanks are due to the anonymous referees and to Mireille Ravassat for their useful feedback on an earlier version of this chapter. We cannot enlarge here on the special but not uncommon case of multilingual wordplay; for this specific variety of punning, see Parker (2001) and Delabastita (2002, 2005). But it might be noted that here again, as for other distinctions in the definition, the model of a continuum is preferable to that of a binary opposition. Given the fact that the integration of loanwords into the lexicon of a language is a gradual process, wordplay involving loanwords can be said to occur in the grey zone between monolingual and bilingual punning. In this and later Shakespearean quotations, the emphasis is mine. I first encountered the terms vertical and horizontal wordplay in a few interesting German studies on Wortspiel from the 1960s and 1970s (Haussman 1974, Wagenknecht 1965) which time and the language barrier should not allow us to forget. For the rather special case of implicit wordplays, where the confrontation takes place neither paradigmatically nor syntagmatically but in absentia, see Mahood (1957: 24): ‘Sometimes a word, the various meanings of which offer the poet a range of images, itself remains unexpressed’. Booth (1977: 13–4, passim) provides examples from the Sonnets. With the help of the model it is of course possible to describe and compare the various historical, conventionalized typologies of wordplay (e.g. rhetorical ones) one may encounter. In this chapter there is room only for a presentation of the model in its generality. The fact that specific varieties of wordplay were recognized as having a specific form and effect is obviously of major historical significance, but I do not believe that we should necessarily follow the terminology and classifications of classical rhetoric in our scholarly descriptions of wordplay. Classical and Renaissance rhetoricians called this type of wordplay antanaclasis. See also my previous note. Inasmuch as horizontal puns are of the ‘long distance’ variety – as with example (5) – this auto-signalling effect tends to dissolve. This has to do with cognitive limitations inherent in our attention span and retention capacity. There is, moreover, a statistical argument. Repetitions of identical or similar discourse units in a short text segment are likely to stand out and to catch our attention, but inasmuch as we consider larger stretches of discourse, such repetitions are bound to become more frequent and their foregrounding effect will diminish correspondingly. Stephen Booth in his edition of the Sonnets (1977: 259) notes that ‘choirs’ can also refer to ‘groups of singers’, anticipating ‘the sweet birds’ later in the same line. This association, says Booth, ‘gives an extra-logical solidity both to the image and to the very vague assertion that the speaker (a poet, a singer), is like a season’ (ibid.). This effect would qualify as a pun inasmuch as the metonymically based polysemy of ‘choir’ (s1: place in a church where singing takes place; s2: group of singers in a church) became a lexical fact long before Shakespeare had ambiguous recourse to it.

Wholes and Holes in the Study of Shakespeare’s Wordplay 10

11

163

Source: www.amazon.com/Filthy-Shakespeare-Shakespeares-OutrageousSexual/dp/B0014E92O6/ref=sr_1_1?ie=UTF8&s=books&qid=1235772996& sr=1–1 – (accessed 27 February 2009). For more recent performances, the resources of theatre criticism and especially the advent of modern recording technologies have made it increasingly possible to turn to effective interpretative use the fact that situational and pragmatic contexts on the stage perform a crucial part in either realizing potential wordplays or blocking them from the semantics of the performance. We can indeed learn a great deal from studying how various actors and directors have managed to make a scene ‘work’ on the stage with or without the potential wordplays being dramatically activated, and how audiences have responded to this. Such theatrical contexts can and should be ‘read’ as highly valuable ‘glosses’ on potential wordplay; as such they are no less useful than the kind of information one may glean from critical works or annotations in text editions.

References The text used throughout is The Arden Shakespeare: Complete Works (revised edition), edited by Richard Proudfoot, Ann Thompson and David Scott Kastan (London: Thomson Learning, 2001). Bather, Dr F. A. 1887. ‘The puns of Shakespeare’, in: Hawkins, Charles Halford (ed.), Noctes Shaksperianae. Papers by late and present members [of the Winchester College Shakspere Society]. Winchester: Warren & Son, 69–91. Boecker, Bettina 2008. ‘Groundling, gallants, grocers: Shakespeare’s Elizabethan audience and the political agendas of Shakespeare criticism’, in: Dirk Delabastita, Jozef De Vos and Paul Franssen (eds), Shakespeare and European Politics. Newark: University of Delaware Press, 220–33. Booth, Stephen (ed.) 1977. Shakespeare’s sonnets. Edited with analytic commentary by Stephen Booth. New Haven and London: Yale University Press. Cercignani, Fausto 1981. Shakespeare’s Works and Elizabethan Pronunciation. Oxford: Clarendon Press. Colman, E. A. M. 1974. The Dramatic Use of Bawdy in Shakespeare. London: Longman. Crane, Mary Thomas 2001. Shakespeare’s Brain: Reading with Cognitive Theory. Princeton and Oxford: Princeton University Press. Crystal, David and Ben 2002. Shakespeare’s Words: A Glossary & Language Companion. London: Penguin Books. Delabastita, Dirk 2002. ‘A great feast of languages. Shakespeare’s multilingual comedy in: King Henry V and the Translator’, The Translator 8(2): 303–40. Delabastita, Dirk 2005. ‘Cross-language comedy in Shakespeare’, Humor: International Journal of Humor Research 18(2): 161–84. Empson, William 1984 [1930]. Seven Types of Ambiguity. London: the Hogarth Press. Fish, Stanley 1980. Is There A Text in This Class? The Authority of Interpretive Communities. Cambridge, MA: Harvard University Press. Haussman, Franz J. 1974. Studien zu einer Linguistik des Wortspiels. Das Wortspiel im “Canard enchaîné”. Tübingen: Max Niemeyer.

164


Kiernan, Pauline 2006. Filthy Shakespeare: Shakespeare’s Most Outrageous Sexual Puns. New York: Gotham Books. Kökeritz, Helge 1953. Shakespeare’s Pronunciation. New Haven: Yale University Press. Mahood, M. M. 1957. Shakespeare’s Wordplay. London: Methuen. Mahood, M. M. 1996–1997. ‘ “Shakespeare’s wordplay – some reappraisals”: a reply’, Connotations 6(2): 135–7. Morgan, Appleton 1900. A Study in the Warwickshire Dialect, with a Glossary and Notes Touching the Edward the Sixth Grammar Schools and the Elizabethan Pronunciation as Deduced from the Puns in Shakespeare’s Plays, and as to Influences Which may have Shaped the Shakespeare Vocabulary. 4th edition. New York: The Shakespeare Press. Parker, Patricia 1996. Shakespeare from the Margins: Language, Culture, Context. Chicago and London: The University of Chicago Press. Parker, Patricia 2001. ‘The novelty of different tongues: polyglot punning in Shakespeare (and others)’ in: François Laroque and Franck Lessay (eds), Esthétiques de la nouveauté à la Renaissance. Paris: Presses de la Sorbonne Nouvelle, 41–58. Partridge, Eric 1968 [1947]. Shakespeare’s Bawdy: A Literary and Psychological Essay and a Comprehensive Glossary. Revised and enlarged edition. London: Routledge & Kegan Paul. Rubinstein, Frankie 1984. A Dictionary of Shakespeare’s Sexual Puns and their Significance. London and Basingstoke: Macmillan. Shewmaker, Eugene F. 2008 [1996]. Shakespeare’s Language: A Glossary of Unfamiliar Words in His Plays and Poems. 2nd edition. New York: Facts on File. Wagenknecht, Christian J. 1965. Das Wortspiel bei Karl Kraus. Göttingen: Vandenhoeck & Ruprecht. West, Gillian 1990. ‘Lost humour in The Comedy of Errors and Twelfth Night’, English Studies 71(1): 6–15. West, Gillian 1998. A Dictionary of Shakespeare’s Semantic Wordplay. Lewiston, Queenston and Lampeter: The Edwin Mellen Press. Williams, Gordon 1994. Dictionary of Sexual Language and Imagery in Shakespearean and Stuart Literature. London: The Athlone Press. Wurth, Leopold 1895. Das Wortspiel bei Shakspere (Wiener Beiträge zur Englischen Philologie, 1). Wien & Leipzig: Wilhelm Braumüller.

Further reading Delabastita, Dirk 1993. There’s a Double Tongue: An Investigation into the Translation of Shakespeare’s Wordplay, with special reference to Hamlet. Amsterdam and Atlanta: Rodopi. Delabastita, Dirk (ed.) 1997. Traductio: Essays on Punning and Translation. Manchester: St Jerome; Namur: Presses Universitaires de Namur. Lopez, Jeremy 2003. Theatrical Convention and Audience Response in Early Modern Drama. Cambridge University Press. Ravassat, Mireille 2007. ‘Assessing and translating the ambiguities of wordplay in Shakespeare’s Macbeth’, Bulletin de la Société de Stylistique Anglaise 29: 51–60. http://stylistique-anglaise.org/document.php?id=64

Chapter 8

‘a thing inseparate/Divides more wider than the sky and earth’ – of Oxymoron in Shakespeare’s Sonnets1 Mireille Ravassat

Some outstanding scholars have carried out most illuminating investigations into the nature and meaning of oxymoron in Shakespeare’s Sonnets, but no comprehensive and systematic study of this figure of speech and thought in this specific literary work has yet been undertaken. Furthermore, oxymoron is often seen as the signature trope of Petrarchism, which makes Shakespeare’s highly personal use of it all the more intriguing and challenging and the relative neglect of it in his work all the more puzzling. The present chapter therefore intends to demonstrate how oxymoron, albeit far less obtrusive2 in the sequence than the distantly close figures of antithesis and paradox, represents the touchstone for this poetic collection. Indeed the 21 oxymoric paradigms in the Sonnets display the pattern of difference within similarity and similarity within difference3 that has always appealed to Shakespeare as dramatist and poet. Also, such ever-renewed stylistic combinations adumbrate both the baroque vein and the modern vision in this specific body of texts. A ‘biformed’ concept, to take up the term coined by Francis Bacon, informed thinking patterns, rhetorical modes, like the habit of arguing on both sides – argumentum in utramque partem – or the medical tradition founded on contradictory humours, thus highlighting a distinctively Early Modern ‘kind of mental promiscuity in spawning contraries’ (Clark 1997: 47/50).4 The culture of Early Modern England, and of Early Modern Europe as a whole, was pre-eminently a rhetorical one. As Rosalie L. Colie puts it in Paradoxia Epidemica (1966: 304) ‘the world was then a discordia concors, a composition to which oxymoron was the most appropriate figure of rhetoric.’ In that era the hallmark of which was ‘[t]he interaction of contraries’ (Grudin 1979: 14), Shakespeare was born, took up and revived a time-honoured literary genre, the sonnet form, and infused it with his

166


inbred habit of seeing ‘the obverse sides of things’ manifesting an unparalleled ‘ability to give us contrasting things without the slightest diminution of either’ (Hubler 1952: 101/48). Faced with the task of tracing and accounting for the presence of oxymora in the Sonnets, the stylistician’s initial assumption is that such a challenging body of poetry, that discloses as much as it conceals, exhibits Shakespeare’s usual bifocal concern with recapturing an undivided state of wholeness which can only be reached by probing an initial conflation of opposites, at once male and female, weighty and volatile, terrestrial and celestial . . . An apt example is the quotation from Troilus and Cressida (5.2.155–6) which gives its title to the present chapter. Revealing and defining the essence of this very moment when ‘things’ are seen ‘with parted eye, / When everything seems double’, as Hermia says in A Midsummer Night’s Dream (4.1.188–9), is one of Shakespeare’s central issues. ‘Preoccupation with joining’, as Patricia Parker has it, ‘is everywhere in Shakespeare, from the “twain” made “one” of The Phoenix and the Turtle . . . to the explorations of the implications of marriage as “one flesh” . . . ’ (1996: 108–9)5. More often than not, in his dramatic production, Shakespeare probes into such instances of joining versus sundering or rather of joining by means of sundering ‘[f]or nothing can be sole or whole / That has not been rent’, as W. B. Yeats puts it in one of the ‘Crazy Jane’ poems (1987 [1933]: 295). ‘Murd’ring impossibility, to make / What cannot be, slight work’ (Coriolanus 5.3.61–2), Shakespeare tirelessly examines the ‘unity’ of ‘contraries’ (The Rape of Lucrece line 1558) exposing and exploring such rationally improbable phenomena and mechanisms as a miraculous shipwreck in Twelfth Night, the chiaroscuro of the tragic psyche in Macbeth, wise folly and the rhetoric of silence in King Lear, the pilgrimage of hate in Timon of Athens, the problematic of natural art versus artistic nature in The Winter’s Tale, avatars of the black pastoral in Titus Andronicus and Cymbeline, to name a few. Now, it turns out that the Shakespeare of the Sonnets instead devises ‘compounds strange’ (Sonnet 76, line 4), themselves bespeaking a ‘twofold truth’ (Sonnet 41, line 12) – more often than not a cleft one coming short of a self-proclaimed ‘mutual render’ (Sonnet 125, line 12). The Sonnets ultimately leave the protagonists of the love/hate triangle, especially the poet/ young man pair, not so much twinned as twinning in the sense of going asunder. Initially intended fusion ends up in dissolution. Behind the protean persona of the poet-speaker, Shakespeare traces and captures the ‘similitude in dissimilitude, and dissimilitude in similitude’ (Wordsworth 1961 [1904/1778]: 740) inherent to the fluctuating moods and facets of love and desire. Amorous moods, like discourse itself, are

Oxymoron in Shakespeare’s Sonnets

167

‘changeable’ like ‘taffeta’ as Feste has it in Twelfth Night (2.4.74). Thus the sonnets straddle the extremities of erotic experience from exhilaration to utter despondency, from ‘love’s delight’ (36.8) to ‘loath’d delight’ already explored in The Rape of Lucrece (line 742). Like Du Bellay, Shakespeare has nearly ‘forgotten the art of petrarchizing’6, and consequently his poems do not rehearse standard worn-out phrases such as the topos of the lady as ‘[s]weet warriour’ in Spenser’s Amoretti (Sonnet 57.1: 138)7. Shakespeare’s oxymora are no mere adornments or colours of rhetoric ‘to paint the blackest face of woe’, to use Sidney’s own self-reflexive satiric phrase in Astrophel and Stella (1.5: 2). Instead I will argue that the 21 oxymoric paradigms to be found in this puzzling collection can be said to map out the narrative of the fiendishangelical triangle. The structure of the Sonnets could be said to be oxymoric throughout with a ‘present absent’ figure ruling the various intertwined plots. Indeed the three main protagonists of the Sonnets are engaged in an ontological progress bearing the stamp of the carnal and the metaphysical – a process of psychomachia highlighted in Sonnet 144. The young man is a ‘master mistress’ (Sonnet 20, line 2), one among a crowd of ‘pitiful thrivers’ (Sonnet 125, line 8), the equally mysterious female protagonist is a ‘black beauty’ (Sonnet 127, line 3). As for the poet-speaker, plagued by a ‘[b]ifold authority’ (Troilus and Cressida, 5.2.151), he is a ‘willing patient’ (Sonnet 111, line 9) wallowing in masochism. Even the ‘moon’ is ‘mortal’ (Sonnet 107, line 5) symbolizing as it does man’s existential progress under the sign of mutability and transience. Finally the narrative strategy at work in the Sonnets is that of a ‘dialogue of one’ (Magnusson 2006 [2004]: 635). Consequently I will argue that oxymoron could very well provide us with the narrative thread intimately connecting a poetic sequence that might just otherwise appear disconnected..I take the term ‘narrative’ in its most basic sense of ‘telling a story’. For even if the plotline of the Sonnets is definitely a stylized one and we get neither a physical portrayal of the three protagonists nor any precisely delineated environment for their actions, still the poems are no pure philosophical or metaphysical epistles and meditations. This study means to focus more precisely on the process of ‘oxymoronic clashing’ (Jeffries 2010: 125), on the figure of speech enacting a clash of contiguity, what Brian Vickers (1984: 70) referring to antithesis, calls ‘a collision’8 of opposites, ‘sold’[ring] close impossibilities, / And [making] them kiss’ (Timon of Athens 4.3.390–1). Oxymoron is by definition a close encounter, and ultimately a welding together, of apparently irreconcilable opposites admitting no syntactic expansion unlike paradox although it is partly attuned to it, acting on one single object, or referent, whereas antithesis involves an unreconciled juxtaposition of contraries.. As John Porter

168


Houston explains: ‘oxymoron’ is a ‘yoking of contrary adjectives and nouns’ (1983: 313), an only partly accurate definition though, as the ‘present absent’ oxymoron of Sonnet 45 shows. Such a definition as is to be found in Sister Miriam Joseph’s Shakespeare’s Use of the Arts of Language (1949: 135), is more satisfactory. She speaks of ‘a composition of contraries, stimulat[ing] attention by the seeming incompatibility of the terms it unites’. But the major point I wish to emphasize here is that oxymoron in the Sonnets, definitely much more low-key than either paradox, the figure of the subversion of doxa, or antithesis, flagging unresolved antinomic parallels, strikingly enough, in this specific body of texts, violates its ethos of reconciliation in favour of division and estrangement. Very much in a baroque fashion, what oxymoron does, it undoes9 thereby expressing a stance of ontological vacillation, but also the use of this particular figure in the Sonnets testifies to the emergence of a modern self, a ‘modern subjectivity’. Shakespeare’s Sonnets, ‘locating themselves within a literary tradition whose contents and contours they both revise and supplant’ articulate the very modern ‘subjectivity effect in western literary tradition’ that Joel Fineman explored in an eponymous collection of texts published posthumously, and this they manage by demonstrating ‘the intentionality of the literary subject, i.e. his desire . . . motivated by strictly psychological, rather than by biological or by theological, motives’ (1991: 222). Or rather, to qualify Fineman’s statement, Shakespeare’s Sonnets articulate a bifocal, partly organic orientation – itself informed by the tenets of Galenic medical theory – and a distinctively individualized discourse of interiority most aptly mirrored in the self-enclosed units of the sonnet form. *

*

*

In the very first of the series of 17 sonnets, to be notably construed as Shakespeare’s own secular rewriting of the Biblical exhortation to procreation, the structure follows a series of paradoxical and antithetical statements paving the way for the ‘tender churl’ oxymoron of line 12: But thou, contracted to thine own bright eyes, Feed’st thy light’s flame with self-substantial fuel, Making a famine where abundance lies, Thyself thy foe, to thy sweet self too cruel. ... Within thine own bud buriest thy content, And, tender churl, mak’st waste in niggarding. (Sonnet 1, 5–8/11–2) (italics mine)


169

As for line 8 – ‘Thyself thy foe, to thy sweet self too cruel’ – the antithesis enhances the polar nature of the young man’s selfhood. When we come to line 12, the fair friend’s youthful carefree attitude makes him spend, or rather, squander his beauty even as he tries to preserve it. He is termed a ‘tender churl’, in other words a youthful miser and a ‘niggard prodigal’ (The Rape of Lucrece line 79).10 For Joel Fineman the sonnets to the young male dedicatee rehearse the tradition of the poetry of praise, or epideictic poetics, whereas those to the dark lady, an unusually promiscuous and accessible creature, at once emulate and imitate this particular poetic genre making it paradoxical. Fineman construes Sonnet 1 within the framework of ‘Christian and Neo-platonic self-fulfillment and unity’ (1986: 246), but perceives the whole development as a distorted mirror image of a time-honoured orthodox ethos. Indeed the young man, perceived through the synecdochic prism of his ‘own bright eyes’ (line 5) – ‘contracted’ with a pun on betrothed to himself and restricted to the specular orb of his gaze – is definitely an enemy to his best, ‘sweet self’ (line 8), rejecting as he does the exhortation to ‘increase’ (line 1) and perpetuation. Nevertheless he remains the paragon of earthly beauty and as such is worthy of respect and admiration. The young man in the speaker’s discourse turns out to be in chiastic terms an embodiment of ‘pure impiety and impious purity’ (Much Ado About Nothing 4.1.103). *

*

*

In Sonnet 4, the poet harps again on the idea of beauty as a treasure that should be invested for profit and on the fair friend’s mission of the perpetuation of beauty: Then, beauteous niggard, why dost thou abuse The bounteous largesse given thee to give? Profitless usurer, why dost thou use So great a sum of sums, yet canst not live? For having traffic with thyself alone, Thou of thyself thy sweet self dost deceive; (Sonnet 4, 5–10) Here the young dedicatee’s refusal to beget a child is likened to his wasting an inheritance on self-centred pleasures. Sonnet 4 makes use of the vocative vein of indictment that characterizes a good number of the Sonnets.11 The poet-speaker’s tone is that of a ‘homily’ that ‘has been

170


secularized’ (Vendler 1999 [1997]: 62). The two oxymora of lines 5 and 7, ‘beauteous niggard’ and ‘[p]rofitless usurer’, come up in vocative form, which imparts an isomorphic structure to the second quatrain. In the first instance, ‘beauteous niggard’ is to be understood as a miser with a fair appearance, to be perceived as a combination of antithetical terms, but the oxymoric tenor is still more obvious, line 7, with ‘[p]rofitless usurer’. The poet thereby paradoxically designates a moneylender who makes no profit. In other words, the friend uses up, exhausts nature’s ‘bounteous’ loan instead of using it to make either money or children. Through his deliberate choice of non-reproductive sexuality, expressed by the masturbatory activity of line 9, the fair friend cuts himself off from any possibility of a future-oriented identity. Moreover, contrary to the mitigated violence of ‘churl’ thanks to the use of the affectionate epithet ‘tender’ in Sonnet 1, the ‘beauteous niggard’ of Sonnet 4, line 5 testifies to sheer provocation on the part of the older man. As John Blades argues: At first it sounds like an insult, a positive and a negative held in tension. ‘Beauteous’ takes up the flattering metonymy of ‘loveliness’ from the first line but ‘niggard’ drives the poem in a new direction. In the opening line the youth is ‘Unthrifty’, a spendthrift of his beauty, but ‘niggard’ is paradoxically the opposite, a miser. (2007: 8) This riddle could be solved by arguing that the young man indulges in bootless sensual pleasures and is thereby unthrifty, but he is also niggardly in that he privileges carnal indulgence over procreation. This is one among the instances of oxymora in the Sonnets relevant to the itself oxymoric rhetorical strategy of ‘negative politeness’ probed into by Lynne Magnusson (2006 [1999]: 50):. ‘Negative politeness in its extreme forms is a rhetoric of contradictions, for it works in such a way as to simultaneously do and undo the speech actions it undertakes’. This particular self-defeating rhetorical strategy is explicitly reminiscent of the Antony and Cleopatra quote (2.2.215); see note 9. Moreover, by ‘avoiding coercing or controlling the action of the other’ (Magnusson 2006 [1999]: 50), the poet-speaker of the Sonnets allows an ever-yawning chasm to divide his own psyche. *

*

*

Sonnet 125 is the last before one poem addressed to the young dedicatee. It is quite dense both in form and content and implicitly harks back to the


171

procreation poems: Have I not seen dwellers on form and favour Lose all, and more, by paying too much rent, For compound sweet forgoing simple savour, Pitiful thrivers, in their gazing spent? (Sonnet 125, 5–8) It starts by distinguishing the poet from the crowd of those who crave his friend’s love by relying on mere appearances – those are ‘dwellers on form and favour’ (line 5) – but they are bound to fail in their efforts and have to express their admiration for the young man from a distance, a conclusion crystallized in the ‘[p]itiful thrivers’ oxymoron. On the contrary, the poetspeaker’s devotion is loyal and heartfelt as opposed to the ‘nothing-gift of differing multitudes’ (Cymbeline 3.7.58). The oxymoric ‘[p]itiful thrivers’, as Thomas M. Greene explains, ‘take their place in a line of disappointed or misguided would-be thrivers distributed throughout the work’, the ‘onanistic friend’ himself being a case par excellence, which sheds new light on the profitless usury and the beauteous stinginess, amounting to barrenness, of the procreation sonnets (2003 [1985]: 231).12 As Greene further argues, the barrenness of the friend’s attitude, paradoxically ‘[m]aking a famine where abundance lies’ (line 7), is translated from the literal, financial and biological, level to the figurative realm of literary creation: ‘[i]n this econonomic system, all value seems to reside in the friend, or in thoughts of the friend, and the poet seems to be a leaky vessel constantly in need of replenishing, his personal and linguistic poverty never definitely abolished’ (2003 [1985]: 232). In other words, in a global structural perspective, the series of ‘[p]itiful thrivers’, ‘beauteous niggard’, ‘profitless usurer’ oxymora can be interpreted as the threat of both a seminal and semantic bankruptcy precariously, if triumphantly, counterbalanced by the metastylistic assertions of Sonnet 76. As usual Shakespeare’s apprehension of things and events is dual. *

*

*

In Sonnet 8, although a new field of imagery is introduced, as in Sonnets 1 and 4, the young man’s insistently solipsistic stance makes him stand above the fray as a very incarnation of ever ‘jarring-concord and discord-dulcet’ (All’s Well that Ends Well 1.1.172): If the true concord of well-tuned sounds By unions married, do offend thine ear,

172

Stylistics and Shakespeare’s Language They do but sweetly chide thee, who confounds In singleness the parts that thou shouldst bear: Mark how one string, sweet husband to another, Strikes each in each by mutual ordering, Resembling sire, and child, and happy mother, Who all in one, one pleasing note do sing: Whose speechless song being many, seeming one, Sings this to thee: ‘Thou single wilt prove none’. (Sonnet 8, 5–14)

Here, the poet observes the young man listening, Orsino-like, to ‘music sadly’ (line 1) and concludes that the fair friend hears in the instruments’ individual yet united harmony a condemnation for his refusing to take a part in the general concord of ‘sire, and child, and happy mother’ (line 11). The poet is engaged, oxymorically speaking, in ‘sweetly chid[ing]’ (line 7) – that is ‘mellifluously rebuk[ing]’ (Burrow 2002: 396) – the tone-deaf handsome fair friend for his refusal to marry. The apparent simplicity, even naïvety, of the gnomic stance of the closing line finds itself rejuvenated by the presence of the philosophical issue of the Many and the One, the very idea of his Oneness having to merge into Manyness – which in fact amounts to both a mathematical and philosophical impossibility13 – being met with a sharp rebuff by the young man. Almost all the editors of the Sonnets focus attention on frequent instances of deliberate point counterpoint echo between subject and numerology. Paul Edmondson and Stanley Wells (2004: 60) endorse this very idea, thus implicitly emphasizing the overall oxymoric substance of the poem: The numerological perspective of Shakespeare’s sonnets, which can often seem to be an imperative which is proved by degrees of coincidences, here seems utterly appropriate as Sonnet 8, the number of musical notes in a complete octave, makes the sweet and bitter concords and discords of desire attune to the procreation of children. *

*

*

Sonnet 18 strikes a very different note, broaching as it does the central issue of the collection, that is vicarious self-perpetuation by means of literary creation: Sometime too hot the eye of heaven shines, And often is his gold complexion dimmed;


173

And every fair from fair sometime declines, By chance, or nature’s changing course, untrimmed: But thy eternal summer shall not fade, Nor lose possession of that fair thou ow’st, (Sonnet 18, 5–10) The poet thereby solves the dilemma of memory versus oblivion by ‘boldly suggest[ing]’, as Michael Schoenfeldt argues ‘that the climatological impossibility of an eternal summer can in fact be attained in the idealizing verse of the poet’ (2007: 137). However triumphant it may seem, the self-confident closing line of Sonnet 18: ‘So long lives this, and this gives life to thee’, the performative vow of staving off the ravages of Time, finds itself jeopardized by the very next line in the sequence, Sonnet 19, line 1 wherein the threat of the arch destroyer is seen to mangle the lover like a lion. This example among many highlights the artistry of the whole as a progressive contradictory experience characterized by ambiguous resolutions. Similarly, in a typically dialectical fashion, poetry’s best is to be ‘a tomb, / Which hides your life, and shows not half your parts’ (Sonnet 17, lines 3–4) whereas in the following sonnet, the literary monument, or eulogistic tribute, becomes the ‘eternal lines’ that warrant the young man’s immortality, promising to get him out of the grip of death’s ‘shade’ (Sonnet 18, lines 12/11), thereby turning him into an oxymoric ‘boy eternal’ (The Winter’s Tale 1.2.64), as Robert Ellrodt puts it (2007: 34). *

*

*

The experience of the Sonnets is definitely placed under the sign of mutability in very much an Ovidian fashion. Sonnets 35 and 40 should now be considered as a pair, notably because they feature an oxymoric vision of robbery in matters of love: No more be grieved at that which thou hast done; Roses have thorns, and silver fountains mud; Clouds and eclipses stain both moon and sun, And loathsome canker lives in sweetest bud. All men make faults, and even I, in this, Authorizing thy trespass with compare, Myself corrupting, salving thy amiss, ... Thy adverse party is thy advocate, And ‘gainst myself a lawful plea commence:

174

Stylistics and Shakespeare’s Language Such civil war is in my love and hate That I an accessory needs must be To that sweet thief which sourly robs from me. (Sonnet 35, 1–7/10–4) I do forgive thy robb’ry, gentle thief, Although thou steal thee all my poverty; And yet love knows it is a greater grief To bear love’s wrong, than hate’s known injury. Lascivious grace, in whom all ill well shows, Kill me with spites; yet we must not be foes. (Sonnet 40, 9–14)

As for the ‘[s]weet thief’ of Sonnet 99, line 2, it will be dealt with on its own, having a totally different thematic and semantic dimension. Both Sonnets 35 and 40 are poems of indictment, or rather mainly of self-indictment in the case of the former. In Sonnet 35, the poet exculpates the fair friend by making a catalogue of exempla of degraded beautiful natural objects. He then accuses himself of being contaminated through the very process of excusing his beloved’s faults. The frontiers of the poet-speaker’s self have become porous and through a process of mute acquiescence – ‘[l]ike the dyer’s hand to which he himself refers in Sonnet 111 – he takes on the faults of those he loves’ (Dubrow 1987: 212). The ambiance is a gloomy one. The poem shifts from proverbial inspiration in the octave to legal terminology in the sestet, but the dominating image is that of the ‘civil’, or internal, ‘war’ of line 12, ‘an insurrection’ within ‘the state of man’, to take up the phrase from Julius Caesar (2.1.69/67). What has been snatched away from the poet in 35 is more abstract and conceptual than is the case in Sonnet 40. Whereas in Sonnet 35 the poet may have been deprived of his reputation, of his friend’s company, or of both, the theme of Sonnet 40 is the poet’s accusation against the young man for having stolen the poet’s mistress, which becomes still more explicit in Sonnets 41 and 42. Acknowledging his own corruption as he does in Sonnet 35, the poet-speaker states he has become an accomplice – he uses the specific legal term ‘accessory’ (line 13) – in the friend’s sin. As the intestine war metaphor referred to above shows, Sonnet 35 is eminently a poem evincing a bifocal perception of things, as can be seen again in the speaker’s bitter examination of poetic commonplaces, all of them amounting to the poet’s bootless attempts at exonerating himself from blame. As a matter of fact, the whole structure of Sonnet 35 is cleft. It is, in fact, as though – in the course of a ‘dialogue of one’ mentioned earlier – the poetic voice of Sonnet 35


175

underwent a splitting of personality with the speaker of the second quatrain chastising the speaker of the first for misbehaviour, hence again the ‘civil war’ metaphor. The image of civil strife, the sweet-sour, or ‘Mel and Sal ’, opposition (Colie 1974: 68–134) epitomizing the theme of divided selfhood, culminates in what Helen Vendler calls the ‘love-hate voice’ the speaker adopts ‘in speaking of the sweet thief that sourly robs’ (1999 [1997]: 188), a metaphor which crops up again in Sonnet 40. Heather Dubrow comments thus upon this odi et amo trend14: [The poet-speaker’s] rhetorical equivocations stem from psychological ones. . . . [The sequence’s] frequent references to civil war, its repeated allusions to a loss of identity, all suggest a character who is losing touch with his own emotions and with the language intended to express them. . . . It is possible to see much of the rhetoric in the sonnets as . . . the product not of a skilled sophister but of a mind diseased. (1987: 201) As such Sonnet 35 adumbrates the ‘willing patient’ pathological oxymoron of Sonnet 111. *

*

*

Sonnet 40 yet evinces a far heavier and gloomier mood than Sonnet 35, signalling as it does a first climactic manifestation of the poet-speaker’s masochistic wallowing in the abjectness of the love feeling, a masochism crystallized in the ‘lascivious grace’ oxymoron of line 13 combining the spiritual and the carnal with a sense of ‘lustful nobleman’ ‘vestigially registered’ (Burrow 2002: 460). This oxymoron shows a poet-speaker helplessly enthralled by beauty, for whom aesthetics is the very be-all and end-all of any human experience worthy of the name. Whatever the infidelity at stake, infatuation prevails and in taking ‘all [his] loves’ (line 1), the addressee has managed to rob the poet of what he paradoxically already lacks, hence his being a ‘gentle thief’. This is a perfect example of what Heather Dubrow, ascribing a mourning function to oxymoron, explains in the following way: ‘[t]he rhetorical figure syneciosis . . . ties together opposites and thus at once gives and snatches away’ (1999: 1–2). Also this oxymoron encapsulates a process of revival of the medieval concept of psychomachia informing Shakespeare’s major tragedies. Thus like Cleopatra, the young dedicatee, even when wanton, is simultaneously to be condemned and praised ‘for vilest things / Become themselves’ in him, just as they do ‘in her’ (Antony and Cleopatra 2.2.248–9).

176


Moreover the ‘lascivious grace’ oxymoron, with its specific blend of the religious and the unchaste, paves the way, in the poet-speaker’s hell-bound heavenly progress, for the ‘willing patient’ oxymoric posture of Sonnet 111, line 9 at the hands of both the ‘pitiful thriver’, the young man, and the ‘gentle cheater’, the dark lady, of 151.3: Pity me, then, and wish I were renewed, Whilst like a willing patient I will drink Potions of eisell’gainst my strong infection; (Sonnet 111, 8–10) Randall McLeod comments thus on ‘such an oxymoron as the “willing pacient” (line 9)’: it is a paradox . . . which we perceive when we remember that the Latin etymon of “pacient” is pati, “to suffer”. . . . The transformational desire to suffer is the prerequisite of “penance” (line 12) and of submitting to a “cure” (line 14). This contrasts with the usual fare of static and unredemptive paradoxes in love sonnets – of hot ice and cold fire. This dynamic paradox of patience must be seen to transvalue the initial and partial assumption of masochism. (1981: 91) Such an unobtrusive alliance of polar twins, showing the poet-speaker trapped in crippling contraries, is certainly a far cry from either Spenser’s showy antithesis: ‘My love is lyke to yse, and I to fyre:’ (Amoretti 30.1: 126) or Drayton’s combination of paradox and antithesis in the closing couplet of his Idea, 62: ‘I have, I want; Despaire, and yet Desire, / Burn’d in a Sea of yce, and drown’d amidst a fire’ (113). The poet-speaker of Shakespeare’s Sonnets drinks ‘most delicious poison’ (Antony and Cleopatra 1.5.28) and experiences a ‘loath’d delight’, which amounts to the Sonnets’ distinctive maddening experience. The physiological terminology of Sonnet 111, just as the explicit humoural lexicon of Renaissance medicine in the twin sonnets, 44 and 45, ‘underscore [a] portrait of desire as a disease that threatens the physical and mental health of [the] self’; ‘[l]yric poet and medical doctor, then, are both students of inwardness’ (Schoenfeldt 1999: 75). *

*

*

Continuing from Sonnet 71, in Sonnet 72 the poet-speaker again indulges in a masochistic mood of self-indictment – under the sway of ‘love’s own sweet constraint’ (All’s Well that Ends Well 4.2.16) – explaining that the fair


177

friend can be paradoxically morally justified by speaking falsely, by giving the poet more credit than he deserves: For you in me can nothing worthy prove; Unless you would devise some virtuous lie To do more for me than mine own desert, (Sonnet 72, 4–6) The general idea of the poem is that the beloved is urged to forget the poet once he is dead, a statement finding itself retrospectively contradicted in Sonnet 71 by the ambiguous exhortation of line 11: ‘Do not so much as my poor name rehearse’, wherein ‘rehearse’ is to be perceived as a subtle antiphrastic pun meaning ‘do not care to repeat my poor name’, but also ‘do not bury it again’ in the tomb of forgetfulness. *

*

*

In Sonnet 99, the oxymoron ‘sweet thief’ of line 2 is only formally attuned to the ‘sweet thief’ of Sonnet 35 and the ‘gentle thief’ of Sonnet 40: The forward violet thus did I chide: ‘Sweet thief, whence didst thou steal thy sweet that smells, If not from my love’s breath? The purple pride Which on thy soft cheek for complexion dwells In my love’s veins thou hast too grossly dyed.’ (Sonnet 99, 1–5) Contrary to what happens in the two earlier bleaker sonnets, the tone is much more light-hearted, and the fair friend, instead of being berated by the lover-poet for misconduct, finds himself the object of unmitigated praise. In this poem, the poet again meditates on the beloved’s absence. Here he accuses spring flowers and herbs of stealing their attractive colours and odoriferous fragrance from the beloved. The violet in particular is accused of being a ‘sweet thief’, amounting to a hypallage meaning either ‘fragrant thief’ or ‘thief of fragrance’. At stake is a myth of origin: the couplet concludes that all the flowers stole their attributes, whether ‘sweet, or colour’ (line 15), from some aspect of the beloved. In other words, Sonnet 99 is oxymorically speaking a chaste blazon, itself a syncretic palimpsest of classical, Biblical and Petrarchan traditions of comparing the beloved’s body to a catalogue of flowers. Sonnet 99 offers one striking example of rejuvenated literary recycling of ‘new and old’ discourse. However close in subject to Petrarch’s Rime Sparse, Henry

178


Constable’s Diana and to Campion’s lyric, ‘There is a garden in her face’ and though ‘[i]t treats a timeworn theme’, as Colin Burrow reminds us (2002: 578), Sonnet 99 encapsulates Shakespeare’s typical attitude when it comes to reading the language and style of his predecessors. Here the specific wit consists in substituting a literal to a figurative use of language in the elaboration of a conceit, traditional in the context of Petrarchan praise. As such the speaker presents himself as one who really believes the totally irrational assertions he is putting forward in making the fair friend the supreme natural nonpareil. But above all, the meaning of Sonnet 99 cannot be understood fully without reference to the Ovidian intertext of metamorphic transformations calling up the indissociable realities of Eros and Thanatos. The poem is first and foremost about transmutation through the process of ‘dying’ in both senses of the word. In that perspective lines 3–5 are reminiscent of the allegory of the little flower once white, ‘now purple with love’s wound’, in A Midsummer Night’s Dream (2.1.167) – an apt emblem for the cruelty of love. *

*

*

Against a background of ‘shining nights’ (Love’s Labour’s Lost 1.1.90), Sonnets 27, 43 and 61 form a trilogy wherein the poet finds himself transfixed by phantasmatic obsession with the beloved. All three poems are about the projected image of the lover keeping the febrile poet awake at night., in a state of ‘[s]till-waking sleep’ (Romeo and Juliet 1.1.181), but among the trio only Sonnet 43 makes use of two oxymora exploring the chiaroscuro of presence in absence. Here, as Paul Hammond puts it, Shakespeare gets away from tradition ascribing no ‘private space to Poet and Boy, be it pastoral arbour or limpid stream’. Instead, ‘[t]he spaces in which they come together are dream spaces, imaginary places in which the Poet contrives to fashion some form of the Boy’s presence, only to leave us (and, often, himself) all the more aware of the Boy’s actual absence,’ (2008: 274): When most I wink, then do mine eyes best see; For all the day they view things unrespected, But when I sleep, in dreams they look on thee, And darkly bright, are bright in dark directed. Then thou whose shadow shadows doth make bright, How would thy shadow’s form form happy show To the clear day with thy much clearer light, When to unseeing eyes thy shade shines so? (Sonnet 43, 1–8)


179

Sonnet 43 is inspired at once by the Renaissance belief that eyes create the light allowing sight and by Shakespeare’s own omnipresent concern with the distinction between appearance and essence. Vivid nocturnal dreams, inevitably part as they are of ‘life’s fitful fever’, as Macbeth puts it (3.2.23), are devoutly to be wished compared to the empty real day. But would the day benefit from the presence of the fair friend, the real night would seem but a bootless, bloodless counterfeit. The ‘darkly bright’ oxymoron of line 4 enhances the paradoxical vein of lines 1 and 2. Shining darkly amounts to a sheer impossibility in the prosaic rational world, but such an incompatibility finds itself resolved in the realm of poetry. Furthermore ‘darkly bright’ should be construed as a pun involving a single oxymoric layer. Indeed the first sense of ‘darkly bright’ is luminous behind the eyelids’ darkened ocular globes whereas the second meaning is blind but seeing, hence oxymoric. The ‘shade’/‘shines’ oxymoric association of line 8 itself is only to be interpreted as such if ‘shade’ is understood as an echo of the ‘shadow’ of line 5, and not as a mere synonym of image. In this specific sonnet, explicitly characterized by copia, hinging on an accumulative rhetoric of binary oppositions and balanced isocolons, what stands out above all is the parodic emphasis on subversion of the norm. As Joel Fineman has it: ‘it is because Sonnet 43 thus deliberately overstylizes its Petrarchist style and overthematizes its conventionally oxymoronic themes that we do not really read the poem as a solemn inquiry into optical epistemology’ (1988 [1986]: 237). In the end, here again, like in the very first sonnet in the procreation sequence, the young man appears as but a distorted reflection of the ideal world of Ideas so that the poem foregrounds an ever-widening discrepancy between a moot, ambiguous, ideal and an oxymoric real exposing the poet-speaker’s self-division and the dedicatee’s alike. ‘Conceit’, in the Shakespearean sense of (brooding) imagination, is more than ever in the Sonnets a double-edged experience, both a ‘comfort’ and an ‘injury’ (The Comedy of Errors 4.2.65), wherein ‘the self is submerged, swallowed up, or obliterated by the other’, which amounts to a ‘[p]sychic death’, itself ‘equivalent to the loss of personhood’ as the psychoanalyst Arnold H. Modell puts it (1996 [1993]: 102). *

*

*

When it comes to the two-part poem made up of Sonnets 44 and 45, the poet-speaker’s and the young man’s ‘separation’ ‘abides and flies’ (Antony and Cleopatra 1.3.104). Both texts trace the eagerly yearned for progress from duality to indivision. In Sonnet 44 the poet wishes that he

180


were thought rather than flesh so that he could constantly abide with the beloved. However the poet, being mortal, is made up of the four elements and the dullest of these, earth and water, constitute the dominant part of his being, leaving him desperately in the doldrums, condemned to weep ‘heavy tears’. The twin poem, Sonnet 45, features the poet’s thoughts and desires as the noblest elements, those of air and fire: The other two, slight air, and purging fire, Are both with thee, wherever I abide: The first my thought, the other my desire, These, present absent, with swift motion slide; (Sonnet 45, 1–4) When the poet’s thoughts and desires dwell with the beloved, the poet, reduced to the base elements, earth and water, sinks into melancholy. But such a dire state is mended when his thoughts and desires return, assuring him of the friend’s ‘fair health’ (line 12). The agony of separation from the beloved is a literary topos if any, but Shakespeare’s oxymoron, ‘present absent’ (line 4), somewhat rejuvenates an old story. The idea is that the volatile elements in the poet, those of air and fire, have been dispatched from the poet to his friend, leaving the former in a state of dejection. The oxymoric combination in ‘present absent’ encapsulates this very process of transference. Harping as it does on the dual, implicitly chiastic, interplay between body and substance and body and thought, Sonnet 45 signals the failure of the poet’s mercurial, Ariel-like, yearning. The devoutly wished for disincarnation as a warrant of ubiquitous reunion with the beloved only indicates one more stage in the process of the poet’s ‘twofold’, ever-more splitting personality. *

*

*

Sonnet 76 is itself the touchstone for Shakespeare’s artistic intent in the Sonnets, devising new meanings by means of oxymoron. Here the poet raises the issue of why his poetry never alters, but keeps repeating the same language and stylistic techniques: So all my best is dressing old words new, Spending again what is already spent: For as the sun is daily new and old, So is my love still telling what is told. (Sonnet 76, 11–4)


181

The answer, he explains, is that his theme remains ever one and the same, always rhapsodizing as he does on the beloved and on love. Positioned almost halfway through Shakespeare’s collection, Sonnet 76 however offers an illuminating comment on the author’s actually ever-renewed methods of poetic variation. As such line 2 most aptly exposes the poet’s false modesty: since ‘quick change’ is itself synonymous with ‘variation’, this line is itself an instance of variation in choice of expression, and thus shows that his verse is all but barren or devoid of inspiration. Here Shakespeare is ‘dressing’, not only ‘old words’, but also ideas, poetic topoi and commonplaces ‘new’ (line 11). Sonnet 76 is intensely self-reflexive and its purpose is definitely a metastylistic one, full as it is of rhetorical terms and devices designating poetic rules and verbal arrangements of all kinds (lines 1–6): ‘invention’ (line 6), that is poetic creation, is what the poem is fundamentally about. But the poem, and even the collection of Sonnets at large derive their meaning from line 11. What the poet does, despite the self-confessed monotony of the theme (line 10), is ‘dress’ – with a pun on the French ‘dresser’, erect (Vendler 1999 [1997]: 345) – a new stylistic system designed to clothe old thoughts in a new garb, to take up a rhetorical topos of the period. Hackneyed as it may seem, the poet’s ‘argument’, or theme, thus finds itself, Phoenix-like, rejuvenated like the sun which daily rises and sets, a poetic topos for man’s terrestrial progress (line 13). Here the coordinator ‘and’, far from lessening the impact of oxymoron, acts as a warrant of simultaneity (Ravassat 2007). Lesley Jeffries comments thus on one of the functions of the conjunction ‘and’ ‘in introducing created opposites’: it is a means to put forward the notion that despite being opposites, and thus stereotypically mutually exclusive, there is sometimes a paradoxical co-existence of the two extremes: . . . This is not the same as categorizing them as converses, since they are not simply two perspectives on the same phenomenon but are conflicting but co-existent states that logically should not be able to co-exist. (2010: 44–5) Since the lexicon of any given language is limited by definition, the poet has to devise new means to say old things and to ‘dress’ old topoi new (Ravassat 2008). Such is the case in Sonnet 1, wherein Shakespeare takes the conventional Petrarchan praise of chaste female beauty and turns it upside down. There the poet-speaker defeats the reader’s expectations by neither engaging nor indulging in commonplace eulogy of the beloved, but instead accuses a young man of ravenous self-consumption in his

182


refusal to beget a ‘tender heir’ in a position to perpetuate beauty’s legacy to the world of men beyond the grave. Another notable example is that of the notorious dark lady. The latter is referred to in terms of her ‘black beauty’ (Sonnet 127, line 3) – she is an icon of oxymoric ‘black aesthetics’, that is the dissonant counterpart of ‘the Petrarchan blazon, the lyrical itinerary of female beauty’ (Callaghan 2007: 52). Sonnet 127 itself ends with a paradoxical epigram turning blackness, rather hitherto the ‘badge of hell’ (Love’s Labour’s Lost 4.3.250), into the very quintessence of beauty. Referring to the dark lady’s eyes, Shakespeare writes: ‘Yet so they mourn, becoming of their woe, / That every tongue says beauty should look so’ (lines 13–4). In both cases, whether the ‘black beauty’ oxymoron or the paradoxical ending of Sonnet 127, Shakespeare definitely goes beyond either Sidney’s own praise of the oxymoric ‘black beames’ (47.2: 20) emanating from Stella’s dark eyes or Berowne’s eulogy of the brunette, Rosaline, ‘born to make black fair’ (Love’s Labour’s Lost 4.3.257). *

*

*

At this stage, my point is to connect the procreation sonnets and the metastylistic sonnet par excellence, Sonnet 76: since the young dedicatee will not reproduce his fair image in the flesh, the poet has to devise ‘new-found methods’ (Sonnet 76, line 4) in order to perpetuate his memory. Of course, as is a well known fact, the Sonnets are primarily meant for that, but more implicitly, oxymoron, the Crosse-couple, in Puttenham’s (1936 [1589]: 206) own jocular and vernacular renaming of syneciosis or contrapositum15, turns out to be a mongrel figure born of a form of semantic miscegenation. And that very peculiar, prodigious figure in turn produces incongruous creatures, ‘compounds strange’ (Sonnet 76, line 4) – purely literary and stylistic artefacts – such as an implicit fair black potential agent of subversion or the explicit ‘master mistress’ oxymoron of Sonnet 20. Margreta de Grazia highlights the spectacular watershed that is Sonnet 127, reversing as it does, in a social perspective, the ideal of the perpetuation of an only ‘fair’ (blond, light-complexioned) commendable race. She tackles the issue of miscegenation, the creation of a new cross-coupled race, a virtually fair black one, itself jeopardizing the very ideological edifice of the Sonnets: . . . after sonnet 17, it is through his own poetic lines rather than the youth’s generational loins that fair’s lineaments are to be reproduced,


183

fair’s lineage extended. The fair line ends, however, at 127 with the shocking declaration that ‘now is blacke beauties successive heire’. As if a black child had been born of a fair parent, a miscegenating successor is announced, one who razes fair’s lineage (‘And Beautie slandered with a bastard shame’) and seizes fair’s language (‘beauty hath no name’) – genealogy and etymology. . . . In praising the youth’s fair lineaments, social distinction had been maintained; in praising the mistress’s dark colours, social distinction is confounded. (de Grazia 2001: 159) Both oxymora – the implicitly derived fair black one and the actual ‘master mistress’ one – however eminently different they may seem, in fact have in common a facet of incongruity which is one of the hallmarks of oxymoron.16 The ‘master mistress’ of Sonnet 20, the most famous oxymoron in the whole collection, itself an arresting fusion of polar twins with a distinctly Ovidian flavour, harking back to the myth of Pygmalion, is construed by John Blades in the following way: ‘The young man is a siren for both sexes . . . a misfit, an oxymoron, a lapse of “doting” Nature’. (2007: 25) The ‘master mistress’ of Sonnet 20 suggests a play on double gender and feudal service, a recurring feature of Shakespeare’s production, also a superimposition of discourses of androgyny and of misogyny alike. As David Schalkwyk demonstrates (2007 [2002]: 224): [t]he poem . . . oscillates between male and female desire and desirability, in both its position of representation and in the position being represented. The object of desire is both male and female; but so is the desiring subject. Only at the very end of the poem is the chiastic confusion of sex and gender “naturalised” through the transformative power of “doting” Nature. In fact, it could be argued that the ‘master mistress’ oxymoron of Sonnet 20 fleshes out the essential chiastic structure of the Sonnets as a whole culminating in a poem like Sonnet 42 wherein the poet-speaker striving to console himself for the loss of both the young friend and the mistress, depicts himself as crucified (line 12) by faintly oxymoric ‘[l]oving offenders’ (line 5) making him fluctuate between unbearable extremes of ‘loss’ and ‘gain’: If I lose thee, my loss is my love’s gain, And losing her, my friend hath found that loss;

184

Stylistics and Shakespeare’s Language Both find each other, and I lose both twain, And both for my sake lay me on this cross: (Sonnet 42, 9–12) *

*

*

The present chapter has throughout aimed at showing that – whatever the persona he hides behind, the moralist, the masochist, the mercurial poet . . . – ‘[r]hetoric matters in Shakespeare’s poems not least because it is the principal means of creating variety’ (Roe 2007: 33), which is the very point of Sonnet 76. Cross-coupling in totally innovative ways, the oxymoric paradigms in Shakespeare’s Sonnets breed new senses, a unique performative language of action, as well as a totally new mode of narrative approach. Finally, oxymoron in the Sonnets is far less often than in Shakespeare’s dramatic production a warrant of either coincidentia oppositorum or reconciliation of opposites. The Sonnets display a deliberately plain vocabulary, the explicitly metastylistic ‘russet yeas and honest kersey noes’ advocated by Berowne in the early Love’s Labour’s Lost (5.2.413). Beyond that, as was demonstrated, Shakespeare innovated in more ways than one. Notably, he managed, to paraphrase his metastylistic Sonnet 76, to dress an old figure new. This implies a redefinition of the sonnet genre proper as well as of its speech quirks and mannerisms, including discarding the too often mawkish Petrarchan posture of languishing. Also, definitely a hallmark of Shakespeare’s Sonnets is the poetspeaker’s thriving on contradictions and oppositions, notably in the poems wherein his solipsistic nature appears most clearly. In such a perspective Shakespeare’s Sonnets tirelessly explore the interface between ‘novelty’ and ‘tradition’, ‘reve[l] in them, infusing the sonnet sequence with unexpected skepticism and satire, bawdiness and bitterness’ (Roberts 2007: 172). But no author – albeit one of the greatest – is an island, immune from (intertextual) influences and born as he was in a literary culture that valued and cultivated imitation, Shakespeare followed in his predecessors’ footsteps. As Jonathan Bate reminds us, ‘[t]hat good imitation involves difference as well as similarity, is a cardinal principle of Renaissance poetics’ (Bate 1993: 87). Refining the ‘Art of Precedent’, Shakespeare explores ‘new’ and ‘old’ oxymoric paradigms, infusing fresh blood into stale topoi, in doing so he relies on ‘the play of linguistic contraries’ not just for ‘ornament’, but ‘for it figures the psychology of contrariness’ (Bate 1993: 69). Also, the very ‘taut and rigid structure’ of the Shakespearean sonnet form – octave + sestet – encapsulates ‘a vocabulary of extravagant longing’ (Schoenfeldt 1999: 76). As such the content and very architectonics of Shakespeare’s sonnets signal a sharp oxymoric contrast. In fact, I would argue, relying on


185

Michael Schoenfeldt’s analysis, that Shakespeare’s Sonnets notably capture the very spirit of that ‘sweet constraint’ of desire, a dialectic wherein the Early Modern period finds itself caught, between a ‘fetish of control’ – the very ethos of the era most forcibly embodied by Sonnet 94 – and a sense of pre-baroque ‘explosive instability’ (Schoenfeldt 1999: 18) which Shakespeare’s updated use of oxymoron brilliantly illustrates. Oxymoron not only enacts the ‘collision’ of opposites mentioned above, but also an explosion of semantic categories. In that, oxymoron both contains and overflows the very limits of the language of the self, thereby bridging, if only partially so, the gap between the Early Modern and the modern self. Above all what I have emphasized is that Shakespeare’s oxymora do have an organic, structural, function, substantiating the poet-speaker’s discourse. Shakespeare is openly anti-Petrarchan in Sonnet 130, less ‘metaPetrarchan’ (Cousins 2000: 117) than Sidney – who explicitly mocks the trite evocations of unrequited lovers, experiencing ‘living deaths’, plagued by ‘deare wounds’ and enduring ‘faire stormes and freesing fires’ (Astrophel and Stella 6.4: 4)17 – but definitely what Shakespeare achieves in his own Sonnets is creating a distinctively jarring ‘strange harmony which must be expressed in . . . discords’ (Hoskins 1935 [ca. 1599]: 36) by means of oxymoron. This study has sought to demonstrate that, far beyond the (meta-) Petrarchan paraphernalia, the young man appears as an at once ambiguously perfect and flawed reflection of an ideal of unity which in fact conceals a ‘circumscribed bisection of the self-contained’ (Fineman 1988 [1986]: 248), an avatar of the ‘bifold authority’ ruling Cressida in Troilus’ perturbed psyche. This pattern of difference within similarity informs the character of the young man defeating in a way the essence of reconciliation which is the very hallmark of oxymoron. This impossible unity, this solipsistic, narcissistic self-absorption of the young man in turn generates a sense of ever-shrinking embedded circularity accounting for the obsessive leitmotiv of ‘withinness’ that Fineman traces in a number of sonnets18 and he applies to the poet-speaker’s own stance even outside the yoked pair of sonnets, 44 and 45: ‘this same chiasmicized “withinness” . . . governs the poet’s present-absent, hither-thither relationship to the young man’ (248). Fineman applies the same train of thought to Sonnet 48: Thee have I not locked up in any chest, Save where thou art not, though I feel thou art, Within the gentle closure of my breast, From whence at pleasure thou mayst come and part; (Sonnet 48, 9–12)

186


The poet-speaker further becomes ensnared within a pattern of ‘oxymoronic yoking’ (Fineman 1988 [1986]: 249); his own identity fractured under the influence of his ‘master mistress’, he can, like a blasphemous Christ, but drink ‘[p]otions of eisell’ like an oxymoric ‘willing patient’ (Sonnet 111, lines 10/9). Doing that the poet’s only ‘sure uncertainty’ (The Comedy of Errors 2.2.184) is that of the ‘dear divorce’ (Timon of Athens 4.3.384) of himself and himself and the sweet friend. This Shakespeare achieved in an unprecented way, at once in a pre-baroque spirit and on the cusp of modernity, acutely aware as he was of the variations of identity caught in a process of ever-fluctuating experience. This notably results in a pre-Hegelian master and slave relationship emphasizing ‘the asymmetry of self and other’ itself resulting from ‘the asymmetry of desire’ (Modell 1996 [1993]: 100), a notion highlighting the lord/vassal relationship ruling the Sonnets. Also when the poet-speaker of Sonnet 121 solemnly asserts ‘I am that I am’ (line 9), duplicating the divine ontological statement par excellence, he could just as well say ‘I am not what I am’ (Othello 1.1.64), although not in a Machiavellian way like Iago, but rather in a vein of ontological vacillation. There lies the rub. A ‘different sameness’ (Fineman 1988 [1986]: 259) plagues the young man and accounts for his cleft identity and in turn for the poet-speaker’s devoted but excruciatingly painful attachment to him. As such the young man fails to be an embodiment of concordia discors contrary to, for example, in a dramatic context, Cleopatra reconciling art and nature in Enobarbus’ ekphrastic evocation of her floating on a barge on the river Nile, or Antony referred to as a ‘heavenly mingle’ (1.5.62) – a possible definition of oxymoron – compounded as he is of human and superhuman parts and attributes, and beyond that being an incarnation of the resolved ‘struggle for the infinite within the finite’ (Davis 2007: 97 n3), thereby adumbrating the radiant achievement of the romances.

Notes 1

2

3

I am grateful to Heather Dubrow, Robert Ellrodt and Jonathan Culpeper for their precious feedback on this chapter. A paradox in itself since oxymoron is by definition ‘the show-off among figures of speech’, as Helen Vendler puts it (cited by Raymond Gibbs 2002 [1994]: 95). The present development will notably highlight the intimate connection between oxymoron and chiasmus. As Joel Fineman puts it: ‘Rhetoric gives many names to this troping of trope – syneciosis, antimetabole, metathesis, contrapostum, conjunctio, commistio, chiasmus – all of which refer to some way that language manages


4

5

6

7

8

noticeably to redouble with a difference the complementary similarities of a figurality based on likeness’ (1988 [1986]: 37). For the distinctively dual Weltanschauung ruling the Early Modern period at all levels, see Stuart Clark’s extensive analysis (2005 [1997]) and also Robert Grudin (1979). For the rhetorical strategy of argumentum in utramque partem see also Stephen Greenblatt (2005 [1980]: 230–1). For a qualified assessment of ‘twain’ in the Sonnets, see Hammond’s commentary of Sonnet 36 (2008: 275–7). ‘De voz doulceurs, ce n’est que sucre & miel, / De voz rigueurs n’est qu’aloës & fiel . . . ’ Œuvres poétiques édition Chamard, iv, 206, cited by Rosalie L. Colie in Shakespeare’s Living Art (note 53, p. 92). As K. K. Ruthven puts it: ‘The conceits themselves are ancient, for the idea of love as warfare stems originally from Ovid (Amores, I.ix), and the burning and freezing go right back to Sappho. Both occur in Petrarch and later Italian Petrarchizers, from whom English writers were able to inherit a ready-made vocabulary of love’ (Ruthven 1969: 20). Brian Vickers writes: ‘. . . if he often unites contraries, Shakespeare much more frequently opposes them. . . . This tremendous density of opposites points less to the binary structure of the human mind than to the endless possibilities of generating energy from collision’ (1984: 70). Such an ‘energy’ ‘resembles the heat generated by the interaction of two bodies that rub or jar against one another’, as Stephen Booth argues (1997: 6). To paraphrase Enobarbus’ pronouncement in his evocation of Cleopatra sailing on the river Nile, her cheeks simultaneously glowed and cooled by her attendants’ fans: ‘And what they undid did’ (2.2.215). Stephen Booth points out the existence of a double-barrelled oxymoron, first ‘tender churl’ meaning ‘gentle boor’ with ‘tender’ echoing and playing on ‘tender’ meaning ‘young’ in line 4, and he goes on to explain that ‘ . . . [s]ince tender carries a reminder of tendering, offering, giving tender churl embodies suggestion of another oxymoron: “generous miser” ’ (2000 [1977]: 136). See David Schalkwyk on the interplay of the contradictory and indissociable ‘poetics of praise’ and ‘of blame’ in the Sonnets and Twelfth Night (2008: 138). Also the use of the vocative here evinces a sense of immediacy, of the poet-speaker being intent on buttonholing his addressee thereby eliciting a performative speech act, which is part of the ethos of oxymoron, and thereby matches the performative nature of the Sonnets underscored by David Schalkwyk (2007 [2002]: 12). For similar conclusions, see Bradin Cormack: ‘Sonnet 126 takes leave of the beloved, while in sonnet 125 the speaker takes leave of his own desire for the beloved. . . . The gaze of the thrivers, although outwardly directed, is self-absorbed and self-destructive. Line 8 thus returns us to sonnet 1 . . . The phrase “pitifull thriuors” suggests abundance and lack’ (2009: 139). Indeed as Paula Blank explains in Shakespeare and the Mismeasure of Renaissance Man (2006: 51): ‘The proverb “one is no number” derived from the Pythagorean idea that One, as the antithesis of Many, cannot itself be subdivided; it cannot be a number because it contains no plurality’. In the combined ‘civil war’ and ‘odi et amo’ vein, see also the ‘pretty wrongs’ oxymoron in Sonnet 41 (line 1) explained in the following terms by Heather Dubrow: .........

9

10

11

12

13

14

187

188

15

16

17

18


‘The oxymoron that opens the poem may indirectly prepare us for the internalized oxymoron in the speaker: to describe wrongs as “pretty” (1) can be the result of an attraction to loving hate or even . . . of a tendency toward another oxymoron, praising blame’ (1987: 208). In Puttenham’s definition, the Crosse-couple ‘takes me two contrary words, and tieth them as it were in a paire of couples, and so makes them agree like good fellowes, as I saw once in Fraunce a wolfe coupled with a mastiffe, and a foxe with a hounde’ (1936 [1589]: 206). As for Hoskins, he defines the scheme syneciosis, also known as contrapositum, as ‘[a] composition of contraries, and by both words intimateth the meaning of neither precisely but a moderation and mediocrity of both;’ (1935 [ca. 1599]: 36). See Raymond Gibbs: ‘[o]ne of the most visible figures of thought and speech that reflect our incongruous understanding of experience is the oxymoron’ (2002 [1994]: 394) and also Javier Herrero Ruiz: ‘In paradox and oxymoron, the incongruity between two or more terms is solved via a projection space that arises from integration and accomodation operations in such a way that the conceptual structure of the apparently contrasting terms is possible in a given cognitive environment’ (2009: 264). Shakespeare happens to be explicitly ‘meta-Petrarchan’ himself in his portrayal of the pre-Juliet Romeo or of Orsino in the throes of the ‘sweet pangs of love’ (Twelfth Night 2.4.16); see Ravassat (2006). See ‘ “within thine own bud buriest thy content, (1)” . . . “within thine own deepsunken eyes” (2,7) or else “[t]hy gifts, thy tables, are within my brain (122)” ’ (Fineman 1988 [1986]: 248).

References Quotations from the Sonnets are from Katherine Duncan-Jones’ edition, 2005 [1997], Shakespeare’s Sonnets. The Arden Shakespeare, Third Series. London: Thomson. Other Shakespeare quotations are from The Complete Works 2007 [2001] The Arden Shakespeare. Revised edition, Richard Proudfoot, Ann Thompson and David Scott Kastan (eds). London: Thomson Learning. Bate, Jonathan 1993. Shakespeare and Ovid. Oxford: Clarendon Press. Blades, John 2007. Shakespeare The Sonnets, Analysing Texts series. Basingstoke and New York: Palgrave Macmillan. Blank, Paula 2006. Shakespeare and the Mismeasure of Renaissance Man. London: Ithaca and London, Cornell University Press. Booth, Stephen (ed.) 2000 [1977]. Shakespeare’s sonnets Edited with analytic commentary. Yale: Nota Bene Books, Yale University Press. Booth, Stephen 1997. ‘Shakespeare’s language and the language of Shakespeare’s time’, in: Stanley Wells (ed.), Shakespeare Survey 50, Shakespeare and Language. Cambridge: Cambridge University Press, 1–17. Burrow, Colin (ed.) 2002. The Complete Sonnets and Poems. Oxford: Oxford World’s Classics, Oxford University Press.


189

Callaghan, Dympna 2007. Shakespeare’s Sonnets. Malden MA, Oxford UK, Carlton, Victoria, Australia: Blackwell Publishing. Clark, Stuart 2005 [1997]. Thinking with Demons: The Idea of Witchcraft in Early Modern Europe. Oxford: Oxford University Press. Colie, Rosalie L. 1966. Paradoxia Epidemica: The Renaissance Tradition of Paradox. Princeton: Princeton University Press. Colie, Rosalie L. 1974. Shakespeare’s Living Art. Princeton: Princeton University Press. Cormack, Bradin 2009. ‘Shakespeare’s Narcissus, Sonnet’s Echo’ in: Leonard Barkan, Bradin Cormack and Sean Keilen (eds), The Forms of Renaissance Thought. Basingstoke and New York: Palgrave Macmillan, 127–49. Cousins, A. D. 2000. Shakespeare’s Sonnets and Narrative Poems. Harlow: Longman. Davis, Philip 2007. Shakespeare Thinking. London: Continuum. De Grazia, Margreta 2001. ‘The scandal of Shakespeare’s Sonnets’, in: Catherine M. S. Alexander and Stanley Wells (eds), Shakespeare and Sexuality. Cambridge: Cambridge University Press, 146–67. First published in Shakespeare Survey 46(1994): 35–49. Dubrow, Heather 1987. Captive Victors: Shakespeare’s Narrative Poems and Sonnets. New York: Cornell University Press. Dubrow, Heather 1999. Shakespeare and Domestic Loss: Forms of Deprivation, Mourning, and Recuperation. Cambridge: Cambridge University Press. Edmondson Paul and Stanley Wells 2004. Shakespeare’s Sonnets. Oxford: Oxford Shakespeare Topics, Oxford University Press. Ellrodt, Robert 2007 (editor and translator). William Shakespeare: Sonnets. Édition bilingue, Babel. Arles: Actes Sud. Evans, Maurice (ed.) 1984 [1977]. Elizabethan Sonnets. London and Melbourne: Dent and Sons Ltd, Everyman’s Library. Fineman, Joel 1988 [1986]. Shakespeare’s Perjur’d Eye: The Invention of Poetic Subjectivity in the Sonnets. Berkeley and Los Angeles: University of California Press. Fineman, Joel 1991. The Subjectivity Effect in Western Literary Tradition: Essays Toward the Release of Shakespeare’s Will (includes: ‘Introduction: Joel Fineman’s “Will” ’by Stephen Greenblatt). Cambridge MA and London UK: The MIT Press, An October Book. Gibbs, Raymond Jr. 2002 [1994]. The Poetics of Mind: Figuration, Thought, Language, and Understanding. Cambridge: Cambridge University Press. Greenblatt, Stephen 2005 [1980]. Renaissance Self-Fashioning: From More to Shakespeare. Chicago: The University of Chicago Press. Greene, Thomas M. 2003 [1985]. ‘Pitiful thrivers: failed husbandry in the Sonnets’, in: Patricia Parker and Geoffrey Hartman (eds), Shakespeare and the Question of Theory. London: Routledge, 230–44. Grudin, Robert 1979. Mighty Opposites: Shakespeare and Renaissance Contrariety. Berkeley, Los Angeles and London: University of California Press. Hammond, Paul 2008. ‘Shakespeare’s male utopias’, in: Études Anglaises 61–3. Paris: Klincksieck, 266–78. Hoskins, John 1935 [ca. 1599]. Directions for Speech and Style, Hoyt H. Hudson (ed.). Princeton: Princeton University Press. Hubler, Edward 1952. The Sense of Shakespeare’s Sonnets. New York: Hill and Wang.

190


Jeffries, Lesley 2010. Opposition in Discourse: The Construction of Oppositional Meaning. Advances in Stylistics (series editor: Daniel McIntyre). London and New York: Continuum. Joseph, Miriam, Sister 1949. Shakespeare’s Use of the Arts of Language. New York: Colombia University Press. Magnusson, Lynne 2006 [1999]. Shakespeare and Social Dialogue: Dramatic Language and Elizabethan Letters. Cambridge: Cambridge University Press. Magnusson, Lynne 2006 [2004]. ‘Shakespeare’s Sonnets: A modern perspective’, in: Barbara A. Mowat and Paul Werstine (eds), Shakespeare’s Sonnets and Poems. New York: Folger Shakespeare Library, Washington Square Press. McLeod, Randall 1981. ‘Unemending Shakespeare’s Sonnet 111’, Studies in English Literature 21. Houston: Rice University, Johns Hopkins University Press, 75–96. Modell, Arnold H. 1996 [1993]. The Private Self. Cambridge MA and London: Harvard University Press. Parker, Patricia 1996. Shakespeare from the Margins: Language, Culture, Context. Chicago and London: The University of Chicago Press. Porter Houston, John 1983. The Rhetoric of Poetry in the Renaissance and Seventeenth century. Baton Rouge: Louisiana State University Press. Puttenham, George 1936 [1589]. The Arte of English Poesie, Gladys Doidge Willcock and Alice Walker (eds). Cambridge: Cambridge University Press. Ravassat, Mireille 2006. ‘The pangs of dispriz’d love – on some discourses of amorous languor and melancholy in Shakespeare’ in: Max Duperray, Adrian Harding and Joanny Moulin (eds), Discourses of Melancholy. Aix-en-Provence: e-rea, 4.1, spring 2006: 51–8. Available at: http://erea.revues.org/index402.html Ravassat Mireille 2007. ‘Oxymoron, hendiadys and coordinate structures – Shakespeare from duality to indivision’, in: Wilfrid Rotgé (ed.), Bulletin de la Société de Stylistique Anglaise 28, Paris, 95–110. Also available at: http://stylistiqueanglaise.org/document.php?id=548. Ravassat, Mireille 2008. ‘ “So all my best is dressing old words new” – l’art de Shakespeare dans les Sonnets’, in: Pierre Kapitaniak and Jean-Michel Déprats (eds), Costume et déguisement dans le théâtre de Shakespeare et de ses contemporains. Paris: Société Française Shakespeare, www.societefrancaiseshakespeare.org/document.php?id=1467 Roberts, Sasha 2007. ‘Shakespeare’s Sonnets and English Sonnet sequences’, in: Patrick Cheney, Andrew Hadfield and Garrett A. Sullivan, Jr. (eds), Early Modern English Poetry: A Critical Companion. Oxford: Oxford University Press, 172–83. Roe, John 2007. ‘Rhetoric, style, and poetic form’, in: Patrick Cheney (ed.), The Cambridge Companion to Shakespeare’s Poetry. Cambridge: Cambridge University Press, 33–53. Ruiz, Javier Herrero 2009. Understanding Tropes at the Cross-Roads between Pragmatics and Cognition. Frankfurt am Main: Peter Lang. Ruthven, K. K. 1969. The Conceit. London: Methuen. Schalkwyk, David 2007 [2002]. Speech and Performance in Shakespeare’s Sonnets and Plays. Cambridge: Cambridge University Press. Schalkwyk, David 2008. Shakespeare, Love and Service. Cambridge: Cambridge University Press.


191

Schoenfeldt, Michael 1999. Bodies and Selves in Early Modern England: Physiology and Inwardness in Spenser, Shakespeare, Herbert, and Milton. Cambridge: Cambridge University Press. Schoenfeldt, Michael 2007. ‘The Sonnets’, in: Patrick Cheney (ed.), The Cambridge Companion to Shakespeare’s Poetry. Cambridge: Cambridge University Press, 125–43. Vendler, Helen 1999 [1997]. The Art of Shakespeare’s Sonnets. Harvard: The Belknap Press of Harvard University Press. Vickers, Brian 1984. ‘Rhetoric and feeling in Shakespeare’s Sonnets’, in: Keir Elam (ed.), Shakespeare Today: Directions and Methods of Research. Florence: La Casa Usher, 53–98. Wordsworth, William 1961 [1904/1778]. The Poetical Works of Wordsworth, Preface to the second edition of the ‘Lyrical Ballads’, Thomas Hutchinson (ed.). London, New York, Toronto: Oxford University Press. Yeats, William Butler 1987 [1933]. ‘Crazy Jane talks with the Bishop’, ‘Words for Music Perhaps’, VI in The Collected Poems of William Butler Yeats (Papermac). London and Basingstoke: Macmillan Publishers.

Chapter 9

‘Rue with a difference’: a Computational Stylistic Analysis of the Rhetoric of Suicide in Hamlet Thomas Anderson and Scott Crossley

1. Introduction Suicide – or ‘self-slaughter’, as Shakespeare describes it in Hamlet (ca. 1601) is the play’s motivating force.1 Hamlet’s famous deliberation over his own fate in Act 3, scene 1 – ‘To be, or not to be; that is the question’ (58) – is the most famous contemplation of suicide that Shakespeare offers the Early Modern audience who would have had conflicting opinions about self-murder because of the shift in the cultural significance of the act taking place at the turn of the seventeenth century. In response to Hamlet’s death, Horatio briefly considers suicide in Act 5 when he grabs the poisoned cup: ‘Never believe it. / I am more an antique Roman than a Dane. / Here’s yet some liquor left’ (5.2.282–4). Horatio’s affiliation with Roman stoicism that tolerated self-murder as heroic action is in marked contrast to the play’s concern with Early Modern accounts of suicide, which describe it as a most heinous sin – ‘an offence against God, against the king, and against Nature’ (Dalton 1626: 234) – that results in certain damnation. Other characters consider the cost of suicide in a culture that prohibited it. In Act 5, a gravedigger poses a question about the religious fate of Ophelia in reference to her drowning: ‘Is [Ophelia] to be buried in Christian burial that wilfully seeks her own salvation?’ (5.1.1–2); the other gravedigger responds that the ‘coroner . . . finds [Ophelia’s interment] Christian burial’ (4–5). In Early Modern England, the coroner headed a jury that tried suicides, and if the death was ruled self-murder, then the victim and her/his family were subject to severe punishment, including a pauper’s burial and loss of family assets (MacDonald and Murphy 1990: 17). Hamlet himself underscores the tension in the play between the desire to represent suicide outside of the religious and legal doctrines that regulate it and the Christian tradition that

The Rhetoric of Suicide in Hamlet

193

precludes this impulse. In Act 1, scene 2, he wishes that ‘the Everlasting had not fixed / His canon ’gainst self-slaughter’ (131–2). This chapter examines the language of suicide, mourning and grief of both Hamlet and Ophelia within the context of the shifting signification of suicide within Early Modern English culture that alters the meaning of the act – transforming a sin into psychological estrangement. While the play’s depiction of Ophelia’s apparent suicide registers the rhetoric of religious sin that offers a narrative of appropriate punishment by Renaissance standards, Shakespeare’s representation of Hamlet’s suicidal impulses and his final enactment of rage highlights the evolving cultural meaning of suicide during the period, removing it from its religious narrative and placing it within an emerging discourse of affect (Shneidman 1996) and societal disengagement (Durkheim 1951, Prezant et al. 1988). In the process of our argument, we demonstrate how a stylistic interpretation can enrich literary approaches to Hamlet and how literary interpretation complements linguistic inquiry. We will use lexico-semantic and corpus approaches to analyse Hamlet’s and Ophelia’s dialogue for suicidal discourse in order to reveal the text’s semantic prosodies and multi-word meaning structures that have remained under-examined by literary critics (Hoover 2007). Central to this analysis will be the computational tool LIWC (Linguistic Inquiry and Word Count; Pennebaker, Francis, and Booth 2001), which provides an efficient method for analysing the emotional, cognitive, structural and processing components of discourse. Of interest to our study is the use of pronominal references, social factors, communicative functions and lexicon related to death, sexuality and religion. A corpus analysis of suicidal discourse can provide an understanding of character that is more complete and underscores the play’s representation of the cultural tension engendered by shifting social attitudes towards suicide. By comparing the linguistic elements of characters’ discourse of suicide, we hope to convey a clear sense of the play’s interest in the emerging secular subject defined by nascent psychological discourses that challenge the hegemony of religious narratives. While implicitly arguing for an expanded notion of textuality that includes the play itself, legal and religious discourses contemporaneous with the play’s first performance, and a lexico-semantic corpus comprised of Early Modern and modern affective indicators of suicidal desire, the aim of this chapter is to suggest that Shakespeare considers suicide ambivalently. MacDonald makes a similar observation about suicide in the play as it pertains to Hamlet and Horatio, arguing that ‘[t]he emergence of new and more tolerant ideas about self-destruction is significant. . . . For these

194


ideas created a new ambivalence about suicide that Shakespeare exploits in Hamlet’ (MacDonald 1986: 315). He uses the shifting notion of Early Modern suicide as a case study to demonstrate what he calls Shakespeare’s ‘prismatic imagination’ (317) that adds complicated hues and shades to seemingly transparent cultural artefacts and concepts. The corpus analysis that helps shape our own interpretations of suicide in Hamlet, we hope, offers even more evidence for Shakespeare’s prismatic imagination – an imagination that includes linguistic discovery as well as literary and historical inquiry. This chapter asks what happens when we wed literary, cultural and linguistic analysis in order to prove that the rhetoric of suicide in Hamlet at once retains the theological and legal implications of a heinous sin while simultaneously registering an emergent discourse that makes self-slaughter less a sin and more a private action with psychological signification. By acknowledging that God’s law (the ‘fixed . . . canon’ [1.2.131–2]) firmly prohibits suicide, Hamlet underscores that the church was the primary social force regulating self-murder. However, not only was God’s law firmly against suicide, but also as the debate over Ophelia’s interment suggests, Tudor and Stuart England had developed a legal system that posthumously regulated acts of self-murder, determining which ones could legally be considered criminal acts. MacDonald and Murphy’s (1990) account of suicide in Renaissance England provides a detailed picture of the legal attempts to control how suicide was understood. According to their study of self-murder between 1485 and 1659, suicides in England peaked between the years 1600 and 1610 around the time that Hamlet was first performed (30). Those who killed themselves in madness, as ruled by the coroner’s jury, were judged to be innocent of a crime and could, therefore, receive a Christian burial. A victim of self-murder was legally categorized as non compos mentis, not of sound mind; a suicide, however, who the jury found to be sane and capable of premeditating the act was categorized as felo de se, a felon himself. Drawing on statistics culled from the Public Records Office, MacDonald and Murphy argue that those who committed suicide were rarely exonerated for their crime: ‘Over 95 per cent of the men and women who killed themselves between 1485 and 1660 were convicted as felones de se; fewer than 2 per cent were excused as persons non compos mentis’. They conclude that ‘the rigour with which the law against suicide was enforced . . . distinguishes this period from the centuries before and afterwards’. Before the Early Modern period, laws against suicide were infrequently enforced, and after 1660, suicide was increasingly ‘secularized and decriminalized’ (16).


195

Despite the draconian enforcement of the laws governing suicide in England, Hamlet captures what Williams (1977) might describe as an emergent discourse about suicide that counters the dominant legal and religious doctrines that attempt to regulate the action. Wymer (1986) has identified the emergence of this competing discourse as a response to the new humanism that sought out ‘valid ideals of human conduct’ (2) outside of ossifying religious categories. The nascent discourse created a new ethic of suicide that would not be fully articulate until the mid-seventeenth century, but early evidence of this new ethic, for example, appears in John Donne’s Biathanatos (2000 [ca. 1608/ca. 1647]), with the subtitle, ‘A Declaration of that Paradox of Thesis, that Self-Homicide is not so naturally Sin that it may never be otherwise’. Targoff points out, however, that Biathanatos was by no means an ‘unqualified endorsement of suicide’ (2006: 218). Indeed, even within Donne’s new ethic regarding self-murder, it was a suicide motivated by the desire to embody God’s glory and not self-interest that he considered theologically and legally acceptable. For Targoff, what is critically new in Donne’s defence of the act is his ‘insistence that the desire to end one’s life is not against the laws of nature’ (219). As an indication that the meaning of self-murder was evolving in ways that Donne perhaps initiated in his writings about death, Wymer argues that the term ‘suicide’, which first appears as a description of the act of self-murder in Sir Thomas Browne’s Religio Medici in 1635, is evidence of a ‘new outlook’ that views suicide as ‘morally neutral’. Wymer writes, ‘Unlike the older terms, “self-murder” and “self-slaughter”, [the term suicide] avoids the implication of violence and criminality’ (2). The character Hamlet, like his eponymous play, appears on the precipice of this discursive shift.2 Read within this cultural context, Hamlet’s most famous contemplative moment, ‘To be, or not to be’, is also a comment on the social, theological and legal forces that he knows will attempt to define him posthumously – a concern that appears most pressing in Act 5, after his death is assured. Concerned about his legacy in a society that can legally control how his ‘wounded name’ (2.286) is remembered, Hamlet enjoins Horatio to ‘tell my story’ (2.291). Upon hearing the ‘warlike noise’ (2.291) of Fortinbras’s arrival, Hamlet implores his surrogate Horatio to give Fortinbras his support and, along with the vote of confidence, the ‘occurents, more and less, / Which have solicited. The rest is silence’ (299–300). Hamlet’s concern that his ‘dying voice’ (298) still has the force to define his legacy during the impending political transition underscores the inescapable reality that in death, he surrenders this ability. His assertion that the ‘rest is silence’ is a powerful rebuke of the theo-juridical forces that seek

196


posthumously to regulate the significance of Hamlet’s suicide by proxy in Act 5. Central to our comparison of the language of suicide and despair in the play is the argument that Hamlet’s death is, in fact, self-slaughter – what we term suicide by proxy. In his analysis of Early Modern melancholy, Trevor (2005) suggests that Early Modern scholars such as John Donne or the fictional Hamlet ‘find themselves in the curious . . . predicament of wanting to be sad at the same time that they recognize that such sadness imperils their lives . . . because it can prompt thought of self-annihilation’ (9). For Trevor, the melancholic’s awareness that sadness can generate painful psychological trauma points to a type of introspection that challenges the assumption that, for instance, there is nothing to Hamlet’s interior. The depths of Hamlet’s interiority might be measured, then, by the extent to which he pursues a suicidal end within a culture that prohibits it. Hamlet acknowledges as much in the play’s first act: O that this too too solid flesh would melt, Thaw, and resolve itself into a dew, Or that the Everlasting had not fixed His canon ‘gainst self-slaughter! O God, O God, (1.2.129–32)3 Hamlet expresses the tension between his desire to commit suicide – that his solid flesh would melt – and the theological injunction against it. Yet, by Act 5, Hamlet’s confusion has vanished. Suicide emerges as an option within a performance – the fencing match – that will disguise his intentions and effectively silence the regulatory power of the church and state, allowing him to define his legacy after death through Horatio’s retelling. Pollin (1965) terms Hamlet’s death ‘a successful suicide’, noting that Hamlet’s ‘intellect, sensitivity, and moral and religious scruples caused him to seek the least reproachable means of terminating an intolerable burden of existence’ (240). Pollin observes that Hamlet repeatedly and intentionally exposes himself to peril: he returns to Denmark’s hostile shore ‘naked’ (4.7.42), or unarmed and without followers. Hamlet also ‘plays into the hands’ (254) of his two adversaries by accepting the challenge from Laertes of the fencing match, although Laertes has already tried to kill him in the graveyard. Pollin notes that Hamlet refuses to postpone the match twice, even as he admits to Horatio: ‘thou wouldst not think how all here about my heart—but it is no matter’ (5.2.149–51). About Hamlet’s choosing first the rapier and dagger, Pollin concludes, ‘In a sense Hamlet chooses


197

the instrument and Claudius through his henchman [Laertes] effects the release which Hamlet’s moral and religious scruples prevent him from grasping himself’ (254). Attaching surrogacy to the relationship between Claudius and Laertes, Pollin’s conclusion insinuates that in the final fencing match the play enacts self-murder by proxy. Far from being the naïve dupe in the Claudius-Laertes plot to kill him, as Pollin suggests, Hamlet willfully directs this play, as he does ‘The Mousetrap’ in Act 2. Where the earlier playlet exposes ‘the conscience of the King’ (2.2.582), Hamlet’s final production in Act 5 disguises his own suicidal intentions and obscures his reservations about engaging in self-murder. Suicide by proxy – as Laertes ushers in Hamlet’s demise with poison – is the play’s loophole that enables Hamlet’s inaugural selfauthorizing act, ironically requiring a second proxy: by telling Hamlet’s story, Horatio arbitrates his friend’s legacy and effectively silences the theo-political discourse against suicide by rendering it moot – ‘the rest is silence’. Although the play does not explicitly show that Ophelia’s death is suicide, there is ample evidence to suggest that she, too, commits self-slaughter. From the churchyard debate over her proper burial to her retreat into insanity that presages her demise, the play hints that her death was suicide. Gertrude’s description of her death personifies ‘her garments’ (4.7.152), granting agency to the clothes that ‘[p]ulled the poor wretch from her melodious lay / To muddy death’ (153–4). Her description that redirects agency away from Ophelia’s will necessarily protects Ophelia from the explicit charge of self-murder and allows for the possibility of a Christian burial; however, before her drowning Ophelia’s concerns about her own post-suicidal legacy attest to her contemplation of suicide. As she gives various plants to several on-stage spectators, including her brother, Ophelia offers rue among the many plants that have symbolic value to her: ‘There’s rue for you, and here’s some for me. We may call it herb-grace o’ Sundays. O, you must wear your rue with a difference’ (4.5.177–9). In her madness, Ophelia articulates the play’s truth that during the cultural moment of its production, suicide produces different types of regret: Ophelia rues the prospect of her suicide through a theo-juridical lens and Hamlet rues his through an emerging personal or psychological lens that subverts theological and legal forms of regulation. Hamlet wants his death to be remembered differently than he assumes the theo-juridical mandates of the court will allow. Ophelia, however, appears resigned to the fact that her suicide is subject to the condemnation that comes with self-murder. Put otherwise, the representation of Ophelia’s apparent

198


suicide includes the threat of legal and theological condemnation, while Hamlet’s death embodies a new cultural understanding of suicide that avoids the stain of criminality and sin – in Ophelia’s words, they ‘rue with a difference’.

2. Methods To demonstrate how a stylistic interpretation can enrich literary approaches to Hamlet, we examined the discourse of Hamlet and of Ophelia for lexico-semantic implications of differences in suicidal rhetoric using LIWC (Pennebaker et al. 2001), which provides an efficient method for analysing the emotional, personal, communicative, sexual and metaphysical components of language. We developed a corpus of discourses for Hamlet and Ophelia that included pre-suicidal and suicidal discourses as well as two control corpora selected from non-suicidal characters (Horatio and Laertes). We predict that the discourse of Hamlet will exhibit lexico-semantic patterns that befit modern narratives of suicide (e.g. more personal pronouns, lack of communicative and social features), while the discourse of Ophelia will exhibit more lexico-semantic patterns that denote a rhetoric of religious sin (e.g. more plural pronouns, words related to death, sex, religion).

3. Linguistic inquiry and word count (LIWC) LIWC was developed as a tool to examine the cognitive, emotional and structural processes found in writing and speech. The tool is a text analysis program that uses word counts to search text for lexical and semantic features that correlate with standard language categories (e.g. articles, prepositions, pronouns), psychological processes (e.g. positive and negative emotion words), relativity-related words (e.g. time, verb tense, spatiality) and content dimensions (e.g. death, sex, home, occupation) (Pennebaker et al. 2001). The theoretical underpinnings of the tool are premised on the notion that words individuals use when writing and speaking provide access to their emotional and cognitive states (Gottschalk and Gleser 1969, Rosenberg and Tucker 1978, Stiles 1992). The words selected for the LIWC categories are based on emotional rating scales, thesauri and dictionaries. These words were then evaluated by independent judges and the words were placed into aligned categories if agreement was met (ranging from 0.86 for the category optimism to 1.00 for relatives). The selected words


199

then came to form the LIWC database, which capture, on average, 80 per cent of the words found in writing and speech. The external validity of LIWC as a general measure of emotional and cognitive lexical and semantic features was explored in the Pennebaker and Francis (1999) study that examined the psychological ratings of first-year college student essays written about the affective toll of the college experience. These essays were first evaluated by human judges in consideration of the psychological, emotional and cognitive categories analysed by LIWC and later by the LIWC tool. The evaluations made by the human judges and LIWC were highly correlated. LIWC has also been used to analyse suicidal vocabularies (Stirman and Pennebaker 2001). In this study, the writing styles and characteristics of writers who had committed suicide were compared to those that had not with the prediction that suicidal writers would exhibit more words of hopelessness in their writings than non-suicidal writers. The LIWC tool was used to analyse 300 poems written by suicidal and non-suicidal poets. Statistical analyses of the LIWC output supported the notion that suicidal writers and non-suicidal writers differed in their use of words indicating the desire for social integration and indicating a sense of ‘hopelessness’. Specifically, suicidal poets used more first person pronouns and words related to death and sexuality and less words related to communication.

4. Corpus To collect our corpus, we used Stephen Greenblatt’s Norton edition of Hamlet (1997), considered a standard, scholarly edition, as well as a hypertextual version of the play from MIT’s edition of the complete works. From these editions of Hamlet, six corpora were created. These were the suicidal discourses for Hamlet and Ophelia, the non-suicidal discourses for Hamlet and Ophelia, and the non-suicidal discourses of Horatio and Laertes. We demarcated the suicidal from the non-suicidal discourses of Hamlet and Ophelia based on the critical opinion that Ophelia’s fall into despair in Act 4 is a prelude to her death by drowning (MacDonald 1986, Wymer 1986) and that Hamlet’s own deliberation about his fate in Act 3 reveals the desire that self-murder governs his actions for the remainder of the play (Spinrad 2005). Ophelia’s discourse in Act 4, scene 5 was considered suicidal. Her discourse before this act was considered non-suicidal. However, the number of lines assigned to Ophelia limited


200 Table 9.1

Corpora descriptive statistics

Character

Discourse

Hamlet Hamlet Ophelia Ophelia Horatio Laertes

Non-suicidal Suicidal Non-suicidal Suicidal Non-suicidal Non-suicidal

Act. Scene

Number of texts

Average number of words

1.2 5.2 1.3–2.1 4.5 1.1 1.3

7 13 5 5 6 4

113.00 113.38 119.80 99.20 97.17 96.25

the size of the analysis. Thus, her discourse was broken into 100-word segments to ensure enough data points were available for a statistical analysis. These segments were then compared with 100-word segments taken from Hamlet’s suicidal discourse in Act 5, scene 2 and non-suicidal discourse taken from Act 1, scene 2. For controls, we selected the beginning scenes for the characters Horatio and Laertes, both non-suicidal characters. Descriptive statistics for the corpora can be found in Table 9.1. The text segments for all the characters were then processed through the LIWC tool and normalized lexico-semantic values were selected for linguistic categories related to personal, communicative, sexual and metaphysical lexical components. In applying LIWC’s modern database of words related to suicide, hopelessness, and estrangement to an Early Modern corpus, we made minor, but important, alterations to the play’s corpus. We made these alterations to the corpora in order to align the language used in the original work to that examined by LIWC. Principal among these was the modification of pronouns (thee, thou, and thy) to the modern equivalents. In addition, the word lord, when used to refer to an aristocrat, was changed to the aristocrat’s title (e.g. prince or king). This was to control for confounds within the religious semantic category.

5. Statistical analysis To compare for differences among the characters, the LIWC values were analysed using Multivariate Analyses of Variance (MANOVA) and follow-up Analyses of Variance (ANOVA) to compare individual differences. These statistical analyses were conducted to demonstrate that differences existed between Ophelia’s and Hamlet’s suicidal discourses and non-suicidal discourses.


201

6. Analysis 6.1 Suicidal discourse: Hamlet and Ophelia We predicted that the suicidal discourse of Hamlet would exhibit more lexico-semantic patterns that befit modern narratives of suicide (e.g. personal pronouns, less communicative functions), while the suicidal discourse of Ophelia would exhibit more lexico-semantic patterns that denote a rhetoric of religious sin (e.g. words related to death, sex and religion). The statistical analyses support these predictions with the suicidal discourses of Hamlet and Ophelia exhibiting significant difference: F(12,5) = 6.78, p < 0.05 (where F represents the measure of distance between individual word choices, followed by the degrees of freedom [between groups = 12 and within groups = 5] and finally by p, which is the confidence level of significant difference

Table 9.2 Mean, standard deviation, and F scores for LIWC values (Hamlet’s versus Ophelia’s suicidal discourse) Category

Character

Mean

Standard deviation

F Score

I

Hamlet Ophelia Hamlet Ophelia Hamlet Ophelia Hamlet Ophelia Hamlet Ophelia Hamlet Ophelia Hamlet Ophelia Hamlet Ophelia Hamlet Ophelia Hamlet Ophelia Hamlet Ophelia Hamlet Ophelia

6.09 3.26 1.14 1.41 7.23 5.05 2.19 4.96 10.66 16.69 1.13 1.98 0.12 0.80 0.82 4.75 0.60 3.33 0.22 1.42 1.14 0.40 0.15 0.78

1.98 2.40 1.96 0.56 3.10 2.92 1.65 1.70 4.54 2.21 0.67 0.59 0.30 0.45 0.85 4.51 0.73 3.53 0.42 1.58 0.92 0.55 0.36 0.83

6.58*

We Self You Social Communication Family Metaphysical Religion Death Body Sexual

* p < 0.05 ** p < 0.001.

0.09 1.84 10.05* 7.88* 6.05* 13.85** 9.93* 7.70* 6.91* 2.79 5.48*

202


between two means). Specifically, Hamlet’s suicidal discourse contains more instances of words related to personal discourse and less to communicative discourse. On the other hand, Ophelia’s suicidal discourse exhibits significantly more words related to religion, death and sexuality (see Table 9.2 for additional information). The increased frequency of these words is evidence that, at the linguistic level, Ophelia articulates the anxiety that suicide has theo-juridical consequences common in Early Modern England. The category ‘metaphysical’ in Table 9.2 and in the other Tables below is a superordinate category that includes within its subordinate categories (‘religion’, ‘death’, ‘body’ and ‘sexuality’). 6.2 Suicidal and pre-suicidal discourse: Ophelia When considering Ophelia’s pre-suicidal and suicidal discourses, no significant differences were noted in the MANOVA analysis F(9,1) = 4.83, Table 9.3 Mean, standard deviation, and F scores for LIWC values (Ophelia’s suicidal discourse versus non-suicidal discourse) Category

Character

Mean

Standard deviation

F Score

I

Pre-suicidal Suicidal Pre-suicidal Suicidal Pre-suicidal Suicidal Pre-suicidal Suicidal Pre-suicidal Suicidal Pre-suicidal Suicidal Pre-suicidal Suicidal Pre-suicidal Suicidal Pre-suicidal Suicidal Pre-suicidal Suicidal Pre-suicidal Suicidal Pre-suicidal Suicidal

8.71 3.26 0.17 1.42 8.88 4.68 2.19 4.98 12.43 16.76 0.50 1.99 1.84 0.81 1.00 4.75 1.00 3.33 0.00 1.42 3.01 0.41 0.00 0.79

3.37 2.40 0.42 0.56 3.43 2.11 2.59 1.70 5.10 2.24 0.55 0.60 2.64 0.45 0.63 4.51 0.63 3.53 0.00 1.58 2.13 0.56 0.00 0.83

9.11*

We Self You Social Communication Family Metaphysical Religion Death Body Sexual

* p < 0.05 ** p < 0.001.

18.14** 5.68* 4.23 3.07 18.46* 0.74 4.15 2.58 4.99* 6.98* 5.55*


203

p > 0.05; however, the associated ANOVAs demonstrate that Ophelia’s pre-suicidal language contains more personal discourse and less occurrences of words related to death, body and sexuality (see Table 9.3 for additional information). This subtle variation in her pre-suicidal and suicidal rhetoric shows that Ophelia’s concerns about death and sexuality emerge as significant once her suicide becomes her primary motivation. As we will discuss in more detail below, our prediction that Ophelia would use more words pertaining to the body in her suicidal language was not statistically significant. While the mean value for religious words was higher in Ophelia’s suicidal discourse (3.33) as compared with her pre-suicidal discourse (1.00), the differences only approached significance (p =.063). The mean value for words related to the body was significantly higher in pre-suicidal discourse (3.01) than suicidal (0.41). Rhetoric about the body may indeed emerge as meaningful using non-statistical analysis; however, this meaningfulness is demonstrated through metaphorical analysis, which proves difficult for computational tools like LIWC.

6.3 Suicidal and pre-suicidal discourse: Hamlet When Hamlet’s pre-suicidal discourse is compared to his suicidal discourse, significant findings do emerge, F(14, 5) = 5.77, p < 0.05, with Hamlet’s earlier discourse containing more incidences of words relating to the metaphysical and the body (see Table 9.4 for additional information). This supports the assumption that the change registered in Hamlet’s suicidal rhetoric corresponds to an emergent cultural understanding of suicide that subverts the formerly dominant model registered in Ophelia’s theo-juridical concerns about the fate of her body and soul.

6.4 Suicidal and pre-suicidal discourse: Ophelia and Hamlet to controls Comparisons between pre-suicidal Hamlet and Ophelia to the controls, Horatio and Laertes, demonstrate no significant differences in word choices, F(15,5) = 1.32, p > 0.05, while comparisons between suicidal Hamlet and Ophelia to the controls exhibited significant differences in word choices, F(15,12) = 9.57, p

Stylistics and Shakespeare's Language: Transdisciplinary Approaches

Landscape in Language: Transdisciplinary perspectives

Assessment methods for infants and toddlers: transdisciplinary team approaches

Shakespeares Planet

Philosophical Shakespeares

Shakespeares Planet

Shakespeares Planet

Literature and Stylistics for Language Learners: Theory and Practice

Local Shakespeares Proximations and Power

Feminist Stylistics

Language, Discourse and Literature: An Introductory Reader in Discourse Stylistics

Approaches and Methods in Language Teaching

Contexts and Constructions (Constructional Approaches to Language)

Approaches to Second Language Acquisition

Language Contact and Grammatical Change (Cambridge Approaches to Language Contact)

Stylistics: A Practical Coursebook

Handbook of Transdisciplinary Research

Stylistics: A Resource Book for Students (Routledge English Language Introductions)

Feminist Stylistics (Interface)

Handbook of Transdisciplinary Research

Historical Corpus Stylistics (Corpus and Discourse)

Etwas ueber William Shakespeares Schauspiele

Spiritual Shakespeares (Accents on Shakespeare)

Mercury Pollution: A Transdisciplinary Treatment

The Ecology of Language Evolution (Cambridge Approaches to Language Contact)

The Ecology of Language Evolution (Cambridge Approaches to Language Contact)

Handbook of Generative Approaches to Language Acquisition

Constructional Reorganization (Constructional Approaches to Language)

Alternative Approaches to Second Language Acquisition

English stylistics. Стилистика английского языка

Communicative competence approaches to language proficiency assessment: research and application

Stylistics and Shakespeare's Language: Transdisciplinary Approaches

Landscape in Language: Transdisciplinary perspectives

Assessment methods for infants and toddlers: transdisciplinary team approaches

Shakespeares Planet

Philosophical Shakespeares

Shakespeares Planet

Shakespeares Planet

Literature and Stylistics for Language Learners: Theory and Practice

Local Shakespeares Proximations and Power

Feminist Stylistics

Language, Discourse and Literature: An Introductory Reader in Discourse Stylistics

Approaches and Methods in Language Teaching

Contexts and Constructions (Constructional Approaches to Language)

Approaches to Second Language Acquisition

Language Contact and Grammatical Change (Cambridge Approaches to Language Contact)

Stylistics: A Practical Coursebook

Handbook of Transdisciplinary Research

Stylistics: A Resource Book for Students (Routledge English Language Introductions)

Feminist Stylistics (Interface)

Handbook of Transdisciplinary Research

Historical Corpus Stylistics (Corpus and Discourse)

Etwas ueber William Shakespeares Schauspiele

Spiritual Shakespeares (Accents on Shakespeare)

Mercury Pollution: A Transdisciplinary Treatment

The Ecology of Language Evolution (Cambridge Approaches to Language Contact)

The Ecology of Language Evolution (Cambridge Approaches to Language Contact)

Handbook of Generative Approaches to Language Acquisition

Constructional Reorganization (Constructional Approaches to Language)

Alternative Approaches to Second Language Acquisition

English stylistics. Стилистика английского языка

Communicative competence approaches to language proficiency assessment: research and application

Recommend Documents