Shakespeare continues to attract a staggering number of new editions, critical commentaries, and discursive essays, but the supportive linguistic literature has been surprisingly sparse. When we were researching the need for this book, we found very little that might be classed as ‘high-quality Shakespearian lexicography’. On the other hand, we found a great deal that could be described as ‘low-quality lexical commercialism’, especially on the Web, in the form of selective word-lists providing crude approximations to the meanings of Elizabethan words, and generalizing about meanings in ways that were often misleading. The tools of enquiry that all students of literature have a right to demand from linguists - in the form of dictionaries, glossaries, thesauri, and concordances - are still remarkably few, by comparison with the literary energy that has been expended on the canon. And there seems to be no let-up in the demand for additional resources, as expressed by teachers, academics, students, actors, journalists, and others, anxious to develop their awareness of Shakespeare’s language.
‘Onions’, of course, is the splendid exception. Having spent three years on the present book with the benefit of modern technology, we cannot but doff our caps in admiration at Charles Talbot Onions’ remarkable feat of compilation, first published in 1911, revised in 1919, and enlarged in a further revision by Robert D Eagleson in 1986. Both the present authors have lived with this book in the literary and theatrical parts of their professional lives, and they have benefitted repeatedly from its content - in the first author’s case, for some 40 years. But when you live with someone for so long, you find out their weaknesses as well as their strengths; and our personal marginalia identifying omissions of coverage and inadequacies of treatment reinforces our feeling that there is a need for a fresh work.
All dictionaries should be regularly revised, to take account of new findings and methods, and Shakespeare is no exception.
The texts are available to study in ways that were not possible before, and new texts have begun to attract attention: in the present case, our corpus includes the vocabulary of The Two Noble Kinsmen and King Edward III, which Onions, for example, does not include. The Oxford English Dictionary (OED), which Onions himself helped to create, is now available on CD-ROM and on the Internet, making it much more practicable to carry out large-scale focused projects of a lexical kind. We have thesaurus and concordance material from the fine project by Marvin Spevack to aid us. And between Hilda Hulme’s Explorations in Shakespeare’s Language (1962) - being written when the first author was one of her students - and Frank Kermode’s Shakespeare’s Language (2000), we have a wealth of individual lexical studies to inform our judgements as lexicographers.
It is, one must always remember, a matter of judgement. Lexicographers are as human and as creative as anyone else - as the comparison of any two dictionaries will show, with thousands of differences of coverage and treatment.
To refer again to Onions: his judgements, made on the basis of an intuition shaped by the lexicon of Victorian England and an educational system in which Latin, in particular, played a major part, will be very different from ours, a century later, where the arrival of new meanings have made old ones less accessible (e.g. portal), and where we have different intuitions about what counts as a ‘difficult’ word. The contrast between our coverage and that of Onions is striking: we include many words that he omitted - a random dozen is behest, beshrew, betimes, cellerage, commonweal, contumely, disrelish, lineaments, incontinency, saws, surfeit, and wormwood, and in all - as we guess from comparing a few pages - some 3000 items. On the other hand, Onions, fascinated as he was with word origins, contains much more on the etymology and cultural history of certain words.
It is not just lexicographers who rely on their intuitions. Editors do too, and in our reading of the notes accompanying the various editions of the plays we have been struck by the remarkable range of opinion about what editors judge to be a word ‘worth glossing’. Some plays in the Penguin series have twice as many glossed lexical items as others; and marked variations are found in other series. Similarly, there are personal practices in relation to lexical explanation. Some editions rely on individual sources: one, for example, relates virtually all its lexical observations to Onions; another only to the senses of the OED. We have been repeatedly struck by the number of difficult words which are not glossed at all - a manifestation, we feel, of the same sort of limitations as everyone encounters when they read an instruction manual written by specialists. It is difficult for an editor to remember just how difficult some words are, with familiarity increasingly breeding contentment. With 35 years separating the ages of the two authors of this book, we found the same problem: the first author’s mark-up of the difficult words in a play would be far exceeded by those identified by the second author - and when this happened, we would always err on the side of inclusiveness. Doubtless some people will query why some words are in this book, thinking them ‘too easy’; our only defence is that not everyone finds them so.
This is, of course, why we have extended the traditional coverage of a dictionary to include the appendices we have selected. In the days when a Classical background and Latin were a routine part of secondary schooling, there would not have been a need to include some of the features listed there. Regrettably even French words cannot be taken for granted any more, in some parts of the world. We have accordingly been eclectic. We know that Shakespeare is read virtually everywhere; but we also know that not all readers will be familiar with the British local geography, so we have included units on locational names. The same principle applies to some of the special features: for example, we include in our ‘plants’ unit some terms which are still in use, but which will be unfamiliar to people from parts of the world where these plants are not known.
On the other hand, a dictionary is not a substitute for editorial notes, which focus on extracting the full meaning of a word in its individual context, and examining its resonance for the passage or text as a whole. Dictionaries do not particularize in this way; rather, they attempt to make generalizations about the meaning of words. They therefore have to ignore many of the sense associations, plays on words, etymological echoes, quibbles, and other effects that comprise the linguistic identity of a word in its full context. There is no place in a dictionary for the kind of extended discussion of a word like prone (such as occurs in Frank Kermode’s insightful appreciation), teasing out its implications. At the same time, when there is a chance to draw attention to the complexity of a lexical situation, it is incumbent on the lexicographer to do so. We therefore regularly alert readers to the existence of other meanings than the one which is the focus of a particular entry, using three types of parenthetic information at the end of examples.
- The ‘[also]’ convention. For example, a quotation assigned to sense 2 of a headword might be followed by ‘[also: sense 4]’. This means that editorial practice generally finds sense 2 the primary sense, but sense 4 is definitely relevant in this context.
- The ‘[or]’ convention. This convention recognizes the fact that editorial opinion is divergent. If a quotation in sense 2 is followed by ‘[or: sense 4]’ it means that our example could just as easily have been placed under sense 4, editorial opinions differing on the point.
- The ‘[pun]’ convention. If we give a quotation in sense 2 and follow it with ‘[pun: 63, sense 4]’, we mean that in line 63 the word is punned in sense 4. The use of ‘[bawdy pun]’ is also needed, from time to time.
It is not our intention to give the sense of a difficult passage as a whole - that is the job of the textual edition. However, from time to time, especially when sentence structure is difficult to follow, we have found it useful to add a further parenthetic explanatory gloss following the quotation, preceded by ‘[i.e.]’. In such ways, our dictionary makes more of a bow in the direction of editorial practice than is usual in glossaries, and we see this as one of its strengths.
We did not base our book on any previous compilation. Rather, we took a set of texts and worked through them from scratch, word by word. As this was a book commissioned by Penguin, the obvious choice for primary material was the New Penguin Shakespeare, under the general editorship of T J B Spencer and associate editorship of Stanley Wells. This provided us with virtually all our primary data; the absence in that series at the time of a text for Cymbeline was made good by the use of the Arden edition (editor J M Nosworthy), and for the Website we used an edition of King Edward III prepared by David Crystal.
Our procedure was straightforward. We went through each text and highlit any word whose form or meaning we felt to be ‘difficult’, either because it was an Elizabethan usage no longer current or because it would pose a problem to a modern readership despite its continued currency. Thus under the first heading we would highlight fardel and tainture; and under the second importune and ordure. We were particularly sensitive to those words which have developed a modern sense which is different from the earlier one (the so-called ‘false friends’), such as doubt [= fear], supervisor [= onlooker] and revolting [= rebellious], and in such cases we sometimes add an example of the modern usage in order to point the contrast with the obsolete one (as in the case of rheumatic).
Any word identified by the Penguin series editors as a problem case, for whatever reason, was automatically included as an entry; and to avoid the biases of those editors, we carried out the same exercise on all the words singled out for special comment in volumes from two other series.
Often, this would make us encounter variants which the Penguin editors had chosen not to include, and we would add these to our corpus too. In particular, we tried to include all First Folio and significant early Quarto variants where these presented more than just an alternative spelling or printer’s error. However, we did not go beyond these basic sources, and we certainly did not try to include the hundreds of individual emendations made by the pantheon of editors over the centuries. As it was, the book ended up with 21,263 entries.
With just a few exceptions, every instance of a problematic item was logged separately, regardless of the number of times it appeared. This eventually produced a very large corpus, of over 50,000 entries, and it is this corpus which is now made available at this Website. We knew we could never fit all these examples into a single book; but the reasons for covering the texts in such detail were twofold.
First, it enabled us to scrutinise all the examples of a word before deciding which one to put into the book: where there was a choice, we selected the example which illustrated a word meaning most clearly. And secondly, it was always our intention to make the entire corpus available electronically. But, for the book, we restricted ourselves as a rule to one quotation illustrating each word, or distinct sense of a word - just occasionally allowing ourselves the luxury of two or three quotations, where it proved necessary to focus on different aspects of usage (alternative spellings, for example).
We then gave up to five other line-references to the word from the corpus, choosing examples which we felt illustrated the diversity of the usage in the clearest way. The aim of multiple examples is to provide enough illustration for a reader to begin to develop a confident intuition about Shakespearian English. That is why we gave several examples even in cases where the usage was not especially complex. Hold 8 [= ‘stop, cease’] is a case in point. Following up the examples listed would illustrate a range of usage from the single ‘Hold, sir’ through the double emphasis of ‘Hold, Richard, hold’ to the fivefold reiteration of the Messenger in Two Noble Kinsmen, as well as its use in a sentence (‘we cannot hold’) and governing a pronoun (‘Hold thee’). In this way one ‘gets to know the word’ more intimately than if just a single example were given.
For the Website, considerations of space do not apply, and the database includes every instance of a usage, with just two types of exception.
- There are several cases where a particular word is used several times in quick succession, in exactly the same sense, within a speech or piece of dialogue. In such cases, we included the first usage only. An example is Falstaff’s rising from the dead at the end of Henry 4 Part 1 (V.iv.110,ff), where he uses the word counterfeit nine times in a dozen lines. We include just one instance, on the grounds that any reader who might use the glossary to check on this word would find that the one look-up would suffice for the whole sequence. On the other hand, if the word were to be used a few pages later, we would once again include it in the database, on the grounds that a reader might value the reminder. This is the way text editors often work.
- We also do not give full coverage of grammatical words. These are prepositions, conjunctions, pronouns, and a few other forms whose function is more to help build a sentence structure than to convey the content of an utterance. In such cases, we restrict ourselves to a single example of the distinctive usage.
By headword we mean any item in bold face - what in lexicology is often referred to as a lexical item or lexeme. This is often a single word, but sometimes it is two words (e.g. come near, give over) or a series of words (e.g. come cut and long tail). The following points should be noted in relation to the way we handle lexical items as headwords.
- The alphabetical arrangement of the headword list is for the most part letter-by-letter - strict alphabetical order. However, we make an exception in the case of phrasal verbs and associated idioms - such as put apart, put back, put by, etc - where it would be silly not to respect the parallels in the way these verbs are used. As a consequence of this decision some lexical items appear out of strict alphabetical order: putter-on, for example, is placed after put upon in the A--Z listing.
- We treat all the senses of a word-form under the same heading, regardless of whether they belong to different etymological paths or not. This is an important difference from, say, Onions’ approach. He has bay [= part of a house] and bay [= noise of a dog] as completely different headwords, on the grounds that they have distinct etymological origins. We simply recognize them as different senses of the headword bay. We find that there are too many cases of uncertain etymology and too many cases where the meaning distinction is unclear for the etymological principle to be consistently implemented.
- We distinguish headwords according to their different grammatical functions - their word-class, or part-of-speech. Many words appear in more than one word-class; indeed, this is one of the features of Shakespeare’s style. A word like wanton is used as noun, adjective, verb, and adverb. When a form appears in several functions, we give noun (n.) uses first, then adjective (adj.), then verb (v.), then adverb (adv.), then others.
- We do not generally include elided forms as headwords - forms where a letter (or letters) has been replaced by an apostrophe, such as o’er, e’en, and ‘lege [= allege]. In such cases, the headword is shown in its full form, to enable readers to find it easily. The alternative, to put words beginning with o’er- at OE- and those with over- at OV- - the procedure followed by Onions, for example, though not by the OED - we find highly confusing. In the quotations, of course, the presence of elided words is shown as they appear in the texts.
- We give only occasional guidance about the pronunciation of a headword. The question of how words were pronounced in Early Modern English is contentious, as is the issue of whether we should represent an older version or a modern equivalent. There can also be several variations in a word’s pronunciation, according to the demands of the metre. We have therefore restricted our guidance to just a few cases (such as disme, foison, guerdon) where we felt an indication of the general character of a pronunciation would be helpful.
The entries in the A--Z section group themselves under 13,626 headwords. Just over 10,000 of these are single-sense entries; but the remainder have anything from 2 to 22 senses. Some of the very common verbs have a remarkable semantic range, just as they do today - break, pass, and draw, for example, with take achieving the highest number, just as it does in modern English dictionaries.
When a headword has more than one sense, there are several ways in which the senses might be ordered. A historical dictionary, for example, puts the oldest senses first. In our case we felt a more useful criterion was frequency: senses which occurred most often in the texts were put first, to give users a better chance of having their enquiry quickly answered. A good example is favour, where we distinguish nine senses, illustrated by 63 textual examples; but 38 of these examples illustrate favour in the sense of ‘facial appearance’ or ‘looks’, so this is given as sense 1. Semantically related senses then follow; and generally we aim to keep clusters of related senses together.
On the other hand, there are many cases where neither frequency nor any kind of semantic logic provides a principled way of ordering senses; and in such cases we have no alternative but to list them arbitrarily. We have, on the whole, opted for recognizing sense distinctions, when these are warranted, even if the distinctions are on occasion somewhat fine. This is because we want our dictionary to preserve as many of the nuances of individual contexts as possible. It would be perfectly possible to group the various adjectival senses of wanton, for example, into just two or three types; but we give twelve. We recognize some quite restricted senses of the word, accordingly, such as its sense of ‘equivocal’ in the collocation of ‘wanton words’ in Twelth Night, or the sense of ‘badly behaved’ in the collocation of ‘wanton boys’ in King Lear. But we feel this level of discrimination is ultimately more useful than lumping together all such examples into a general-purpose entry with some vague sense of, say, ‘irresponsible’. In this way, we believe our approach is closer to the approach of text editors, who pay a great deal of attention to individual nuance in this way. Indeed, our procedure, relying as it does on the observations of many individual editors, could hardly be otherwise.
We were scrupulous in checking the sense of each word against the range of usages listed in the OED, insofar as this is possible. It is not uncommon to find editors ascribing a sense to a Shakespearian word which, on the best evidence, would not have been available in Elizabethan times. Although the first recorded usage of a word or sense is never an infallible guide - most words will have been in spoken currency for several years before they turn up in a written text - there is nonetheless something unsettling when we see words being ascribed senses whose first recorded use in the OED is a generation or so after Shakespeare’s death, or even later. We recommend standing on the shoulders of the OED editors at all times, for one can definitely see so much further as a result. On the other hand, this kind of checking is extremely time-consuming, especially for those words which have dozens of potentially relevant senses. It is the main reason why our own project took so long to complete.
At the same time, one must use the OED resource critically. To begin with, not all of Shakespeare’s word-forms are listed in that work (especially a number of compounds, such as after-hours, and various prefixed forms) and some of Shakespeare’s usages conflate several of the senses that the OED distinguishes. But the point is a more general one. No dictionary should be treated as god: all dictionaries are compiled by people who have good days and bad days (we have first-hand experience, in this respect), and whose definitional results range from brilliant to inadequate. We quickly found that we had to devise our own procedure for definitions. The single-synonym approach used in several glossary lists is patently inadequate when we are dealing with someone whose vocabulary is so allusive; and the long list of glosses which the OED gives to many word meanings introduces a degree of irrelevance which we need to eliminate. We needed something in between.
In the event we opted for a system which we call lexical triangulation - adapting a standard mapping technique to linguistic use. For most headwords we have provided three glosses. Because there are no such things as perfect synonyms, each gloss provides a slightly different slant on the sense of the headword we are trying to capture and thus enables us to focus more sharply on the word’s semantic ‘core’. Bale, for example, is glossed as ‘sorrow, pain, misfortune’; fardel as ‘burden, load, bundle’. To provide a concrete analogy from the world of lighting: one torch shone at a person highlights one side of the face, but a combination of three torches from different angles produces a brightly lit subject. Our hope is that after reflecting on all three glosses, it will be possible to see the meaning of a word more clearly, and develop a sharper sense of where it is located in the semantic scheme of things. The approach works best when comparing a succession of senses within a single headword, where the extra words bring the contrasts in meaning into sharper relief.
We have found this approach illuminating, and we hope others do too. Offering a series of glosses gives people a choice when paraphrasing a piece of text - something which both students and editors need to do. Our three glosses are sometimes very close in meaning, but each word has its individual aesthetic properties and linguistic associations (collocations) which could make it the synonym of choice for someone engaged in interpreting a text. It also allows people to reflect on the direction in which a piece of textual interpretation should go: Hamlet’s fardel is more in the direction of ‘burden’; Autolycus’s is more ‘bundle’. In many cases, it is not a question of choice: two or three of the nuances turn out to be equally relevant - or indeed, as editorial notes frequently say, the whole point of the usage is to see the interaction between the constituent senses. We have frequently been struck by the way in which multiple glosses suggest directions of interpretation which had not previously occurred to us, and we hope that the approach will prove to be a helpful editorial tool.
No lexical approach works perfectly all the time with an author whose vocabulary is as diverse as Shakespeare’s, and in several cases the triangulation technique is simply irrelevant. Sometimes it is impossible or undesirable to do anything other than give a single gloss for a headword, especially for very specific and concrete entities, such as the name of a plant. Faced with daffadilly, all we need to know is that it means ‘daffodil’, and no other glosses are necessary. Similarly, lip is glossed as ‘kiss’, upreared as ‘standing on end’, and seniory as ‘seniority’. And of course, for a book we were always constrained by length. Many items turn out to require not single-word but longer two-word glosses, such as shore 3 meaning ‘waterside dump, sewage channel’, and if two of these provided the degree of precision required, we would drop our search for a third item. Sometimes we had to drop the synonym-search completely, and go for whole phrases or sentences to explain a meaning. Words like harpy and country base need descriptions and explanations rather than lexical equivalents.
There are many variants in our glossing pratice, accordingly; but the majority of our definitions conform to the triangulation model. Finally, in relation to definitions, we sometimes have to acknowledge lexicographical defeat, in the form of such comments as ‘[unclear meaning]’. It is well-recognized that several Shakespearian words have senses which are either hotly disputed or totally opaque, and we have tried to reflect this in our label. But when we call a word or usage ‘unclear’ or ‘debated’, it refers to any of three possibilities: (i) nobody has a clue what the item means; (ii) some people think they know but are uncertain; (iii) an individual editor is strongly in favour of a particular meaning, but another editor - equally strongly - disagrees. In short, our label is no more than a warning to readers: it identifies a word which needs a fuller discussion than a glossary can provide. In such cases, readers need to go to the editorial notes of individual editions for a real grasp of the situation.
All examples of Shakespearian usage are treated in the same way.
- They begin with a text identifier, using one of our recognized abbreviations.
- This is followed by a line reference, using the corpus texts described above. Where a quotation involves more than one line of text, the line reference is to the line containing the headword. For readers using this glossary in relation to other editions than our chosen texts, some variation is of course to be expected, especially in the prose sections of plays, where - depending on the typographical setting - a line reference may differ by several lines. Paying attention to the context (see below) can be a useful means of quick location, in such cases.
- Care needs also to be taken with those editions where significantly different editorial decisions have been made about scene divisions (such as the opening scenes of Cymbeline) or about the balance of content taken from Folio and Quarto texts. Where we include variant forms used by editors other than those found in our text corpus, we give the line-reference to our corpus edition, but put the variant into the quotation, noting the alternative in following parentheses.
- Words which appear in stage directions are given the line number of the first line of speech following the direction.
- We provide context for each quotation, in the form of information in square brackets immediately following the line reference. The context tells you who is speaking to whom, and - where this is unclear from the quotation - about whom or what. Many quotations contain such context-dependent words as he, that, and there (what linguists refer to as deictic forms), and these make it difficult to develop a full sense of what is being said. It is all very well illustrating the use of hallow [= bless] by ‘I will hallow thee for this thy deed’ (2H6 IV.x.65), but the quotation means so much more when you know the context - ‘[Iden to his sword, of killing Cade]’. We therefore give as much context as is needed to make a quotation self-contained. We have also found these contexts helpful in providing a mnemonic for recalling the location of the word being glossed, in the setting of a play.
- Certain other contextual conventions are used. In particular, when a character is the only person on stage, we use the convention illustrated by: ‘[Hamlet alone]’. When a character is speaking to himself/herself, and other people are on stage, we use the convention: ‘[Hamlet to himself]’. When characters speak to each other so that others should not hear them, we use the aside convention from the texts themselves, as illustrated by: ‘[Smith aside to his companions]’.
Finally, entries make limited use of cross-referencing, always signalled by the >> symbol. Cross-references are of three kinds:
- a cross-reference to an alternative headword, where information can be found, used especially in cases of spelling variants, as in: bankrout (n.) >> bancrout (n.)
- a cross-reference to a semantically related word, as in: bate >> abate; best >> meanest
- a cross-reference to a panel, appendix, or the FEW (frequently encountered words) section, in which other information of the same kind is collected.
This Website is not (yet) a guide to Shakespeare’s language as a whole. In particular, it contains little information about pronunciation and grammar. Its focus is exclusively on vocabulary, and on those aspects of usage which influence its choice and form (such as the topics covered in the panels). It is moreover a site compiled for a particular purpose, to aid those who want to explore the richness of meaning found in the texts.
It contains a great deal of detail. But we would not want the existence of our glossary to be interpreted as support for the view that Shakespeare’s language is intrinsically difficult or impenetrable, as we see stated in the media from time to time.
It is perfectly possible to go to a Shakespeare play, with a little or no awareness of Early Modern English vocabulary, and have a great time. There are many stretches of text where the vocabulary is virtually identical with that used today, and many more where the presence of the occasional ‘difficult word’ is not noticed because the context makes the meaning of the utterance perfectly clear. Even when the distinctive vocabulary begins to pile up, as in some of the more complex poetic passages, the quality of the acting can transcend the limitations of lexical obscurity. We are reminded of our experience of operas sung in a foreign language (in the days before surtitles) where we have been profoundly moved despite the unintelligibility. There are actually very few passages in Shakespeare where the combination of alien grammar and vocabulary makes the text comparable to it being in a foreign language.
Some of the insult-sequences are perhaps the nearest (such as Doll’s ‘basket-hilt stale juggler’), but such utterances never leave the audience in any doubt as to their pragmatic force.
On the other hand, there are many many places where our appreciation of what is happening in a play can be immensely increased by a sharpened awareness of the language used. We have found this ourselves, in the dual experience of reading the texts as literature and as being involved in the texts as drama. Since beginning to write this book, we have between us watched several dozen plays at the various seasons of Stratford, the Globe, and elsewhere, and we have noticed the way our lexical investigations have increased our sensitivity to so much of what we see happening on stage. In particular, Ben Crystal has repeatedly found that his work on the book has added a fresh dimension to his work as an actor. We therefore hope that other people using this book will also find their enjoyment of the plays, as literature or drama, enhanced by its use.